Identifying Strategies for Strengthening the Health Care Workforce in Virginia (RAND RR-A2903-1)
Summary
A RAND Health Care research report, RR-A2903-1, titled Identifying Strategies for Strengthening the Health Care Workforce in the Commonwealth of Virginia, dated October 2023 and marked as a prepublication copy of September 29, 2023. It was prepared for the Virginia Health Workforce Development Authority by Megan Andrew, Brian Briscombe, Raffaele Vardavas, Nazia Wolters, Nabeel Qureshi, Wilson Nham and Mahshid Abir. The report assesses shortages in the nursing, primary care and behavioral health workforces in Virginia and identifies potential interventions to address them.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Cite this document
“2023 Rd481.”
Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/2023-rd481-e78ff599c0/
, SHA-256 c2fb4564b90e2515a4f72eaa1c7a404b018df232e166172220dfea852871c920.Full text
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Research Report
Identifying Strategies for Strengthening the
Health Care Workforce in the
Commonwealth of Virginia
Megan Andrew, Brian Briscombe, Raffaele Vardavas, Nazia Wolters, Nabeel Qureshi,
Wilson Nham, Mahshid Abir
RAND Health Care
RR-A2903-1
October 2023
Prepared for the Virginia Health Workforce Development Authority
This document has not been formally reviewed, edited, or finalized for public release. It should not be cited without the permission of
the RAND Corporation. RAND’s publications do not necessarily reflect the opinions of its research clients and sponsors. is
a registered trademark.
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
ii
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
About This Report
Shortages in the nursing, primary care, and behavioral health workforces are an ongoing and
widespread issue in the United States. This report assesses shortages in these health care
workforces in the Commonwealth of Virginia and identifies potential interventions to address
these shortages. The report is the culmination of the second and third phases of a larger study. In
the first phase of the study, the Virginia Health Workforce Development Authority and its
partners focused on initial data collection and analysis related to challenges faced by Virginia’s
health care workforce. This report extends the first phase to identify specific interventions for (1)
retaining existing health care workers, (2) recruiting first-time future health care workers, and (3)
maximizing the ability of the Virginia health care workforce to meet the Commonwealth’s needs
via structural efficiencies—for example, better geographic distribution of the Commonwealth’s
workforce via economic and social policy interventions or telehealth practice. We identified
interventions to improve retention, recruitment, and structural efficiency based on an
environmental scan of peer-reviewed and grey literature; primary multi-stakeholder conference,
interview, focus group, and survey data; and statistical analysis and simulation based on data
from the Commonwealth and other relevant sources. Importantly, a wide variety of Virginia
health care workforce stakeholders participated in and supported the research by attending the
study conference, participating in interviews, focus groups, and serving on the study’s advisory
board. The participation of these Virginia stakeholders was critical to identifying solutions that
better fit Virginia’s specific health care workforce landscape and population needs.
Appendixes A–H to this report, which provide details on the methods and analysis used for
this research, are contained a separate annex, available at www.rand.org/t/RRA2093-1.
This research was funded by the Virginia Health Workforce Development Authority
(VHWDA) and carried out within the Access and Delivery Program in RAND Health Care. The
Virginia General Assembly established VHWDA in 2010 in order to identify and address health
workforce issues in the Commonwealth. As a public entity, VHWDA exercises public and
essential governmental functions to secure the health, welfare, convenience, knowledge, benefit,
and prosperity of Virginians. VHWDA’s mission is to “facilitate the development of a statewide
health professions pipeline that identifies, educates, recruits, and retains a diverse,
geographically distributed and culturally competent quality workforce” (VHWDA, undated-a).
VHWDA accomplishes this through core functions outlined in the Code of Virginia (Virginia’s
Legislative Information System, undated-c).
RAND Health Care, a division of the RAND Corporation, promotes healthier societies by
improving health care systems in the United States and other countries. We do this by providing
health care decisionmakers, practitioners, and consumers with actionable, rigorous, objective
iii
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
evidence to support their most complex decisions. For more information, see
www.rand.org/health-care, or contact
RAND Health Care Communications
1776 Main Street P.O. Box 2138
Santa Monica, CA 90407-2138
(310) 393-0411, ext. 7775
RAND_Health-Care@rand.org
Acknowledgments
We would like to thank the VHWDA’s Harrison Hayes, Amy Paul, and Art Kellermann for
facilitating our acquisition of the needed data and connecting our team with the key stakeholders
in Virginia for this study.
This report would not have been possible without the collaboration and assistance of the
many practitioners, educators, and other experts who participated in the study’s multi-
stakeholder conference, interviews, and focus groups. We thank Carrie Sutter and Peggy
Maddox from George Mason University’s Center for Healthcare Workforce; Yetty Shobo and
James Jenkins from the Virginia Department of Health Professions’ Healthcare Workforce Data
Center; Olivette Burroughs from the Virginia Department of Health, Office of Health Equity;
and Sharon Alexander, Andre Tolleris, and David Vaamonde from the Virginia Hospital and
Healthcare Association for sharing their data for use in this study. We are grateful to Melody
Eaton from the Virginia Nurses Association for sharing her expertise and insights.
We thank the members of the study’s advisory board for all their contributions and insights:
Marcelle Davis, William Harp, William Hazel, Jr., James Jenkins, Cynthia Lawrence, Nathalie
Molliet-Ribet, Sandra Serna, and Caroline Sutter.
We are grateful to RAND’s Pedro Nascimento de Lima for his assistance with analyses and
Emily Lawson and Rachel Widom for their help with preparing this report. Thank you to
RAND’s Kristin Leuschner, David Adamson, and Lisa Turner for their editorial assistance with
the report.
We extend our appreciation to reviewers of this report, Christina Brooks and RAND’s
Federico Girosi, Jeanne Ringel, and Paul Koegel, whose expertise and insights helped improve
this work.
We dedicate this report to the health workers of the United States—a national treasure. We
hope that the results of this work can help health care workers in the Commonwealth and other
states thrive in their work lives and pave the way for others to join their ranks.
iv
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Summary
Issue
Like the United States as a whole, Virginia faces a significant shortage of health care workers
in primary care and behavioral health care across a range of health care settings. For example,
102 of 133 Virginia counties and equivalents are federally designated primary care Health
Professional Shortage Areas (HPSAs), accounting for 29 percent of the Commonwealth’s
population. Many of these shortage areas are in urban and rural regions, as opposed to suburban
areas (U.S. Department of Health and Human Services, 2023). Similarly, 93 of Virginia’s 133
counties (and equivalents) are federally designated mental health care HSPAs, and about 32
percent of Virginians live in these areas (Virginia’s Legislative Information System, 2023;
Virginia Health Care Foundation, 2022a; Virginia Health Care Foundation, 2022b).
In response to these health care workforce shortages, Virginia has passed nearly 60 related
bills over the past ten years, with a notable increase in the number of health care workforce bills
in 2023–2023 (BillTrack50, undated; National Conference of State Legislatures, 2023). Recently
enacted legislation includes interstate compacts (i.e., agreements) for licensing and scope-of-
practice regulations for behavioral health professionals, medical assistance payment for remote
patient monitoring, expanded midwife practice agreements, and more.
Because of the limited evaluation research on existing practices, programs, and policies
(hereafter “interventions”), it is unclear whether past and current interventions are likely to
succeed in addressing the workforce shortages in Virginia. An evidence-based approach that
carefully evaluates interventions across different contexts is needed to identify the most
promising interventions to help retain and expand the health care workforce in Virginia.
The objective of this study was to identify interventions that can help the Virginia Health
Workforce Development Authority (VHWDA) address nursing, primary care, and behavioral
health workforce challenges and to define the role that the authority can play in facilitating such
solutions. To accomplish this goal, we applied an analytic framework to existing or potential
interventions for retaining, recruiting, and improving the structural efficiency (i.e., the efficient
allocation of workers within a given health care system and efficient health care output using
technology, policy changes, or other structural adjustments) of the nursing, primary care, and
behavioral health workforces in Virginia. In this report, we highlight which interventions
VHWDA should prioritize based on its desired outcomes and policy goals.
Overarching findings are shown in Box S.1.
v
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Box S.1. Overarching Findings
1. Team-based care needs to be promoted and supported in Virginia through regulatory
and reimbursement reform.
2. To help populate teams for team-based care, “health care highways” need to be
embraced to help create career ladders for upward movement.
3. Barriers to pursuing health care careers need to be removed, including the number of
supervised training hours and related trainee expenses—particularly in behavioral
health professions.
4. Tuition support and other incentives need to be expanded—especially for students in
most need—and effectively advertised.
5. Support for faculty and clinical preceptors needs to be enhanced and more flexibility
needs to be built in such positions.
Approach
We used results from a review of peer-reviewed and grey literature and analysis of
multidisciplinary conference, interview, focus group, and quantitative data sources to arrive at
recommendations for promising interventions to be implemented in Virginia for health care
workforce retention, recruitment, and structural efficiency. Furthermore, we developed a system
dynamics economic model to assess the shortages of nurses, primary care practitioners, and
behavioral health providers in the next five, ten, and 15 years at baseline (i.e., if the status quo
persists) and if (1) retention barriers (such as fatigue, distress, and disengagement) are decreased,
(2) recruitment is increased, (3) wages are increased, or (4) all three interventions are
implemented.
Key Findings
The results from the system dynamics model show that, if current trends persist, nursing,
primary care, and behavioral health workforce shortages will increase across Virginia.
We forecast that, by reducing the barriers that cause reluctance to remain working in the field
of nursing by half—barriers such as workplace violence, trauma, and stress—Virginia can effect
a rapid increase in the number of registered nurses (RNs) retained in the field. However, over a
period of 15 years, increased recruitment will have a more dramatic effect on workforce
numbers, surpassing the effect of decreased barriers to retention. Increasing nurse wage growth
from an average 2 percent annual growth to 3 percent annual growth also increases the number
of nurses retained and recruited, though not as much as the other interventions do. As expected, a
combination of all three interventions has the largest effect on the forecast of nurse employment,
boosting the total number of full-time equivalent nurses employed in 2038 by a little more than
10,000 above the baseline projection.
vi
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Unlike our forecasts for RNs, our forecast for the number of active primary care workers in
the Commonwealth shows a decline under baseline conditions, making the future shortfall in
primary care workers even more severe than for nursing alone. Similar to our model for nursing,
the largest immediate boost to the workforce would come from cutting in half the number of
barriers that make workers reluctant to remain working in the field. Over the course of 15 years,
the effect on total employment of increasing wage growth catches up to the effect of decreasing
retention barriers. The effect of increasing recruitment is more modest because baseline
recruitment begins at very modest levels. An ambitious combination of all three interventions
(the multi-intervention scenario) would boost total employment by more than 2,000 workers,
forecast to slowly decline in the coming years. Implementing any one of the three interventions
promises to keep the existing workforce numbers at 2023 levels, but only by combining all three
interventions can Virginia produce a sustainable increase in the number of behavioral health
workers employed.
Recommendations
VHWDA should convene and lead a coalition of key stakeholders to implement the
following recommendations to retain and expand the nursing, primary care, and behavioral
health workforces in Virginia. These stakeholders include, but are not limited to, the Virginia
legislature, state agencies, health care organizations, and educational entities. Implementation of
many of the proposed interventions will require cross-sector, including public-private,
collaboration. To ensure effective and efficient implementation of these recommendations,
VHWDA should serve as the coordinating center for the coalition to help gather the needed
collaborations, data, and funds and to leverage other political and social capital in the
Commonwealth.
The guide to intervention recommendations by strength of evidence found in this study is
outlined in Table S.1. Tables S.2 and S.3 list retention and recruitment interventions,
respectively.
Table S.1. Guide to Intervention Tiers
Tier 1 interventions are strongly supported by the analyses. The interventions are supported by the
literature review (evaluation studies) and were identified multiple times across the multi-stakeholder
conference, interviews, and/or focus group data.*
Tier 2 interventions are moderately supported by the analyses. These interventions are supported by
the literature review (non-evaluation studies) and were identified multiple times across the multi-
stakeholder conference, interviews, and/or focus group data.
Tier 3 interventions have some support in the literature review or were identified in the multi-
stakeholder conference, interviews, and/or focus group data.
* Support for an intervention in the qualitative work indicates that the intervention has valence—or is deemed an
attractive option by interviewees—but it does not indicate whether the intervention will have a positive effect on
retention, recruitment, or productivity.
vii
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Table S.2. Interventions for Health Care Workforce Retention
Retention
Intervention Category
(Key Implementation
Tier Stakeholders) Specific Intervention
Tier 1 Resiliency and well-being 1. Implement programs to address mental health and well-being
(health care organizations) among health care workers—especially in behavioral health
professions
2. Address underlying causes of health care worker stress and
burnout, such as inappropriate patient-provider ratios, throughput
requirements, documentation requirements, low and/or differential
compensation for care provided, and lack of basic supplies.
Tier 1 Scope of practice and 1. Reevaluate license restrictions for physician assistants (PAs)
practice agreement and NPs to increase the appeal of primary care for these health
(Legislature) professionals and to expand primary care capacity.
Tier 2 Team-based care 1. Implement team-based care to mitigate the stresses of high patient-
(Health care organizations) to-provider ratios and large patient panels and take patient complexity
and acuity into account when assigning patients to providers.
Tier 2 Pay increase and other 1. Increase pay for nursing, primary care, and behavioral health
incentives providers and provide appropriate benefits—including health
(health care organizations) insurance coverage commensurate with risks involved in health care
delivery (e.g., exposure to disease, despair, and violence). Provide
assistance with transportation, child care, and housing.
Tier 2 Documentation burden 1. Evaluate technologies—such as voice recognition and artificial
(health care organizations, intelligence—that may assist with reducing documentation burden.*
systems, legislature) 2. Revisit documentation requirements to identify areas of redundant
and/or unnecessary documentation that could be removed from
documentation requirements.
Tier 2 Workplace diversity 1. Develop retention strategies centered around creating more
(health care organizations, diverse workplaces for nursing, primary care, and behavioral
educational entities) health.*
Tier 2 Personal/professional 1. Invest in making professional development opportunities
development available to staff and increase recognition efforts.
(health care organizations)
Tier 3 Workplace violence 1. Evaluate legislation to increase legal consequences related to
(health care organizations, violence toward health care providers—on par with those currently
legislature) in place for assaulting prehospital providers and first responders.*
2. Evaluate health care organization policies around violence toward
health care providers and identify strategies for more strict
enforcement of those policies.
Tier 3 Cross-sector collaboration 1. Identify and leverage opportunities for cross-sector
(health care organizations, collaboration between the health care sector, educational
educational entities) organizations, and public and private entities to increase opportunities
and funding for nursing, primary care, and behavioral health exposure
and training.
Tier 3 Continuing education 1. Evaluate mechanisms to increase continuing education
(health care organizations, opportunities and funding—especially through leveraging
educational entities) collaboration between health care organizations and educational
entities.
* Designates an innovative intervention.
viii
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Table S.3. Interventions for Health Care Workforce Recruitment
Recruitment
Intervention Category
(Key Implementation
Tier Stakeholders) Specific Intervention
Tier 1 Scholarship, loan forgiveness, 1. Increase scholarships opportunities in Virginia.
and other support 2. Raise public awareness of loan forgiveness and scholarship
(private entities, health care opportunities.
organizations, educational 3. Develop mechanisms for rural applicants and first-generation
entities) students to receive loan forgiveness up front.*
4. Include support for housing, transportation, and food for rural and
first-generation students.
Tier 1 Wages and regulations for 1. Increase wages for nursing faculty and preceptors to levels
educators similar to engineering, business, and other faculty that receive higher
(private entities, health care wages.
organizations, educational 2. Loosen strict behavioral health preceptorship/licensed
entities, Virginia Department supervising requirements, such as educational degree
of Health Professions) requirements for preceptors, preceptor-to-student ratios, and need for
preceptors in community settings.*
3. Create mechanisms to pay for behavioral health preceptorship
and licensed supervisors to remove this financial burden from
students/trainees.*
Tier 1 Work-based learn and earn 1. Offer work-based learning, paid and unpaid. Health care
(health care organizations, occupations have historically included more work-based learning in
educational entities) education and training than other occupations (e.g., clinical rotations,
clinical intern or clerkships, and residencies). Health care
stakeholders are increasingly interested in expanding these options
across different health care occupations—particularly in nursing.
Tier 2 Recruiting youth and 1. Develop (and/or expand on current programs) to expose K–12
members of underserved children to health professions. Such programs should include
communities collaboration between educational and health care entities.
(health care organizations, 2. Specific outreach strategies for youth from underrepresented
educational entities) communities are needed. Such interventions should aim to include
both children and their parents in outreach efforts and should, ideally,
be delivered by health care professionals from similar backgrounds.*
Tier 3 Preceptor/licensed supervisor 1. Leverage nursing and behavioral health retirees to increase pool
capacity of preceptors/licensed supervisors.*
(health care organizations, 2. Create mechanisms to compensate preceptors/licensed
educational entities) supervisors—including ability to bill for supervised visits.*
Tier 3 Outreach and public image 1. Change the image of health professions through public
(health care organizations, education campaigns. Such efforts should focus on highlighting the
educational entities) rewarding nature of nursing and behavioral health work, and
removing the stigma associated with becoming a behavioral health
professional.
Tier 3 Online instruction 1. Increase opportunities for online learning to increase access and
(health care organizations, flexibility to local and nonlocal clinical introductory level education.
educational entities)
Tier 3 New pathways 1. Create multiple entry ramps and educational pathways to nursing
(health care organizations, and behavioral health careers as a recruitment strategy. This includes
educational entities) developing academic programs at different levels, such as associate
and bachelor’s degrees for behavioral health, to provide more
immediate relief to the workforce shortage.
2. Offer paid internships at the high school level so students can be
exposed to health professions; match undergrads with medical
students and promote mentorship alongside internship.
ix
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Recruitment
Intervention Category
(Key Implementation
Tier Stakeholders) Specific Intervention
Tier 3 Licensing exam and licensure 1. Significantly truncate the duration of application for licensure in
(Department of Health behavioral health fields and allow students that need to retake the
professions) licensure exam to have more frequent opportunities to sit for the
examination.
Tier 3 Advanced education/training 1. Increase access to simulation-based learning in nursing.
technology 2. Increase access to training opportunities in telepsychology.
(health care organizations,
educational entities)
* Designates an innovative intervention.
Interventions for Increasing Structural Efficiency and other Innovations
In addition to retention and recruitment interventions, the Commonwealth should consider
strategies that optimally use the current workforce through increasing structural efficiency. Some
of the interventions that we identified have little evidence behind them to determine whether they
increase the number of available health care professionals or otherwise address critical health
care workforce issues (e.g., burnout). Nonetheless, some are innovative and worth considering:
• Incentivizing existing nursing, primary care, and behavioral health workers to move from
areas of more provider capacity to those of less capacity in Virginia may help to achieve
an allocation of labor that maximizes overall health outcomes.
- In the case of primary care and behavioral health, some of this reallocation of labor
may be achieved by increasing opportunities for telehealth for providers, effectively
(though not physically) moving labor to the locations with the highest need.
• Removing work barriers—such as reducing documentation, reporting, and regulatory
requirements and/or leveraging technologies such as voice recognition and artificial
intelligence to help with these tasks—can both create efficiency and potentially increase
worker job satisfaction to reduce attrition.
• One strategy that can help increase structural efficiency in both primary care and
behavioral health is to increase locations in the Commonwealth that offer both services in
the same setting—implementing a team-based approach to more efficiently address
related mental health illnesses.
Transitioning service members (as well as military spouses)—many of whom are qualified
health care experts—into civilian employment and anchoring them in Virginia is a strategy that
may help retain talent in the Commonwealth and expand the health care workforce.
Furthermore, regular health care workforce data-tracking may be an effective strategy to
inform interventions for retention, recruitment, and structural efficiency and to pivot as needed to
ensure that the health care needs of Virginia residents are met. Health care workforce data exist
in Virginia Commonwealth University’s Virginia Ambulatory Care Outcomes Research Network
(ACORN) (Virginia Commonwealth University, undated-a), in the Virginia Longitudinal Data
System (Virginia Longitudinal Data System, 2023), and at the George Mason University Center
x
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
for Health Workforce (Mason Center for Health Workforce, 2023). Similar data-tracking should
be implemented by health professions colleges and universities to inform strategies to retain
graduates in the Commonwealth.
Also, although there are promising programs in Virginia focused on health care workforce
retention and recruitment, there is currently no mechanism for tracking and disseminating these
programs. A clearinghouse for tracking and sharing current practices and programs can facilitate
dissemination of promising interventions. The newly established Virginia Department of
Workforce Development and Advancement will play a key role in supporting workforce data
infrastructure across different industries.
Health Care Workforce Retention and Recruitment in Rural Settings
Many of the identified interventions apply to urban, suburban, and rural settings in Virginia.
However, there are some unique considerations for health care workforce development in
Virginia’s rural communities.
• Recruiting of local students from these communities as a retention strategy to bolster
rural workforces—rather than convincing providers from elsewhere to move. Such
recruitment would be most effective if it is paired with providing up-front financial
support for education and life-expenses.
• Developing infrastructure, such as housing, transportation, and child care, in rural areas
as both a retention and recruitment tool.
VHWDA should consider convening and leading a coalition of key stakeholders to
implement the above recommendations to retain and expand the nursing, primary care, and
behavioral health workforces in Virginia. These stakeholders include, but are not limited to, the
Virginia legislature, state agencies, health care organizations, and educational entities.
Implementation of many of the proposed interventions will require cross-sector, including
public-private, collaboration. To ensure effective and efficient implementation of these
recommendations, VHWDA should serve as the coordinating center for the coalition to help
gather the needed collaborations, data, and funds and to leverage other political and social capital
in the Commonwealth.
In closing, Virginia can address these shortages by implementing practical strategies to
enhance retention and recruitment, and by changing (or removing) rules, regulations, and
policies that degrade efficiency, make it difficult for eager students to enter the workforce, or
pull dedicated and experienced nurses, primary care, and behavioral health professionals away
from what they do best: caring for patients.
xi
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Contents
About This Report.......................................................................................................................... iii
Summary ......................................................................................................................................... v
Figures, Tables, and Boxes .......................................................................................................... xiii
Chapter 1. Background, Objectives, and Approach ........................................................................ 1
Introduction ............................................................................................................................................... 1
Overview of Approach .............................................................................................................................. 5
Key Definitions and Scope ........................................................................................................................ 5
The Health Care Workforce as a System: Supply Outflows, Inflows, and Structural Efficiency ............ 6
Methods ..................................................................................................................................................... 8
Organization of this Report ..................................................................................................................... 10
Chapter 2. Key Findings ............................................................................................................... 11
Interventions for Workforce Retention ................................................................................................... 29
Interventions for Workforce Recruitment ............................................................................................... 39
Interventions That Augment the Workforce by Increasing Efficiency ................................................... 56
Chapter 3. Recommendations for Strengthening the Virginia Health Care Workforce ............... 59
Key Findings ........................................................................................................................................... 60
Interventions by Strength of Evidence and Key Stakeholders for Implementation ................................ 64
Chapter 4. Intervention Implementation ....................................................................................... 69
Implementation Timeline ........................................................................................................................ 69
Communication Tools ............................................................................................................................. 69
Abbreviations ................................................................................................................................ 71
References ..................................................................................................................................... 72
Appendixes A–H to this report, which provide details on the methods and analysis used for
this research, are contained a separate annex, available at www.rand.org/t/RRA2093-1.
xii
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figures, Tables, and Boxes
Figures
Figure 1.1. Percentage Share of Populations Living in a Primary Care Profession Shortage Area,
Mid-South Atlantic States Plus West Virginia and the District of Columbia, 2022............... 2
Figure 1.2 Percentage Share of Populations Living in a Mental Health Care Profession Shortage
Area, Mid-South Atlantic States Plus West Virginia and the District of Columbia, 2022 ..... 3
Figure 1.3. Number of Bills Passed in Virginia Related to the Health Care Workforce, by
Calendar Year ......................................................................................................................... 4
Figure 1.4. The Health Care Workforce System: How Inflows, Outflows, and Structural
Efficiency Relate to Health Outcomes .................................................................................... 7
Figure 2.1. Full-Time Equivalent Nurses Employed in Virginia 2023–2038 Median Forecast in
Five Scenarios ....................................................................................................................... 16
Figure 2.2. Gap Between Employment and Demand for Nurses Employed in Virginia 2023–2038
Median Forecast in Five Scenarios ....................................................................................... 17
Figure 2.3. Full-Time Equivalent Primary Care Workers Employed in Virginia 2023–2038
Median Forecast in Five Scenarios ....................................................................................... 21
Figure 2.4. Gap Between Employment and Demand for Primary Care Workers in Virginia,
2023–2038 Median Forecast in Five Scenarios .................................................................... 21
Figure 2.5 Full-Time Equivalent Behavioral Health Workers Employed in Virginia, 2023–2038
Median Forecast in Five Scenarios ....................................................................................... 24
Figure 2.6 Gap Between Employment and Demand for Behavioral Health Workers in Virginia,
2023–2038 Median Forecast in Five Scenarios .................................................................... 25
Figure 2.7. Statewide Nursing, Behavioral Health, and Primary Care Workforce Shortages as a
Share of Each Respective Workforce, 2023–2038 Baseline Projections ............................. 27
Tables
Table S.1. Guide to Intervention Tiers.......................................................................................... vii
Table S.2. Interventions for Health Care Workforce Retention .................................................. viii
Table S.3. Interventions for Health Care Workforce Recruitment ................................................ ix
Table 2.1. Identified Retention Challenges in Nursing, Primary Care, and Behavioral Health in
Virginia ................................................................................................................................. 13
Table 2.2. Identified Recruitment Challenges in Nursing, Primary Care, and Behavioral Health in
Virginia ................................................................................................................................. 14
Table 3.1. Guide to Intervention Tiers .......................................................................................... 65
Table 3.2. Interventions for Health Care Workforce Retention.................................................... 66
xiii
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Table 3.3 Interventions for Health Care Workforce Recruitment ................................................ 67
Boxes
Box S.1. Overarching Findings...................................................................................................... vi
xiv
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Chapter 1. Background, Objectives, and Approach
Introduction
The United States needs more than 17,000 additional primary care health workers and 8,000
additional behavioral health care workers just to meet current demand (U.S. Department of
Health and Human Services, 2023). According to the American Medical Association, the United
States will need an additional 17,800 to 48,000 primary care physicians by 2034. By 2030, there
will likely be a national nursing shortage of over 63,000 registered nurses (National Center for
Health Workforce Analysis, 2022). Behavioral health worker shortages are less severe, in
absolute numeric terms, in comparison to those projected for physicians and nurses; however, the
U.S. Department of Health and Human Services’ Health Resources and Services Agency
highlights shortages in psychiatrists and addiction counselors, in particular (Health Resources
and Services Administration, 2022). Shortages in the U.S. health care workforce will likely pose
an ongoing challenge into the near future (Institute of Medicine, 2011; Zhang et al., 2020;
Duquesne University School of Nursing, undated; Mensik, 2023). Rural areas and areas with
high proportions of children and aging adults are particularly susceptible to the effects of these
shortages (Nguyen et al., 2016; Bennett et al., 2021).
The Commonwealth of Virginia faces a significant shortage of health care workers in
nursing, primary care, and behavioral health across a range of settings. For example, 102 of 133
counties and equivalents are federally designated primary care Health Professional Shortage
Areas (HPSA), accounting for 29 percent of the Commonwealth’s population. Many of these
locations are urban and rural areas (U.S. Department of Health and Human Services, 2023).
Compared with other mid-Atlantic states and West Virginia (Figure 1.1), Virginia has a smaller
percentage share of its population living in a primary care HPSA (Kaiser Family Foundation,
undated-b).
1
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure 1.1. Percentage Share of Populations Living in a Primary Care Profession Shortage Area,
Mid-South Atlantic States Plus West Virginia and the District of Columbia, 2022
SOURCE: Features data from Kaiser Family Foundation, undated-b.
Similarly, 93 of Virginia’s 133 counties (and equivalents) are federally designated mental
health care HSPAs, and about 32 percent of Virginians live in these areas (Virginia’s Legislative
Information System, 2023; Virginia Health Care Foundation, 2022a; Virginia Health Care
Foundation, 2022b).
A notably larger share of Virginia’s population lives in a mental health care HPSA than in
Georgia and South Carolina; the share of Virginia’s population living in a mental health care
HPSA is comparable to that of Florida and North Carolina; and the share of Virginia’s
population in a mental health care HPSA is much lower than that of the District of Columbia
(Figure 1.2). A similar percentage share of Virginians lives in a mental health care HPSA as in a
primary care HPSA (Kaiser Family Foundation, undated-b).
2
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure 1.2 Percentage Share of Populations Living in a Mental Health Care Profession Shortage
Area, Mid-South Atlantic States Plus West Virginia and the District of Columbia, 2022
SOURCE: Features data from Kaiser Family Foundation, undated-b.
In response to its health care workforce shortage, the Commonwealth of Virginia has passed
nearly 60 related bills in the past ten years (Figure 1.3) (BillTrack50, undated), with a notable
increase in the number of health care workforce bills in 2022–2023 (National Conference of
State Legislatures, 2023). Recent legislation includes interstate compacts for licensing and scope
of practice regulations for behavioral health professionals, medical assistance payment for
remote patient monitoring, and expanded midwife practice agreements, among others.
3
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure 1.3. Number of Bills Passed in Virginia Related to the Health Care Workforce, by Calendar
Year
SOURCE: Features data from BillTrack50, undated.
The Virginia budget also included important measures to expand support to the health care
workforce, including an increase in the Medicaid reimbursement rate from 70 to 80 percent of
the federal rate for primary care services in 2022 (Masters, 2022). This measure was adopted to
improve the state’s low spending rank in comparison with other Medicaid state programs and to
address the concern that Virginia tends to allocate 3 to 5 cents of every dollar spent on health
care to primary care, compared with an average of 7 to 8 cents in other states (Masters, 2022).
Other stakeholders, including private foundations, have taken similar steps to address the
health care worker shortage through scholarships and other means. For example, Carilion Clinic,
a health care system, has initiated a training program in Virginia to support junior nursing staff in
registered nurse (RN) training (Carilion Clinic, undated).
Because of the limited evaluation research on existing practices, programs, and policies
(hereafter “interventions”), it is unclear whether past and current interventions are likely to
succeed in addressing the workforce shortages in Virginia. An evidence-based approach that
carefully evaluates interventions across different contexts is needed to identify the most
promising interventions to bolster the health care workforce in Virginia.
The objective of this study is to identify interventions that can help the Virginia Health
Workforce Development Authority (VHWDA) address primary care, behavioral health, and
4
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
nursing workforce challenges and to define the role the authority can play in facilitating such
solutions. To accomplish this goal, we apply in this report an analytic framework to existing or
potential interventions for retaining, recruiting, and improving the structural efficiency (i.e., the
efficient allocation of workers within a given health care system and efficient health care output
using technology, policy changes, or other structural adjustments) of the nursing, primary care,
and behavioral health workforces in Virginia. We then highlight which interventions VHWDA
should prioritize based on its desired outcomes and policy goals.
Overview of Approach
We employed a mixed methods approach that included a review of peer-reviewed and grey
literature (i.e., environmental scan) and the analysis of qualitative interview, focus group, and
quantitative data. We first identified a broad array of potential interventions in Virginia and
beyond, based on the environmental scan, multi-stakeholder conference, interview, focus group,
and survey data collection and analysis. Additionally, we reviewed readily available federal data
and data collected by other organizations in the Commonwealth to evaluate and project regional
and statewide supply of several occupational categories in the nursing, primary care, and
behavioral health workforces. We used a systems framework (discussed in greater detail below)
to organize our analyses, findings, and recommendations. This framework, a system flows
model, is rooted in demography, economics, and systems science and provides a structure for
thinking about and understanding the current stock or supply of health care worker output and
the ways in which this stock or supply might change over time with new workers entering health
care, existing workers leaving health care, or structural changes that affect the health outcomes
(“output”) per hour worked. Retaining existing workers, recruiting new workers, and increasing
structural efficiency are all often cited by analysts as distinct factors critical for helping to ensure
that the Commonwealth’s health care workforce can meet Virginians’ health care needs. For
example, when modeling how “better health care” can be delivered in the primary care setting,
systems researchers Homer and Hirsch have separately modeled “recruitment” and “increased
efficiency” (Homer and Hirsch, 2006).
In the remainder of this chapter, we discuss key definitions used in the research and provide
an overview of the health care workforce as a system.
Key Definitions and Scope
Definitions of Key Occupations Evaluated
Primary care refers to health services that emphasize “health promotion, disease prevention,
health maintenance, education and rehabilitation” (Starfield, Shi, and Macinko, 2005; van Weel
and Kidd, 2018; American Nurses Association, undated). In this report, the primary care
workforce includes internal medicine, family medicine, internal medicine, pediatrics, and
5
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
obstetrics and gynecology. For the primary care workforce, we include physicians (medical and
osteopathic doctors), nurse practitioners (NPs), and physician assistants (PAs).
Behavioral health refers to health services that emphasize the “prevention, diagnosis, and
treatment of mental health and substance use disorders, life stressors and crises, and stress-
related physical symptoms” (American Medical Association, 2022). In this report, we include
psychiatrists, psychologists, counselors, psychiatric nurses, and social workers in defining the
behavioral health workforce.
In this report, nursing is mainly focused on RNs.
Scope of Analyses and Recommendations
This study’s analyses focus on both Commonwealth-wide and region-specific insights and
interventions. However, the recommendations in this study primarily focus on those that confer
Commonwealth-wide net benefits. In other words, if an intervention helps one region by moving
personnel from another region, we consider this to be a neutral intervention from the state
perspective (no net gain of workers), unless that redistribution of labor carries a net statewide
efficiency gain, such as instances in which workers can be encouraged to move from areas of
more supply (e.g., urban and suburban communities) to areas of less supply (e.g., rural
communities).
The Health Care Workforce as a System: Supply Outflows, Inflows, and
Structural Efficiency
Any workforce, including the health care workforce, can be understood as a dynamic system
that evolves over time, depending on key characteristics of the system, such as the number of
worker hours in a given occupation group and the structural context (organization, technology,
and policies) of that system (Homer and Hirsch, 2006; Rutter et al., 2017; El-Sayed and Galeo,
2017). The structure of a system refers to the existing distribution of occupations and staff, how
they work within their organizations’ team structures, protocols, and what technology and other
infrastructure facilitates their work (Dussault and Dubois, 2003; Yeow and Goh, 2015;
Jithitkulchai, 2022). In Figure 1.4, we present a schematic model of the health care workforce
system that emphasizes three key dynamics that determine the amount of health care delivered
by a given number of workers functioning within the existing structure during a given period of
time:
• Worker outflow: the exit of workers from a given health care workforce geography or
specialty
• Worker inflow: the inflow or entry of workers into a given health care workforce
• The structural environment that influences worker productivity: structural factors that can
increase or decrease the health care outcomes produced by the workforce, even if the
number of full-time equivalent (FTE) workers remains the same.
6
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Note that the first two bullets refer to decreases or increases of FTE employees, who are
assumed to increase that workforce’s health care output when labor is added or to decrease
health outcomes when less labor is available. We define worker output as the positive effect the
workforce has on improving Virginians’ health, not necessarily workers’ ability to see more
patients.
Figure 1.4. The Health Care Workforce System: How Inflows, Outflows, and Structural Efficiency
Relate to Health Outcomes
Describing the health care workforce system in this way elucidates the main challenges
policymakers and other stakeholders face:
• The Retention Challenge: The more workers who exit a particular workforce geography
or specialty (by reducing hours worked, moving out of state, leaving their field of
practice, etc.), the larger the reduction in that workforce’s health care output.
• The Recruitment Challenge: A workforce’s inflow mainly originates from newly
trained health care workers but also includes workers moving to Virginia from another
location or switching from part-time work to full-time work. The larger the inflow of
worker FTEs, the larger the workforce’s health care output.
• The Efficiency Challenge: The third dynamic, structural efficiency, refers to increasing
the efficient allocation of workers within a given health care system or increasing their
health care output using technology, policy changes, or other structural adjustments.
Structural efficiency may be produced when some retention or recruitment challenges are
addressed directly through changes to the context of a health care workforce and its
structural characteristics.
The number of current worker hours (i.e., the workforce supply) depends on both inflow and
outflow of workers in a health care workforce system. Currently, many health care workforce
systems face a large number of retiring workers, and too few individuals are in training, either to
7
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
replace retiring workers or to serve the growing health care needs of the U.S. population
(Gamble, 2021).
The number of needed workers (i.e., workforce demand) partially depends on the structural
efficiency of the system in which they operate and also depends on the population’s need for
health care (i.e., the demand for health care). If new technologies reduce the physical demands of
a health care occupation, or if technologies increase efficiency in other ways, fewer individuals
could perform the same amount of work. The reduced burden on individuals could also inspire
more workers to remain in or enter the health care workforce. For example, technologies to assist
in lifting patients can reduce physical strain on personal care aides, potentially allowing some
workers to remain in the field longer with fewer injuries or allowing individuals with physical
impediments to provide more personal care. Such technologies can allow a given number of
workers to produce greater health outcomes—an increase in structural efficiency that allows
more “output” per worker.
Many other structural improvements can potentially affect the supply of and demand for
health care workers, including improvements in the efficiency of licensing and regulation,
improved education and training, and increased financial compensation. The extent to which a
community values a workforce (i.e., the “social value” of health care) is also part of the system
or structure in which the health care workforce operates.
Methods
Our research was determined to be exempt from oversight by the RAND Corporation Human
Subjects Protection Committee. As a first step, we convened a nine-person multi-stakeholder
advisory board—including representatives of nursing and labor/workforce and a past
Commonwealth health official, among others—who provided input into the study focus,
interview and focus group guide, and other aspects of the study.
Environmental Scan
We conducted an environmental scan following methods defined by the Institute of Medicine
(2011). We followed six steps: (1) define the research team for the environmental scan, (2)
collect user and stakeholder input, (3) identify the topic and research questions in the scan, (4)
develop and implement a review protocol (including the study screening and selection process),
(5) screen and select studies, and (6) appraise and synthesize selected studies, including
evaluating the quality of evidence. We reviewed peer-reviewed and grey literature related to
strategies for health care workforce retention and recruitment. for the years 2013–2023. The
environmental scan search strategy, search terms, and other methodological details are included
in Appendix B.
8
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Multi-Stakeholder Conference
In collaboration with the VHWDA, we held an in-person, two-hour, multi-stakeholder
conference in Richmond, Virginia, on June 9, 2023. The conference was attended by over 40
invited attendees, including nursing, primary care, and behavioral health providers;
administrators; educators; current and past Commonwealth health officials; and other key
stakeholders. The conference included five PowerPoint presentations in which we shared
preliminary findings from the study, including promising interventions for health care workforce
retention and recruitment. Each 10-to-15-minute presentation was followed by up to 10 minutes
of question-and-answer (Q&A). We recorded the Q&A sessions and prepared transcriptions
based on the recordings. We analyzed the Q&A session data using rapid qualitative analysis
techniques (Watkins, 2017), identifying challenges and potential strategies for health care
workforce retention and recruitment in the Commonwealth.
Interviews and Focus Groups
We conducted and analyzed data from 15 interviews and group discussions that involved a
total of 34 individuals, including Virginia nursing, primary care, and behavioral health
practitioners; educators; labor experts; and other key stakeholders. Interviews included between
one and three people, and focus groups included between four and ten people. One focus group
was dedicated to nursing workforce development, one to behavioral health, and one to workforce
development in Virginia rural settings. We developed one guide for both interviews and focus
groups, which was informed by the results of the study environmental scan, experts on our
research team, and the study’s advisory board (Appendix C). We used snowball sampling to
identify and recruit participants with practical experience and/or expertise in health care labor,
administration, and education for these discussions. Some of the interviewees also attended the
June 9, 2023, multi-stakeholder conference. All sessions were conducted remotely over the
Microsoft Teams platform. Interviews spanned about an hour, and focus groups between one and
two hours. All sessions were recorded and transcribed for analysis. We used rapid qualitative
analysis techniques (Watkins, 2017) to analyze the transcriptions and generate themes and
subthemes based on topics represented in the interview/focus group guide.
Survey
Between July and August 2023, we fielded an electronic survey to 20 hospitals in the
Commonwealth over a four-week period, with assistance from the Virginia Hospital and
Healthcare Association. The survey included two questions (Appendix D): one regarding
strategies hospitals use to retain nurses and another focused on use of technology, or other
practices, to support flexible nurse staffing models.
9
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Quantitative Analyses
Utilizing data from multiple sources (refer to Appendixes E, F, G, and H), we constructed a
system dynamics economic model to assess the shortages of nurses, primary care practitioners,
and behavioral health providers in Virginia at both statewide and regional levels. The model
employs a set of differential equations that are numerically integrated to track the dynamics of
supply, demand, wages, and the impacts of perceived barriers on health care workforce
engagement. The model assumes that wages adjust slowly to changes in labor market conditions
and incorporates the interaction between work reluctance (because of fatigue, distress, and
disengagement) and wage (compensating wage differential theory) to address the necessity of
raising wages in response to heightened perceived barriers, ensuring a sustained supply. Our
model is designed to help Virginia policymakers better understand the relative scale of the
workforce shortfall by region and by job category over the next five, ten, and 15 years. In
addition to illustrating the current and future scale of the workforce shortfalls, we model the
extent to which various interventions can contribute to reducing these shortfalls. These analyses
can provide insight into the relative importance of different categories of interventions (e.g.,
wage effects on supply compared with recruitment effects on supply).
Additionally, we analyzed available surveys of health care sector students and workers, labor
data, and student loan repayment data (see Appendix A) to better understand the economic and
demographic context of Virginia’s workforce challenges.
Organization of this Report
The remainder of this report consists of three chapters. In Chapter 2, we present key findings
from the environmental scan, multi-stakeholder conference, interviews and focus groups, and
quantitative analyses. We start by outlining health care workforce development challenges and
potential solution levers for nursing, primary care, and behavioral health. Then we list and
discuss specific interventions for retention, recruitment, and increasing structural efficiency. In
Chapter 3, we offer recommendations by strength of evidence, and in Chapter 4 we outline an
implementation strategy for the recommendations. We provide details on our methods and
analysis in Appendixes A–H, which are contained in a separate annex, available at
www.rand.org/t/RRA2093-1.
10
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Chapter 2. Key Findings
Overall, Virginia has a robust menu of existing interventions designed to improve retention,
recruitment, and structural efficiency in the health care workforce. Current interventions include
student loan forgiveness, expanded scope of practice, license transfer, more flexible practice
models (e.g., increasing the types of medical professionals who can supervise practice of
midwives), more flexible clinical training models (increasing the number of students a
preceptor/licensed supervisor can supervise) (BillTrack50, 2021), early health care training and
career development (e.g., secondary school scholarships and support for students interested in
health care professions such as the Claude Moore Behavioral Health Scholars; Claude Moore
Charitable Foundation, 2023), and higher Medicaid reimbursement rates. Additionally, the newly
established Department of Workforce Development and Advancement will serve as a
coordinating body for workforce support across different industries in the Commonwealth.
However, attendees of the study’s June 9, 2023, conference indicated that, while many
innovative health care workforce development interventions are being implemented in the
Commonwealth, they are not disseminated effectively.
Virginia is also beginning to implement or propose new ways to support health care
workforce retention, recruitment, and structural efficiencies. These include expanded student
loan forgiveness, use of primary care technicians and associate physicians, increased license
transfer, and a nursing apprenticeship. These efforts involve significant investments in the
Commonwealth’s broadband infrastructure based on the Infrastructure Investment and Jobs Act
Fund (Virginia Department of Housing and Community Development, 2023). Such investments
can support telehealth, helping the roughly one-third of Virginians living in HPSAs to access
medical care in other areas of the Commonwealth and, if need be, beyond Virginia.
In the context of Virginia’s ongoing efforts to address health care workforce shortages, we
summarize findings from our mixed methods analysis in this chapter, including a description of
challenges regarding nursing, primary care, and behavioral health workforce. We also discuss
interventions that the Virginia legislature, state agencies, health care organizations, educational
entities, VHWDA, and other key stakeholders in Virginia might consider implementing in the
near term to enhance retention, recruitment, and structural efficiency in these fields. The list of
interventions below is not exhaustive—there are others that Virginia could consider to address
health care workforce shortages—but we focus on those that are most pertinent to retention,
recruitment, and structural efficiency.
Importantly, the discussion below includes interventions whose effectiveness has not been
previously evaluated using rigorous methods and relevant workforce outcome metrics.
Conducting such evaluations would provide the evidence necessary to assist the Commonwealth
11
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
in planning and implementing targeted and cost-effective strategies for strengthening its health
care workforce (Harvard Catalyst, undated).
Retention interventions mainly aim to enhance the number, composition, and career
duration of licensed professionals working in a given health care occupation. Recruitment
interventions focus on enhancing the number and/or composition of individuals who enter into
and complete education and training for a given health care occupation prior to practicing as a
(licensed) professional in that occupation. Interventions focused on structural efficiency—that
is, the efficient allocation and use of resources (Jithitkulchai, 2022)—attempt to fully leverage
the existing workforce pool through use of technology (e.g., telehealth) (Yeow and Goh, 2015),
geographic redistribution of the workforce, or other changes (Dussault and Dubois, 2003) to the
structural context in which workers operate.
Health Care Workforce Development Challenges and Solution Levers
Stakeholders representing Commonwealth nursing, primary care, and behavioral health
workforces described many challenges to retaining, recruiting, and enhancing structural
efficiency of the health care workforce. In addition to the challenges outlined below, nursing,
primary care, and behavioral health in Virginia’s rural communities face very specific retention
and recruitment challenges because of insufficient infrastructure, including housing,
transportation, and child care services. Detailed data and analyses on employment, wages, and
other statistics—including quantitative forecasts based on those data—provide additional context
for understanding workforce challenges in these fields. Identified challenges for retention and for
recruitment that underlie workforce shortages are summarized, respectively, in Tables 2.1 and
2.2. Additional details and supporting data are then presented for nursing, primary care, and
behavioral health challenges.
12
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Table 2.1. Identified Retention Challenges in Nursing, Primary Care, and Behavioral Health in
Virginia
Primary Behavioral
Challenges Nursing Care Health
Regulations and documentation
Insurance documentation and billing ✓ ✓
Provider licensing and documentation ✓ ✓ ✓
Care documentation (e.g., electronic health ✓ ✓ ✓
record)
Provider licensing fees ✓ ✓
Regulation and documentation of care site ✓
Public health insurance reimbursement ✓ ✓
Compensation
Compensation differentials, providers ✓ ✓
Business costs of a practice ✓
Wage/salary compensation ✓ ✓ ✓
Insurance benefits ✓
Other benefits (e.g., flexible scheduling, ✓ ✓ ✓
housing)
Workload
Patient throughput rates ✓
Overall patient acuity ✓
Patient-provider ratios ✓ ✓ ✓
Work-life balance ✓ ✓ ✓
Burnout ✓ ✓ ✓
Workplace setting
Violence ✓
Insufficient supplies ✓
Provider turnover ✓ ✓ ✓
Retirements, early and overall ✓ ✓
SOURCE: Multidisciplinary conference, interview, and focus group data for Virginia stakeholders, RAND Corporation,
July–August 2023.
13
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
NOT CLEARED FOR PUBLIC RELEASE. DO NOT CITE.
Table 2.2. Identified Recruitment Challenges in Nursing, Primary Care, and Behavioral Health in
Virginia
Nursing Primary Behavioral
Challenges Care Health
Training
Insufficient clinical placements ✓ ✓ ✓
Insufficient number of educators, nonclinical ✓ ✓
Insufficient number of educators, clinical ✓ ✓ ✓
Insufficient number of preceptors/licensed ✓ ✓
supervisors
Cumbersome and rigid education ✓
requirements
Compensation differentials, educators ✓ ✓
Lack of career pathways ✓ ✓ ✓
Time to initial certification ✓
Public perceptions and opinions
Poor public image of work in health care ✓ ✓
Public awareness of financial assistance ✓ ✓
Public awareness of health care careers ✓ ✓
Financial aid
Costs of education, initial ✓ ✓ ✓
Available financial assistance, initial ✓ ✓
education
Expensive licensing fees, initial ✓ ✓
Costs of clinical supervision ✓
SOURCE: Multidisciplinary conference, interview and focus group data for Virginia stakeholders, RAND Corporation,
July–August 2023.
Nursing Challenges and Solution Levers
Analysis of a variety of data sources confirms that Virginia will continue to face a shortage
of RNs even as the number of active RNs working in the Commonwealth continues to climb.
According to National Nursing Workforce Surveys (2017, 2020, 2022), the number of active
RNs working in Virginia has steadily increased in each year since at least 2017. Their data
indicate that the number of RNs active in the Commonwealth rose by almost 1 percent per year
from 2017 to 2020, and then rose at more than twice that rate from 2020 to 2022 (at a 2.3 percent
annual rate of increase). Despite this rate of increase, Nursing Solutions Inc. (NSI) Retention &
RN Staffing Reports indicate that nationwide vacancy rates for RN positions continued to
increase, suggesting that demand for RNs is rising even faster than the available RN workforce is
growing. The 2019 NSI report showed an RN vacancy rate of 8 percent, which rose to 9.9
percent by 2020 and to 15.7 percent in 2022. These most recent vacancy statistics partially
14
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
reflect a spike in demand for RNs caused by COVID-19, but they also reflect an increasing nurse
attrition. A September 2022 McKinsey Frontline Workforce Survey showed that 31 percent of
surveyed nurses indicated they intend to leave their current direct care positions within the next
year, implying a continuing retention crisis.
Looking beyond 2022, our model forecasts nurse employment in Virginia to shed light on
whether this nursing labor shortage will worsen and the degree to which various interventions
might help. (See Appendixes E and F for more information about the model, including the data
and assumptions used to build the model and its scenarios.) As illustrated in Figure 2.1, if current
trends continue (the baseline scenario), the Commonwealth may continue to add nurses to the
workforce (Figure 2.1) but at a rate that may prove insufficient to meet demand for nursing labor
(Figure 2.2).
Both figures also display how our forecasts change if ambitious interventions related to
wages, recruitment, and the work environment are implemented. Each intervention represents
improvements that we assessed to be ambitious, but possible:
• Reduced barriers: In the baseline scenario, we modeled three barriers to retention and
recruitment: fatigue, distress, and disengagement as a single parameter named work
reluctance. In this scenario, we assume that each of these three barriers is cut in half
(during the first year and continues at that level for all subsequent years), which results in
more retention and more recruitment mostly accomplished early in the simulated period.
• Increased recruitment rate: In this scenario, we assume that the rate of recruitment
rises from the 2,000 FTE nurses per year (baseline) to 2,500 FTE nurses per year, a 25
percent increase compared with the baseline.
• Elevated wage growth rate: Here we assume that average real wages rise for nurses in
Virginia by 3 percent per year instead of 2 percent per year (baseline), which results in
more retention and more recruitment.
• The multi-intervention scenario combines all three of the above interventions.
15
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure 2.1. Full-Time Equivalent Nurses Employed in Virginia 2023–2038 Median Forecast in Five
Scenarios
Nurses Employed
Year
Figure 2.1 suggests that the number of nurses entering the Virginia health care workforce
will continue to grow even under current conditions, and that ambitious interventions could
speed up this increase in nurses employed in the Commonwealth. We forecast that a 50 percent
reduction of the barriers that cause reluctance to work in the field of nursing will result in a rapid
increase in the number of nurses employed, mainly because of the positive effects on retention.
However, over a period of 15 years, increased recruitment has a more dramatic effect on
workforce numbers, surpassing the effect of decreased barriers. Increasing nurse wage growth
from an average of 2 percent annual growth to 3 percent annual growth also increases the
number of nurses retained and recruited, but not as much as the other interventions do. As
expected, a combination of all three interventions has the largest effect on the forecast of nurse
employment, boosting the total number of FTE nurses employed in 2038 by a little more than
10,000 above the baseline projection.
16
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure 2.2. Gap Between Employment and Demand for Nurses Employed in Virginia 2023–2038
Median Forecast in Five Scenarios
Additional Nurses Needed
to Meet Demand
Year
Figure 2.2 suggests that, even if Virginia aggressively intervenes to boost the number of
nurses employed in the Commonwealth, the current shortfall of nurses (the gap between
employment and demand for nurses) is likely to persist in the coming 15 years. If current
“baseline” trends continue, the shortfall will continue to expand, but it will expand more slowly
if any one of the interventions we have modeled is aggressively implemented. The model
suggests that if all three interventions are implemented (the multi-intervention scenario), the
nursing workforce shortfall can be entirely eliminated, at least through 2030, at which time even
more aggressive recruitment, wage growth, or reduction of barriers would need to be
implemented for supply to match demand.
Our interviews and focus groups with nurses and other health care professionals confirm that
Virginia is experiencing a shortage of nurses, which places additional pressures on the existing
workforce in nursing and beyond. Various challenges appear to drive Virginia’s nurse workforce
shortage. During and in the aftermath of the pandemic, many nurses left the bedside because of
stressful experiences and burnout. As one practicing nurse remarked, “Nurses survived some
awful stuff.” Older nurses in the Commonwealth retired early during the pandemic, costing the
profession many experienced hands that might have helped guide newer nurses with more
limited experience. This follows a national trend (Weston, 2022). As one practicing nurse
indicated, “When I was a newbie nurse there were a lot of people looking out for me. The
younger nurses have no one looking out for them. They don’t know what they don’t know.”
Multiple interviewees cited high patient-to-nurse ratios as a major challenge for retention.
Many indicated that, because patient-to-nurse ratios typically do not consider patient acuity (i.e.,
17
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
how sick a patient is) or the level of experience and expertise of the nurse, patient-to-nurse ratios
are a “farce” as a tool to determine nurse staff workload.
Participants cited both high patient-to-nurse ratios and high documentation burdens as
contributing to nurse burnout. Interviewees said that the documentation requirements take nurses
away from the bedside and patient interaction—the reasons many people go into nursing—with
nurses spending “70 to 80 percent of their shift in front of the screen.”
Fewer experienced nurses in the face of early retirements or other nursing workforce exits
(overall and in acute care settings like emergency departments and hospitals), high patient-to-
nurse ratios, and high documentation requirements all contribute to burnout. Burnout was cited
as “the number one driver” for staff leaving. As one participant put it,
What [nurses] soon discover is that they are entering an overburdened, heavily
compressed profession where being perpetually short-staffed means juggling
unsafe nurse/patient ratios while maintaining safety expectations.
—Nursing representative
The likelihood of burnout only increases in a work context characterized by a rise in
throughput and patient acuity. Some of the strategies that hospitals use to deal with these
challenges raised concerns among nursing stakeholders. One practicing nurse shared an example
of her hospital directing floor nurses to take care of critically ill patients in intensive care units
even though the nurses did not have the expertise or experience to care for such patients. The
pandemic also revealed significant pay differences in nursing, leading nurses to take positions
with higher pay. Although nursing stakeholders noted that “you can’t just throw money at the
problem,” nurses need supplies, equipment, and training to do their jobs well. Nursing
stakeholders expressed frustration that newly hired nurses with many fewer years of experience
were getting paid much more than more experienced nurses. In reference to traveling nurses, one
participant expressed that the “traveling nurse has been renumerated at unsustainable levels.”
Hospitals have been forced to adopt an unsustainable model of hiring more expensive travel
nurses to ease staffing shortages caused by the pandemic (Coombs, 2023; Lucas, 2022).
Recently, travel nurse salaries are trending toward a decrease, but they are still generally higher
than the average salaries of registered nurses (Kusterbeck, 2022). The growing gap in salary
between travel and staff nurses has led to frustrations, increased turnover, and exacerbated
shortages (Coombs, 2023; Lucas, 2022).
Two interviewees indicated that health insurance coverage for nurses has become more
restrictive and expensive. A nurse stakeholder reported that
A for-profit hospital chain with a large footprint in Virginia modified its health
insurance plan to require employees to cover 20 percent to 40 percent of their in-
network health care expenses on top of their co-pays. Hourly wages can be
18
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
increased, and bonuses can be doled out, but it doesn’t take long for staff to
realize that their earning power is lessening.
—Nurse representative
The rise in workplace violence—especially since the start of the COVID-19 pandemic, with
many people having a “shorter fuse”—was also cited as a major barrier to retention. This
sentiment is shared by nurses more generally, with significant consequences for retention (Lin et
al., 2023). The following comment by a nurse stakeholder captures the sentiments expressed by a
number of interviewees:
As an ER [emergency room] nurse, I was punched, slapped, bit, kicked, spit on,
thrown into a wall hard enough to crack the sheetrock, had urine, blood, and
feces thrown at me, and confronted multiple knives and guns. With every one of
these situations, although hospital leadership preached a zero-tolerance policy to
workplace violence, I was discouraged from reporting (a common practice, per
nursing organizations and social media posts). Nursing retention is difficult if the
threat of physical harm is present every day.
—Nurse representative
The “bad public image” that nursing and other health care professions gained during the
pandemic—given the public discourse around the stresses and burnout related to the work—does
not help with recruitment efforts.
The nursing education and training pipelines also pose significant challenges to recruitment.
There are not enough clinical placements for nursing students—largely because of the limited
number of preceptors—and some clinical instructors have limited experience. Furthermore, there
is a lot of emphasis on simulation in education and training and not enough in-person time in the
clinical setting. As an expert in labor issues in the Commonwealth put it:
[There is a] shortage of faculty preceptors to train nurses in the classroom. It is
not a paid opportunity, so to get people to volunteer is challenging. This is a huge
break in the system because we cannot get people to licensing because there is
nobody to supervise them, leading to staff shortage and burnout.
—Labor representative
Nursing stakeholder interviewees expressed that some nursing schools have to turn away
“thousands of qualified and eligible applicants” because of shortages in nurse faculty and/or
space in clinical facilities. They also said that other faculty (for example, engineering and
business school faculty) get paid “way more” than nursing faculty, making preceptorship “not a
particularly enticing career opportunity.” On the other hand, some interviewees said that more
consistent messaging on how one can become a successful nursing health care worker (and how
to access related educational funds) is needed in the Commonwealth. This suggests that such
efforts may currently vary across regions.
19
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Primary Care Challenges and Solution Levers
Greater supply availability of primary care has been associated with multiple benefits,
including better health outcomes, lower mortality, reduced disparities, and lower health care
costs (Basu et al, 2019; Shi, 2012), and while research supports increasing the supply of primary
care physicians, it may be necessary to pair this investment with other interventions, such as
payment reforms for physicians and reduced administrative burden (Song and Gondi, 2021). As
with nursing, national data from HRSA provided in Chapter 1 (Figures 1.1 and 1.2) confirm that
Virginia struggles to employ a sufficient number of primary care physicians (medical and
osteopathic doctors), NPs, and PAs to meet demand. Based on our study’s Virginia-specific
simulations, which evaluated primary care workforce projections under different intervention
scenarios, Virginia will likely continue to face primary care workforce shortages in the coming
years. Unlike our forecasts for RNs, our forecast for the number of active primary care workers
in the Commonwealth shows a decline (Figure 2.3) under baseline conditions, making the future
shortfall (Figure 2.4) in primary care workers even more severe than for nursing alone. Similar to
our model for nursing, the largest immediate boost to the workforce would come from cutting in
half the barriers that make workers reluctant to work in the field. Over the course of 15 years, the
effect on total employment of increasing wage growth catches up to the effect of decreasing
barriers. The effect of increasing recruitment is more modest because baseline recruitment begins
at very modest levels, and the lag time for recruitment is greater in primary care than it is for
nursing, given longer related education and training times for primary care. An ambitious
combination of all three interventions (the multi-intervention scenario) would boost total
employment by more than 2,000 workers.
20
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure 2.3. Full-Time Equivalent Primary Care Workers Employed in Virginia 2023–2038 Median
Forecast in Five Scenarios
Primary Care Workers Employed
Year
Figure 2.4. Gap Between Employment and Demand for Primary Care Workers in Virginia, 2023–
2038 Median Forecast in Five Scenarios
Additional Primary Care Workers
Needed to Meet Demand
Year
Interviewees indicated that, since the pandemic, many health care workers have left the
(aging) primary care workforce because of being overworked in already understaffed health care
delivery settings. They also mentioned that the lower the pay for a primary care provider, the
21
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
easier it is to leave for a less demanding health care setting, another career, or caring for their
children, for example. Importantly, simulations of the supply of available primary care providers
in Virginia indicate that increased compensation is more important for increasing the number of
available primary care workers (relative to other individual simulations) than it is for increasing
the overall number of available RNs.
Other key issues raised by stakeholders were pay parity and cost inflation. One participant
noted the disparity in pay between male and female providers as a barrier to retention.
Richmond, Virginia, was recently identified as one of the ten cities where female physicians earn
the least annual compensation (Doximity, 2023). Similarly, Virginia Beach ranked as having one
of the lowest average annual salaries for female physicians (Doximity, 2020). Nationally, gender
pay differences for physicians may be declining—especially in primary care (Kane, 2023).
However, large gender pay differences remain, as do other gender differences related to work-
life balance and other complex factors (Lo Sasso et al., 2020). Another cited barrier to retention
was provider expectation for pay, which often outpaces employer organization revenue,
especially for smaller health care organizations.
Interviewees raised concerns about the underfunding of Medicaid programs (Cover Virginia,
undated; Kaiser Family Foundation, undated-a) for all providers in Virginia in spite of “well over
a million” residents being on Medicaid. This is in spite of a recent increase in Virginia Medicaid
reimbursement rates (Masters, 2023). Interviewees also said that Medicare, Medicaid, and
private insurance are difficult to bill—so more and more physicians are opting for a sliding scale
model or going into concierge practice, “impacting how many providers are in play.” We also
heard from participants that the shortage of PAs and NPs leads to increasing patient-to-provider
ratios (i.e., larger patient panels per provider).
Participants said that public primary care clinics have additional regulatory requirements that
create more retention challenges. As one interviewee explained,
There is a lot of regulation and red tape that involves not enough people doing it,
not enough pay, and it becomes a cumbersome process to get reimbursed. The
pay disparity between public and private entities was also noted as a major issue
contributing to high provider vacancy rates in public settings.
—Community health representative
Interviewees indicated that lack of preceptorship/licensed supervisor capacity contributes to
retention and recruitment challenges. As one interviewee put it,
Before things became primarily revenue driven, it was a privilege to coach and
mentor students. There has been a shift from the stress and burnout in health care.
The pressure is on to do more visits, generate revenue. No one has time for
mentoring or preceptorships, or they are exhausted.
—Community health representative
Primary care stakeholders raised concerns over the high cost of education and low return on
investment when entering the field. This was cited as one cause of not having enough people
22
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
entering the field. Low compensation in primary care compared with medical specialty and
surgical fields was also cited as a recruitment challenge.
Stakeholders said that another root cause of people not pursuing primary care is lack of
awareness of primary care professions in communities and particularly in schools. Participants
noted that this is a missed opportunity. Primary care stakeholders also noted that the departure of
providers from the field has reduced the supply of practitioners. As one primary care provider
and educator noted, “People are leaving at an increased rate and not enough people to backfill.”
One primary care provider/educator also noted, “We don’t have nurses and physicians in settings
where they are most needed.”
Interviewees cited a lack of career pathways as a barrier to recruitment of younger personnel:
The younger generation wants to know what the [health care career] trajectory
extends to five to ten years from now; Gen Z is definitely asking these questions.
—Labor representative
One of the biggest barriers noted was the lack of an equitable pipeline.
Behavioral Health Challenges and Solution Levers
Virginia health care faces similar challenges in employing a sufficient number of behavioral
health workers to meet rising demand. If current conditions persist (the baseline scenario), the
number of behavioral health workers employed in Virginia is forecast to slowly decline in the
coming years. Implementing any one of the three interventions promises to keep the existing
workforce numbers at 2023 levels, but only by combining all three interventions can Virginia
produce a sustainable increase in behavioral health workers employed, as illustrated in Figure
2.5.
23
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure 2.5 Full-Time Equivalent Behavioral Health Workers Employed in Virginia, 2023–2038
Median Forecast in Five Scenarios
BH Workers Employed
Year
Because demand for behavioral health care is rising, all of our forecasts indicate an
increasing shortfall of behavioral health workers during the next 15 years (Landi, 2023). Figure
2.6 shows that each intervention produces similar modest improvements, and the combination of
all three interventions still leads to a widening gap between employment and demand for
behavioral health workers.
24
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure 2.6 Gap Between Employment and Demand for Behavioral Health Workers in Virginia,
2023–2038 Median Forecast in Five Scenarios
Additional BH Workers
Needed to Meet Demand
Year
Interview and focus group participants also reported that there is a significant shortage of
behavioral health providers, including psychiatrists, psychologists, social workers, and
counselors, making it difficult to meet Virginia’s demand for behavioral health services. They
indicated that a large percentage of behavioral health providers are at retirement age and
attributed some of this shortage to the high cost of graduate degrees in mental health professions
and the need to take out significant loans to afford these programs. Furthermore, participants
indicated that the lack of funding and limited financial support for students pursuing mental
health degrees can deter potential professionals from entering the field.
Interviewees also noted that the compensation and reimbursement rates for behavioral health
services may not be sufficient to attract and retain qualified professionals, leading to a higher
turnover rate and understaffed facilities. The administrative and regulatory burdens faced by
behavioral health professionals—including “excessive” paperwork and documentation—were
noted by participants as a barrier to retention. Also, secondary trauma and burnout among the
behavioral health workers because of the nature of the work was also noted as one of the causes
of high turnover in the field by interviewees. One interviewee noted that “mental health deserts”
are leading to delayed care and sicker patients, inflicting “moral injury” to providers who
practice in those settings.
Behavioral health stakeholders indicated that there are not enough faculty in behavioral
health because they do not get paid enough. They also noted that the revenue-generating
requirements of existing faculty in larger institutions create problems because faculty cannot
generate sufficient revenue while overseeing students and trainees. The limited number of
25
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
psychiatry residency slots was also cited as a barrier. One behavioral health stakeholder put it
this way:
The national bodies only give you certain slots. The general assembly got
focused on child psychiatry and ten more seats opened up for residencies. But it
is piecemeal. It is not a silver bullet. . . . Paying for more slots is helpful, but
getting national boards to allow more slots is important.
—Behavioral health stakeholder
Behavioral health stakeholders also identified overregulation of education and training as
another significant barrier to recruitment. For example, at the bachelor of science education
level, faculty have to be masters-prepared or ask for a temporary exception.
Participants said that stigma surrounding mental health also serves as a barrier to recruitment.
They noted that this stigma not only affects the willingness of individuals to seek care but also
affects the desire of certain communities, particularly communities of color, to pursue careers in
behavioral health. One participant expressed that, in community settings, experiences should not
be limited by ratios, as these are important opportunities for student learning. They also noted
that there is a requirement of a minimum of 500 practice hours but that some people will reach
competency at fewer hours and some will need more hours to reach competency.
Interviewees highlighted challenges related to regulations around becoming a licensed
clinical social worker, noting that “there are a lot of hoops to jump through.” One interviewee
said,
For people who want to become LCSWs [licensed clinical social workers] and
licensed professional counselors, they are thwarted by paying for the required
supervision on their own that they have to go through to get a license. . . . A lot
of people who want to become therapists and are motivated, and have the
aptitude, cannot run that financial gauntlet while we have a mental health crisis.
—Behavioral health stakeholder
Social workers need to complete a significant portion of clinical hours before they can take
the licensing exam. If an examinee fails, they will have to wait 90 days before they can retake the
test. Any clinical work until that time involves paid supervision. One interviewee indicated that
“some work multiple jobs to pay for supervision which is a big financial and mental health
burden.”
Participants also noted that the licensing process for behavioral health professions is very
cumbersome and very expensive. By the time many graduate from their program, they may still
need a few thousand hours of supervised work, for which they have to pay out of pocket.
Furthermore, interviewees noted that the time it takes from licensure application submission to
getting licensed is unreasonably long. Interviewees also noted the importance of loan forgiveness
for retention and recruitment but expressed that many people are not aware of behavioral health
loan forgiveness programs (Guan et al., 2023).
26
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Nursing, Primary Care, and Behavioral Health Workforce Shortages
Our projections, based on nursing, primary care, and behavioral health workforce data from
the United States, Virginia, and Virginia Area Health Education Centers (AHEC) regions,
illustrate the scope of Virginia’s health care workforce shortages and the extent to which three
interventions could reduce these shortages during the next five, ten, and 15 years. Our model
suggests that nursing, primary care, and behavioral health workforce shortages will increase
across Virginia if current conditions persist (Figure 2.7) and that the percentage shortfall in
behavioral health personnel is even more severe than the shortfalls in primary care and nursing.
The behavioral health shortfall is also the most difficult one to reverse by applying the
interventions we modeled.
Figure 2.7. Statewide Nursing, Behavioral Health, and Primary Care Workforce Shortages as a
Share of Each Respective Workforce, 2023–2038 Baseline Projections
50%
Shortfall as Percent
45%
40%
35%
30%
Primary Care
25%
of Employed Workers
Nursing
20%
Behavioral Health
15%
10%
5%
0%
2023 2028 2033 2038
Year
As illustrated in Figures 2.5–2.7, our statewide model forecasts suggest that the shortfall of
behavioral health workers is also more difficult to reverse by raising the rate of wage growth,
reducing barriers, and increasing the rate of recruitment.
Analysis of regional differences in the same three health care workforces does not indicate
significant differences in the trajectory of the workforce shortages among Virginia regions (see
Appendixes E, F, and G for regional model results), but this finding is likely the result of the
dearth of region-specific data feeding our models. Currently, this model differentiates only
regional wage levels for each job category and does not include other regional variations, such as
demographic trends and supply shortfalls. We would need more data to better differentiate
region-specific differences in the numbers of workers employed, as well as differences in
perceived barriers, as measured by our work reluctance metric, to gain greater region-specific
27
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
insights into health care workforce shortages. Data related to age demographics, interstate
workforce flows, perceived benefits, regional nuances, interconnections between regions, and
current work allocation efficiencies will significantly enhance our model’s ability to provide
insights and projections to regional differences in health care workforce shortages and related
dynamics.
We modeled the extent to which real wage increases could change the relative trajectories of
workforce shortages and found that speeding up real wage increases from the national average of
2 percent per year to 3 percent per year would decrease the nursing shortage in Virginia by 21.7
percent in five years, 23.8 percent in ten years, and 22.3 percent in 15 years, compared with the
baseline. Compared with 2023, the gap between employment and demand would increase to 291
percent in the baseline scenario and 226 percent in the increased wage scenario in 15 years.
Similar wage increases among primary care physicians appear to have a similar relative
effect on reducing their workforce shortages (20.5 percent in five years, 23.7 percent in ten
years, and 23.3 percent in 15 years) but are less significant for behavioral health specialists (7.9
percent, 8 percent, 5.9 percent, respectively). Compared with 2023, the gap between employment
of and demand for primary care providers would increase to 459 percent in the baseline scenario
and 352 percent in the increased wage scenario in 15 years; for behavioral health specialists, the
gap would increase to 198 percent in the baseline scenario and 186 percent in the increased wage
scenario in 15 years.
We also modeled the effect of “cutting in half” the number of barriers to working in nursing.
Unlike modeling the effects of wage increases, changes in barriers are more challenging to
quantify. For our reduced barriers scenario, we assumed that each of the three barriers (fatigue,
distress, and disengagement) is cut in half, which results in more retention and more recruitment.
The result of decreasing those barriers by 50 percent is to reduce the nursing workforce shortage
by 49.1 percent in five years, 33.2 percent in ten years, and 25.5 percent in 15 years, compared
with the baseline. Compared with 2023, the gap between employment and demand would
increase to 291 percent in the baseline scenario and 217 percent in the increased wage scenario in
15 years.
Similar reductions in barriers among primary care physicians (25.9 percent in five years, 17.9
percent in ten years, and 14.2 percent in 15 years) and behavioral health specialists (8.8 percent,
5.5 percent, 3.5 percent, respectively) appear to have a smaller relative effect on reducing their
workforce shortages. Compared with 2023, the gap between employment of and demand for
primary care providers would increase to 459 percent in the baseline scenario and 394 percent in
the increased wage scenario in 15 years; for behavioral health specialists, the gap would increase
to 198 percent in the baseline scenario and 191 percent in the increased wage scenario in 15
years.
Interventions targeting retention, recruitment, and structural efficiencies are key to
addressing demonstrated current and projected health care workforce shortages in Virginia. In
the sections that follow, we evaluate specific interventions for each, in Virginia and beyond.
28
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Interventions for Workforce Retention
Retention at all stages of a health care professional’s career is critical to ensuring a robust
health care workforce (Sermo, 2023). For example, the first five years of practice are the most
critical for retention for all physicians, regardless of practice specialty, gender, race, urban-rural
status, or HPSA status (Boydstun and Cossman, 2016). One interviewee described the
importance of attention to, and investment in, retention as follows,
Speaking as a bedside RN, I would certainly like for more high school and
college students to consider a career in health care, particularly nursing; however,
recruitment is one thing, retention is another. An organization can engage in
aggressive hiring practices to fill the void, but if equal measures are not applied
to retention, it will continue to bleed personnel. You cannot right a sinking vessel
by bailing water without plugging the leak.
—Nursing stakeholder
Improving retention in the health care workforce depends on interventions that address key
challenges to retention in Virginia, including the financial burdens of education and training,
resiliency and well-being, limited professional development, and siloed and unrealistic work
demands (Brennan and McGrady, 2015; Eden, Jabbarpour, et al., 2020; Eden, Chesluk, et al.,
2020). Model projections for nursing, primary care, and behavioral health care workforce under
different types of interventions presented above similarly demonstrate the importance of
removing work barriers and improving wage compensation to ensure that the future supply of
health care workers meets the needs of Virginians (see Figures 2.5–2.8).
In the subsections below, we discuss different interventions in Virginia and elsewhere that
address challenges to health care workforce retention.
Pay Increases and Other Incentives
Nursing, primary care, and behavioral health stakeholders repeatedly emphasized the need
for better pay and other incentives as a key strategy for retention. Increased pay, benefits
(including good health insurance coverage), and assistance with transportation and housing were
all mentioned as important incentives. Stakeholders shared,
29
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Offer good health care, good pay/benefits, good working conditions. I will go to
wherever has the best of these.
—Practicing nurse
The Medicaid population in Virginia is well over a million and is underfunded
for all providers.
—Labor representative
If you work for the X County, incentive is if you buy a house up to a certain
amount, they pay for the down payment. Assist with down payment, tax credit
for X, etc.
—Community health representative
The electronic survey that we administered in partnership with the Virginia Hospital and
Healthcare Association was sent to 20 hospitals, and five completed the survey (20 percent
response rate). We asked about strategies that hospitals use to retain nurses. All respondents
indicated that they offer bonuses; more flexibility in work hours; convenient and/or free parking;
mental health services for staff; diversity, equity, and inclusion strategies; and changed patient
flow or staffing models. Eighty percent of respondents indicated that they offer on-site child
care, job training to improve and/or expand skillsets, educational advancement opportunities as
part of employment, and supervisory training or an ombudsman program to address employee
concerns. Sixty percent of respondents indicated that they offer educational loan repayment and
implemented strategies to promote teamwork. Forty percent of respondents indicated that they
increased nurse salaries, and 20 percent indicated that they offer transportation vouchers. None
of the respondents indicated that they offered subsidized housing costs, meal coverage, or
covering the cost of maintaining licensure.
A survey respondent indicated that their hospital implemented scheduling software that
allows for a flexible staffing schedule, giving users the “ability to schedule via multiple
technologies (e.g., mobile versus PC) and provides greater visibility into open shifts in four-,
eight-, and ten-hour segments.
Interviewees shared concerns around Medicaid being underfunded for primary care and
behavioral health and identified challenges associated with these as barriers to workforce
retention:
They should pay (or pay more) for mental health through Medicaid. Has worked
well for post-partum care.
—Primary care physician/educational leader
States and the federal government [need to] create rates that cover the cost to
deliver the service. You can increase compensation, do one time funding for
recruitment and retention bonuses. All those kinds of things . . . . stops the
30
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
bleeding right now, but the sustainable, long-term solution is to actually increase
rates, so that it makes sense from a business perspective.
—Community health/behavioral health representative
A primary care physician shared concerns about the disparity in pay between primary care
and specialty care as a retention challenge:
Hurts to see other specialties getting so much more than primary care right out of
residency.
—Primary care physician/educational leader
Team-Based Care
Our nursing, primary care, and behavioral health interviewees identified high patient-to-
provider ratios as a major barrier to retention in these workforces and noted that this needs to be
replaced with team-based nursing. Nursing stakeholders indicated that patient-to-nurse ratios
need to be replaced with patient-to–provider team ratios and stressed that both patient acuity and
a nurse’s experience level need to be factored into patient assignment to nurses. A nursing
stakeholder suggested recruiting ancillary staff—such as respiratory therapist and
phlebotomists—to decrease nurse workload. Interviewees shared,
Look more at “team nursing” enhance techs, phlebotomy, unit clerk. . . . Look at
scope and practice of LPNs [licensed practical nurses] in Virginia (“how can we
use LPNs again”) so RNs can practice at the top of their license.
—Nurse leader/educator
The pure nurse-to-patient ratio is a farce. Staff up and down based on patient
acuity.
—Practicing nurse
I can’t take on one more patient because it’s not safe for my other five.
Similar concerns regarding the negative impact of high patient to provider ratios and the need
for team-based care were expressed by primary care and behavioral health stakeholders:
My second-year resident in primary care went from a panel of 120 patients to 150
overnight. Need support staff to get their work done.
—Primary care physician/educational leader
When you are fully staffed, people are happier, retention relies on recruitment.
Lower patient to provider ratio would make a difference.
—Community health/behavioral health representative
A hospital respondent to the electronic survey that we administered in partnership with the
Virginia Hospital and Healthcare Association indicated that they were using robotics, virtual
nursing, and “adding back” LPNs for team-based nursing.
31
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Documentation Burden
All stakeholder groups—nursing, primary care, and behavioral health—indicated that
tremendous documentation requirements result in burnout (National Academy of Medicine,
2022) and are a major barrier to retention. Many indicated that documentation requirements are
often excessive and redundant and that they take providers away from direct patient care.
Furthermore, some interviewees expressed the need to leverage technology to reduce
documentation burdens. Interviewees shared,
Reduce the documentation expectations. Going through inboxes doesn’t add
value—it’s the interactions with people. EHR [electronic health record] was
developed as a billing device and not for patient care. The amount of clicking in
each shift . . . each click takes a bit of your soul from you.
—Primary care provider/academic leader
Reprioritize what we really want to know. What is truly important? What do we
truly need? Weigh benefit against costs in time and level of burden. Ask why you
need to assess for bed sores every shift as a nurse. Having policies has not
allowed for variation SUD [substance use disorder] screening is a good example.
“Horrifically inefficient” . . . Need to pass the sniff test of “Why am I doing
this?”
—Primary care physician/Educational leader
Another interviewee had a different take on documentation challenges:
It is the way it is, until we harness the true power of technology. It is a part of the
job. If you like the job and serving people, it comes with the job. Of course, you
don’t want to duplicate work.
—Community health center leader
This interviewee also added,
A company in Nashville—Prediction Health—makes technology for dictation
using voice recognition that providers can use during bedside assessment. We
stopped using them. We’re going to be using scribes. People weren’t happy with
that. There’s still a human component that needs to check [the transcription].
—Community health center leader
Regulatory and Reporting Burden
A 2022 report from the National Academies flagged reducing compliance, regulatory, and
policy barriers in the daily work of health care providers as a key mechanism for addressing
health care workforce shortages in the United States (National Academy of Medicine, 2022).
Virginia stakeholders echoed this objective in interviews and focus groups. Primary care
stakeholders noted the need to reduce reporting and regulatory burdens, especially in the public
sector. One interviewee noted,
There are regulation and reporting requirements specific to public entities that the
private sector is not responsible for. Also, the pay disparity between public and
32
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
private is huge, and as a result there is a high vacancy rate for those [public]
positions.
—Community health representative
One interviewee shared insights on the implications of reporting burden and red
tape, particularly in reference to reimbursement:
Medicare/Medicaid/private insurance are difficult to bill so more and more
physicians are opting for sliding scale model. We are losing a lot of qualified
clinicians to concierge practices. Lot of primary care [providers] are going into
concierge practice where they have more autonomy over deciding number of
case load and how much time they will spend, and how they bill. That can impact
how many providers are in play.
—Labor representative
Workplace Violence
Nursing stakeholders in our interviews and focus groups expressed significant concern about
workplace violence as a barrier to retention. Interviewees shared,
Implement and adhere to zero tolerance for workplace violence.
—Nursing stakeholder
Implement state-level policies that penalize those that assault health care workers
similarly to prehospital workers and first-responders.
—Nursing stakeholder
We need to have legislation with teeth behind it. Many in the general assembly
do not want to tell judges what to do. If you aren’t protecting the team, why
would you want to come to work? Why would I stay in this environment? I can
go to the clinic, telework, do research.
—Nursing stakeholder
Cross-Sector Collaboration
Nursing, primary care, and behavioral health stakeholders identified cross-sector
collaboration as key to workforce retention. Nursing stakeholders noted that health system–
academic partnerships can be leveraged to create pathways to careers in nursing. An interviewee
noted,
I think we will draw more nurses into the bedside and recruit more nurses into
the profession if they see multiple pathways over the life of their career,
33
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
educational opportunities and career advancement by pursuing further education
and certifications.
—Nursing stakeholder
Earn as you Learn retains practice-ready nurses because they feel more prepared.
VA can attract and retain more if they pay for time commitment.
—Nursing stakeholder
Behavioral health stakeholders noted that building cross-sector collaborations can work to
strengthen mental health services delivery. One interviewee commented,
Need to build collaborative structures that bring together various stakeholders,
including public and private providers, businesses, schools, and concerned
citizens, to strengthen the mental health workforce and support the community.
—Behavioral health representative
Efforts to create more entryways and on-ramps into the behavioral health
workforce should be accompanied by partnering with employers to ensure there
are sufficient job opportunities for newly trained professionals.
—Behavioral health representative
The multi-stakeholder conference attendees discussed the importance of rebranding and
improving the perception of current health care jobs among young people through partnership
between the Virginia Department of Education; school-based providers, such as high school
counselors; and community colleges.
Workforce Diversity
Interview and focus group participants stressed the importance of having a diverse workforce
in order to increase retention—particularly with respect to populations in health care from
underrepresented communities—and in order to arrive at health care solutions that may be
achieved by including varied perspectives in the health care workplace. Participants shared many
thoughts on the importance of a diverse health care workforce:
A workforce that mirrors the community can improve the care of everybody.
Goes back to the pathways. Why don’t we have enough diversity in residency?
We still don’t have enough diversity in med schools. You have to make it a core
value and truly believe.
—Primary care physician/educational leader
Some interviewees noted that addressing diversity, equity, and inclusion issues in Virginia
may require different approaches for different regions of the Commonwealth. One interviewee
noted,
We need to treat everyone with dignity. Northern Virginia is different from other
parts of the state as far as tolerance. What does dignity mean? Detail it to where
34
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
you are. It’s an easier conversation to have if you approach it from a civility
standpoint. Meeting individual people where they are at.
—Primary care physician/educational leader
Continuing Education
As in other states, Virginia interventions typically do not address the financial burdens of
continuing education, training, and certification. Prior research suggests that certification and
recertification exams pose particular difficulties for health care workers. Primary care physicians
reported that recertification board examinations cause work-life imbalances, especially in the
case of female physicians (Eden, Jabbarpour, et al., 2020; Eden, Chesluk, et al., 2020).
Qualitative data collected as part of this study show that license renewal presents a significant
financial barrier to nursing and behavioral health providers, especially low-income and/or rural
individuals. Research elsewhere suggests this is a national challenge in retaining the health care
workforce (Orr, 2020; Fast, 2022).
Given the challenges that licensing and associated fees present, it is noteworthy that most
European Union countries regulate health care professions much less than the United States
does: Just over 70 percent of all health care professions are regulated in the United States,
compared with just under 50 percent in the European Union (Hermansen, 2019). Virginia’s
recent efforts to reduce licensing barriers focus mainly on recognizing out-of-state licenses but
still require those individuals to pay Virginia licensing fees (e.g., out-of-state counselors under
VA HB1433) (Virginia’s Legislative Information System, 2023).
Professional/Personal Development, Recognition, and Decisionmaking Capacity
One study showed that residents (i.e., physicians in training) value education that includes
strategies on managing debt, retirement planning, and information about the business of
medicine (Garrett et al., 2022). This is important because residents’ framing of educational debt
and their level of financial literacy were found to affect their sense of agency and well-being.
Resident participants in this study indicated that residency programs can facilitate stress
mitigation by providing residents with skills to help manage debt and plan for retirement. This
study also noted that, in order to reduce clinician debt, such education needs to occur in tandem
with systemic changes to financing medical education.
In our interviews and focus groups, nursing, primary care, and behavioral health stakeholders
noted the importance of professional development and staff recognition as a retention strategy.
Interviewees noted,
35
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
It’s about creating opportunities, scheduling fairly, and considering employee’s
needs, thinking about reducing burn out . . . for longer term, retention bonuses
and valuing staff are critical.
—Nursing stakeholder
In reference to health care workplace retention lessons learned from the pandemic,
interviewees shared,
Employers think about employees in a new way, not just some number. It has
paid dividends for those employers who are thinking like this.
—Labor representative
Training and mentorship programs have been on the rise. Companies [used to]
spend a lot more on recruiting talent than training up and retaining the talent they
have. There’s been a shift in this since the Great Resignation. Gen-Z and
millennials are not afraid to leave their employers and negotiate higher salary.
—Labor representative
Wage salary compression and retention bonuses are great strategies, but this is
what people expect and deserve. Pay is the short answer.
—Community health/behavioral health representative
Nursing stakeholders particularly stressed the need to give bedside nurses decisionmaking
authority and including them in shared governance plans. Interviewees noted,
Have nurses on the table instead of on the menu . . . make them a part of
decisionmaking tree
—School of nursing representative
I am surprised health care hasn’t learned from travel agency on shared
governance, how they want to work. We are meeting with Virginia Hospital and
Healthcare Association and the nursing association on staffing committees to
improve them, so they have teeth and are meaningful. Not just inundated with
reporting.
—Nursing stakeholder
One measure that I have seen is effective in delaying departure is track
scheduling. The nurse can design the schedule or pick your option from a list of
options. It renews at the next session. You can plan your life.
—Nursing stakeholder
Scope of Practice and Practice Agreements
Restrictive licensure for PAs and NPs can inhibit the benefits of increasing the supply of PAs
and NPs and reducing the reliance on primary care physicians (that are in short supply) in
meeting patient needs. A study showed that changing NP and PA scope of practice legislation in
Alabama to match states in the upper quartile of those that use a collaborative practice model,
such as Washington and Arizona, would increase both the employment and distribution of NPs
and PAs (Hooker and Muchow, 2015). Other studies have shown that hiring NPs and PAs may
36
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
be a strategy to accommodate new Medicaid patients at lower cost (Cawley et al., 2016; Barnes
et al., 2022; Burt, Sparbel, and Corbridge, 2021). Over half of U.S. states have adopted a more
relaxed NP licensure model than Virginia, including some of its neighboring states, such as
Maryland and West Virginia, and NP licensure plays an important role in retention of NPs
(Kueter, 2022; Wiesen, undated). The NP scope of practice in Virginia is restricted practice by
state law, which means that NPs must enter a collaborative agreement with physicians,
autonomous practice is also restricted, and Virginia does not recognize NPs as primary care
providers (McGhee, 2022). In contrast, in Maryland, NP scope of practice is full practice—that
is, NPs can diagnose and treat the patient independently, practice autonomously, and deliver care
in the primary care setting (Wiesen, undated). Although in comparison with Maryland, West
Virginia is reduced practice—meaning that NPs have to be in a collaborative agreement with a
physician—they are still able to be a primary care provider (Wiesen, undated).
Modernization of licensure laws can remove unnecessary barriers to access to NP health care
services (Kapu and Pitts, 2023). Studies show that the NP-physician co-management care
delivery model did not adversely affect patient outcomes and, in some cases, was more beneficial
in reaching practice and clinical targets (Norful et al., 2019). Efforts should be undertaken to
implement the law in each organization by engaging leadership, increasing awareness about the
positive impact of the law promoting NP independence, and promoting relationships between
NPs and physicians (Poghosyan, Norful, and Laugesen, 2018). In 2015, New York State adopted
the Nurse Practitioners Modernization Act to remove required written practice agreements
between physicians and NPs with at least 3,600 hours of practice experience. Such
modernization of licensure laws also applies to physician’s assistants, dentists, pharmacists,
dieticians, and community health workers as a strategy to expand the health care workforce
(National Academy of Medicine, undated).
New Jersey’s COVID-19 Temporary Emergency Reciprocity Licensure Program provided
temporary licenses to more than 3,000 out-of-state health care providers. Over a quarter of these
providers were mental health providers. As the need for mental health care increased during the
pandemic, expanding mental health provider capacity was critical to ensure access to the needed
care. As demonstrated by one study, temporary out-of-state mental health providers helped
ensure mental health care continuity and access (Nguyen et al., 2023).
Seventeen states currently allow for licensure of psychologists at the master’s level, and the
American Psychological Association has developed accreditation standards for such graduate
programs. To integrate authorized subdoctoral psychology practitioners into the mental health
services system, models such as collaborative care or service extender models could be useful to
increase mental health services provided by doctoral-level psychologists (Thompson, 2020).
In our interviews and focus groups, interviewees expressed the need to expand representation
from NPs and PAs as a way of expanding primary care capacity. One interviewee noted,
37
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
NPs are left out of solutions for primary care shortages. Nursing, especially
advanced practice roles, can be part of that solution.
—Nursing representative
Behavioral health stakeholders noted the need to expand the scope of work of nonlicensed
providers as a strategy to increase behavioral health capacity. One interviewee said,
Look at licensing and regulation and streamline [the] process, what you want to
achieve and decide what nonlicensing people can be doing but need to change the
regulation around that as well.
—Community health/behavioral health representative
Behavioral health stakeholders also expressed the need to “broaden” the definition of a
behavioral health provider. One interviewee noted,
Exploring broader roles for health care providers, such as community health
workers, in addressing behavioral health issues . . . can help alleviate workforce
challenges.
—Behavioral health representative
Role of Technology
The use of physician order entry clerical personnel is a simple, cost-effective intervention
that can improve the work experience among primary care physicians (Contratto et al., 2017).
Rural stakeholders indicated that technology (i.e., IT solutions) can play in important role in
retention in Virginia’s rural areas. Some of the issues pertinent to workforce development in
rural areas are unique to these settings. A leader from a rural hospital noted,
EPIC has some solutions but too expensive for a rural hospital like ours—any
technology-based solution for documentation would be too expensive.
—Rural hospital leadership
Resiliency and Well-Being
In our interviews and focus groups, we heard about the need to address stress and burnout
among the behavioral health workforce. Participants noted,
Address the issue of workforce burnout and stress by providing resources and
support for self-care and mental health among mental health professionals.
—Behavioral health representative
Invest in resilience and well-being programs for existing workforce to prevent
burnout and attrition.
—Behavioral health representative
Stress and burnout are common across health care occupations. Building resiliency can
prevent burnout and may improve a resident’s quality of life and health behavior. Research to
date has often focused on providing health care professionals with additional education and
38
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
training on identifying stress and burnout, as well as strategies to address these conditions (e.g.,
Brennan and McGrady, 2015; Couser, Chesak, and Cutshall, 2020). These courses can be
supplemented with online anonymous interactive screening programs targeting well-being and
suicide and/or professional coaching (Norcross et al., 2018; Dyrbye et al., 2019).
However, such courses and support services have not been extensively evaluated for
improvements in health care provider well-being and resilience. Additionally, courses like these
likely provide a partial solution, at best, to health care provider resiliency and well-being, given
the complex myriad of individual, institutional, and other factors driving stress and burnout
(National Academy of Medicine, 2019). For example, participants in qualitative interviews and
focus groups noted that patient-on-provider violence, inappropriate patient-provider ratios
and throughput requirements, burdensome documentation requirements, low and/or
differential compensation for care provided, and lack of basic supplies all impact stress and
well-being in their day-to-day work; all of these factors are at least partially, if not mostly,
driven by institutional factors beyond individual health care providers.
Interventions for Workforce Recruitment
Scholarships, Loan Forgiveness, and Other Support
Currently, Virginia’s support for the health care workforce mainly addresses the financial
burdens related to education and training via scholarships and loan forgiveness. There is
evidence that scholarships and free tuition programs in particular encourage individuals to obtain
additional education and training. Research evaluating programs that offer general scholarships
or free tuition in a field experiment found that individuals attending four-year institutions of
higher education are more likely to enroll and complete a course of study—especially low-
income individuals (Burland et al., 2022; Dynarski et al., 2021).
A similar pattern likely holds for individuals attending technical colleges as well (Anderson,
Monaghan, and Richardson, 2023). There are likely two main reasons for the apparent
effectiveness of scholarships and free-tuition programs. Besides significantly reducing (or
eliminating) the costs of attendance, they are simpler than federal financial aid programs that
require students to navigate processes and submit paperwork each academic year while taking on
sometimes significant amounts of student loan debt. Low-income and otherwise disadvantaged
populations are less likely to take out loans more generally (Waddell and Singell, 2011; Black,
Denning, and Rothstein, 2023; Burland et al., 2022). While the Commonwealth offers general
scholarships that can be applied to health care professions (e.g., G3, FastForward) (Virginia G3,
2023; Virginia Foundation for Community College Education, undated), Virginia legislation and
government websites and interview data suggest there are many fewer scholarships targeted
towards the health care professions than loan forgiveness programs in Virginia (Virginia
39
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Department of Health, undated; VHWDA, undated-b; Claude Moore Charitable Foundation,
2022).
This is in contrast to other states, such as neighboring West Virginia, that have targeted
scholarship programs specifically for individuals entering health care professions. West Virginia
and other states specifically target health care shortage areas, such as rural and other underserved
areas, by offering financial supports for educating and training health care workers (Vestal et al.,
2022). Some states report success with targeted scholarship programs (Hu, Dill, and Conrad,
2022). For example, research suggests that a rural physician scholarship program increased the
number of rural physicians (e.g., (Vestal et al., 2022)). However, evaluation research to date is
relatively sparse, and other studies suggest rural physician scholarship programs do not lead
physicians to change their preferences (Spiers and Harris, 2015). Instead, individuals with
preexisting preferences for rural practice and/or who originally hail from a rural area are the
main participants in such programs, and they remain in rural areas after any terms of service
associated with the scholarship have ended (Rural Health Information Hub, 2022).
Virginia has many more loan forgiveness programs than scholarship programs focusing on
health care occupations. Virginia’s loan repayment programs (Virginia Department of Health,
undated) include a variety of health care professionals (e.g., physicians, nurses, social workers)
and use a mix of Commonwealth, employer, and federal funds (e.g., the Virginia State Loan
Repayment Program [VA-SLRP]) or strictly Commonwealth funds (e.g., Virginia Physician
Loan Repayment Program) (Virginia Department of Health, undated). Unlike its scholarship
programs, at least some of Virginia’s loan repayment programs target underserved areas facing a
shortage of health care providers (e.g., VA-SLRP). Virginia has recently expanded the VA-SLRP
program to include certified nursing aids (Virginia’s Legislative Information System, undated-a).
Prior research suggests that loan forgiveness programs might increase the number of individuals
who become physicians, but the evaluation research to date on physician or other health care
worker loan forgiveness programs is unclear (Bärnighausen and Bloom, 2009; Guan et al., 2023)
regarding the amounts of debt relief for higher-paid health care workers (e.g., physicians)
(Catherine and Yannelis, 2023). Additionally, cited research on scholarship and free tuition
programs suggests that scholarships are likely more effective for propelling disadvantaged
populations into postsecondary education and training.
Scholarships and loans are also used to reallocate Virginia’s existing health care workforce,
not just to recruit new workers. For more discussion on improving systemic efficiency by
reallocating workers, please see the last section of this chapter.
Stakeholders noted the importance of offering loan repayment and scholarship programs as a
recruitment tool in Virginia. In reference to behavioral health, interviewees commented,
Offer loan repayment, scholarships, and incentives for going into behavioral
health, especially to younger, more diverse population. If you come from a rural
or socioeconomically challenged area, the idea of a four-year institutional degree
and borrowing loans seems insurmountable. Federal money can often help pay
40
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
back much of that debt. Loan repayment is helpful for those in the field, but not
for attracting college students.
—Behavioral health representative
In southwest Virginia, people do not like to take on debt, or people of color, or
bilingual people, you need to provide money at the front end.
—Behavioral health representative
Loan repayment is essential. In Virginia, in addition to the regular state-funded
loan repayment, we just created a behavioral health loan repayment program two
years ago. If we pass a budget, the Governor has increased the money in it and
increased the amount available.
—Behavioral health representative
Health care organizations offer loan repayment programs but with high stakes (3-
year commitment), current generation is not into this and [they] don’t take
advantage of those.
—Behavioral health representative
There is a scholarship program for psych NPs. It is a post-certificate program
which has been attractive to older folks who may have some debt but have made
progress in their career. They want to go back to school but they don’t want to
incur more debt. Scholarship pays for the full freight.
—Scholarship program representative
Interviewees also noted a role for private funding for behavioral health education and
training. One interviewee expressed,
Private business funding has not been tapped [into], [we] don’t do it enough. Go
Virginia is already in place but . . . it has not taken off as it should have. It’s in
place already so it’s a shame.
—Behavioral health representative
Attendees of the study’s multi-stakeholder conference discussed industry-supported tuition
models in Virginia where the employer supports their employees’ school tuition. They
mentioned that such models can potentially create an opportunity to move the current generation
into the health care workforce. For example, Bon Secours Health in Virginia pays for
individuals’ education upfront (Bon Secours Mercy Health, undated). They need to make a
commitment on students’ part to stay with industry for a specific amount of time (recently
changed it from three years to 18 months). Then it is on the industry to make students realize
they want to stay.
Behavioral health representatives cited payment for preceptorship/licensed supervision
as a major barrier to behavioral health recruitment and expressed a need for creating
mechanisms to help students pay for preceptorship/licensed supervision. Interviewees
commented,
For people who want to become LCSWs [licensed clinical social workers] and
licensed professional counselors, they are thwarted by paying for the required
41
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
supervision on their own that they have to go through to get a license. …a lot of
people who want to become therapists and are motivated, and have the aptitude,
cannot run that financial gauntlet.
—Behavioral health representative
The licensing process for behavioral health is very cumbersome and expensive.
By the time I graduated I still needed 3,000 hours that had to be supervised for
which I had to pay somebody…This is not a good system. This was 12 years ago
but I am still hearing the same concerns.
—Labor representative
Stakeholders also expressed the need for more effective advertisement around scholarship
and loan repayment opportunities in Virginia. Interviewees shared,
Talking to the psych group, they didn’t know the loan repayment was there to
help recruit faculty or [options to] keep people in Virginia.
—Behavioral health representative
It is shocking how few people know about the behavioral health loan repayment
program. Some of it has to be because it started during the pandemic, and a lot
has to do with...loan repayment in the department [being short-staffed]. We need
to figure out more creative ways to make sure the key folks are aware of the
program.
—Behavioral health representative
Rural stakeholders indicated the need for providing social support to students and trainees as
part of recruitment efforts in rural areas:
Children from rural areas are struggling with complicated issues (food insecurity,
insufficient class supplies, incarcerated parents). Need more resources for
families with K–12 students.
—Rural community health/education representative
Many students are first generation so don’t have the proper support in place.
They play a lot of catch-up...Any financial cost that is incurred (for licensing,
registration, etc.) is going to be a problem, so vouchers and offers along those
lines can go a long way. Cost of education, licensing, and continuing education is
typically a huge barrier for first generation [students].
—Rural community health/education representative
Rural stakeholders also stressed the importance of securing proper housing for students and
trainees in rural areas:
Housing will resolve many of the issues. . . . Medical students have struggled to
find a place to live [in Southwest Virginia] that is of sufficient quality, safe, and
comfortable. . . . There is a sense of “otherness”—inability to find a place to live
42
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
makes them feel unwelcome. A feeling of belonging is critical from the
institution as well as the community side.
—Rural community health/education representative
Wages and Regulations for Educators
The number of available educators has direct implications for whether existing education and
training programs in health care can train a sufficient number of health care workers, especially
in the case of nurse education and training.
Prior research notes that educator shortages in nursing schools limit the number of students a
program can accept and can also degrade the overall experience of students admitted to the
program (Margolis et al., 2023; American Association of Colleges of Nursing, 2022). In
Virginia, the number of nursing educators limits the number of nursing students that institutions
of higher education can accept as well as the ability of these students to find appropriate clinical
placements and preceptors during their education (Zhavoronkova et al., 2023).
Nurse preceptors are a particular pain point in Virginia nurse education and training. Nurse
preceptors must complete an application and pay a fee, regularly renew their preceptor license,
provide on-site and face-to-face instruction, and can supervise no more than two trainees at a
time (Virginia’s Legislative Information System, 2023). Additionally, interviews and focus
groups emphasized revised preceptorship requirements and regulation. Participants shared the
following comments,
Some education . . . is very highly regulated . . . on ratios of faculty/preceptors.
For example, for students at BS [bachelor of science] level, faculty have to be
masters prepared or ask for temporary exception. In community settings, we
should not be limited to ratio, may miss out on opportunity for student learning.
A guidance document that allows us to be more flexible, students can be in
different settings, this is not in place.
—Behavioral health representative
Faculty have to be on site with students. This limits where students can be. For
example, need a faculty to learn about food pantry—does not make sense.
Precious Time Program in James Madison University, which helped disabled
children at home . . . used to count as clinical hours, but now does not count
because faculty is not there.
—Behavioral health representative
Focus group data also suggest that traveling nurses cannot serve as preceptors—even if
travelers have sufficient experience in Virginia health care facilities, currently work in a Virginia
facility, and otherwise meet all preceptor requirements. Notably, preceptors receive little to no
compensation (e.g., University of Virginia, 2023). The recently established Virginia Nursing
Preceptor Incentive Program provides $1,000 for 250 hours of supervision for a maximum of
$2,000 per school year (Virginia Department of Health, 2023).
43
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Nurse preceptor survey data in published research reinforce these challenges to recruiting
new nursing professionals to Virginia’s workforce. Those data indicate compensation and
faculty-preceptor communication are top preceptor concerns. Preceptors also indicated the need
for protected teaching time, preceptor training, and meaningful preceptor recognition (Boyce et
al., 2022; Burt, Sparbel, and Corbridge, 2021).
The Virginia Nursing Stakeholder Workgroup recently published a plan to support Virginia’s
nursing workforce (2023), noting the importance of financial incentives for clinical nurse
preceptors. The group endorsed tax credits as an important way to incentive nurse preceptors.
Many of the same issues for nurse educators’ wages and regulation are evident in behavioral
health as well. Behavioral health stakeholders also noted the importance of paid
preceptorship/licensed supervision and an overall shortage of faculty. Interviewees commented,
I have been against paying for preceptorship, but given the world we live in, we
might have to think that way. Or pay the organization to free up the preceptor.
—Behavioral health representative
Is there enough faculty? No. Why? Because faculty don’t get paid enough, so
they don’t want to be faculty. There are required ratios. The revenue generating
requirements of existing faculty in larger institutions create problems because
they cannot generate revenue while overseeing people. There are a lot of
unreasonable requirements there.
—Behavioral health representative
Preceptor Capacity
Nursing stakeholders expressed low capacity for preceptors and faculty as a significant
barrier to recruitment. Specifically, nursing stakeholders indicated that some nursing schools
have to turn away qualified applicants because of insufficient faculty and preceptor capacity.
According to one interviewee,
The number of students that come out of high school, who want to be a RN, we
are turning them away because of faculty and space on campus shortages. It’s not
a simple fix but still needs to be addressed in some way; it takes a lot of funding.
Scholarships will not make a difference in this case because classes are full.
—Community health representative
Behavioral health stakeholders also indicated the need for creating a pool of behavioral
health preceptors/licensed supervisors by leveraging retirement age providers. They noted that
about 60 percent of providers were retirement age or older in 2019, and that pool can be a source
to help increase behavioral health preceptor/licensed supervisor capacity.
Outreach and the Public “Image” of Health Care Professions
Nursing stakeholders indicated a need for raising public awareness and improving the
“image” of nursing careers through a public campaign. Interviewees identified the negative
44
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
image of health care professions—especially since the COVID-19 pandemic—as a key barrier to
effective recruitment. Interviewees shared that,
Other industries are marketing in elementary and high school. Health care has a
bad PR situation. The headlines around being in health care in general, being
overworked, nursing in particular.
—Behavioral health representative
Johnson & Johnson marketing strategy does a very good job about getting the
right stories out. Be better about sharing stories about the joy of the [health care
provider] jobs, not just entice with benefits and salary.
—Nursing stakeholder
Behavioral health stakeholders shared similar sentiments regarding the need for more
effective outreach and raising public awareness of behavioral health professions for successful
recruitment. Interviewees said,
You have to convince the parents it’s a viable career option. We do a terrible job
at career paths. . . . [High schoolers] can talk to professionals who are in that
field. . . . We just can’t show them a pathway unless we create one.
—Community health/behavioral health
Engage in creative events such as such as college fairs, city buses wrapped in
advertising, career fair targeting just bachelors to get them into the door as case
managers. Offer unique internships etc.
—Behavioral health representative
Need to raise public awareness about the importance of mental health and
behavioral health services, as well as the value of the professionals who provide
these services.
—Behavioral health representative
Implement campaigns and initiatives to combat the stigma around mental health
professions, highlighting the meaningful work they do and the positive impact
they have on individuals and communities.
—Behavioral health representative
We need to make sure there is clear and effective communication to the right
population. What will grab the attention of a 17-year-old?
—Behavioral health representative
The multi-stakeholder conference attendees discussed the importance of rebranding and
improving the perception of current health care jobs among young people through partnership
between the Department of Education; school-based providers. such as high school counselors;
and community colleges.
45
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Health Care Workforce Data-Tracking
During our study interviews we heard from a participant regarding the importance of data-
tracking by colleges and universities in Virginia in order to better understand where students are
going after completion of their programs and to inform efforts to keep students in the
Commonwealth. This interviewee shared,
The colleges and universities do not see it as their role to educate the people in
front of them. They don’t see it as their role to identify positions to help
graduates, encourage graduates to stay in the state. They don’t track where they
go. You would think the alumni office does it, but not the behavioral health
departments. There is a role for the universities where they can be much more
engaged.
—Behavioral health representative
During the study’s multi-stakeholder conference, participants indicated the importance of
leveraging data to track Virginia’s health care workforce needs on an ongoing basis to inform
interventions to address any ongoing staffing shortages.
Virginia has a number of data sources that can help facilitate workforce data-tracking. The
Virginia Commonwealth University’s Department of Family Medicine and Population Health
houses the Virginia Ambulatory Care Outcomes Research Network (ACORN) (Virginia
Commonwealth University, undated-a). ACORN partnered with Virginia Department of Medical
Assistance Services in 2018 to evaluate primary care capacity in the context of Medicaid
expansion in the Commonwealth. Through this work, ACORN identified every primary care
physician in the Commonwealth and conducted extensive interviews and surveys focused on
understanding primary care capacity in both 2018 and 2022. Using its robust Virginia-based
datasets, ACORN has facilitated a deep understanding of primary care clinician panel size, scope
of practices, and communities served (Huffstetler et al., 2022).
The George Mason University Center for Health Workforce (Mason Center for Health
Workforce, undated) conducts workforce data collection and analysis and leads collective efforts
in health workforce planning and development in Virginia. Notably, the center houses health
care workforce supply and demand dashboards.
Data sources such as those utilized by ACORN and Center for Health Workforce can be
leveraged to conduct surveillance on the health care workforce needs of the Commonwealth to
inform needed interventions and policies on an ongoing basis.
Recruiting Youth and Members of Underrepresented Communities
African-American/Black physicians are more likely to practice primary care and care for
underserved populations than other physicians and remain underrepresented in medicine.
Because Historically Black Colleges and Universities (HBCUs) have a legacy of matriculating
students with lower standardized test scores, majority medical schools may be reluctant to form
partnerships with HBCUs. Despite lower average Medical College Admissions Test (MCAT)
46
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
scores, groups did not differ significantly in terms of graduating, matching into residencies,
selecting primary care, or achieving board certification (Capers and Way, 2015).
Nursing, primary care, and behavioral health providers in our interviews and focus groups
indicated a need for outreach to school-aged children regarding health professions. Interviewees
shared,
Need to go into schools early . . . even 5th grade. Middle school kids are not
encouraged to go into health care. . . . A lot of people who are awesome at
[providing health care] got Bs in high school. GMEC runs a summer camp and
will use Claude Moore money to support health care counselors for overnight to
encourage kids re health care careers.
—Rural provider
Get to them in high school and in training, too. You need to expose them to long
term settings. Incentive-wise show high school graduates, if you become a LPN,
you can go up the ranks (many examples of starting off work at front desk and
now own nursing homes). Pathways are there but need to make clear what the
opportunities are.
—Community health representative
Dedicated recruiters need to fan out and tap into underserved communities to get
the message out . . . need to think holistically and strategically, not just post it.
[Adopt] a grassroot messaging approach for local communities in addition to
providing support such as child care, transportation etc. This is where gains can
be made where historically it has not been.
—Labor workforce representative
Primary care stakeholders noted that AHEC can probably play more of a role in outreach to
underrepresented communities for health care professions recruitment. One interviewee said,
Not enough being done with AHEC to introduce health careers to students/people
in different communities. Need recruitment strategies for people of color. They
need more information, introduction of public health in that community,
specifically. We need to garner a lot of ideas from those populations, talk to
white kids, Black kids, brown kids. . . . You’ll get different solutions.
—Community health center leader
Behavioral stakeholders also stressed the need for specific recruitment strategies for
communities underrepresented in health care professions. Interviewees said,
Another way to motivate people of color who are leaning towards the behavioral
health profession is to see people who look like them in those roles, like a
counselor or similar capacity. Build pipelines way back down into grammar
school, middle school, involve the guidance counselors, particularly in
underserved populations.
—Behavioral health representative
First generation students need extra support. That is a strategy that brings them
along. That leads to them becoming licensed and staying practicing. For our
47
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
programs, we recruit by saying we will pay for things. That removes a financial
obligation.
—Behavioral health representative
Online Instruction
Health care education experts have highlighted the need to center clinical education around
online learning (e.g., Emanuel, 2020). Online instruction can increase access and flexibility to
local and nonlocal clinical education, especially at the introductory level (Goldberg and
Crocombe, 2017). This was demonstrated during the COVID-19 pandemic when most health
care education was forced online. However, a relatively robust body of research on the
effectiveness of online postsecondary education generally suggests that online instruction can
negatively impact student learning, with some important caveats. Online education does not
seem to support student learning any more or less than large, in-person classes. Also, mixed
instruction with in-person and online learning does not negatively affect student outcomes.
Lastly, online education that uses peer-to-peer learning or other collaborative and social
components may improve the effectiveness of online learning (Zhao et al., 2018; Alpert, Couch,
and Harmon, 2016; Bowen et al., 2014; and Joyce et al., 2015).
Virtual instruction can support new approaches to enhance conventional learning. During the
COVID-19 pandemic, interactive virtual clinical teaching was routine, including virtual rounds
on hospital wards, virtual clinical clerkships, virtual emergency department callbacks where
emergency department providers contact discharged patients for follow up evaluation, virtual
medical response teams for supporting community health teams, and greater access to instruction
from renowned clinicians (Wilcha, 2020).
In spite of its promises, online instruction is not without weaknesses. Online instruction can
contribute to isolation and disengagement, is associated with lower grades, and might not confer
attitudinal growth critical for skillful clinical practice, such as willingness to conduct routine
screening for illicit drug use (Lee et al., 2022; Knopf-Amelung et al., 2018). That being said, the
COVID-19 pandemic led to natural experiments in online instruction, often for medical students,
that can be built upon for all health care professions to address shortages in experienced
instruction in the classroom and in clinical settings. Combined with conventional instruction in a
blended learning environment (McCutcheon, O’Halloran, and Lohan, 2018; Terry et al., 2018)
and perhaps using advanced training technology such as augmented or virtual reality, online
instruction is an important tool for innovation in health care education and training (National
Academies of Science, Engineering, and Medicine 2021).
Pathways to Health Care Professions
Nursing stakeholders discussed the importance of creating multiple pathways to nursing
careers as a recruitment strategy. Interviewees shared,
48
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
One of the things we’re doing at my institution, we are on a pilot…to help build
this pipeline, where students take particular curricula and they get dual credit and
they have a pathway so they get exposed to the health care careers, not just
nursing but all the health care careers and have shadowing opportunities.
—Nursing stakeholder
[Offer] tuition assistance and demonstrate it’s an ascending staircase and have
some sort of mentorship support. I was in your shoes just six years ago… Plug
away, and there is a future for you here. Not all organizations can lean towards
that but demonstrate that there is a career.
—Labor workforce representative
In Minnesota, the governor said we are going to recruit by paying for tuition, pay
for working time, pay for uniform, etc. We need to get 1,000 CNAs [certified
nursing assistants] within 90 days, and they were immensely successful. It cost
the state $5–6 million. CNAs can be trained very quickly, that’s why it was the
focus. Retained 20 percent of those that were trained.
—Community health representative
Other stakeholders shared similar thoughts about the importance of developing
the primary care pipeline. They shared the following comments:
Offer paid internships at high school level so students can be exposed to the
reality of health professions; match undergrads with medical students and
promote mentorship alongside internship.
—Labor representative
One of the biggest barriers is lack of equitable pipeline. There is a well-earned
mistrust in certain communities towards medicine and health care. Look at
pipeline overall and create something that garners trust.
—Labor representative
In-depth understanding of the pipeline is needed as well as an understanding of
capacity of the programs. Building comprehensive, in the weed maps of what is
happening and what the need is. A three-year goal is a sweet spot for identifying
concerns and launching pilot programs to address sub concerns, followed by
evaluation of programs and what can be scaled up.
—Labor representative
Behavioral health stakeholders raised the need for implementing alternative pathways for
behavioral health licensure. They noted that in order to increase the number of individuals
entering behavioral health fields, there is a need for more entry ramps and educational pathways
into the behavioral health workforce. This includes developing academic programs at different
levels, such as associate and bachelor’s degrees, to provide more immediate relief to the
workforce shortage. Interviewees shared these comments:
49
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Implement expedited and streamlined licensure processes for in-demand areas of
specialties facing workforce shortages so that qualified professionals can enter
the field quickly.
—Behavioral health representative
Offer structured and supervised training programs to individuals to gain practical
experience while working towards licensure.
—Behavioral health representative
Establish interstate telehealth licensure compacts that allow professionals to
practice across state lines without requiring additional licenses.
—Behavioral health representative
Participants in the study’s multi-stakeholder conference stressed the importance of creating
intermediary on-ramps for behavioral health workers given the current path to the field is entry
level and it takes six years to reach licensure. They offered focusing on growing the pool of NPs
and PAs in primary care and behavioral health as a strategy to increase additional capacity in
these fields. Programs such as those offered by George Mason University (Nursing Process,
undated) can help accelerate family NPs into psychiatric NPs to help fill the dual void.
Multi-stakeholder conference participants raised the need to expand on innovative programs
such as I2CRP and fmSTAT offered at Virginia Commonwealth University (Virginia
Commonwealth University, undated-b; Virginia Commonwealth University, undated-c) that
attract students into the health care workforce and incentivizes them to stay in Virginia. They
also mentioned the MedServe model from North Carolina (MedServe, undated), which takes
advantage of gap year and allows for fellowship opportunities whereby students get residency in
North Carolina and typically chose to stay there because the tuition is less.
Conference participants also offered engaging existing community health care workers and
strategizing to transition them into licensed/specialized roles in primary care and behavioral
health as a potentially promising solution for increasing the workforce capacity in these fields.
They raised transitioning service members (as well as military spouses)—many of whom are
qualified health care experts transitioning from the military—into civilian employment,
anchoring them in Virginia as a strategy to retain talent. Participants recommended evaluating
recruitment opportunities such as military medical school physician preparatory program
(College Recon, 2021), designed to give promising enlisted service members interested in
becoming military physicians a pathway to medical school.
Rural stakeholders stressed the importance of creating pathways for rural residents to enter
nursing, primary care, and behavioral health careers:
50
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Recruitment locally is important. Especially in rural areas. Recruit from rural
high schools. True for all three fields.
—Primary care physician/educational leader
Students from local community college . . . plug into a nurse externship and they
work as CNA until they finish their program and they retain about 50 percent of
them. There’s a nurse in the hospital that is the key liaison to community college.
—Rural hospital leadership
There are only a few med schools in rural settings in the U.S., and most are D.O.
programs.
—Primary care physician/educational leader
Competency-Based Education
Behavioral health stakeholder discussed how practice hours for behavioral health training
should be set based on the level of competency. They noted that individuals should be assessed
based on their skills and competencies rather than solely relying on their level of education and
training. One interviewee shared,
There is a limit, 500 minimal of practice hours. Some people will be competent at
300 but still have to do 500; others will not be competent even at 700.
—Behavioral health representative
Work-Based Learning for Pay
Work-based learning, paid and unpaid, is not a new concept—although it is one that has
garnered increased attention of late as a tool for addressing a variety of education, training, and
workforce issues in health care and non–health care occupations. Health care occupations have
historically included more work-based learning in education and training than other occupations
(e.g., clinical rotations, clinical intern or clerkships, and residencies). However, stakeholders are
increasingly interested in expanding these options across different health care occupations,
particularly in nursing (Goolsbee, Hubbard, and Ganz, 2019; Love and McCarthy, 2018; U.S.
Department of Labor, undated). Residencies and apprenticeships are two primary forms of work-
based learning under consideration by health care workforce stakeholders in Virginia and beyond
(Wilkinson, 2023).
Virginia is currently proposing to build upon its existing 12-month apprenticeship or “learn
and earn” programs for certified nursing assistants to include nursing staff at all levels based on
collaborative partnerships between K–12 and postsecondary educators and health systems and
other employers (Commonwealth of Virginia, 2023). Virginia joined a number of other states
that offer general nursing apprenticeships (Love and McCarthy, 2018; U.S. Department of Labor,
undated). The Council on Adult Education and Learning partnered with nine states under a U.S.
Department of Labor grant to offer an apprenticeship that combines online academic instruction
and work-based training and mentoring while the student earns a paycheck (U.S. Department of
51
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Labor, undated). Nevada also offers a nurse apprenticeship program (Figure 2.1), including a
transportation voucher for students/apprentices who travel at least 50 miles one way in this
largely rural state (Nevada Rural Hospital Partners, undated).
In the study interviews, we heard about the importance of paid internships in behavioral
health. One interviewee commented,
Paid internship for behavioral health should be an option instead of frontloading
all the debts and then getting the hours. Behavioral health apprenticeship exists in
the country . . . and Virginia has just authorized its first one.
—School of nursing representative
Licensing Exam and the Licensure Process
Behavioral health stakeholders expressed a lot of frustration with the licensing exam and
licensure processes in the Commonwealth, indicating that examinees should have more frequent
opportunities to retake the exam and that the licensure timeline needs to be truncated.
Interviewees commented that,
The amount of time that it takes once you submit your application [for licensure]
to get that turned around, it’s criminal. [The Department of Health Professions] is
working to bring that down.
—Community health/behavioral health representative
Social workers have to get a huge portion of clinical hours before they can take
the exam. If you fail, you have to wait 90 days [before retaking].
—Community health/behavioral health representative
Importantly, many nursing apprenticeship programs center around career pathways with
defined training levels that build on each other, such that an apprentice has a clear pathway(s)
from nursing aid to nurse leader positions (U.S. Department of Labor, undated). This follows
current practice in postsecondary education and training around “guided pathways” that aim to
clarify and support student pathways through postsecondary enrollment to credential completion
(Bailey, Jaggars, and Jenkins, 2015).
In general, research suggests that apprenticeship and earn-and-learn models might positively
impact individual learning and wage outcomes (Reed et al., 2012; Kemple 2008), but the size of
these potential positive impacts likely depends on the economic and other contextual factors of
apprenticeships (Novella and Pérez-Dávila, 2017; Dougherty and Ecton, 2021). However,
evidence on the effectiveness of apprenticeships generally, and particularly in health care, is still
developing (Rosen, Visher, and Beal, 2018; Goolsbee, Hubbard, and Ganz, 2019). To date,
research on nursing apprenticeships is highly descriptive but suggests financial incentives and
mentoring activities are central to training and retaining new nurses (Brook et al., 2019).
Apprenticeships are also a potentially important mechanism by which to increase and diversify
the recruitment pool for new nursing staff (Love and McCarthy, 2018).
52
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
While physicians complete an obligatory residency before becoming fully certified to
practice, residencies are not common in other health care professions. Moreover, it is noteworthy
that policy perspectives on residencies for physicians and other health care professionals are
trending in opposite directions. While residencies are often seen as a barrier to practice in
primary care for physicians who have completed medical school (Boyle, 2020), residencies are
seen as an answer to enhanced recruitment of new nursing and other health care professionals
(Brook et al., 2019; Goodwin et al., 2021; Finneran and Kreye, 2019; Goolsbee, Hubbard and
Ganz, 2019).
Nursing residencies vary in duration but appear to uniformly include work-based learning
upon completion of a regular degree program and under the supervision and support of a clinical
preceptor and/or mentor. Evidence is limited on the effectiveness of residency programs for
nursing professions. However, it appears that programs 12 months in duration can reduce nursing
staff turnover in the first year after initial hire (Perron et al., 2019). Residency programs might
also improve confidence, competence, job satisfaction, critical thinking, and mental well-being
for a nurse (Edwards et al., 2015).
In contrast to policy momentum supporting the expansion of residencies for nursing
professionals (especially new RNs and NPs), policy efforts related to the medical residency have
focused on ensuring the residency is not a barrier to general practice. Virginia, for example,
introduced legislation in 2023 to allow individuals who successfully completed medical school
but did not immediately match to a residency to practice as an associate physician in the interim.
Though this legislation ultimately did not pass, other states have established laws to allow
physicians to practice without completing a residency, including Missouri, Arizona, Florida,
Washington, Arkansas, and Utah (Sawhney, 2022). States have reduced residency requirements
in other ways. For example, Tennessee now allows foreign-trained physicians to practice
medicine without completing a residency in the United States. (Hollowell, 2023). Others go
further, compressing physician education overall to reduce time to full independent practice
while still providing standardized training under medical experts in a given field. In 2009,
Shanghai, China began piloting a program to train fully practicing physicians in eight years—
five years for a medical bachelor’s degree and a three-year residency (Zheng et al., 2020). This
represents a reform over the traditional Chinese model of a bachelor’s in medicine followed by
an apprenticeship with a local, practicing physician. Under Shanghai’s pilot, a three-year
residency aimed to provide more standardized training with national experts in a field to improve
the overall quality of the physician workforce while still keeping a new physician’s time to
independent practice minimal. Others have simply established new medical schools to provide
additional residencies or funded new residency positions (e.g., Brazil, Hone, et al., 2020; Centers
for Medicare and Medicaid Services, 2021). As is typically the case in health care and other
workforce research, little research evaluates the impacts of these interventions on physician
training, overall supply, retention, quality care provided, or other outcomes (Hone et al., 2020).
53
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Education/Training Curricula
In response to whether stakeholders believe that medical school education and/or primary
care training should be shortened as a recruitment strategy, stakeholder shared,
Revamp the [primary care] curriculum instead of shortening the training period
for providers. Students benefit from exercising knowledge in real world
environment and see what they learn in the classroom does not jive with real
world, it’s a lesson to learn. Get the book knowledge, and then learn the practical
experience. Shortening education will make them less prepared; curriculum
needs more updating. Programs that offer immersion with the community, such
as Community Nourishment Project, relieves the jaded-ness and preserves
empathy. But if data is showing no measurable change or improvement in
outcome from shortening the education/training period, then it should be
explored.
—Rural community health/education representative
There is a generational change, current generation is more aware of the world, in
tune with emotional needs of people, so a revamping [of the curriculum] will
need to address things like multiple disabilities, insurance barriers. . . . Social
determinants of health need to be looked at from modern perspective.
—Rural Community health/education representative
While I do think that there is a lot of value to shortening the curriculum between
high school and college, doing that kind of dual program, I just don’t think that it
will take root in Southwest Virginia until we address the infrastructure issues that
we see with K through 12 education in this area.
—Rural community health/education representative
Rural stakeholders stressed the importance of providing cultural awareness training to
students and trainees in rural areas:
There is an element of mistrust running through the population. It’s an
underserved community, coal mining camps, etc., so the providers need to be
aware of the [need for] cultural competency.
—Rural community health/education representative
When you have health care professionals making assumptions about patients and
lack cultural competency of that local region, there is a problem. Jobs held by
local leaders can help alleviate this concern.
—Rural community health/education representative
Advanced Education/Training Technologies
An example from midwifery demonstrates innovations in clinical practice and education
through the use of web-based technologies, such as three-dimensional (3D) virtual learning
environments in interprofessional education. A nurse-midwifery education program at a
Southeastern U.S. university provides web-based interprofessional education for nurse-
midwifery and third-year medical students using the Virtual Community Clinic Learning
54
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Environment (VCCLE). The VCCLE is a 3D, web-based, asynchronous, immersive clinic
environment where students interact with instructor-controlled virtual patients and virtual
preceptor avatars. They move through diagnostic sequences and are asked to develop a care plan
for women throughout the lifespan. Through this interface, students learn both clinical
competencies and competencies for interprofessional collaborative practice (Reis, Faser, and
Davis, 2015).
More students, including those who live in rural and underserved regions and who otherwise
may not be able to attend a traditional onsite campus, are provided the opportunity to complete
quality midwifery education through online programs, which in turn may help expand the
women’s health care provider workforce (Arbour, Nypaver, and Wika, 2015). Furthermore,
simulation-based learning is increasingly seen as an effective strategy to develop student nurse-
midwives into effective health care team members (Warren et al., 2016; Ruyak et al., 2018).
Training programs that integrate the use of technology are needed to prepare students for the
future of service delivery in psychology. Training opportunities in telepsychology can equip
students to develop, implement, and evaluate innovative modalities for mental health care.
However, few options are available for trainees seeking to acquire these experiences, and while
guidelines for service provision using technology are emerging, guidelines for and documented
examples of training programs in telepsychology are virtually nonexistent. The Telehealth
Counseling Clinic, located at the Texas A&M Health Science Center, has developed a training
model to prepare the next generation of psychologists to work with new technology to provide
counseling services to low-income and uninsured individuals living in designated Mental Health
Professional Shortage Areas in rural Texas (McCord et al., 2015).
The U.S. Department of Veterans Affairs Connecticut Healthcare System (VACHS) is one of
five Centers of Excellence in Primary Care Education pilot sites. The overall goal of the
CoEPCE program, which is funded by the Office of Academic Affiliations, is to develop and
implement innovative approaches for training future health care providers in postgraduate
education programs to function effectively in teams to provide exceptional patient care. This
longitudinal study works to understand the effect of a combined nursing and medical training
model on professional identity and team development at the VACHS CoEPCE site. The study
findings suggest that internal medicine residents initially struggled to understand NPs’ roles and
responsibilities, whereas NP trainees doubted their ability to work alongside physicians. At the
end of one academic year, these uncertainties disappeared, and a strong interprofessional team of
health providers developed. This study provides early evidence of successful interprofessional
collaboration among NPs and internal medicine residents in a primary care training program
(Meyer, Zapatka, and Brienza, 2015).
Number of Residency Slots
Behavioral health stakeholders indicated the need for increasing the number of psychiatry
residency slots. One interviewee said,
55
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
NOT CLEARED FOR PUBLIC RELEASE. DO NOT CITE.
The national bodies only give you certain slots, that’s where faculty ratios come
in. The general assembly got focused on child psych, so ten more seats opened up
for residencies. But it is piecemeal. It is not a silver bullet. It is all intertwined.
Paying for more slots is helpful, but getting national boards to allow more slots is
important.
—Behavioral health representative
Notably, there is recent precedence for Congress approving 1,000 new residency
slots targeted at training in rural and other underserved areas (Rains et al., 2023).
Behavioral Health and Primary Care
Behavioral health stakeholders in our interviews indicated that behavioral health is often seen
as separate from primary care, leading to a fragmented approach to health care. They noted that
integrating behavioral health services into primary care settings and training primary care
providers to address behavioral health issues can help improve access and quality of care.
Interventions That Augment the Workforce by Increasing Efficiency
Geographic Reallocation of the Health Care Workforce
As mentioned above when discussing financial incentives for recruitment, the Virginia State
Loan Repayment program (VA-SLRP), administered by Virginia Department of Health, Office
of Health Equity, incentivizes health care workers to serve for two years in underserved locations
in return for $7,000 to $20,000 of state-funded student loan repayments (up to $40,000 per
recipient when including matching Federal and employer funds).
Although moving a health care worker to an underserved location in Virginia often produces
no net gain for Virginia recruitment, it may still enhance the workforce by increasing structural
efficiency. Workers functioning in a more efficient system may produce improved health
outcomes in Virginia.
For example, one physician interviewee mentioned how addressing workforce shortages in
underserved areas of Virginia can increase the overall efficiency of the system by reducing the
number of emergency cases in underserved areas. This interviewee noted that,
We don’t have nurses and physicians and primary care people in the right places.
. . . We don’t have enough people in rural areas, underserved areas and that just
puts pressure on everybody that this makes patients sicker. So, when they go to
Regional Medical centers, they’re sicker.
-Physician stakeholder
The Virginia Department of Health has collected some preliminary evidence that recipients
of SLRP often stay in underserved areas even beyond their years of required service. Data from
97 SLRP recipients who completed their obligatory service during 2021 or 2022 show that 51
percent remained in their original obligation post as of 2023. Future analysis of how many SLRP
56
NOT CLEARED FOR PUBLIC RELEASE. DO NOT CITE.
recipients move to locations that are not HPSAs will reveal the return on investment that SLRPs
provide. If, for example, the state provides $10,000 to help persuade a health care worker to
serve in a HPSA for one year, that may be a steep price to pay for this improvement in Virginia’s
labor allocation ($10,000 per year). If, on the other hand, the program expects 50 percent of
SLRP recipients to serve in a HPSA for decades after their one or two-year obligation, then the
investment may be well worth the cost (less than $1,000 per year of HPSA service, assuming the
worker wouldn’t have moved to the HPSA without the incentive).
Technology Interventions That Improve Health Outcomes Produced by a Given
Workforce
Some of the technologies mentioned earlier in this report may increase retention or
recruitment of health care workers, and they can also increase the structural efficiency of any
given workforce. For example, even if a word recognition or artificial intelligence (AI)
technology that facilitates notetaking has little or no effect on retention or recruitment, it may
improve the health outcomes produced by any given workforce by freeing up worker hours.
Hours spent performing notetaking or physically demanding tasks can then be spent providing
other services that improve health outcomes. Clinical decision support and other digital tools
such as a patient’s risk calculator can reduce physician cognitive burden and allows for more
informed decisionmaking, better tracking, and improved communication with patients and across
the care team (Mujumdar and Jeffcoat, 2022). Even during the upcoming 15-year window that is
the focus of this study, hospitals are likely to deploy machines that lift patients, perform
sanitation tasks, and fetch supplies (Wang, 2022), which should free up significant health care
worker time to help patients in other ways.
The health care sector is unlikely to see the decline in demand for workers that many other
sectors of the economy have experienced, but even marginal use of machine learning has already
been demonstrated to augment the amount of diagnostics a single worker can perform
(Henriksen, 2022). These applications could eventually empower any given number of primary
care physicians to diagnose a greater number of patient conditions, and may also improve the
accuracy of diagnoses, both of which will improve health outcomes.
Rural stakeholders noted the importance of technology in creating work efficiency and for
improving patient outcomes:
Data tracking/health information exchange system can help with patient care in
rural areas. If we had a good data tracking system, the health information
exchange—like ConnectVirginia, which is still not operationalized after 3–5
57
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
years—we could facilitate patient-centered care plans. . . . This exchange can
facilitate patients receiving care anywhere in Virginia.
—Rural representative
Interventions That Improve Health Outcomes Produced by a Given Workforce
Removal of reporting, licensing, and other administrative burdens placed on health care
workers can increase health outcomes produced by any given health care workforce.
Interventions that reduce these burdens are discussed above in the context of retention and
recruitment, and it is also worth noting that such interventions can bolster any existing workforce
by improving the structural efficiency of the system in which they operate.
58
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Chapter 3. Recommendations for Strengthening the Virginia
Health Care Workforce
Shortages in nursing, primary care, and behavioral health workforces are an on-going
challenge across most of the United States and in Virginia. In the previous chapter, we identified
potential interventions to address key drivers of Virginia health care workforce shortages, while
in this chapter we recommend prioritizing a subset of those interventions based on their potential
impact and their practical accessibility.
Our results show that multiple interventions involving all key stakeholders in the
Commonwealth are needed to reduce or eliminate the health care worker shortages. We
forecasted these shortages using a model of supply and demand that quantifies the impact one or
several interventions may have on the future supply of workers. Stakeholders that may
implement these interventions include health care and educational organizations, public and
private enterprises, and lawmakers—all of whom may leverage innovative methods and
technology to bolster the health care workforce. Our labor supply and demand forecasts indicate
that if these stakeholders do not intervene, workforce shortages will only worsen over the next 15
years. Our models also help to quantify the extent to which various interventions may reduce or
eliminate workforce shortages in the coming years.
This chapter highlights the most promising interventions for improving retention,
recruitment, and the efficient use of Virginia’s health care workforce. Some of the interventions
we identify help to bolster the workforce in more than one way (i.e., retention-, recruitment-,
and/or efficiency). However, each recommended intervention is assigned to the facet of
workforce improvement that it affects the most. For example, an intervention classified under
“Retention” can also impact recruitment, but it primarily impacts retention of the existing
workforce. If retention interventions result in creating better jobs in these health care fields and,
consequently, happier workers, this may impact the reputation of the field, which may encourage
others to join those workforces.
Within each grouping, we also categorize interventions by whether they are financial,
academic, nonacademic, or institutional. These categorizations facilitate mapping interventions
to the key actors or stakeholders that are best placed to implement them. For example, a state
government representative working on the budget may wish to focus on the interventions that are
financial.
Although we identify interventions specific to nursing, primary care, and behavioral health,
some of the interventions are cross-cutting and pertinent to all three fields. Increasing pay and
other financial incentives, reducing the burden of documentation, investing in professional
development, implementing team-based care, increasing workforce diversity, leveraging cross-
sector collaboration, reevaluating scope of practice, and leveraging technology are important for
59
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
workforce retention in all three fields: nursing, primary care, and behavioral health. Providing
scholarships, loan forgiveness, and other support, conducting more effective outreach to
communities, correcting the public image of the health professions, data-tracking, and targeted
recruitment strategies for underrepresented communities are also important for recruitment in all
three fields. Interventions that increase workforce productivity, such as reallocation of workers to
where they are most needed, leveraging technology to improve patient outcomes, and removing
work barriers that impede worker efficiency and productivity are also pertinent to all three fields.
Our quantitative forecasts support the conclusion that many interventions are applicable to all
three workforces, but their effect size can vary depending on the workforce. For example, raising
real wages always has at least some positive effect on retention and recruitment, but our model
suggests that the same percent increase in pay may have a greater effect on expanding the
primary care and nursing workforces than it has on expanding the behavioral health workforce.
Although our model assumes real wages increase by 2 percent each year for all baseline
scenarios, and real wages increase 3 percent each year for all wage increase scenarios, the effect
of wage increases after 15 years on each workforce’s number of FTE workers (relative to
baseline forecasts) varies: Our 2038 primary care workforce supply forecast in the wage increase
scenario is 4.4 percent larger than our baseline forecast of the 2038 primary care workforce,
while nursing is 3.2 percent larger, and behavioral health is 2.7 percent larger than the baseline
forecasts.
Key Findings
The results from the system dynamics model show that if current trends persist, nursing,
primary care, and behavioral health workforce shortages will increase across Virginia.
By reducing the barriers that cause reluctance to work in the field of nursing by half—
barriers such as workplace violence, trauma, and stress—we forecast a rapid increase in the
number of nurses employed. However, over a period of 15 years, increased recruitment will have
a more dramatic effect on workforce numbers, surpassing the effect of decreased barriers.
Increasing nurse wage growth from an average 2 percent annual growth to 3 percent annual
growth also increases the number of nurses retained and recruited, but not as much as the other
interventions do. As expected, a combination of all three interventions has the largest effect on
the forecast of nurse employment, boosting the total number of FTE nurses employed in 2038 by
a little more than 10,000 above the baseline projection.
Unlike our forecasts for RNs, our forecast for the number of active primary care workers in
the Commonwealth shows a decline under baseline conditions, making the future shortfall in
primary care workers even more severe than for nursing alone. Similar to our model for nursing,
the largest immediate boost to the workforce would come from cutting in half the number of
barriers that make workers reluctant to work in the field. Over the course of 15 years, the effect
on total employment of increasing wage growth catches up to the effect of decreasing barriers.
60
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
The effect of increasing recruitment is more modest because baseline recruitment begins at very
modest levels. An ambitious combination of all three interventions (the multi-intervention
scenario) would boost total employment by more than 2,000 workers.
If current conditions persist, the number of behavioral health workers employed in Virginia is
forecast to slowly decline in the coming years. Implementing any one of the three interventions
promises to keep the existing workforce numbers at 2023 levels, but only by combining all three
interventions can Virginia produce a sustainable increase in behavioral health workers employed.
The following sections include key takeaways from the study’s environmental scan and
qualitative and quantitative analyses regarding solutions for nursing, primary care, and
behavioral health workforce retention, recruitment, structural efficiency, and related innovations.
Nursing Retention and Recruitment
• To successfully retain the current nursing workforce in Virginia, the Commonwealth
needs to ensure safe and supportive workplaces that prioritize nurse autonomy in
scheduling and enforce zero tolerance for workplace violence.
• Given the level of risk involved in nursing work (e.g., exposure to infectious diseases,
mental trauma, and potential violence), especially in inpatient and emergency department
settings, nurses at all levels need to have health insurance that adequately covers such
risks.
• Nurses need to spend more time at bedside and less on documentation (i.e., “charting”) at
the computer. Voice-recognition and artificial intelligence (AI) technologies may help
facilitate reducing documentation burden for nurses.
• Patient-to-nurse ratios need to be replaced with patient to “care team” ratios and need to
factor in patient acuity and complexity and the team’s combined level of experience and
expertise to inform patient assignments. Such an approach will allow providers to work at
the top of their license by including and shifting tasks that do not require as much
expertise to other health care workers on the team.
• Bedside nurses need to have an active voice in governance, and their professional growth
needs to be supported by employers.
• Resources are needed to address nursing staff mental health and to mitigate burnout.
• Recruiting a diverse nursing workforce that reflects the communities served by a health
system can help retain nurses from historically underrepresented communities.
• Scholarships, loan forgiveness, and other forms of financial support need to be expanded
and robust advertisement strategies implemented so that those interested in nursing
professions can see a path forward in the field.
• Investments in expanding nursing slots at colleges and universities need to occur
concurrently with expanding clinical training facilities and the number of nursing faculty
and preceptors.
• Recruiting retired nurses to serve as faculty and preceptors in sufficiently paid positions
could help increase nursing faculty and preceptor capacity.
• More pathways need to be created for individuals to enter nursing fields and work-
based/learn and earn opportunities should be leveraged—especially through health
system nursing college/university partnerships.
61
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
• To improve recruitment, the public image of nursing professions needs to be improved,
and specific strategies need to be developed for outreach to school-aged children and
communities underrepresented in nursing through collaboration between hospitals and
health systems, nursing colleges and universities, and K–12 schools.
Primary Care Retention and Recruitment
• Payment increases (especially Medicaid payments), loan forgiveness, and other financial
incentives are important for retaining primary care providers. Virginia should join the
ranks of 21 other states that currently mandate a minimum state primary care spend to
ensure that the Commonwealth has the resources needed to retain and expand this
workforce (Johnson et al., 2022; National Conference of State Legislatures, 2023).
• Reducing patient-provider ratios by implementing team-based care in primary care
settings should be evaluated as a strategy to reduce the burden of high patient volumes on
providers.
• Implementing strategies for provider professional and personal development will be
important to ensure continued professional growth and retention of the primary care
workforce.
• Increasing representation from PAs and NPs in primary care should be considered as a
cost-effective approach to increase primary care capacity.
• Virginia should provide scholarships for underrepresented groups as opposed to loan
forgiveness.
• Furthermore, creating opportunities for medical students to do rotations in federally
qualified health centers (or requiring such rotation as a core elective) can increase
provider capacity in these settings and can serve as a recruitment tool.
• Given that African American/Black physicians and Doctors of Osteopathic Medicine
(DOs) are more likely to enter primary care fields and practice in underserved
communities, targeted recruitment strategies to attract physicians from these communities
to primary care professions may be an effective recruitment strategy for primary care.
• The Virginia legislature should consider adopting legislation, similar to that adopted by
Tennessee (National Academies of Sciences, Engineering, and Medicine, undated),
allowing physicians trained in residencies outside of the United States to practice in
Virginia without having to repeat a residency.
• Virginia should adopt legislation to allow medical school graduates to practice primary
care under supervision while waiting for a medical residency match. This policy
intervention can help recruit additional primary care capacity in the Commonwealth.
Behavioral Health Retention and Recruitment
• Increasing payment for behavioral health services is a critical step for both retention and
recruitment of workers in the behavioral health professions. Increasing Medicaid
payment, in particular, can help with retention. Increasing payment can help reduce
provider patient volume (or panels) by ensuring that behavioral health services are
supported through adequate staffing so the needed care can be delivered to every patient.
• Documentation and regulatory and reporting requirements should be reevaluated in order
to reduce redundant and unnecessary tasks for providers that lead to stress, burnout, and
inability to focus more on direct patient care.
62
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
• Addressing the mental health needs of behavioral health providers who work in a field
that is often very stressful can help to mitigate burnout and attrition.
• To recruit new behavioral health providers, several actions will be critical, including
creating new (and shorter) pathways to behavioral health professions and expanding the
pool of preceptors and licensed supervisors by leveraging retired providers and creating
mechanisms to pay these individuals to reduce the financial burden on students and
trainees (for example, by allowing providers to bill for supervised visits).
• The Commonwealth should evaluate mechanisms for competency-based certification so
that persons can get certified regardless of the number of hours of preceptorship/licensed
supervision they receive if they demonstrate proficiency in the required skillsets.
• In addition to shortening the licensing process for behavioral health professions, the
Virginia Department of Health Professions should consider allowing examinees that do
not initially pass the exam to take the licensure exam more frequently.
• Create mechanisms to help students pay for preceptorship/licensed supervision.
• Loosen strict behavioral health preceptor and licensed supervisor requirement such as
educational degree requirements for preceptors /licensed supervision, preceptor/licensed
supervisor-to-student ratios, and need for preceptors/licensed supervisors in community
settings.
• A public education campaign that both helps counter the stigma associated with working
in behavioral health fields and highlights the rewarding aspects of the work—especially
in communities underrepresented in behavioral health—should be prioritized as a
recruitment strategy.
Increasing Structural Efficiency and other Innovations
In addition to retention and recruitment interventions, the Commonwealth should consider
strategies that optimally use the current workforce through increasing structural efficiency. Some
of the interventions that we identified have limited evidence behind them to determine if they
increase the number of available health care professionals or otherwise address critical health
care workforce issues (e.g., burnout). Nonetheless, some are innovative and worth considering:
• Incentivizing existing nursing, primary care, and behavioral health workers to move from
areas of more provider capacity to those of less capacity in Virginia can help to achieve
an allocation of labor that maximizes overall health outcomes.
- In the case of primary care and behavioral health, some of this reallocation of labor
may be achieved by increasing opportunities for telehealth for providers, effectively
(though not physically) moving labor to the locations with the highest need.
• Removing work barriers—such as reducing documentation, reporting, and regulatory
requirement and/or leveraging technologies such as voice recognition and AI to help with
these tasks—can both create efficiency and potentially increase worker job satisfaction to
reduce attrition.
• One strategy that can help increase structural efficiency in both primary care and
behavioral health is to increase locations in the Commonwealth that offer both services in
the same setting—implementing a team-based approach to more efficiently address
related mental health illnesses.
63
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
• The Commonwealth should consider offering loan forgiveness for all individuals who go
into primary care and not just those that practice in federally qualified health centers and
HPSAs.
Furthermore, regular health care workforce data-tracking may be an effective strategy to
inform interventions for retention, recruitment, and structural efficiency, and to pivot as needed
to ensure that the health care needs of Virginia residents are met. Health care workforce data
exist in Virginia Commonwealth University’s ACORN, in the Virginia Longitudinal Data
System, and at the George Mason University Center for Health Workforce. Similar data-tracking
should be implemented by health professions colleges and universities to inform strategies to
retain graduates in the Commonwealth (Mason Center for Health Workforce, 2023).
Also, although there are promising programs in Virginia focused on health care workforce
retention and recruitment, there is currently no mechanism for tracking and disseminating these
programs. A clearinghouse for tracking and sharing current practices and programs can facilitate
dissemination of promising interventions. The newly established Virginia Department of
Workforce Development and Advancement will play a key role in supporting workforce data
infrastructure across different industries. In close collaboration with this department, VHWDA
could take on such role for health care workforce data.
Transitioning service members (as well as military spouses)—many of whom are qualified
health care experts—into civilian employment and anchoring them in Virginia is a strategy that
may help retain talent in the Commonwealth and expand the health care workforce.
Health Care Workforce Retention and Recruitment in Rural Settings
Many of the interventions we identified apply to urban, suburban, and rural settings in
Virginia. However, there are some unique considerations for health care workforce development
in Virginia’s rural communities.
• One of the most important factors for retention and recruitment in Virginia’s rural areas is
developing infrastructure such as housing, transportation, and child care.
• Another important intervention in these settings is providing cultural awareness training
to incoming students, trainees, and providers so that they are able to better understand the
local community and integrate.
• Recruitment of local students from these communities is an effective retention strategy to
bolster rural workforces—rather than convincing providers from elsewhere to move. This
recruitment would be most effective if it is paired with providing upfront financial
support for education and life-expenses.
Interventions by Strength of Evidence and Key Stakeholders for
Implementation
Table 3.1 is a guide to interventions reflected in Tables 3.2 and 3.3 to help rank each
intervention by strength of evidence based on our analyses. Table 3.2 includes interventions that
64
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
primarily impact retention, and Table 3.3 includes interventions that primarily impact
recruitment. The key actors/stakeholders important for implementing each intervention are noted
in each table.
As mentioned above, simulations that forecast increases in workforce supply caused by wage
increases indicate that all three health care workforces would increase if real wages increase.
Additionally, retention and recruitment interventions were modeled, but because of the lack of
evidence in the available data and peer-reviewed literature on the effectiveness of most barrier
reduction and recruitment interventions, the simulation model does not support definitive
statements regarding specific interventions for barriers and retention. This underscores the
critical need for rigorous evaluation research—not only for understanding a single intervention’s
effects on the workforce in a given context but to understand that intervention’s potential effects
in other contexts given changing empirical conditions and in tandem with other interventions in
simulation/systems models. Given these features of the simulation model, we note evidence for a
specific barrier and recruitment intervention (e.g., preceptor capacity) in Tables 3.2 and 3.3
based on the environmental scan and the qualitative data analysis below, with one exception: It is
important to note that evidence from simulations for wage-specific data also support wage-
related interventions for health care workers in Tables 3.2 and 3.3.
Table 3.1. Guide to Intervention Tiers
Tier 1 interventions are strongly supported by the analyses. The interventions are supported by the
literature review (evaluation studies) and were identified multiple times across the multi-stakeholder
conference, interviews, and/or focus group data.*
Tier 2 interventions are moderately supported by the analyses. These interventions are supported by
the literature review (non-evaluation studies) and were identified multiple times across the multi-
stakeholder conference, interviews, and/or focus group data.
Tier 3 interventions have some support in the literature review or were identified in the multi-
stakeholder conference, interviews, and/or focus group data.
* Support for an intervention in the qualitative work indicates that the intervention has valence—or is deemed an
attractive option by interviewees—but it does not indicate whether the intervention will have a positive effect on
retention, recruitment, or productivity.
65
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Table 3.2. Interventions for Health Care Workforce Retention
Retention
Intervention Category
(Key Implementation
Tier Stakeholders) Specific Intervention
Tier 1 Resiliency and well-being 1. Implement programs to address mental health and well-being
(health care organizations) among health care workers—especially in behavioral health
professions
2. Address underlying causes of health care worker stress and
burnout such as inappropriate patient-provider ratios, throughput
requirements, documentation requirements, low and/or differential
compensation for care provided, and lack of basic supplies.
Tier 1 Scope of practice and practice 1. Reevaluate license restrictions for PAs and NPs to increase
agreement the appeal of primary care for these health professionals and to
(Legislature) expand primary care capacity.
Tier 2 Team-based care 1. Implement team-based care to mitigate the stresses of high
(health care organizations) patient to provider ratios and large patient panels and take patient
complexity and acuity into account when assigning patients to
providers.
Tier 2 Pay increase and other 1. Increase pay for nursing, primary care, and behavioral health
incentives providers and provide appropriate benefits—including health
(health care organizations) insurance coverage commensurate with risks involved in health
care delivery (e.g., exposure to disease, despair, and violence).
Provide assistance with transportation, child care, and housing.
Tier 2 Documentation burden 1. Evaluate technologies—such as voice recognition and AI—that
(health care organizations, may assist with reducing documentation burden.*
systems, legislature) 2. Revisit documentation requirements to identify areas of
redundant and/or unnecessary documentation that could be
removed from documentation requirements.
Tier 2 Workplace diversity 1. Develop retention strategies centered around creating more
(health care organizations, diverse workplaces for nursing, primary care, and behavioral
educational entities) health.*
Tier 2 Personal/professional 1. Invest in making professional development opportunities
development available to staff and increase recognition efforts.
(health care organizations)
Tier 3 Workplace violence 1. Evaluate legislation to increase legal consequences related to
(health care organizations, violence toward health care providers—on par with those
legislature) currently in place for assaulting prehospital providers and first
responders.*
2. Evaluate health care organization policies around violence
toward health care providers and identify strategies for more strict
enforcement of those policies.
Tier 3 Cross-sector collaboration 1. Identify and leverage opportunities for cross-sector
(health care organizations, collaboration between the health care sector, educational
educational entities) organizations, and public and private entities to increase
opportunities and funding for nursing, primary care, and
behavioral health exposure and training.
Tier 3 Continuing education 1. Evaluate mechanisms to increase continuing education
(health care organizations, opportunities and funding—especially through leveraging
educational entities) collaboration between health care organizations and educational
entities.
* Designates an innovative intervention.
66
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Table 3.3 Interventions for Health Care Workforce Recruitment
Recruitment
Intervention Category
(Key Implementation
Tier Stakeholders) Specific Intervention
Tier 1 Scholarship, loan forgiveness, 1. Increase scholarship opportunities in Virginia.
and other support 2. Raise public awareness of loan forgiveness and scholarship
(private entities, health care opportunities.
organizations, educational 3. Develop mechanisms for rural applicants and first-generation
entities) students to receive loan forgiveness upfront.*
4. Include support for housing, transportation, and food for rural
and first-generation students.
Tier 1 Wages and regulations for 1. Increase wages for nursing faculty and preceptors to levels
educators similar to engineering, business, and other faculty that receive
(private entities, health care higher wages.
organizations, educational 2. Loosen strict behavioral health preceptorship requirements,
entities, Virginia Department of such as educational degree requirements for preceptors,
Health Professions) preceptor to student ratios, and need for preceptors in community
settings.*
3. Create mechanisms to pay for behavioral health
preceptorship and licensed supervisors to remove this financial
burden from students/trainees.*
Tier 1 Work-based learn and earn 1. Offer work-based learning, paid and unpaid. Health care
(health care organizations, occupations have historically included more work-based learning
educational entities) in education and training than other occupations (e.g., clinical
rotations, clinical intern or clerkships, and residencies). Health
care stakeholders are increasingly interested in expanding these
options across different health care occupations—particularly in
nursing.
Tier 2 Recruiting youth and members 1. Develop (and/or expand on current programs) to expose K–12
of underserved communities children to health professions. Such programs should include
(health care organizations, collaboration between educational and health care entities.
educational entities) 2. Specific outreach strategies for youth from underrepresented
communities are needed. Such interventions should aim to
include both children and their parents in outreach efforts and
should, ideally, be delivered by health care professionals from
similar backgrounds.*
Tier 3 Preceptor/licensed supervisor 1. Leverage nursing and behavioral health retirees to increase
capacity pool of preceptors/licensed supervisors.*
(health care organizations, 2. Create mechanisms to compensate preceptors/licensed
educational entities) supervisors—including ability to bill for supervised visits.*
Tier 3 Outreach and public image 1. Change the image of health professions through public
(health care organizations, education campaigns. Such efforts should focus on highlighting
educational entities) the rewarding nature of nursing and behavioral health work, and
removing the stigma associated with becoming a behavioral health
professional.
Tier 3 Online instruction 1. Increase opportunities for online learning to increase access
(health care organizations, and flexibility to local and nonlocal clinical introductory level
educational entities) education.
67
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Recruitment
Intervention Category
(Key Implementation
Tier Stakeholders) Specific Intervention
Tier 3 New pathways 1. Create multiple entry ramps and educational pathways to
(health care organizations, nursing and behavioral health careers as a recruitment strategy.
educational entities) This includes developing academic programs at different levels,
such as associate and bachelor’s degrees for behavioral health, to
provide more immediate relief to the workforce shortage.
2. Offer paid internships at the high school level so students can
be exposed to health professions; match undergrads with medical
students and promote mentorship alongside internship.
Tier 3 Licensing exam and licensure 1. Significantly truncate the duration of application for licensure
(department of Health in behavioral health fields and allow students that need to retake
professions) the licensure exam to have more frequent opportunities to sit
for the examination.
Tier 3 Advanced education/training 1. Increase access to simulation-based learning in nursing.
technology 2. Increase access to training opportunities in telepsychology.
(health care organizations,
educational entities)
* Designates an innovative intervention.
68
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Chapter 4. Intervention Implementation
Implementation of the listed interventions in Chapter 3 by VHWDA will require multiple
Virginia stakeholders to be involved, including those representing health care, education, labor,
both public and private entities, and local and state government and agencies, among others.
Importantly, VHWDA will need to function as a coordinating entity bringing together a coalition
of key stakeholders and guiding Virginia in taking the steps needed to implement interventions
for retention, recruitment, and creating structural efficiency. These steps include the following:
• Identify the list of key stakeholders that need to be part of the health care workforce
coalition.
• Under the coalition, form a number of committees tasked with considering the various
interventions for implementation (for example, separate committees focused on retention,
recruitment, and structural efficiency, respectively).
• Determine the financial cost, feasibility, equity and other important characteristics of
each intervention and what funding source can and should be used to pay for them.
• Identify the key partnerships that need to be developed in order to successfully
implement the interventions.
• Identify any specific executive, legislative, agency and budgetary action that needs to be
considered as part of the implementation process for each intervention.
• Prioritize which interventions to implement (and in what order) based on input from the
above-mentioned stakeholders.
• Determine the timeline for implementation of various interventions.
• Determine what data to collect on each program as success metrics to use in
intervention/program evaluation.
Implementation Timeline
We recommend that VHWDA develop the health care workforce coalition within three
months and form the needed committees under the coalition within six months from the time this
report is published. Overall, given the urgency to retain and expand the health care workforce in
Virginia, efforts should be made to implement retention recommendations in one to three years
and the recruitment interventions in three to five years from the time the work of the coalition
begins.
Communication Tools
Part of successful implementation of any of the recommended interventions by VHWDA is
engaging the key stakeholders in the Commonwealth. A first step in engagement is effectively
communicating the need to develop the health care workforce (and coalition) in the
69
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Commonwealth by sharing findings from this report. This can be accomplished by widely
disseminating the report or pertinent sections of it and planning webinars and other interactive
sessions open to stakeholders and the public, among other efforts.
70
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Abbreviations
ACORN Ambulatory Care Outcomes Research Network
AHEC Area Heath Education Centers
AI artificial intelligence
COVID-19 coronavirus disease 2019
FTE full-time equivalent
HPSA Health Professional Shortage Area
HRSA Health Resources and Services Administration
LPN licensed practical nurse
NP nurse practitioner
PA physician assistant
RN registered nurse
SLRP state loan repayment program
VA-SLRP Virginia State Loan Repayment Program
VHWDA Virginia Health Workforce Development Authority
71
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
References
Alpert, William T., Kenneth A. Couch, and Oskar R. Harmon, ”A Randomized Assessment of
Online Learning,” American Economic Review, Vol.106, No. 5. 2016.
American Association of Colleges of Nursing, “Nursing Faculty Shortage Fact Sheet,” October
2022. As of September 22, 2023:
https://www.aacnnursing.org/news-data/fact-sheets/nursing-faculty-shortage
American Medical Association, “What Is Behavioral Health?” 2022. As of September 22, 2023:
https://www.ama-assn.org/delivering-care/public-health/what-behavioral-health
American Nurses Association, “What Is Nursing?” undated. As of August 10, 2023:
https://www.nursingworld.org/practice-policy/workforce/what-is-nursing/
Anderson, Drew M., David B. Monaghan, and Jed Richardson, Can the Promise of “Free” Raise
College Attainment? Lessons from the Milwaukee Area Technical College Promise,
Annenberg Institute at Brown University, EdWorkingPaper No. 22-614, January 2023.
Arbour, Megan W., Cynthia F. Nypaver, and Judith C. Wika, “Innovative Uses of Technology in
Online Midwifery Education,” Journal of Midwifery and Women’s Health, Vol. 60, No. 3,
May–June 2015.
Bailey, Thomas R., Shanna Smith Jaggars, and Davis Jenkins, Redesigning America’s
Community Colleges: A Clearer Path to Student Success, Harvard University Press, 2015.
Barker, Timothy Hugh, Jennifer C. Stone, Kim Sears, Miloslav Klugar, Jo Leonardi-Bee, Catalin
Tufanaru, Edoardo Aromataris, and Zachary Munn, “Revising the JBI Quantitative Critical
Appraisal Tools to Improve Their Applicability: An Overview of Methods and the
Development Process,” JBI Evidence Synthesis, Vol. 21, No. 3, March 1, 2023.
Barnes, Hilary, Michael R. Richards, Grant R. Martsolf, Sayeh S. Nikpay, and Matthew D.
McHugh, “Association Between Physician Practice Medicaid Acceptance and Employing
Nurse Practitioners and Physician Assistants: A Longitudinal Analysis,” Health Care
Management Review, Vol. 47, No. 1, January–March 2022.
Bärnighausen, Till, and David E. Bloom. “Financial Incentives for Return of Service in
Underserved Areas: A Systematic Review,” BMC Health Services Research, Vol. 9, No. 86.
2009.
Basu S., S. A. Berkowitz, R. L. Phillips, A. Bitton, B. E. Landon, and R. S. Phillips,
“Association of Primary Care Physician Supply With Population Mortality in the United
States 2005–2015,” JAMA Internal Medicine, Vol. 179, No. 4, 2019.
72
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Bennett, C. L., A. F. Sullivan, A. A. Ginde, J. Rogers, J. A. Espinola, C. E. Clay, and C. A.
Camargo Jr., “National Study of the Emergency Physician Workforce,” Annals of Emergency
Medicine, Vol. 76, No. 6, December 2020.
Bennett, Christoper L., Janice A. Espinola, Ashley F. Sullivan, Krislyn M. Boggs, Carson E.
Clay, Moon O. Lee, Margaret E. Samuels-Kalow, and Carlos A. Camargo, Jr., “Evaluation of
the 2020 Pediatric Emergency Physician Workforce in the US,” JAMA Network Open, Vol.
4, No. 5, 2021.
BillTrack50, homepage, undated. As of September 15, 2023:
https://www.billtrack50.com/
BillTrack50, “VA HB1817,” webpage, 2021. As of August 15, 2023:
https://www.billtrack50.com/BillDetail/1307619
Black, Sandra E., Jeffrey T. Denning, and Jesse Rothstein, “Winners and Losers? The Effect of
Gaining and Losing Access to Selective Colleges on Education and Labor Market
Outcomes,” American Economic Journal: Applied Economics, Vol. 15, No. 1, 2023.
Bon Secours Mercy Health, “Explore Guild, Bon Secours Mercy Health’s Education Benefit,”
webpage, undated. As of August 15, 2023:
https://bsmh.guildeducation.com/partner/
Boyce, Delaney J., Megan M. Shifrin, Shelley R. Moses, and Colleen R. Moss, “Perceptions of
Motivating Factors and Barriers to Precepting,” Journal of the American Association of
Nurse Practitioners, Vol. 34, No. 11, 2022.
Boydstun, Jamie, and Jeralynn S. Cossman, “Career Expectancy of Physicians Active in Patient
Care: Evidence from Mississippi,” Rural Remote Health, Vol. 16, No. 2, April–June 2016.
Bowen, William G., Matthew M. Chingos, Kelly A. Lack, and Thomas I. Nygren, “Interactive
Learning Online at Public Universities: Evidence from a Six-Campus Randomized Trial,”
Journal of Policy Analysis and Management, Vol. 33, No. 1, 2014.
Brennan, Julie, and Angele McGrady, “Designing and Implementing a Resiliency Program for
Family Medicine Residents,” International Journal of Psychiatry in Medicine, Vol. 50, No.
1, 2015.
Brook, Judy, Leanne Aitken, Rebecca Webb, Julie MacLaren, and Debra Salmon,
“Characteristics of Successful Interventions to Reduce Turnover and Increase Retention of
Early Career Nurses: A Systematic Review,” International Journal of Nursing Studies, Vol.
91, March 2019.
Burland, Elizabeth, Susan Dynarski, Katherine Michelmore, Stephanie Owen, and Shwetha
Raghuraman, The Power of Certainty: Experimental Evidence on the Effective Design of
Free Tuition Programs, National Bureau of Economic Research, 2022.
73
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Burt, Leah, Kathleen Sparbel, and Susan Corbridge, “Nurse Practitioner Preceptor Resource
Needs and Perceptions of Institutional Support,” Journal of the American Association of
Nurse Practitioners, Vol. 34, No. 2, June 15, 2021.
Capers, Quinn, and David P. Way, “Academic and Post-Graduate Performance of African
American Medical Students by Category of Premed Institution: Historically Black vs.
Predominantly White Institutions,” Journal of Health Care for the Poor and Underserved,
Vol. 26, No. 3, August 2015.
Carilion Clinic, “Nursing at Carilion,” webpage, undated. As of August 10, 2023:
https://www.carilionclinic.org/nursing
Catherine, Sylvain, and Constantine Yannelis, “The Distributional Effects of Student Loan
Forgiveness,” Journal of Financial Economics, Vol. 147, No. 2, 2023.
Cawley, James F., Steven Lane, Noel Smith, and Elizabeth Bush, “Physician Assistants in Rural
Communities,” JAAPA, Vol. 29, No. 1, January 2016.
Center for Medicare and Medicaid Services, “CMS Funding 1,000 New Residency Slots for
Hospitals Serving Rural & Underserved Communities,” press release, December 17, 2021.
Claude Moore Charitable Foundation, “Claude Moore Scholars,” webpage, undated. As of
August 15, 2023:
https://claudemoorefoundation.org/claudemoore-scholars/
Clay, Moon O. Lee, Margaret E. Samuels-Kalow, and Carlos A. Camargo, Jr., “Evaluation of the
2020 Pediatric Emergency Physician Workforce in the US,” JAMA Network Open, Vol. 4,
No. 5, 2021.
Columbia University Irving Medical Center, “Systems Science,” webpage, undated. As of
August 25, 2023:
https://www.publichealth.columbia.edu/research/population-health-methods/systems-science
Commonwealth of Virginia, “Budget Amendments: HB1400 (Floor Approved),” webpage, 2023.
As of August 25, 2023:
https://budget.lis.virginia.gov/amendment/2023/1/HB1400/Introduced/FA/287/1h/
Contratto, Erin, Katherine Romp, Carlos A. Estrada, April Agne, and Lisa L. Willett, “Physician
Order Entry Clerical Support Improves Physician Satisfaction and Productivity,” Southern
Medical Journal, Vol. 110, No. 5, May 2017.
Coombs, B., “With Travel Nurses Making $150 an Hour, Hospital Systems Are Forced to
Innovate,” CNBC, Health and Science, March 28, 2023. As of August 31, 2023:
https://www.cnbc.com/2023/03/28/with-travel-nurses-making-150-an-hour-hospital-systems-
innovate.html
74
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Couser, Greg, Sherry Chesak, and Susanne Cutshall. “Developing a Course to Promote Self-Care
for Nurses to Address Burnout,” OJIN: The Online Journal of Issues in Nursing, Vol. 25, No.
3, 2020.
Cover Virginia, “Return to Normal Enrollment,” webpage, Virginia Department of Medical
Assistance Services, undated. As of August 10, 2023:
https://coverva.dmas.virginia.gov/covid-19/
Dougherty, Shaun, and Walter G. Ecton, “The Economic Effect of Vocational Education on
Student Outcomes,” Oxford Research Encyclopedia of Economics and Finance, 2021.
Doximity, 2020 Physician Compensation Report: Fourth Annual Study, 2020.
Doximity, 2023 Physician Compensation Report. Physicians Thrive, 2023.
Duquesne University School of Nursing, “The Shortage of Healthcare Workers in the U.S.,”
webpage, undated. As of August 25, 2023:
https://onlinenursing.duq.edu/post-master-certificates/shortage-of-healthcare-workers/
Dussault, Gilles, and Carl-Ardy Dubois, “Human Resources for Health Policies: A Critical
Component In Health Policies,” Human Resources for Health, Vol. 1, No. 1, April 14, 2003.
Dynarski, Susan, C. J. Libassi, Katherine Michelmore, and Stephanie Owen, “Closing the Gap:
The Effect of Reducing Complexity and Uncertainty in College Pricing on the Choices of
Low-Income Students,” American Economic Review, Vol. 111, No. 6, 2021.
Dyrbye, Liselotte N., Tait D. Shanafelt, Priscilla R. Gill, Daniel V. Satele, and Colin P. West,
“Effect of a Professional Coaching Intervention on the Well-Being and Distress of
Physicians: A Pilot Randomized Clinical Trial,” JAMA Internal Medicine, Vol. 179, No. 10,
October 1, 2019.
Eden, Aimee R., Benjamin J. Chesluk, Elizabeth Rose Hansen, Audrey Brock, Elizabeth C.
Bernabeo, and Lars E. Peterson, “The Role of Gender in the Experience and Impact of
Recertification Exam Preparation: A Qualitative Study of Primary Care Physicians,” Journal
of Women’s Health (Larchmont), Vol. 29, No. 11, November 2020.
Eden, Aimee R., Yalda Jabbarpour, Zachary J. Morgan, Mingliang Dai, Megan Coffman, and
Andrew Bazemore, “Gender Differences in Personal and Organizational Mechanisms to
Address Burnout Among Family Physicians,” Journal of the American Board of Family
Medicine, Vol. 33, No. 3, May–June 2020.
Edwards, Deborah, Clare Hawker, Judith Carrier, and Colin Rees, “A Systematic Review of the
Effectiveness of Strategies and Interventions to Improve the Transition from Student to
Newly Qualified Nurse,” International Journal of Nursing Studies, Vol. 52, No. 7, July 2015.
75
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Emanuel, E. J., “The Inevitable Reimagining of Medical Education,” JAMA, Vol. 323, No. 12,
March 24, 2020.
Fast, Austin, “Nurses Are Waiting Months for Licenses as Hospital Staffing Shortages Spread,”
NPR, March 10, 2022.
Finneran, Janet M., and Judy M. Kreye, “Empowering the Novice Family Nurse Practitioner:
Promoting Readiness to Practice Through Immersion in an Innovative Educational Strategy,”
Journal of the American Association of Nurse Practitioners, Vol. 33, No. 5, November 21,
2019.
Gamble, Molly, “Early Retirements Could Be Drastic in Healthcare,” Becker’s Hospital Review,
April 30, 2021.
Garrett, Cameryn C., Ronda L. Doonan, Casey Pyle, and Michelle B. Azimov, “Student Loan
Debt and Financial Education: A Qualitative Analysis of Resident Perceptions and
Implications for Resident Well-Being,” Medical Education Online, Vol. 27, No. 1, December
2022.
Goldberg, Lynette R., and Leonard A. Crocombe, “Advances in Medical Education and Practice:
Role of Massive Open Online Courses,” Advances in Medical Education and Practice, Vol. 8
August 21, 2017.
Goodwin, Miki, Marianne Fingerhood, Eric Slade, and Patricia Davidson, “Development of an
Innovative Curriculum-to-Career Transition Program for Nurse Practitioners in Primary
Care,” Nursing Outlook, Vol. 69, No. 3, May–June 2021.
Goolsbee, Austan, Glenn Hubbard, and Amy Ganz, A Policy Agenda to Develop Human Capital
for the Modern Economy, Aspen Institute, February 2019.
Guan, Ting, Brad Zebrack, Shirley Otis-Green, and Grace DesJardins, “Salary and Student Loan
Debt for Oncology Social Workers: Findings from the Oncology Social Work Competencies,
Opportunities, Roles and Expertise (CORE) Survey,” Journal of Psychosocial Oncology,
Vol. 41, No. 2, 2023.
Harvard Catalyst, “Mixed Methods Research,” webpage, undated. Accessed August 15, 2023:
https://catalyst.harvard.edu/community-engagement/mmr/
Health Resources and Services Administration, “Health Workforce Shortage Areas,” webpage,
U.S. Department of Health and Human Services, August 7, 2023. As of June 19, 2023:
https://data.hrsa.gov/topics/health-workforce/shortage-areas
Henriksen, Christian, “Machine Learning in Radiology: Is AI Coming for My (Future) Job?”
American College of Radiology, November 28, 2022.
76
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Hermansen, Mikkel, “Occupational Licensing and Job Mobility in the United States,”
Organisation for Economic Co-operation and Development, December 12, 2019.
Hollowell, Ashleigh, “Tennessee Law Lets International Medical Grads Skip US Residency,”
Becker’s Hospital Review, June 27, 2023.
Homer, Jack B., and Gary B. Hirsch, “System Dynamics Modeling for Public Health:
Background and Opportunities,” American Journal of Public Health, Vol. 96, No. 3, 2006.
Hone, Thomas, Timothy Powell-Jackson, Leonor Maria Pacheco Santos, Ricardo de Sousa
Soares, Felipe Proenço de Oliveira, Mauro Niskier Sanchez, Matthew Harris, Felipe de
Oliveira de Souza Santos, and Christopher Millett, “Impact of the Programa Mais Médicos
(More Doctors Programme) on Primary Care Doctor Supply and Amenable Mortality: Quasi-
Experimental Study of 5565 Brazilian Municipalities,” BMC Health Services Research, Vol.
20, No. 873, September 2020.
Hooker, Roderick S., and Ashley N. Muchow, “Modifying State Laws for Nurse Practitioners
and Physician Assistants Can Reduce Cost Of Medical Services,” Nurse Economic$, Vol. 33,
No. 2, March–April 2015.
Hu, Xiaochu, Michael J. Dill, and Sarah S. Conrad, “What Moves Physicians to Work in Rural
Areas? An In-Depth Examination of Physician Practice Location Decisions,” Economic
Development Quarterly, Vol. 36, No. 3, 2022.
Huffstetler, Alison N., Roy T. Sabo, Martin Lavallee, Ben Webel, Paulette Lail Kashiri,
Jacquelyn Britz, Mark Carrozza, Michael Topmiller, Elizabeth R. Wolf, Beth A. Bortz,
Ashley M. Edwards, and Alex H. Kris, “Using State All-Payer Claims Data to Identify the
Active Primary Care Workforce: A Novel Study in Virginia,” Annals of Family Medicine,
Vol. 20, No. 5, 2022.
Institute of Education Sciences, homepage, undated. As of August 10, 2023:
https://ies.ed.gov/ncee/wwc/
Institute of Medicine, Finding What Works in Health Care: Standards for Systematic Reviews,
National Academies Press, 2011.
Jacob, Brian A., “What We Know About Career and Technical Education in High School,”
Brookings Institute, October 5, 2017.
Jithitikulchai, T., “Improving Allocative Efficiency from: Network Consolidation A Solution for
the Health Workforce Shortage,” Human Resources for Health, Vol. 20, No. 1, July 15,
2022.
77
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Johnson, Emily K., Matthew A. Wojtesta, Sawyer W. Crosby, Herbert C. Duber, Eunice Jun,
Haley Lescinsky, Phong Nguyen, Maitreyi Sahu, Azalea Thomson, Golsum Tsakalos,
Maxwell S. Weil, Annie Haakenstad, Ali H. Mokdad, Christopher J. L. Murray, and Joseph
L. Dielema, “Varied Health Spending Growth Across US States Was Associated with
Incomes, Price Levels, and Medicaid Expansion, 2000–19: Study Examines Factors
Associated with Health Spending Growth Across the US,” Health Affairs, Vol. 41, No. 8,
2022.
Joyce, Ted, Sean Crockett, David A. Jaeger, Onur Altindag, and Scott D. O’Connell, “Does
Classroom Time Matter?” Economics of Education Review, Vol. 46, 2015.
Kaiser Family Foundation, “Medicaid Enrollment and Unwinding Tracker,” webpage, undated-a.
As of August 10, 2023:
https://www.kff.org/medicaid/issue-brief/medicaid-enrollment-and-unwinding-tracker/
Kaiser Family Foundation, “Mental Health Care Health Professional Shortage Areas (HPSAs),”
webpage, undated-b. As of October 21, 2022:
https://www.kff.org/other/state-indicator/mental-health-care-health-professional-shortage-
areas-hpsas
Kane, Leslie, “Medscape Physician Compensation Report 2023: Your Income vs Your Peers’,”
Medscape, April 14, 2023.
Kapu, April, and Courtney J. Pitts, “Nurse Practitioners Are Stepping Up to Meet Access to
Primary Care Services Across the United States: A Time for Policy Change,” Journal of
Nursing Administration, Vol. 53, No. 4, April 1, 2023.
Kemple, J. J., W. Corrin, E. Nelson, T. Salinger, S. Herrmann, and K. Drummond, The
Enhanced Reading Opportunities Study: Early Impact and Implementation Findings,
Institute of Education Sciences, National Center for Education Evaluation and Regional
Assistance, NCEE Report No. 2008-4015, 2008.
Knopf-Amelung, S., H. Gotham, A. Kuofie, P. Young, R. Manney Stinson, J. Lynn, K. Barker,
and J. Hildreth, “Comparison of Instructional Methods for Screening, Brief Intervention, and
Referral to Treatment for Substance Use in Nursing Education,” Nurse Educator, Vol. 43,
No. 3, May/June 2018.
Kueter, C. P., “For Nurse Practitioners Seeking Autonomy, State Application Deadline Looms,”
UVA Today, June 21, 2022.
Kusterbeck, S, “Travel Nursing Demand Is Waning: Now Hospitals Want Full-Time Nurses
Back,” Nurse.com, November 8, 2022.
Landi, Heather, “Demand for Behavioral Health Services Outstrips Supply of Providers, Driving
Higher Costs, Analysis Finds,” Fierce Healthcare, June 13, 2023.
78
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Lee, Briana E., Brooke Zlotshewer, Rebecca Cruz Mayeda, and Lawrence I. Kaplan, “Impact of
Online-Only Instruction on Preclinical Medical Education in the Setting of COVID-19:
Comparative Analysis of Online-Only Vs. Hybrid Instructions on Academic Performance
and Mental Wellbeing,” Medical Science Educator, Vol. 32, 2022.
Lin, T. K., K. Werner, M. Kak, and C. H. Herbst, “Health-Care Worker Retention in Post-
Conflict Settings: A Systematic Literature Review,” Health Policy and Planning, Vol. 38,
No. 1, 2023.
Lo Sasso, A. T., D. Armstrong, G. Forte, and S. E. Gerber, “Differences in Starting Pay for Male
and Female Physicians Persist; Explanations for the Gender Gap Remain Elusive: An
Examination of How the Gender Gap in Total Starting Pay Has Evolved and the Extent to
Which Preferences over Work-Life Balance Factors Affect the Gap,” Health Affairs, Vol. 39,
No. 2, 2020.
Love, Ivy, and Mary Alice McCarthy, Apprenticeship and the Future of Nursing: An Equity-
Based Strategy to Address the Bachelor’s Degree Gap, New America, September 2018.
Lucas, L., “Rising Cost of Traveling Nurses Having Ripple Effect on Area Hospitals,” 11alive,
February 32, 2022.
Margolis, Marilyn, Carolyn Clancy, Rose Hayes, Eileen Sullivan-Marx, James G. Wetrich, and
Marion Broome, “How Academia Can Help to Grow—and Sustain—a Robust Nursing
Workforce,” Nursing Outlook, 2023.
Mason Center for Health Workforce, homepage, undated. As of August 14, 2023:
https://vahlthwf.gmu.edu/mason-center-for-health-workforce-home-page-1/
Masters, Kate, “Virginia Lawmakers Boosted Medicaid Payments for Primary Care. Providers
Say It’s Not Enough,” Virginia Mercury, July 8, 2022.
McCord, Carly E., Jeremy J. Saenz, Trey W. Armstrong, and Timothy R. Elliott, “Training the
Next Generation of Counseling Psychologists in the Practice of Telepsychology,”
Counselling Psychology Quarterly, Vol. 28, No. 3, 2015.
McCutcheon, Karen, Peter O’Halloran, and Maria Lohan, “Online Learning Versus Blended
Learning of Clinical Supervisee Skills with Pre-Registration Nursing Students: A
Randomised Controlled Trial,” International Journal of Nursing Studies, Vol. 82, June 2018.
McGhee, Moira K., “2023 State by State Scope of Practice: Nurse Practitioner,” Vivian Health,
August 1, 2022.
MedServe, homepage, undated. As of August 17, 2023.
https://www.med-serve.org/
79
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Mensik, Hailey, “Lawmakers Stress Urgency of Healthcare Worker Shortage,” HealthCareDive,
February 16, 2023.
Meyer, Emily M., Susan Zapatka, and Rebecca S. Brienza, “The Development of Professional
Identity and the Formation of Teams in the Veterans Affairs Connecticut Healthcare
System’s Center of Excellence in Primary Care Education Program (CoEPCE),” Academic
Medicine, Vol. 90, No. 6, June 2015.
Mujumdar, V., and H. Jeffcoat, “How Clinical Decision Support Tools Can Be Used to Support
Modern Care Delivery,” American College of Surgeons, September 1, 2022.
National Academy of Medicine, “A Population Health Workforce to Meet 21st Century
Challenges and Opportunities: A Workshop,” webpage, undated. As of August 25, 2023:
https://www.nationalacademies.org/our-work/a-population-health-workforce-to-meet-21st-
century-challenges-and-opportunities-a-workshop
National Academy of Medicine, Taking Action Against Clinician Burnout: A Systems Approach
to Professional Well-Being, National Academies Press, 2019.
National Academy of Medicine, National Plan for Health Workforce Well-Being, National
Academies Press, 2022.
National Center for Health Workforce Analysis, Nurse Workforce Projections, 2020–2035, U.S.
Department of Health and Human Services, Health Resources and Services Administration,
Bureau of Health Workforce, November 2022.
National Conference of State Legislatures, “Database Health Costs, Coverage and Delivery State
Legislation,” webpage, August 11, 2023.
Nevada Rural Hospital Partners, “Nevada Nurse Apprentice Program,” webpage, undated.
Nguyen, Anh T., Kate E. Trout, Li-Wu Chen, Lynda Madison, Katherine L. Watkins, and
Shinobu Watanabe-Galloway, “Nebraska’s Rural Behavioral Healthcare Workforce
distribution and Relationship Between Supply and County Characteristics,” Rural Remote
Health, Vol. 16, No. 2, April–June 2016.
Nguyen, Ann M., Magda Schaler-Haynes, Jolene Chou, Molly Nowels, Danielle H. Llaneza, and
Elissa Kozlov, “Increasing Access to a Diverse Mental Health Workforce Through
Emergency Reciprocity Licensure,” Journal of Medical Regulation, Vol. 109, No. 1, 2023.
Norcross, William A., Christine Moutier, Maria Tiamson-Kassab, Pam Jong, Judy E. Davidson,
Kelly C. Lee, Isabel G. Newton, Nancy S. Downs, and Sid Zisook, “Update on the UC San
Diego Healer Education Assessment and Referral (HEAR) Program,” Journal of Medical
Regulation, Vol. 104, No. 2, 2018.
80
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Norful, Allison Andreno, Kyleen Swords, Mickaela Marichal, Hwayoung Cho, and Lusine
Poghosyan, “Nurse Practitioner-Physician Comanagement of Primary Care Patients: The
Promise of a New Delivery Care Model to Improve Quality of Care,” Health Care
Management Review, Vol. 44, No. 3, July–September 2019.
Novella, Rafael, and Yocelyn Samantha Pérez-Dávila, “Are Apprenticeships Programs
Effective? Lessons for Latin America and the Caribbean,” Inter-American Development
Bank, Technical Note No. IDB-TN-1319, 2017.
Nursing Process, “9 Best Nurse Practitioner Programs in Virginia—2023,” webpage, 2023. As of
August 17, 2023:
https://www.nursingprocess.org/nurse-practitioner-schools/virginia/
Orr, Robert, “U.S. Health Care Licensing: Pervasive, Expensive, and Restrictive,” Niskanen
Center, May 8, 2023.
Page, Lindsay C., and Judith Scott-Clayton, Improving College Access in the United States:
Barriers and Policy Responses, National Bureau of Economic Research, December 2015.
Perron, Tracy, Mackenzie Gascoyne, Teena K. Kallakavumkal, Meghan Kelly, and Nina
Demagistris, “Effectiveness of Nurse Residency Programs,” Journal of Nursing Practice
Applications and Reviews of Research, accepted for publication on June 2, 2019.
Poghosyan, Lusine, Allison A. Norful, and Miriam J. Laugesen, “Removing Restrictions on
Nurse Practitioners’ Scope of Practice in New York State: Physicians’ and Nurse
Practitioners’ Perspectives,” Journal of the American Association of Nurse Practitioners,
Vol. 30, No. 6, June 2018.
Rains, Jacob, George Mark Holmes, Shweta Pathak, and Emily M. Hawes, “The Distribution of
Additional Residency Slots to Rural and Underserved Areas,” JAMA, Vol. 330, No. 10,
August 9, 2023.
Reed, Debbie, Albert Yung-Hsu Liu, Rebecca Kleinman, Annalisa Mastri, Davin Reed, Samina
Sattar, and Jessica Ziegle, An Effectiveness Assessment and Cost-Benefit Analysis of
Registered Apprenticeship in 10 States, Mathematica Policy Research, July 25, 2012.
Reis, Pamela J., Karl Faser, and Marquietta Davis, “A Framework for Web-Based
Interprofessional Education for Midwifery and Medical Students,” Journal of Midwifery and
Women’s Health, Vol. 60, No. 6, 2015.
Rosen, Rachel, Mary Visher, and Katie Beal, Career and Technical Education: Current Policy,
Prominent Programs, and Evidence, MDRC, 2018.
Rural Health Information Hub, “Rural J-1 Visa Waiver,” webpage, March 3, 2022. As of August
25, 2023:
https://www.ruralhealthinfo.org/topics/j-1-visa-waiver
81
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Rutter, Harry, Natalie Savona, Ketevan Glonti, Jo Bibby, Steven Cummins, Diane T. Finegood,
Felix Greaves, Laura Harper, Penelope Hawe, Laurence Moore, Mark Petticrew, Eva
Rehfuess, Alan Shiell, James Thomas, and Martin White, “The Need for a Complex Systems
Model of Evidence for Public Health,” Lancet, Vol. 390, No. 10112, December 9, 2017.
Ruyak, Sharon L., Laura Migliaccio, Amy Levi, and Sarah Patel, “Role Development in
Midwifery Education: A Place for Simulation,” Midwifery, Vol. 59, April 2018.
Sawhney, Henna, “Licensed Without a Residency,” Inside the Match, April 28, 2022.
Sermo, “Employee Retention in Healthcare,” March 29, 2023.
Shi, Leiyu, “The Impact of Primary Care: A Focused Review,” Scientifica (Cairo), December 31,
2012.
Song, Z., and S. Gondi, “Will Increasing Primary Care Spending Alone Save Money?” JAMA,
Vol 322, No.14, October 8, 2019.
Spiers, M. C., and M. Harris, “Challenges to Student Transition in Allied Health Undergraduate
Education in the Australian Rural and Remote Context: A Synthesis of Barriers and
Enablers,” Rural Remote Health, Vol. 15, No. 2, April–June 2015.
Starfield, Barbara, Leiyu Shi, and James Macinko, “Contribution of Primary Care to Health
Systems and Health,” Milbank Quarterly, Vol. 83, No. 3, 2005.
Terry, Victoria R., Peter C. Terry, Clint Moloney, and Les Bowtell, “Face-to-Face Instruction
Combined with Online Resources Improves Retention of Clinical Skills Among
Undergraduate Nursing Students,” Nurse Education Today, Vol. 61, February 2018.
Thompson, Matthew F., “Subdoctoral Practitioners in Psychology: Implications for Master’s-
Level Psychologists,” Practice Innovations, Vol. 5, No. 2, 2020.
Tovar, Molly, Kellie Thompson, and Shanondora Billiot, “Training American Indian and Alaska
Native Social Workers for Indian Country,” Journal of Evidence-Based Social Work (2019),
Vol. 17, No. 1, January–February 2020.
U.S. Department of Labor, Using Registered Apprenticeship to Build and Fill Healthcare Career
Paths, undated. As of August 25, 2023:
https://www.doleta.gov/oa/pdf/HIT_HealthCare_Report.pdf
van Weel, Chris, and Michael R. Kidd, “Why Strengthening Primary Health Care Is Essential to
Achieving Universal Health Coverage,” Canadian Medical Association Journal, Vol. 190,
No. 15, April 16, 2018.
Vestal, April, Fotinos Panagakos, Treah Haggerty, and Ashley Brianna Sheppard, “Longitudinal
Assessment of a Dental and Medical Student Service Program in Appalachia: Successes and
Lessons Learned,” Journal of Public Health Dentistry, Vol. 82, No. 1, January 2022.
82
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
VHWDA—See Virginia Health Workforce Development Authority.
Virginia Commonwealth University, “ACORN,” webpage, undated-a. As of August 17, 2023:
https://familymedicine.vcu.edu/research/family-medicine-research/acorn/
Virginia Commonwealth University, “International/Inner City/Rural Preceptorship (I2CRP),”
webpage, undated-b. As of August 17, 2023:
https://familymedicine.vcu.edu/education-programs/i2crp---underserved-program/
Virginia Commonwealth University, “Family Medicine Scholars Training and Admission
Program (fmSTAT),” undated-c. As of August 17, 2023:
https://familymedicine.vcu.edu/education-programs/fmstat---scholars-program
Virgina Department of Health, “Virgina State Loan Repayment Program,” webpage, undated. As
of August 17, 2023:
https://www.vdh.virginia.gov/health-equity/virginia-loan-repayment-programs-2/
Virginia Department of Housing and Community Development, “Virginia to Receive $1.4
Billion in Federal Funding for Universal Broadband Access,” PR Newswire, June 26, 2023.
Virginia Foundation for Community College Education, “Fastforward Scholarship Fund,”
webpage, undated. As of August 15, 2023:
|https://vfcce.org/fastforward/
Virginia G3, homepage, undated. As of August 15, 2023:
https://virginiag3.com/
Virginia Health Care Foundation, “Assessment of the Capacity Virginia’s Licensed Behavioral
Health Workforce,” January 2022a.
https://www.vhcf.org/wp-content/uploads/2022/01/BH-Assessment-Slides-for-VHCF-
Website-1.2022.pdf
Virginia Health Care Foundation, “New Assessment Finds Alarming Shortage of Virginia
Licensed Behavioral Health Professional and Outlines Attainable Solutions,” press release,
January 19, 2022b.
Virginia Health Workforce Development Authority, homepage, undated-a. As of September 25,
2023:
https://www.vhwda.org/
Virginia Health Workforce Development Authority, “Grants & Scholarships,” webpage,
undated-b. As of August 15, 2023:
https://www.vhwda.org/about/partnership-and-resources/grants-and-scholarships
Virginia Longitudinal Data System, homepage, undated. As of September 15, 2023:
https://vlds.virginia.gov/
83
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Virginia’s Legislative Information System, “2021 Special Session I: SB 1147 Nurse Loan
Repayment Program; Expands Eligibility for Program to Include Certified Nurse Aides,”
webpage, undated-a. As of August 15, 2023:
https://lis.virginia.gov/cgi-bin/legp604.exe?212+sum+SB1147
Virginia’s Legislative Information System, “2023 Session: Chapter 684,” webpage, undated-b.
As of August 15, 2023:
https://lis.virginia.gov/cgi-bin/legp604.exe?231+ful+CHAP0684
Virginia’s Legislative Information System, “Code of Virginia: Title 32.1. Health; Chapter 4.
Health Care Planning; Article 6. Primary Health Care System; § 32.1–122.7. Virginia Health
Workforce Development Authority; purpose,” undated-c. As of September 15, 2023:
https://law.lis.virginia.gov/vacode/title32.1/chapter4/section32.1-122.7/
Waddell, Glen R., and Larry D. Singell Jr, “Do No-Loan Policies Change the Matriculation
Patterns of Low-Income Students?” Economics of Education Review, Vol. 30, No. 2, 2011.
Wakefield, Mary, David R. Williams, and Suzanne Le Menestrel, The Future of Nursing 2020–
2030: Charting a Path to Achieve Health Equity, National Academy of Sciences, 2021.
Wang, Yulin, “Will Tesla’s Humanoid Robot ‘Optimus’ Really Be Able to Serve People?”
IDTechEx, October 4, 2022.
Warren, Jessie N., Marian Luctkar-Flude, Christina Godfrey, and Julia Lukewich, “A Systematic
Review of the Effectiveness of Simulation-Based Education on Satisfaction and Learning
Outcomes in Nurse Practitioner Programs,” Nurse Education Today, Vol. 46, November
2016.
Watkins, Daphne C., “Rapid and Rigorous Qualitative Data Analysis: The ‘RADaR’ Technique
for Applied Research,” International Journal of Qualitative Methods, Vol. 16, No. 1, 2017.
Weston, M. J., “Strategic Planning for a Very Different Nursing Workforce,” Nurse Leader, Vol.
20, No. 2, 2022.
Wiesen, K., “Nurse Practitioner Scope of Practice by State—2023,” Nursing Process, undated.
As of August 31, 2023:
https://www.nursingprocess.org/nurse-practitioner-scope-of-practice-by-state.html
Wilcha, Robyn-Jenia, “Effectiveness of Virtual Medical Teaching During the COVID-19 Crisis:
Systematic Review,” JMIR Medical Education, Vol. 6, No. 2, November 18, 2020.
Yeow, Adrian, and Kim Huat Goh, “Work Harder or Work Smarter? Information Technology
and Resource Allocation in Healthcare Processes,” MIS Quarterly, Vol. 39, No. 4, 2015.
84
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Zhang, Xiaoming, Daniel Lin, Hugh Pforsich, and Vernon W. Lin, “Physician Workforce in the
United States of America: Forecasting Nationwide Shortages,” Human Resources for Health,
Vol. 18, No. 1, 2020.
Zhao, Fei, You Fu, Qi-Jie Zhang, Yue Zhou, Peng-Fei Ge, Hua-Xing Huang, and Yuan He, “The
Comparison of Teaching Efficiency Between Massive Open Online Courses and Traditional
Courses in Medicine Education: A Systematic Review and Meta-Analysis,” Annals of
Translational Medicine, Vol. 6, No. 23, December 2018.
Zhavoronkova, Marina, Bradley D. Custer, Anona Neal, Justin Schweitzer, and Marcella
Bombardieri, “Fact Sheet: Easing the American Nursing Shortage,” Center for American
Progress, 2023.
Zheng, Jilian C., Han Zhang, Beiqing Wu, Zenghan Tong, Yingbo Zhu, Matthew S. Mitchell,
Xioating Sun, Yuhong Yang, Kan Zhang, Ly Fang, and Eli Adashi, “Medical Education
Reform in China: The Shanghai Medical Training Model,” Journal of Graduate Medical
Education, Vol. 12, No. 6, December 2020.
85
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Annex
Identifying Strategies for Strengthening the
Health Care Workforce in the
Commonwealth of Virginia
Annex
Megan Andrew, Brian Briscombe, Raffaele Vardavas, Nazia Wolters, Nabeel Qureshi,
Wilson Nham, Mahshid Abir
RAND Health Care
RR-A2903-1
October 2023
Prepared for the Virginia Health Workforce Development Authority
This document has not been formally reviewed, edited, or finalized for public release. It should not be cited without the permission of
the RAND Corporation. RAND’s publications do not necessarily reflect the opinions of its research clients and sponsors. is
a registered trademark.
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
ii
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
About This Report
This document contains appendixes to the main report, Identifying Strategies for
Strengthening the Health Care Workforce in the Commonwealth of Virginia, available
www.rand.org/t/RRA2093-1. These appendixes provide details on the methods and analysis used
for the research summarized in the main report. The citations in this document point to
references listed in the main report.
This research was funded by the Virginia Health Workforce Development Authority
(VHWDA) and carried out within the Access and Delivery Program in RAND Health Care. The
Virginia General Assembly established VHWDA in 2010 in order to identify and address health
workforce issues in the Commonwealth. As a public entity, VHWDA exercises public and
essential governmental functions to secure the health, welfare, convenience, knowledge, benefit,
and prosperity of Virginians. VHWDA’s mission is to “facilitate the development of a statewide
health professions pipeline that identifies, educates, recruits, and retains a diverse, appropriately
geographically distributed, and culturally competent quality workforce.” VHWDA accomplishes
this through core functions outlined in the Code of Virginia (Virginia’s Legislative Information
System, undated-c).
RAND Health Care, a division of the RAND Corporation, promotes healthier societies by
improving health care systems in the United States and other countries. We do this by providing
health care decisionmakers, practitioners, and consumers with actionable, rigorous, objective
evidence to support their most complex decisions. For more information, see
www.rand.org/health-care, or contact
RAND Health Care Communications
1776 Main Street P.O. Box 2138
Santa Monica, CA 90407-2138
(310) 393-0411, ext. 7775
RAND_Health-Care@rand.org
iii
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Contents
About This Report.......................................................................................................................... iii
Figures and Tables .......................................................................................................................... v
Appendix A. Resources on Virginia’s Health Care Workforce ...................................................... 1
Data Related to Health Care Education..................................................................................................... 1
Data Related to Workforce Personnel ....................................................................................................... 1
Recruitment and Retention ........................................................................................................................ 2
Appendix B. Environmental Scan Methods.................................................................................... 3
Methods ..................................................................................................................................................... 3
Appendix C. Interview and Focus Group Guide ............................................................................ 8
Appendix D. Hospital Survey Questions ...................................................................................... 10
Appendix E. The Mathematical Formulation of our System Dynamics Simulation Model ......... 11
Introduction ............................................................................................................................................. 11
The System Dynamic Model and its Differential Equations................................................................... 12
Modeling the rate of change of the qualified pool .................................................................................. 15
Wage Dynamics ...................................................................................................................................... 16
Health Care Work Reluctance ................................................................................................................. 17
The Experimental Design for Model Calibration .................................................................................... 26
Sensitivity Analysis ................................................................................................................................. 31
Limitations and Extensions ..................................................................................................................... 31
Appendix F. System Dynamics Model Results for Registered Nurses ........................................ 33
Model Experimental Design.................................................................................................................... 33
Calibration Phase Trajectories ................................................................................................................ 33
Sensitivity Analysis ................................................................................................................................. 35
Policy Experiments ................................................................................................................................. 36
Appendix G. System Dynamics Model Results on Primary Care Providers ................................ 40
Model Experimental Design.................................................................................................................... 40
Calibration Phase Trajectories ................................................................................................................ 40
Policy Experiments ................................................................................................................................. 42
Appendix H. System Dynamics Model Results on Behavioral Health Specialists ...................... 46
Model Experimental Design.................................................................................................................... 46
Calibration Phase Trajectories ................................................................................................................ 46
Policy Experiments ................................................................................................................................. 47
Abbreviations ................................................................................................................................ 51
References ..................................................................................................................................... 52
iv
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figures, Tables, and Boxes
Figures
Figure E.1. Impact of Health Care Work Reluctance on Nurse Supply and Real Wages Illustrated
Through Sigmoidal Curves ................................................................................................... 14
Figure F.1 Full Calibration Phase Trajectories for Registered Nurses ......................................... 34
Figure F.2. Final Calibrated Model Trajectories for Registered Nurses....................................... 35
Figure F.3. Comparative Trajectories of Policy Effects for Registered Nurses ........................... 37
Figure F.4. Map of Virginia and How it is Divided into Eight Regions of Analysis ................... 38
Figure F.5. Results of Policy Interventions on Employment by AHEC Region for Registered
Nurses ................................................................................................................................... 39
Figure F.6. Results of Policy Interventions on the Gap Between Supply and Demand by AHEC
Region for Registered Nurses ............................................................................................... 39
Figure G.1. Final Calibrated Model Trajectories for Primary Care Providers ............................. 41
Figure G.2. Comparative Trajectories of Policy Effects for Primary Care Providers .................. 43
Figure G.3. Results of Policy Interventions on Employment by AHEC Region for Primary Care
Providers ............................................................................................................................... 44
Figure G.4. Results of Policy Interventions on the Gap Between Supply and Demand by AHEC
Region for Primary Care Providers....................................................................................... 45
Figure H.1. Final Calibrated Model Trajectories for Health Behavioral Specialists .................... 47
Figure H.2. Comparative Trajectories of Policy Effects for Behavioral Health Specialists......... 48
Figure H.3. Results of Policy Interventions on Employment by AHEC Region for Behavioral
Health Specialists .................................................................................................................. 49
Figure H.4. Results of Policy Interventions on the Gap Between Supply and Demand by AHEC
Region for Behavioral Specialists......................................................................................... 50
Tables
Table B.1. Environmental Scan Topics and Research Questions ................................................... 4
Table B.2. Key Concepts and Related Search Terms ..................................................................... 6
Table E.1. Trajectories of Output Variable ................................................................................... 26
Table E.2. Input Model Parameters, their Values, Ranges, and Sampling Methods .................... 27
v
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Appendix A. Resources on Virginia’s Health Care Workforce
In addition to other sources of information, this research relied on various data collection and
aggregation efforts to glean insight into the national, state, and regional health care workforces.
We used some of the data resources listed here to inform our model of nursing and physician
workforce supply and demand (see Appendix E for more on that model). We used other data
resources listed here to provide context and background for understanding Virginia’s health care
workforce challenges and interventions designed to overcome those challenges.
Data Related to Health Care Education
We reviewed data from the State Council of Higher Education for Virginia (SCHEV) and the
National Student Clearinghouse to better understand the pipeline of graduates in health care
fields. The SCHEV data can be accessed in the Virginia Longitudinal Data System, and we thank
the Mason Center for Health Workforce for helping us to access and process these files.
Mason also collects education data directly from schools or programs, including data related
to the career and technical education pipeline from high schools and college preparatory schools,
by defined types of academic programs. We did not review these data, but we mention them to
raise awareness that data exist to facilitate a more detailed and robust study of the health care
education pipeline.
Data Related to Workforce Personnel
Various publicly available datasets attempt to measure the past and present supply and
demand for nursing, primary care, behavioral health, and other health care workforces. We have
relied primarily on the U.S. Bureau of Labor Statistics’ (BLS’s) Current Employment Statistics
State and Local Areas to quantify the number of personnel and average wages for each job
category in each Virginia locality.
Additionally, a pre-pandemic study conducted by the U.S. Department of Health and Human
Services projected demand for registered nurses (RNs) in the United States (U.S. Department of
Health and Human Services et al., 2017), which we used as a starting point for projecting
demand for nursing labor by observing the ratio of nationwide bed occupancy to national
demand, then using the same ratio for Virginia-specific bed occupancy to scale down to the state
level. Although this technique only roughly approximates demand for nurses in Virginia, the
main purpose of our model was to forecast the relative change in demand and supply (from any
given starting point), so the starting point need not be precise. Furthermore, quantifying demand
is by nature a highly imprecise endeavor, given the fact that there exists a numeric difference
between minimum number of personnel to operate a hospital, desired number of personnel,
1
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
optimal number of personnel from a financial perspective, and optimal number of personnel from
the perspective of maximizing health outcomes.
The Virginia Employment Commission (VEC) also has data on health care workforce supply
and demand, which the Virginia Hospital and Healthcare Association supplied to us through
2019, and we supplemented these data with additional VEC data through 2021 from
https://virginiaworks.com/download-center.
Recruitment and Retention
Preexisting survey data helped us to better understand workers’ motivations for joining or
leaving the Virginia health care workforce. As mentioned in more detail in Appendix E, we
relied on analyses conducted by McKinsey on longitudinal national surveys on nurses carried out
since the onset of the pandemic (Berlin, Lapointe, and Murphy, 2022; Berlin et al., 2023) to
enumerate and weight factors that affect the retention of workers. Additionally, to provide
context on recruitment, retention and the composition of the Virginia health care workforce, we
reviewed survey data collected by the Virginia Department of Health Professions’ Healthcare
Workforce Data Center, which conducts voluntary surveys on the department’s licensees through
the department’s online application and renewal processes. We reviewed the Virginia Hospital
and Healthcare Association’s Survey on Graduating Nurses for additional insights into the
pipeline of nursing recruits. Lastly, we analyzed Virginia State Loan Repayment Program data
from the Virginia Department of Health, Office of Health Equity, for insight into the effects,
costs, and efficacy of state loan repayment programs.
2
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Appendix B. Environmental Scan Methods
The environmental scan entailed a review of the peer-reviewed and gray literature related to
interventions for health care workforce retention and recruitment published from 2013 to 2023.
The scan includes interventions both within and beyond Virginia and the United States.
Methods
The research team conducted its environmental scan following methods defined by the
Institute of Medicine (2011). Based on those methods, we engaged in six steps to conduct the
environmental scan: (1) defining the research team for the environmental scan, (2) collecting
user and stakeholder input, (3) identifying the topic and research questions in the scan, (4)
developing and implementing a review protocol (including the study screening and selection
process), (5) screening and selecting studies, (6) appraising and synthesizing selected studies,
including quality of evidence.
Defining the Environmental Scan Research Team
The environmental scan research team is a subset of the overall study team and included a
social scientist with expertise in quantitative research in education, labor markets, and workforce
development and a policy researcher with expertise in qualitative and mixed methods, health care
and health policy, and environmental scans. The team has extensive experience in environmental
scans as well as the methodological and substantive expertise to evaluate relevant research and
other literature. The larger research team discussed and informed the environmental scan,
including the selection process and preliminary findings.
Collecting User and Stakeholder Input
To elicit user and stakeholder input, the environmental scan research team first identified key
health care domains based on the original study proposal—primary and behavioral health care.
The team then developed preliminary definitions for these domains based on the National
Academy of Medicine methods and identified different occupations and sites of care for each
health care domain. Based on email feedback from the larger research team and key stakeholders
over a two-week period, the environmental scan research team finalized the health care domain
definitions, focal occupations, and sites of care in the environmental scan. Key stakeholder
feedback was especially critical for flagging occupations and sites of care of particular relevance
in the Commonwealth of Virginia.
3
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Identifying Topics and Research Questions
Parameters of the scan, including focal health care domains, occupations in those domains,
and sites of care for focal health care domains were iteratively defined by the environmental scan
team with review and discussion by the larger research team and Virginia stakeholders, including
the Virginia Health Authority, a study advisory board comprised of various Virginia
stakeholders, and the larger Virginia health care practitioner and policy community at the study
multi-stakeholder conference. From this process, five basic topics emerged, outlined in Table
B.1. Stakeholders also helped identify search engines and websites used in this environmental. A
professional librarian identified the initial set of articles based on pre-defined search terms and
processes.
Table B.1. Environmental Scan Topics and Research Questions
Topic Research Questions
Virginia health • What are basic demographic characteristics of the health care workforce in Virginia?
care workforce • How does it compare to other Mid-Atlantic states and the United States as a whole?
Education and • For a given health care occupation, what are the basic education and training requirements?
training • What are areas of concern in relevant education and training programs?
• How does education and training compare to other Mid-Atlantic states and the United States
as a whole?
Legislation and • For a given health care occupation, what are the main governing and regulatory bodies in
regulation Virginia?
• What are the main regulations for each health care occupation in Virginia?
• What are recent proposed and instated changes in a given occupation’s regulation in Virginia?
• How does legislation and regulation compare to other Mid-Atlantic states and the United States
as a whole?
Recruitment and • What factors facilitate/impede recruitment into health care occupations in Virginia overall?
retention • What factors facilitate/impede recruitment into health care occupations in Virginia for key
subpopulations?
• What factors facilitate/impede retention in health care occupations in Virginia overall?
• What factors facilitate/impede retention in health care occupations in Virginia for key
subpopulations?
• How does recruitment and retention compare to other Mid-Atlantic states and the United States
as a whole?
Promising • What are the best or promising programs and practices for supporting the health care
programs and workforce in Virginia?
practices • How do best or promising programs and practices compare to other Mid-Atlantic states, the
United States as a whole, and globally?
• What is the quality of evidence vis-à-vis the effectiveness of these programs and practices?
• How do best or promising programs and practices compare to other Mid-Atlantic states, the
United States, and globally?
For each topic, the environmental scan team identified specific research questions, also
shown in Table B.1. Overall, the environmental scan was intended to provide a detailed
description of the Virginia health care workforce over time, the current education and training
landscape, recent and pending legislation and regulations related to the Virginia health care
4
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
workforce, and promising workforce development programs, policies, and practices to support
the health care workforce in Virginia and elsewhere—particularly those dealing with recruitment
and retention.
Developing and Implementing a Review Protocol
The professional librarian formulated detailed search strategies for PubMed (NIH/NLM),
CINAHL (EBSCOhost), and American Psychological Association PsycINFO (EBSCOhost). We
applied the following limits for each database: articles in peer-reviewed literature, published in
English language, published in the last ten years (2013–2023), and U.S. studies only. We
conducted searches for state legislation in Nexis Uni and BillTrack50. Additionally, we
conducted searches for policy documents and other grey literature (2013–2023) in Google. All
search strategies were based on the defined focal health care domains, occupations in those
domains, and care settings for focal health care domains. The following health care domains,
occupations, and settings further bounded the search:
• health care domains
- primary care, including
§ internal medicine
§ family medicine
§ pediatrics
§ obstetrics and gynecology
- behavioral health, including
§ psychotherapy
§ psychiatry
§ social work
§ drug and alcohol rehabilitation
• occupations
- primary care, including
§ physicians
§ physician assistants
§ licensed midwives
§ nursing
- behavioral health, including
§ psychiatrists
§ licensed alcohol and drug counselors
§ board-certified behavior analysts
§ licensed mental health counselors
§ licensed professional counselors
§ licensed clinical social workers
§ certified substance abuse counselors
§ psychiatric nurse mental health clinical specialists
5
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
§ licensed independent clinical social workers
§ licensed marriage and family therapists
• care settings
- hospitals
- health systems
- nursing homes/long-term residential care
- clinics and medical offices (outpatient, ambulatory)
- community health centers (federally qualified health centers)
- mental health treatment centers, addiction treatment centers.
Table B.2. summarizes key topics and associated search terms.
For each identified topic of interest, the array of search terms was based on (1) basic research
questions associated with each identified topic (e.g., “What are basic demographic characteristics
of the health care workforce in Virginia?”), (2) a preliminary search of related research articles
and grey literature on a given topic by the environmental scan team, and (3) search terms based
on environmental scan and larger research team discussions for a given topic.
After each initial search, the librarian and environmental scan team met to briefly review the
search process and ensure the executed process and search terms followed the review protocol.
The librarian conducted the initial search over a three-week period, addressing one topic
before moving to the next. The librarian limited a given search to a single health care workforce,
including primary care and behavioral health care for physicians, nursing, counselors and social
workers. The librarian provided a detailed Excel database of search results for each topic,
including the article title, authors, journal or other place of publication, abstract, and URL.
Searches were conducted in May and June 2023.
Table B.2. Key Concepts and Related Search Terms
Topic Key Concepts
Virginia health care • Health care workforce statistics including:
workforce – Demographics
– Workforce trends + projections
– Staffing retention rates
– Retirement rates
Legislation and • State regulations/federal regulations /current legislation/ policy + affecting health care
regulation workers including:
– Virginia code + regulation + health care workforce
– Virginia Department of Health Professions + regulation/licensing
– Virginia + regulatory review + health care workforce
– Virginia + legislation + health care workforce
– Virginia + alternative licensing + health care workforce
– Current funding environment + health care workforce development
– State/ federal financial aid and loan forgiveness programs
6
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Education and • Education/training/licensing/registration/accreditation/curriculum
training • Continuing education/professional development/license renewal
• Education/training/licensing/registration/accreditation/curriculum + postsecondary
• Education/training/licensing/registration/accreditation/curriculum + K–12
• Education/training/licensing/registration/accreditation/curriculum + Area Health Education
Center
Recruitment and • Continuing education/professional development/license renewal
retention • Continuing education/professional development/license renewal + Area Health Education
Centers
• Higher education student financial assistance + health care workforce
• Education/training + faculty
• Education/training + clinical placement
• Support/ continuing education/training for providers
• Promoting/impeding factors + health care professions for immigrants, veterans, military
spouses—disparities/inequities/underrepresented, underserved
• Promoting/impeding factors + health care professions + immigrants, veterans, military
spouses
• Barriers/facilitators + health care professions + immigrants, veterans, military spouses
• Incentives/incentivizing + health care education + training + professions
Promising programs • Best practices/promising programs/successful programs + health care
and practices workforce/jobs/occupation/front line health workers
Screening and Selecting Studies
Members of the environmental scan team independently screened and selected articles across
two iterations. In the first iteration, a team member reviewed the full array of articles identified
in an initial search, using specified databases and search terms (see “Developing and
Implementing a Review Protocol” and Table B.2). Across all health care occupations and
searches, a total of 1,405 articles were included in the full, initial array. In the second iteration of
screening and selection, a team member reviewed articles for topical relevance based on the title
and abstract. If the article topically matched specified research questions, it was retained—
regardless of research methods used or the geographic focus of the research. Team members
reviewed each other’s work and discussed and resolved any discrepancies in retained articles.
This topically relevant but methodologically and geographically broad array of articles (N = 357)
served as the core set of articles for the environmental scan.
Appraising and Synthesizing Selected Studies
To analyze and synthesize selected studies in our environmental scan, the environmental scan
team engaged in four steps: (1) an initial overview of articles to understand broad patterns, (2)
coding and otherwise assessing articles according to research questions and identified patterns or
themes in Step 1, and (3) drafting and discussing text regarding main findings, within the
environmental scan team and the larger research teams. In Step 4, the team presented preliminary
findings to Virginia stakeholders for comment and discussion.
7
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Appendix C. Interview and Focus Group Guide
1. From your perspective, what are the main causes of shortages among Virginia’s primary care
workforce?
- Behavioral health?
- Nursing?
2. What strategies and/or policies have you observed in Virginia to retain its primary care
workforce? Which ones are effective (or ineffective) in your view? Please elaborate.
- Behavioral health?
- Nursing?
3. What strategies would you like to see your community or state government use to boost
retention of its existing primary care workforce?
- Behavioral health?
- Nursing?
[Probe on role of strategies such as: increased compensation/monetary incentives,
supporting work-life balance, and transportation, etc.]
4. What strategies might Virginia use to help recruit new practitioners to expand its existing
health care workforce in:
- Primary care?
- Behavioral health?
- Nursing?
5. A few states have created alternative pathways to licensure for international doctors,
veterans, military spouses, and other populations seeking work in in health care. Should
Virginia pursue similar strategies to boost workforce in the following fields? Describe the
risks and/or benefits of doing so.
- Primary care?
- Behavioral health?
- Nursing?
6. How much of a barrier do the costs associated with education and/or licensure (and re-
licensure) present for people seeking practice in health care, nursing, and behavioral health?
- How about requirements for long periods of supervised practice?
- How about stresses related to work in these fields (e.g., workload, paperwork/charting
burden, stressful work environment)?
7. Should Virginia aim to incentivize more high school and/or college students to enter the
following fields? If yes, how would you do it?
- Primary care?
- Behavioral health?
8
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
- Nursing?
8. What steps need to be taken in order for state, federal, and private funding to be allocated to
workforce development programs?
9. What strategies do you recommend be used to boost the number of training positions for
health professionals in
- Primary care?
- Behavioral health?
- Nursing?
10. We’ve heard that there are not enough nursing faculty, clinical preceptors, and training sites.
What ideas do you have to address this challenge?
11. Should Virginia change its processes for awarding and managing educational scholarships
and/or loan forgiveness to incentivize individuals to pursue careers in primary care,
behavioral health, and nursing? What ideas do you have to accomplish this?
12. Based on your expertise, name an actionable step or strategy you believe Virginia should and
can take in the next 6 months to address workforce shortages in primary care, nursing, and
behavioral health?
- In one year?
- Next two years?
13. What strategies can be used to ensure that efforts around health care workforce development
are sustained over time?
9
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Appendix D. Hospital Survey Questions
1. Which of the following strategies has your hospital/health system used to retain nurses?
Mark all that apply.
[ ] Increased salary
[ ] Bonuses
[ ] Educational loan repayment
[ ] More flexibility in work hours
[ ] Transportation vouchers
[ ] Convenient and/or free parking
[ ] On-site childcare
[ ] Subsidizing cost of housing
[ ] Meal coverage
[ ] Mental health services for staff
[ ] Job training to improve and/or expand skillsets
[ ] Diversity, equity, and inclusion strategies
[ ] Strategies that promote teamwork
[ ] Covering cost of maintaining licensure
[ ] Changed patient flow or staffing models
[ ] Offering educational advancement opportunities as part of employment
[ ] Supervisory training and/or ombudsman program to address employee concerns
14. How have you changed practices and/or technology to support flexible staffing models?
10
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Appendix E. The Mathematical Formulation of our System
Dynamics Simulation Model
Introduction
In early 2022, we developed a simulation model focusing on the short and medium-term
impact of the COVID-19 pandemic on the nurse labor market. Our model incorporated a wide
range of factors, including epidemiological trends, behavioral dynamics, hospitalization patterns,
relief programs, and wage considerations., In this appendix, we present a revised model that,
building on this precursor model, simulates and tracks the dynamics of key variables related to
the health care workforce in Virginia over time. While our revised model does incorporate the
effects of the COVID-19 pandemic, our primary objective here is to simulate the medium- to
long-term supply and employment trends within Virginia’s health care workforce. We explore
scenarios under both baseline (status quo) conditions and various policy interventions. The
central aim is to offer a comprehensive set of projections for Virginia’s health care workforce,
moving beyond the sole assessment of the pandemic’s impact over time.
Although our model encompasses workforce dynamics among nurses, physicians, and
behavioral health specialists within Virginia, our focus in this appendix will predominantly
revolve around nurses. The foundational framework remains consistent for physicians and
behavioral health specialists, albeit with variations in initial state variables and input parameters.
Our model provides insights into the health care workforce in Virginia, encompassing both state-
level and regional perspectives, spanning the eight Area Heath Education Centers (AHEC)
regions of Virginia. It calculates medium- to long-term dynamics, yielding output projections for
the next 15 years.
Our model employs a system dynamics approach. This approach ensures a deterministic
representation of aggregate population dynamics. The system dynamics method proves
advantageous when studying labor markets during prolonged non-equilibrium periods, such as
those induced by pandemics. It offers a macroscopic perspective for analyzing research questions
and facilitates scenario exploration and sensitivity testing.
The choice of the system dynamics approach is driven by its ability to specify causal
mechanisms driving system dynamics and its suitability for examining labor markets impacted
by extraordinary events like pandemics, which often operate outside of equilibrium for extended
durations. Furthermore, the system dynamics approach is computationally efficient, quicker to
develop, easier to understand, and faster to execute compared with individual-level
microsimulation or agent-based models, particularly when dealing with a substantial number of
parameters.
11
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
The system dynamics approach involves formulating the model using coupled ordinary
differential equations (ODEs) and numerically integrating them using specialized solvers. The
numerical integration poses a stiff problem, where the step size of the numerical solution is
limited more by the stability of the numerical technique than by its accuracy (Hindmarsh and
Petzold, 1995). For our implementation, we used the R programming language with the deSolve
package (Soetaert, Petzoldt, and Setzer, 2010). DeSolve employs FORTRAN solvers of the
Livermore family to solve initial value problems for stiff ODEs, and it offers various types of
solvers for this purpose. We used the default integration method, LSODA, that switches
automatically between using methods for stiff and nonstiff systems.
Our model operates with a weekly timestep for generating outputs, although the solver
typically uses much smaller time steps during numerical integration. Since many input model
parameters and rate values are provided on a yearly basis, we convert these values into weekly
rates to maintain consistency with the model’s weekly timestep.
The System Dynamic Model and its Differential Equations
The state variables include the supply of nurses (𝑞! (𝑡)), the qualified pool of potential nurses
(𝑛! (𝑡)), the yearly wage for nurses (𝑤(𝑡)), and the perceived barriers affecting their willingness
to work as nurses relative to a baseline (𝑧(𝑡)). We also consider nurse demand (𝑞" (𝑡)), which
represents the need for nurses and is externally specified as an input to the model.
The nurse supply variable (𝑞! (𝑡)) represents the proportion of available nurses actively
choosing and desiring to work as nurses, including those currently employed or actively seeking
nursing positions. On the other hand, the qualified pool variable (𝑛! (𝑡)) includes individuals who
possess the necessary qualifications to work as nurses but may have temporarily chosen not to do
so. This category encompasses retired nurses who could potentially return to the workforce,
individuals in less demanding professions with lower pay, and nurses experiencing hidden
unemployment. However, nurses who have pursued completely different career paths outside of
nursing, such as becoming doctors or researchers, are not considered. The actual number of
available nurses is determined as the minimum between demand and supply, given by 𝑞(𝑡) =
min[𝑞" (𝑡), 𝑞! (𝑡)].
Our model assumes that the supply of nurses (𝑞! (𝑡)) depends on the qualified pool of
potential nurses (𝑛! (𝑡)), the yearly wage for nurses (𝑤(𝑡)), and the perceived barriers affecting
their willingness to work in the nursing field relative to a baseline, summarized by one variable
𝑧(𝑡) whose size is proportional to the strength of said barriers.
In order to derive an explicit expression for 𝑞! (𝑡) as a function of these quantities, we
assume that the decision to work made by the nurses in the qualified pool follows a simple utility
maximization principle. We assume that nurses receive a utility ln(𝑤) from working for wage w,
and a utility ln(𝑤# ) if they do not work, where 𝑤# has the interpretation of a reservation wage,
since nurses only work if 𝑤 > 𝑤# . We assume that there is variation in the reservation wage in
12
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
the nurses’ population, and therefore we model 𝑤# as a random variable with a log normal
distribution. Therefore, the total number of nurses working is:
1
𝑞! = 𝑛! 𝑃3ln(𝑤) > ln(𝑤# )4 = 𝑛! [1 + erf(𝑠! (ln(𝑤) − ln(𝑤# ))]
2
where 𝑠! is a parameter that is inversely proportional to the variance of the reservation wages
𝑤# . Qualitatively, the supply of nurses is an increasing function of wages, which is 0 when
wages are 0, is exactly one half when wages are equal to the reservation wage (because nurses
are then indifferent to the two options) and becomes equal to the entire pool 𝑛! when wages tend
to infinity.
What is missing from this formulation is the reluctance to work z. We assume that the
reluctance to work influences the reservation wage 𝑤# and that higher levels of reluctance to
work raise the reservation wage. We model the relationship between reluctance to work and
reservation wage with an S-shaped function:
1
𝑤# (𝑧) = 𝐾 >1 + [1 + erf(𝑠$ (ln(𝑧) − ln(𝑚$ ))]@
2
When the reluctance to work is 0, the reservation wage is 𝐾 and the supply of nurses is
maximum. As the reluctance to work grows to infinity, the reservation wage reaches its
maximum value (2 𝐾), and the supply reaches its minimum value. We assume that 𝐾 =
𝜌% 𝑤# (1). This is because we choose to set to the reluctance value z equal to 1 to represent our
baseline setting.
In our model, 𝑞! (𝑡) adheres to a specific functional form represented by a sigmoidal S-
surface specified by the equations 𝑞! and 𝑤# (𝑧) given above. It is important to note that this
specific function for 𝑞! (𝑡) is just one of many possible choices or models to describe how supply
reacts to changes in wages and perceived barriers to work. Essentially, it is just one way we
model supply changes. In a more general context, 𝑞! (𝑡) can be characterized using alternative
functional forms, potentially dependent on a different set of parameters. For clarity, we denote
this specific model as 𝑞! (𝑡) = 𝑆& (𝑛! , 𝑤, 𝑧), where 𝑆& (𝑛! , 𝑤, 𝑧) represents the particular
functional form we’ve chosen to represent supply. Here, 𝜃 is a parameter set that defines the
specific shape of the sigmoidal function we’ve selected. The parameter set 𝜃 includes of
𝜌% , 𝑤# (1), 𝑚$ , 𝑠! , and 𝑠$ . In the subsequent text, we will use 𝑞! (𝑡) as a general reference to the
nurse supply and use 𝑆& (𝑛! , 𝑤, 𝑧) to denote the specific functional form we use to model supply.
Our sigmoidal S-surface for supply is visually presented in Figure E.1. The contour plot beneath
the S-surface demonstrates that as health care work reluctance increases, real wages must
increase to sustain the nurse supply. This aligns with the concept of compensating-wage
differentials in economic theory.
13
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure E.1. Impact of Health Care Work Reluctance on Nurse Supply and Real Wages Illustrated
Through Sigmoidal Curves
The time evolution of supply of nurses is determined by the interplay of various variables
that change over time. To capture these dynamics, we consider a differential equation describing
the rate of change of the supply function 𝑞! (𝑡) with respect to time 𝑡. The differential equation is
given by:
𝑑𝑞! ∂𝑞! ∂𝑞! 𝑑𝑛! ∂𝑞! 𝑑𝑤 ∂𝑞! 𝑑𝑧
= + + + .
𝑑𝑡 ∂𝑡 ∂𝑛! 𝑑𝑡 ∂𝑤 𝑑𝑡 ∂𝑧 𝑑𝑡
In this equation, the total derivative of supply with respect to time is expressed as a sum of
terms containing partial derivatives of the supply with respect to its controlling variables. The
'(
first term ')! represents the long-term supply trend independent of changes in real wages,
pandemic-related outcomes, and the qualified pool of registered nurses. We have assumed that
variations in supply are solely attributed to changes in the qualified pool (𝑛! ), wages (𝑤), and
reluctance (𝑧), and as a result, we have omitted this term from consideration. The second term
captures the long-term supply dynamics of the qualified pool of registered nurses, considering
changes due to recruitment and expansion policies, as well as contractions such as permanent
retirements. The medium to short-term dynamics are mainly described by the last two terms,
quantifying how the supply changes with wages and work reluctance. These terms depend on the
14
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
'( '( "%
derivatives '%! and '$! , as well as the rates of change of real wages and work reluctance ( ") and
"$ '( '( '(
). By applying equations d, we derive the expressions for '*! , '%! , and '$! .
") !
The differential equation for supply is a general equation that applies to our overall
expression of supply, denoted as 𝑞! (𝑡). Now, we move forward by calculating the partial
differentials that contribute to the equation for the total derivative of supply. In this context, we
will be dealing with our specific representation of supply, 𝑆& (𝑛! , 𝑤, 𝑧). The derivatives of 𝑆&
correspond to scaled log-normal probability density functions (PDFs) since 𝑆& (𝑛! , 𝑤, 𝑧)
represents a scaled log-normal cumulative distribution function. The expressions are as follows:
∂𝑆& 𝑠! exp{−𝑠!+ [ln𝑤 − ln𝑚! (𝑧)]+ }
= = 𝑆& (1, 𝑤, 𝑧).
∂𝑛! 𝑤 √𝜋
∂𝑆& 𝑛! 𝑠! exp{−𝑠!+ [ln𝑤 − ln𝑚! (𝑧)]+ }
= .
∂𝑤 𝑤 √𝜋
', ', '-! ($)
The expression for '$" is obtained by taking the product of '- "($) and '$
, given by:
!
∂𝑆& 𝑛! 𝑠! exp{−𝑠!+ [ln𝑤 − ln𝑚! (𝑧)]+ }
=− ,
∂𝑚! 𝑚! √𝜋
∂𝑚! 𝑚! (𝑧)𝑠$ exp{−𝑠$+ [ln𝑚$ − ln𝑧]+ }
= .
∂𝑧 𝑧 √𝜋
', ', ',
With the expressions for '*" , '%" , and '$" derived, we proceed to specify the models
!
describing the rates of change of 𝑛! , 𝑤, and 𝑧 over time in the next sections.
Modeling the rate of change of the qualified pool
"*
Here we present our model for ")! , which describes the changes in the overall qualified pool
of workers over time. The model takes into account factors that increase the workforce, such as
enrollment and graduation rates from nursing schools, which contribute to the growth of the
nurse supply. It also encompasses factors that decrease the workforce, including rates of
permanent retirement or nurses transitioning to other career paths. The basic form of the model
subtracts the outflows from the inflows to describe how 𝑛! changes over time. We assume that
the inflow of qualified nurses is proportional to the current supply, represented by the constant 𝜈
, (* ,%,$)
scaled by the normalized scaling factor 𝑆& (1, 𝑤, 𝑧) = " *! . On the other hand, the outflow is
!
determined by a fixed yearly probability 𝜔 of permanently leaving the qualified pool because of
"*
retirement or irreversible decisions. Thus, the simplest model for ")! is:
15
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
𝑑𝑛!
= 𝜈𝑆& (1, 𝑤, 𝑧) − 𝜔𝑛! .
𝑑𝑡
"*! 1," (2,%,$)
The stationary solution is obtained by setting = 0, resulting in 𝜔 = . It is worth
") *!
noting that the depletion of the qualified pool may vary between nurses in the supply and those
in the qualified pool but not in the supply. We assume that the per-person depletion rate for
nurses in the qualified pool but not in the supply (𝜔2 ) is greater than the rate for nurses in the
supply (𝜔+ ) by a constant factor 𝜅3 . Therefore, 𝜔2 = 𝜅3 𝜔+ . Combining these considerations,
we have:
𝜔𝑛! = 𝜔2 𝑛! [1 − 𝑆& (1, 𝑤, 𝑧)] + 𝜔+ 𝑛! 𝑆& (1, 𝑤, 𝑧),
which leads to:
𝜔 = 𝜔2 [1 + 𝑆& (1, 𝑤, 𝑧)(𝜅3 − 1)].
It is important to note that the outflow, representing the permanent retirement of nurses from
the qualified pool, implicitly depends on wages and health care work reluctance through the term
𝑆& (𝑛! , 𝑤, 𝑧) that multiplies 𝜔2 and 𝜔+ . As wages decrease or work reluctance increases, the
supply increases. Since 𝜔2 > 𝜔+ , nurses retire at a higher rate or switch to alternative
occupations or careers. Similarly, the inflow depends on wages and work reluctance through
𝑆& (1, 𝑤, 𝑧). As real wages increase or work reluctance decreases, the normalized supply
increases, signaling to prospective students that nursing is an abundant and desirable occupation
to pursue.
"*!
Returning to the steady-state solution ( = 0), we find:
")
𝜈𝑆& (1, 𝑤, 𝑧)
𝜔2 = .
𝑛! [1 + 𝑆& (1, 𝑤, 𝑧)(𝜅3 − 1)]
Wage Dynamics
In our model, nurse wages are influenced by the imbalance between demand and supply, and
they are assumed to have limited flexibility, resulting in asymmetric adjustments. When wages
are below the market-clearing rate, hospitals need to increase real wages to fill vacancies. On the
other hand, when wages exceed the market-clearing rate, hospitals may desire to reduce them,
but contractual arrangements prevent such adjustments. The real rate of change of nurse wages,
"%
denoted as ") , is described by the equation:
𝑑𝑤
= [(𝛼" − 𝛼! )𝐻(𝑞" − 𝑞! ) + 𝛼! ] ⋅ (𝑞" − 𝑞! ) 𝑤 ,
𝑑𝑡
16
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Here, 𝐻(𝑥) represents the Heaviside step function, which we approximate using an inverse
logistic function for integration in our model implementation. The parameters 𝛼" and 𝛼! ,
determine how the wage gap between demand and supply affects wage changes. We assume that
real wages increase more rapidly when there is excess demand for nurses than when there is
excess supply. Hence, the rate 𝛼" is numerically greater than 𝛼! to reflect this assumed faster
growth rate of wages.
Health Care Work Reluctance
A key component of our model is the health care work reluctance measure 𝑧(𝑡) and its
evolution over time. Working during a pandemic is undeniably more challenging than under
normal circumstances, and 𝑧 increases based on various pandemic-related factors, including
fatigue from COVID-19 hospitalization surges, increased risk of contracting the virus at work,
unclear quarantining guidelines, increased proximity to preventable deaths, shortages of personal
protective equipment, and higher patient-to-nurse ratios during surges.
Health care work reluctance 𝑧(𝑡) is assumed to change according to an exponential weighted
moving average (EWMA) process over 𝛥(𝑡), which captures nurses’ weekly evaluations
affecting their reluctance to work. The variable 𝛥(𝑡) quantifies nurses’ weekly assessments
regarding their willingness to work. One way of expressing the dynamics of the EWMA by a
differential equation is as follows:
𝑑𝑧
= (1 − 𝑠)[𝛥(𝑡) − 𝑧(𝑡)]
𝑑𝑡
Here, 𝑠 is a weekly discount parameter and plays a crucial role in determining the relative
significance of past nurses’ weekly assessments compared with the current ones in forming a
nurse’s overall reluctance in health care work. The discount parameter can be calculated by
specifying a half-life duration 𝜏 using the formula ln(2)/(1 − s). The half-life represents the
duration for past evaluations to contribute half as much to the work reluctance measure.
A broader range of perceived barriers impact 𝛥(𝑡). We assume that these barriers can be
grouped into three main categories, namely distress, disengagement, and fatigue. From a
mathematical perspective, we express these factors as components influencing the assessment of
𝛥(𝑡). Our assumption is as follows:
𝛥(𝑡) = 𝛽4 𝜙4 (𝑡) + 𝛽5 𝜙5 (𝑡) + 𝛽6 𝜙6 (𝑡),
Where 𝛽7 𝜙7 (𝑡), with 𝑋 ∈ 𝐷, 𝐸, 𝐹, represent the contributions from three different
components: distress, disengagement, and fatigue, respectively. The coefficients 𝛽 represent
weights that sum up to one. Each of the components 𝜙 represents evaluations. Specifically, 𝜙6
measures nurses’ fatigue, while 𝜙4 quantifies their distress.
17
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Fatigue (𝜙6 (𝑡)) is assumed to be fully determined by the demand level, which can be
measured using the patient-to-nurse ratio (𝜂) ) and its changes over time (𝑡). We assume that a
fixed proportion (1 − 𝜉8 ) of distress depends on the demand level. Therefore, 𝜙4 =
𝜂# c31 − 𝜉8 4𝜂) /𝜂# + 𝜉8 𝜓4 e, where 𝜓4 is considered a fixed effect that does not depend on
demand and can be influenced by policy interventions unrelated to demand. The default value for
𝜓4 is one, representing the baseline fixed effects.
Disengagement (𝜙5 ) is assumed not to change with demand and is considered as a fixed
effect that can be influenced by policy interventions unrelated to demand. Combining all the
above points, we have the following expression:
𝛥(𝑡) = 𝛽6 𝜂) /𝜂# + 𝛽4 c31 − 𝜉8 4𝜂) /𝜂# + 𝜉8 𝜓4 e + 𝛽5 𝜓5 .
For the initial condition at 𝑡 = 0, the patient-to-nurse ratio 𝜂) is set to its baseline value 𝜂# .
Therefore, by design the baseline default value for 𝛥(𝑡) is set equal to one. As a result, the
baseline value for health care work reluctance is represented by 𝑧 = 1. This is because from the
"$
differential equation, under the stationary condition (i.e., ") = 0), health care work reluctance
does not change and remains at its baseline value, 𝑧(𝑡) = 𝛥(𝑡).
Informing the Model Inputs
Our model draws from a variety of inputs, including parameter values and their ranges, initial
conditions for state variables, and historical time-series data for variables such as occupied
hospital beds, nurse employment numbers, and real wages. These inputs guide the model’s
behavior and serve as targets for calibration. In this section, we offer a concise overview of the
model inputs and the data sources that informed our parameter estimation.
The State Variables’ Initial Conditions
To start, we describe how we computed our state variables’ initial conditions. Our model
encompasses several key state variables, namely: demand, supply, and the qualified number of
nurses within the system. All these variables are quantified in units of nurses. Additionally, our
state variables encompass the average annual wage of nurses (measured in dollars), the health
care work reluctance metric, and the proportion of hospitalizations attributed to COVID-19
patients.
The model relies on input values to establish the initial conditions for these six state
variables. We work with two distinct sets of initial conditions. The first set is employed for
calibration purposes, defining the initial states just prior to the onset of the pandemic. This
calibration enables our model to operate throughout the pandemic years, facilitating a
comparison between projected trajectories of select state variables and the observed data from
our time series. Notably, we focus on trajectories related to annual wages and the number of
18
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
employed nurses. By contrasting these trajectories with empirical data, we refine the input
parameter values via our model calibration process.
The second set of initial conditions delineates the starting states during the current post-acute
pandemic phase. These are used for forecasting policy interventions and their relative impacts. It
enables us to predict potential outcomes and make comparisons among different policy
strategies. When new initial conditions for this phase are absent, we adopt the final conditions
derived from our model calibration runs as the basis for the forecast phase.
Here, we delve into the specifics of the first set of initial conditions. Our baseline for analysis
is derived from our time series data for Virginia, specifically focusing on the employment figures
for nurses in hospitals as of April 5, 2020. On this date, the number of nurses employed in
hospitals across the state of Virginia totaled 66,450, with an average annual wage of $74,380.
We assume that demand exceeded supply on this date, enabling us to estimate supply as the
existing number of nurses employed in Virginia’s hospitals. Given the nascent stage of the
pandemic in the United States at that time, we consider the proportion of COVID-19
hospitalizations to be negligibly small, approximating it to zero. Additionally, we initialize the
health care work reluctance metric at its baseline value of one. Using these foundational
estimates, we proceed to determine the range of values for nurse demand in Virginia on this
pivotal date.
Nationally, the demand for nurses has been assessed at 3,154,000 (Haines, 2022). To localize
this demand to the Commonwealth of Virginia, we employ a scaling methodology. This
approach capitalizes on data pertaining to the count of staffed beds and the average bed
occupancy rate in Virginia, facilitating the alignment of national demand with the state’s
contextual factors.
Specifically, within Virginia, the inventory of inpatient beds aggregates to 15,177, with an
average occupancy rate of 70 percent. To contextualize the national demand within Virginia’s
framework, we necessitate insights into the total count of staffed beds at the national level and
the corresponding average bed occupancy rate. As per recent data, the cumulative tally of staffed
beds across U.S. hospitals reaches 919,649, accompanied by a bed occupancy rate of 64 percent
(American Hospital Association, undated; Michas, 2022).
Calculating the ratio between the product of bed occupancy and staffed beds in Virginia and
their national equivalents enables us to ascertain the scaling factor. The multiplication of this
scaling factor by the national demand yields an approximation of the initial demand tailored to
Virginia’s circumstances. This methodology customizes the demand projection by integrating the
state’s health care infrastructure and bed utilization dynamics.
An alternative estimation for demand in Virginia involves assuming that its demand-to-
supply ratio mirrors the national level. Consequently, we simply need to multiply this ratio by
our supply estimation for Virginia. Nationally, demand surpasses supply, with supply closely
aligned to the estimated number of employed nurses at 3,072,000 (Haines, 2022), marginally
lower than the 3,154,000 demand estimate. This results in a demand-to-supply ratio of
19
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
approximately 1.027. Assuming that initial demand surpasses supply in Virginia, we employ this
ratio to upscale the estimated count of employed nurses in the state. This furnishes an estimation
for Virginia’s demand.
By employing both of these methodologies, we arrive at distinct estimations for the pre-
pandemic initial demand scenario in Virginia. The estimated demand ranges from 59,800 to
76,600, with a central value of 68,224 that we consider the most plausible. The lower bound
approximates the estimate from our first methodology, and the central value is derived from the
second methodology. To establish the upper bound, we apply the same difference between the
central and lower values to the difference between the upper bound and the central value. Within
this range of uncertainty lies a diversity of values, collectively assuming a critical role within our
model calibration procedure.
Informing the Behavioral Model for Health Care Work Reluctance
Here, we outline the parameters that shape our behavioral model for health care work
reluctance. We begin by estimating the 𝛽 weights, which quantify the significance of each
behavioral component in influencing evaluations. The most reliable source for estimating these 𝛽
weights is through surveys that assess work-related attitudes and the desire to continue working,
specifically targeting nurses. For this purpose, we utilized the analyses conducted by McKinsey
on longitudinal national surveys on nurses carried out since the onset of the pandemic (Berlin,
Lapointe, and Murphy, 2022; Berlin et al., 2023).
In exploring the drivers of nurses’ intentions to leave their positions, we focused on
respondents who expressed a likelihood of leaving within a year. These respondents were
required to provide scores for 15 specific factors and 24 barriers, such as insufficient staffing
levels, emotional toll, approaching retirement age, fear of COVID-19, lack of career prospects,
uncertainty, unmanageable workload, and no work-life balance. We categorized each factor and
barrier under fatigue, distress, or disengagement, depending on their nature. While some factors,
such as no development opportunities, clearly fell under disengagement, others, such as move to
desirable location, did not fit neatly into these categories and were excluded from analysis.
Our examination of the McKinsey survey data revealed that fatigue was the most influential
factor, accounting for approximately 50 percent of the importance. Distress followed closely,
constituting around 30 percent of the total importance. Based on the data and respondent
numbers, we estimated an accuracy range of approximately ±15 percent of their values (i.e., ±7.5
percentage points for the case of fatigue) for these importance weights.
Given the constraint that the weights must sum to 100 percent, we derived the importance
weight for disengagement by considering the sampled values of importance weights associated
with fatigue and distress. This allowed us to establish a comprehensive understanding of the
impact of these behavioral components on health care work reluctance.
Alongside the 𝛽 weights, our behavioral model introduces two additional parameters: 𝜉8 ,
which represents the portion of distress that does not depend on the demand level, and the
20
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
discount parameter 𝑠. Regrettably, despite extensive searches through the McKinsey surveys and
existing literature, we were unable to discover any data or sources to aid in estimating 𝜉8 —the
parameter describing the fixed proportion of distress independent of perceived fatigue. In
response, we made an assumption that 𝜉8 falls within the range of 50 percent ±7.5 percent.
However, in stark contrast, the outcomes derived from the McKinsey surveys provide
invaluable insights into the temporal evolution of nurses’ intentions to leave their positions.
These insights can be effectively leveraged to estimate the discount parameter 𝑠.The percentage
of nurses contemplating leaving their jobs surged from 22 percent in February 2021 to 32 percent
by November 2021. This heightened level endured until September 2022, signifying a substantial
increase by a multiplicative factor of 1.45, likely attributed to the pandemic-induced
circumstances.
Given our analysis that points to fatigue as the predominant influencer of health care work
willingness and considering our access to national weekly data on both COVID-19 and non-
COVID-19 hospitalizations, we opted to merge the McKinsey survey findings with
hospitalization data. This fusion allowed us to estimate the value of the discount parameter 𝑠, or
equivalently, the half-life duration of work reluctance evaluations. The underlying concept is that
heightened hospitalizations and an increased proportion of COVID-19 patients among
hospitalizations impact fatigue, which manifests with a lag in time following peaks in
hospitalizations and COVID-19 proportion. This enables us to evaluate the perceived increased
fatigue after a certain lag period. By interlinking these datasets, we aimed to unravel the dynamic
interplay between work reluctance and hospitalization trends.
However, it is important to consider the limitations of our assumption. Initially, the pandemic
saw a decrease in overall hospitalizations even as the proportion of COVID-19 hospitalizations
increased. During this period, nurses experienced significant fatigue because of increased
workloads. The decline in non-COVID-19 hospitalizations contributed to COVID-19
hospitalizations forming a larger proportion of total hospitalizations.
To ensure our analysis yields meaningful results and our simulation model is realistic, we
must quantify the additional effort COVID-19 patients demand from nurses in comparison to
non-COVID-19 patients. Our approach involves considering factors such as evolving nurse
demand, increased ICU patient ratios during the pandemic, and varying levels of effort needed
for COVID-19 patients versus non-COVID-19 patients. Our goal is to calculate and utilize an
effective patient-to-nurse ratio 𝜂) that captures the added efforts due to COVID-19 patients,
serving as a proxy for effort that scales linearly with fatigue.
We start by assessing the pre-pandemic and baseline efforts. Nationally, the nurse-to-patient
ratio during this period ranged from 1:4 to 1:5 for medical-surgical units, and 1:2 for ICUs (“The
Importance of the Optimal Nurse-to-Patient Ratio,” n.d.). For our calculations, we consider a
baseline nurse-to-patient ratio of 1:4. The national proportion of patients admitted to the ICU
pre-pandemic had a median of 12 percent with an interquartile range of 9 percent to 17 percent
21
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
(Seymour et al., 2012). This information aligns with our data from Virginia on hospitalizations.
With these values, our baseline patient-to-nurse ratio 𝜂# is 3.76.
To estimate the added effort for COVID-19 patients, we consider multiple factors. These
include comparing the average length of stay between COVID-19 and non-COVID-19 patients,
the proportion of COVID-19 patients requiring ICU care, and the increased daily effort required
for COVID-19 patients both inside and outside ICU settings.
In 2018, the average length of stay for a hospitalization in the United States was 5.5 days
based on 36.4 million inpatient stays (Tipton et al., 2021). In contrast, the estimated length of
stay for COVID-19 hospitalizations was 15.35 days, with a 95 percent confidence interval of
13.47 to 17.23 days (Alimohamadi et al., 2022). This suggests that the effort required for a
COVID-19 patient is roughly three times that of a non-COVID-19 patient based on length of
stay.
During the pandemic, the national rate of ICU admission for COVID-19 patients was initially
reported as 32 percent (Abate et al., 2020). However, our analysis of COVID-19 data for
Virginia suggests this proportion to be closer to 20 percent.
By integrating these factors, we reevaluated the time-series data on total and COVID-19
hospitalizations, correlating them with the McKinsey survey results when the percentage of
nurses considering leaving their jobs peaked. This analysis allowed us to estimate the half-life
duration of work reluctance evaluations, indicating a range of three to 12 months. The substantial
uncertainty stems from relying on only two time points from the McKinsey surveys. While the
accuracy is limited, we chose to work with the available data and exercise our judgment in
interpreting plausible outcomes from the analysis.
Informing How Supply Responds to Changes in Health Care Work Reluctance
Having elucidated our behavioral model for work reluctance and its parameter estimation
process, our focus now shifts to estimating how changes in health care work reluctance impact
nurse supply. This entails the estimation of the supply parameters 𝜃, which govern the
',
relationship '$" . Specifically, we delve into 𝑚$ and 𝑠$ , which define the supply’s response to
increasing work reluctance. These parameters, being intrinsic to our model’s intricacies, require
an indirect estimation method based on the model calibration process, which is detailed in a
subsequent section of this appendix.
However, in a bid to enhance the model’s input intuitiveness, we’ve opted to derive and
compute the values of 𝑚$ and 𝑠$ from more readily understandable input parameters. While
these parameters still pose challenges for estimation, they boast more intuitive definitions, which
could aid experts in behavioral health economics to provide better assessments of their values
and the associated ranges of uncertainty.
These intuitive input parameters include the initial proportion of nurses within the supply
relative to the total pool of qualified nurses. This proportion acts as our supply’s baseline when
22
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
health care work reluctance assumes its baseline value of one. Nationally, approximately 84.1
percent of licensed RNs are employed in the nursing sector (American Association of Colleges
of Nursing, 2023a). We consider an uncertainty range for this proportion around our central
estimate of 84.1 percent spanning 80 percent to 88 percent. This estimate serves as our baseline
proportion of nurses in the supply.
Furthermore, we need a second estimation of the proportion of nurses within the supply
concerning the total qualified nurse pool. This estimation comes into play when health care work
reluctance surpasses its baseline value, while keeping other factors, including the annual wage,
constant. Our methodology involves setting this second supply proportion to 50 percent of the
baseline supply value. Subsequently, we attempt estimating the range of health care work
reluctance values denoted as 𝑧+ , which would trigger this particular level of supply reduction.
Our estimation for 𝑧+ spans a wide range from 1.1 to 3, with the most plausible and central
value hovering around 1.5. This extensive range reflects the significant uncertainty associated
with this parameter. Nevertheless, this range does encompass a specific estimate derived from
the analysis of the McKinsey surveys.
The surveys highlighted a range of factors influencing nurses’ decisions to leave during the
pandemic, varying from 22 percent to 32 percent. This variability suggests that the supply of
nurses during the same period might have varied between 68 percent and 78 percent of the
qualified nurse pool 𝑛! . A simple analysis or estimation reveals that when the leave factor
increases by 45 percent (i.e., a factor of 32/22), the supply diminishes by approximately 13
percent (i.e., 1 – 68/78). It is important to note that this provides a lower bound of the response,
as the quoted 22 percent factor was not a baseline estimate but was also estimated within the
pandemic period.
The final input parameter essential for estimating how changes in health care work reluctance
affect nurse supply pertains to the percentage of the current wage required to sustain the existing
workforce if all factors causing work reluctance were eradicated (i.e., 𝑧 = 0). This concept
proves challenging to source directly, given its hypothetical nature. The scenario delves into the
hypothetical realm of nurses potentially trading a portion of their wages in return for the
complete removal of perceived barriers in their work environment. The crux of the matter lies in
identifying the minimum portion of their current wage that nurses would be amenable to accept
before it significantly influences their likelihood of opting to leave their nursing profession.
Our estimation for the wage that nurses might accept in exchange for the complete
elimination of perceived barriers in their workplace spans a broad spectrum from 70 percent to
99 percent of their current wage, with the most plausible and central value gravitating around 80
percent because of the highly skewed distribution of results. This parameter encompasses a
diverse range, reflecting the substantial uncertainty inherent in this scenario. Nonetheless, the
central estimate aligns with a plausible value that nurses might be willing to consider in a
scenario devoid of the barriers influencing their work reluctance.
23
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Sampling the values of these more intuitive parameters within their respective uncertainty
ranges leads to the specification of the values for 𝑚$ and 𝑠$ used in the simulation model.
Informing the Model Describing the Wage Dynamics
Our model component that describes how real wages change over time is guided by two key
parameters, 𝛼" and 𝛼! . They control how much real wages increase each week. The larger rate,
𝛼" , applies when there is more demand for nurses than there is supply. The smaller rate, 𝛼! ,
applies when there is more supply than demand. We derive these weekly changes from the yearly
rates. When there is a higher demand for nurses than there is supply, the real wages increase
more rapidly each year. We expect this increase to be somewhere between 0 percent and 10
percent every year, but our best estimate is centered at 2% percent. When there is more supply
than demand, the real wages still go up, but at a slower rate, and our estimate can be anywhere
from 0 percent to 2.5 percent every year, and our best estimate is centered at 0.5 percent.
Informing How Supply Responds to Changes in Wage
After discussing our wage increase model and how we estimate its parameters, our attention
now shifts towards determining the impact of wage changes on nurse supply. This involves the
',
estimation of the supply parameters, denoted as 𝜃, that govern the relationship '%" . Specifically,
we delve into two key parameters, 𝑚! (0), representing the wage when health care work
reluctance is zero, and 𝑠! , which defines the supply’s responsiveness to wage changes. Drawing a
parallel with our previous parameter estimation for supply response to health care work
reluctance, these parameters are intrinsic to our model’s complexities and are challenging to
estimate directly. However, there is a greater potential for accurate estimation in this case
compared with the parameters linked to health care work reluctance. This is because the
relationship between wage changes and nurse supply has been extensively studied in economics
over the years. Researchers have attempted to estimate the elasticity of nurse supply in response
to wage rates, which measures the sensitivity of the quantity of nurses supplied to wage
variations. In simpler terms, this elasticity quantifies how much an increase or decrease in real
wages influences the influx or exodus of nurses from the workforce that are employed or are
actively looking for employment as a nurse. A high elasticity implies that even minor wage
adjustments can lead to substantial shifts in nurse availability, whereas a low elasticity suggests
that wage changes have limited impact on nurse supply.
Nevertheless, it is important to note that, in line with our sigmoid supply surface, the
elasticity of nurse supply in response to wage rates is itself contingent on the prevailing wages
and supply levels. Therefore, the elasticity estimation is a localized measure provided under
baseline conditions.
Let’s begin by introducing the local elasticity of supply equation:
24
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
(𝑞!92 ∂𝑞! )
𝜖! = 92 ,
(𝑤 ∂𝑤)
where 𝑞! represents the local supply of nurses. This equation can be rearranged as follows:
∂𝑞!
𝑤 = 𝜖! 𝑞! .
∂𝑤
Moving forward, let’s delve into the differentiation of the function 𝑆& (𝑛! , 𝑤, 1) concerning 𝑤
and evaluating it using wage and employed staff nurses’ values before the pandemic. This yields:
#
∂𝑆& (𝑛! , 𝑤, 1) 𝑛! 𝑠! 𝑒 9%:
𝑤 = .
∂𝑤 √𝜋
In this context, 𝑤 h = 𝑠! cln(𝑤) − ln3𝑚! (1)4e. Moreover, based on the
h is defined as 𝑤
equation defining our sigmoid supply surface, 𝑤 h can also be expressed as erf 92 [2𝑆& (1, 𝑤, 1) −
1], where erf 92 is the inverse error function, and 𝑆& (1, 𝑤, 1) denotes the baseline proportion of
the supply of nurses from the qualified pool of nurses. We determined this baseline proportion to
be 84.1% based on a previous study (American Association of Colleges of Nursing, 2023b) This
allows us to compute the value of 𝑤 h and subsequently utilize it to determine the values of 𝑠! and
𝑚! (0). The expressions for these parameters are provided by:
2𝑞!
𝑤h = erf 92 i − 1j ,
𝑛!
#
𝜖! 𝑞! √𝜋𝑒 %:
𝑠! = ,
𝑛!
𝑚! = 𝑤𝑒 9%:/!! .
Thus, we successfully derive the values of 𝑚! (0) and 𝑠! , which are presumed to characterize
the pre-pandemic supply curve.
Informing the Model Describing the Dynamics of the Qualified Pool of Nurses
The component of the model that describes how the number of nurses in our qualified pool
changes due to recruitment and permanent retirement is influenced by three key factors: 𝜈, 𝜔2 ,
and 𝜅3 . These parameters determine the amount of new nurses entering and nurses leaving the
qualified pool.
Looking at the national context, we observe that there are approximately 155,000 new
registered nurse (RN) graduates each year (Salsberg, 2018). This number accounts for around 5
percent of the total employed nurse population of 3.07 million. To estimate the suitable value of
𝜈 for Virginia, we apply a similar growth rate that considers the portion of the supply within the
qualified pool. This suggests that 𝜈 for Virginia should fall between 3,500 and 7,000 new
graduates annually.
To gauge the rate at which nurses leave the qualified pool, we assume that this pool changes
"*
slowly and remains relatively steady ( ")! ≈ 0). We also assume that the rate of nurses leaving
the workforce supply, but not belonging to the qualified pool, ranges from being equal to the
25
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
leave rate of employed nurses, up to twice that rate. Consequently, we estimate that 𝜅3 varies
within the range of [1, 2].
By using our determined values for 𝜈 and 𝜅3 , along with the assumption of steady changes,
we can derive an estimated value for 𝜔2 . Within our analysis, we vary 𝜔2 between 0.75 percent
and 1.25 percent. This range encompasses the value that we found during our estimation process.
The Experimental Design for Model Calibration
Having established the model inputs, estimated their values, and detailed the data sources we
relied on for both estimation and uncertainty assessment, we move on to describing the process
of model calibration. This step involves leveraging the collected information to refine our
model’s performance. Calibration is a crucial step in optimizing our model’s performance. It
entails a systematic exploration of various combinations of parameter values within the
predefined uncertainty range. To achieve this, we conducted a total of 40,000 distinct parameter
value combinations, with each serving as an independent simulation, referred to as a “case run.”
This selection of parameter values and the quantity of case runs collectively define our
experimental approach. Our calibration strategy involves aligning the trajectories of output state
variables, particularly those related to nurse supply and wages, with our predefined calibration
time-series targets. These calibration targets are summarized in Table E.1 for reference.
Table E.1. Trajectories of Output Variable
Date Average Wage ($) Number Employed
Registered nurses
2020-04-05 74,380 66,450
2021-04-04 76,680 66,980
2022-04-03 81,860 69,510
Primary care providers
2020-04-05 159,021 15,080
2021-04-04 163,543 16,890
2022-04-03 155,095 16,880
Behavioral health specialists
2020-04-05 58,618 18,350
2021-04-04 63,301 18,560
2022-04-03 64,099 20,450
This comparison enables the identification of a subset of case runs whose trajectories most
closely align with our calibrated targets. Notably, many input parameter values were
intentionally selected with substantial uncertainty ranges around their estimated central values.
26
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
This deliberate choice provides the model with flexibility during the calibration process. While
numerous input parameters possess significant uncertainty ranges, they may have distinct
sampling schemes. These schemes can encompass either a uniform distribution spanning the
uncertainty range or a PERT distribution. The latter, known as the beta PERT distribution,
introduces a bias toward the provided central value or mode, reflecting desired probabilities.
Consequently, even when parameters have broad uncertainty ranges, if they are sampled with a
beta PERT distribution, most of the sampled values cluster around the central value.
Table E.2. Input Model Parameters, their Values, Ranges, and Sampling Methods
Variable Description Mode Lower Upper PDF
initial.demand (RN) Initial demand of nurses 68,200.000 59800.000 76,600.000 PERT
initial.supply (RN) Initial supply of nurses 66,400.000 NA NA Fixed
initial.demand (PCP) Initial demand of primary 15,793.735 13572.000 18,015.471 PERT
care providers
initial.supply (PCP) Initial supply of primary care 1,5080.000 NA NA Fixed
providers
initial.demand (BH) Initial demand of behavioral 21,671.000 16515.000 26,826.161 PERT
specialists
initial.supply (BH) Initial supply of behavioral 18,350.000 NA NA Fixed
specialists
supply.prop Initial proportion of 0.840 0.798 0.882 Uni
providers belonging to the
supply out of the whole
qualified pool of providers
initial.work.reluctance Initial work reluctance 1.000 NA NA Fixed
measure (defined in report);
the initial pre-pandemic
work reluctance is equal to
1 by definition
initial.yearly.wage (RN) Initial yearly wage of nurses 74,400.000 NA NA Fixed
as defined by the BLS
initial.yearly.wage (PCP) Initial yearly wage of 159,021.000 NA NA Fixed
primary care providers as
defined by the BLS
initial.yearly.wage (BH) Initial yearly wage of 58,618.000 NA NA Fixed
behavioral specialists as
defined by the BLS
num.staffed.bed Total staffed beds in all 15,200.000 NA NA Fixed
U.S. hospitals
bed.occup.rate Bed occupancy rate as 0.699 NA NA Fixed
defined by the curative care
hospital bed occupancy rate
in the U.S. from 1960 to
2019
covid.prop Proportion of all NA NA NA Derived
hospitalizations that are due
to COVID-19
27
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Variable Description Mode Lower Upper PDF
yearly.growth.demand Yearly growth in demand as 0.060 0.045 0.075 Uni
a proportion of the current
demand
patient.to.nurse.nonICU Initial patient-nurse ratio for 4.000 NA NA Fixed
non-ICU setting
patient.to.nurse.ICU Initial patient-nurse ratio for 2.000 NA NA Fixed
ICU setting
ICU.rate Proportion of all 0.121 NA NA Fixed
hospitalizations that require
the ICU
yearly.covid.prop.growth Yearly growth of COVID-19 0.350 0.250 0.400 Uni
cases as a proportion of
total hospitalizations
ICU.covid.rate Proportion of all COVID-19 0.198 NA NA Fixed
hospitalizations that require
the ICU
extra.effort.covid.nonICU Multiplicative scaling factor 1.500 NA NA Fixed
of the effort required by a
COVID-19 patient
extra.effort.covid.ICU Multiplicative scaling factor 2.000 NA NA Fixed
of the effort required by a
COVID-19 patient in the
ICU
LOS Average length of stay of 5.500 3.800 7.000 PERT
patients in a hospital
LOS.covid Average length of stay of 15.400 13.500 17.200 Uni
COVID-19 patients in a
hospital
𝑛! (RN) Yearly number of new 7,910.000 5930.000 9,890.000 Uni
graduating nurses that join
the supply of nurses under
baseline conditions
𝑛! (PCP) Yearly number of new 165.000 110.000 247.500 Uni
graduating primary care
providers that join the
supply of primary care
providers under baseline
conditions
𝑛! (BH) Yearly number of new 218.452 145.635 327.679 Uni
graduating behavioral
specialists that join the
supply of behavioral
specialists under baseline
conditions
𝜅" Multiplicative scaling factor 1.500 1.000 2.000 Uni
describing the increase rate
in workforce depletion
among nurses not
belonging to the supply
28
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Variable Description Mode Lower Upper PDF
𝜔# Yearly probability of exiting 0.010 0.007 0.013 Uni
and no longer participating
in the workforce among
those belonging to the
supply
yearly.growth.wage Net yearly percentage 0.020 0.000 0.100 PERT
increase in wages when the
demand for providers is
greater than the supply
after discounting for
inflation
reduced.yearly.growth.wage Yearly percentage increase 0.005 0.000 0.025 PERT
in wages when the supply
for providers is greater than
the demand
𝛽$ Proportion of the change in 0.500 0.425 0.575 Uni
work reluctance that is
attributable to increased
fatigue
𝛽% Proportion of the change in 0.300 0.255 0.345 Uni
work reluctance that is
attributable to increased
distress
𝛽& Proportion of the change in NA NA NA Derived
work reluctance that is
attributable to decreased
engagement
CF Proportion of the distress 0.500 0.375 0.625 Uni
that responds linearly to
increased demand
𝜓% Initial distress component 1.000 NA NA Fixed
attributable to the portion
that does not depend on
demand
𝜓& Initial engagement 1.000 NA NA Fixed
component attributable to
the portion that does not
depend on demand
tS The half-life parameter 0.250 0.020 1.000 PERT
represents the time-scale in
years in an exponential
weighted moving average
model, used to discount
past experiences in work
reluctance, indicating the
time it takes for the
weighting of past
observations to reduce to
half of its original value
𝜖 Elasticity of supply with 0.644 0.250 1.000 PERT
wages
29
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Variable Description Mode Lower Upper PDF
wage.prop The percentage of the 0.950 0.700 0.990 PERT
current wage needed to
maintain the existing
workforce if all factors
causing work reluctance
were eliminated
𝑧2 The level of the work 1.500 1.100 3.000 PERT
reluctance metric required
to reduce the nurse supply
to the value specified by
parameter “supply.prop2”
while keeping wages
unchanged
supply.prop_2 The value of the supply that 0.500 NA NA Fixed
will be achieved when the
work reluctance metric
reaches the level specified
by the parameter 𝑧2,
assuming there are no
changes in wages
NOTE: PCP = primary care physician.
Our input model parameters, their values, ranges, and sampling methods are summarized in
Table E.2. It is important to note that while some parameters in the table are fixed and not
altered, others undergo variation and are thus sampled. To efficiently create our experimental
designs encompassing a wide array of parameter values, we employ Latin Hypercube Sampling
(LHS) (Iman, Helton, and Campbell, 1981). LHS empowers the exploration of unique parameter
input combinations, effectively capturing the inherent variability and uncertainty within the
model. The LHS process involves dividing each parameter’s range into intervals and randomly
selecting a value from each interval. This strategy results in the creation of unique combinations
of parameter values. This approach ensures that each parameter is sampled across its entire
range, preserving the overall distribution and inter-parameter correlation. By carefully
constructing this sampling design, LHS facilitates a streamlined exploration of the parameter
space, providing valuable insights into how the simulation model behaves and performs. This
strategic approach enhances our ability to consider various parameter combinations that might
influence the model’s outcomes and behavior.
Pairing the LHS scheme with Bayesian inference methods offers an efficient iterative
calibration approach. In this context, Iterative Monte Carlo Approximate Bayesian Computation
(IMABC) methods can be employed, progressively refining parameter estimates through a
comparison of simulated and observed data, thereby enhancing model accuracy (Rutter et al.,
2019; Lima et al., 2021). In contrast, a single LHS calibration lacks this iterative refinement,
potentially leading to less precise parameter estimates and a limited ability to capture intricate
parameter-data relationships. However, constrained by time and resources, we chose a single
LHS method involving a large number of sampled case runs. Our approach calculates the error
for each model run trajectory based on target output alignment and then selects the top one
30
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
percentile of trajectories closest to the target values. This method efficiently identifies promising
model runs, streamlining calibration within our time constraints—although, in the future, the
consideration of IMABC methods remains a possibility.
Sensitivity Analysis
We systematically adjusted input values to discern the inputs wielding the most substantial
influence on the outcomes of interest (labor force supply and demand). This exploration of
various parameter values yielded insights into how different factors affect model outputs. The
process also illuminated interactions among parameters, enhancing our comprehension of the
model’s behavior.
Additionally, we produced an Excel-based tool that allows policymakers to more easily view
how adjusting scenario inputs affects outcomes. For example, the tool allows the user to toggle
between wage increase scenarios including baseline (2 percent real annual wage growth), 3
percent, and 4 percent real annual wage growth. By toggling between various scenarios and
immediately viewing the effects on outcomes of interest, the user gains better insight into how
sensitive the model is to each input.
Limitations and Extensions
In the realm of simulation modeling, simplifying assumptions and reliance on challenging-to-
estimate parameter values often become essential because of data limitations. Consequently,
model inputs frequently harbor significant uncertainty, emphasizing the need for sensitivity and
robustness analyses to gauge policy effectiveness across a variety of scenarios.
Despite its apparent mathematical intricacy, our model actually incorporates several
simplifications and disregards various complexities intrinsic to the nurse supply, retention, and
recruitment challenge. Acknowledging these limitations, many of which arise from deliberate
strategic simplifications to balance model accuracy and simplicity, we have engaged in
sensitivity analyses and devised experiments to span a range of potential outcomes. However, it
is crucial to acknowledge key constraints intrinsic to our model. Here we list a few examples of
the limitations of our model:
1. Our current simulation model focuses on retention and recruitment at a broad level, without
accounting for distinct health care worker population groups categorized by demographic
indicators or pertinent characteristics, such as age strata. This omission overlooks the impact
of aging nurses on wage trends and nurse supply dynamics, underscoring the need to stratify
the nurse population by age to better capture these effects.
2. The model primarily addresses the influx of new health care workers into the state, but it
doesn’t consider the inflows and outflows from neighboring states. To enhance the model, it
is essential to include factors such as perceived benefits and wages in Virginia compared
with nearby states, offering insights into where graduating health care workers opt to begin
their careers.
31
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
3. Because of temporal and data constraints, we calibrated the model at the state level and then
extending predictions to regional levels using the calibrated parameters. For more accuracy,
calibration should also be done at the regional level.
4. Our existing model simulates regions in isolation, disregarding interconnections between
them. For a more accurate representation, regions should be simulated together, enabling the
exploration of worker redistribution policies across different regions.
5. An expansion of the model could spotlight the potential gains from reallocating time spent on
paperwork to patient care, leading to improved patient and worker satisfaction. Further
exploration could delve into how efficiency changes and alterations in work types directly
and indirectly impact health care worker supply. Additionally, the concept of “output” in
health care should encompass improved health outcomes and care quality, transcending mere
patient numbers. Structural changes should prioritize increased output without overwhelming
health care providers, achieved by optimizing efficiency and work distribution.
32
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Appendix F. System Dynamics Model Results for Registered
Nurses
Model Experimental Design
This appendix focuses on the analysis of outputs from our dynamic model concerning nurses
in Virginia. We’ll complement our findings with informative plots derived from these outputs.
Our approach begins with conducting a sensitivity analysis and model calibration. For these
tasks we devise an experimental design by utilizing our parameter value uncertainty range and
generating 40,000 distinct case runs. Each case run encompasses a unique parameter
combination derived from our sampling approach. For every case, we simulate trajectories for
real wages, nurse supply, the qualified nurse pool, and health care work reluctance, serving as
our calibration phase and spanning from the pre-pandemic initial conditions on April 5, 2020, to
June 4, 2023. These wage and number of nurses employed outputs are then compared against our
calibration targets. The trajectories of output state variables, encompassing supply and wages,
are aligned with our calibration time-series targets.
Calibration Phase Trajectories
In a single iteration, we calculate the trajectory error for each model run, aligning with the
target outputs. We then narrow down our calibrated subset of case runs by selecting the top one
percentile of trajectories that closely match the target values. This efficient method enables us to
identify the most promising model runs, optimizing the calibration process within our time
limitations.
Figure F.1. displays the outcomes of our calibration phase trajectories. There are four outputs
showing the number employed, the qualified pool, the wage and the health care work reluctance.
We use a pink-to-purple scale to represent the distribution of model runs over time. The purple
bands indicate distributions closer to the median, while the pink bands represent distributions
further away from the median.
33
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure F.1 Full Calibration Phase Trajectories for Registered Nurses
Within the plot, we can observe blue curves representing selected sample trajectories and red
curves showcasing randomly chosen trajectories that align with our calibration targets. These
trajectories are considered “calibrated” because they closely align with the wage and employed
nurse calibration targets. These calibration targets are represented as black points on the plot, and
the black lines above and below the points indicate the calibration window. The calibration
window represents the tolerance we typically allow for a trajectory to pass through in terms of its
distance to the calibration target point.
In Figure F.2, we present the same results, but this time we only consider the set of calibrated
trajectories. The four outputs are again displayed, and the color scale now ranges from yellow to
orange, showcasing the distribution of calibrated model runs over time. The orange bands
represent distributions closer to the median, while the yellow bands represent distributions
further away from the median.
34
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure F.2. Final Calibrated Model Trajectories for Registered Nurses
Sensitivity Analysis
Before finalizing the subset of calibrated runs and proceeding to policy experiments, we
conduct a sensitivity analysis. This analysis explores how different parameter values influence
model outputs using our full set of 40,000 case runs. Here we summarize the main findings of
our analysis.
During the calibration phase, demand remains an exogenous input, leading to no significant
correlations with inputs aside from the initial demand value. The wage trajectory is mainly tied
to yearly wage growth and initial demand. An intriguing finding is the weak yet meaningful
negative correlation between wages and parameter 𝑧+ , indicating that as nurses become less
responsive to increased work reluctance, the necessity for wage hikes diminishes. The qualified
nurse pool is primarily influenced by initial demand and the influx of new nurses, holding more
sensitivity to new nurse inflow relative to workforce depletion, as suggested by its correlations
with 𝜔2 and 𝜔< . Health care work reluctance aligns with our behavioral model parameters,
including fatigue importance (𝛽6 ), fixed effect proportion (𝜉8 ), and the discount parameter (or
35
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
half-life 𝜏! ). A notable observation is its strong negative correlation with LOS, as shorter LOS
diminishes fatigue impact.
Supply dynamics involve multiple factors like wages, new nurse inflow, and parameters
influencing work reluctance. Notably, supply acts as a pivotal mediating output influenced by a
diverse range of inputs, shaping the dynamics of other outputs in our model.
Policy Experiments
Starting from our selected calibrated set of case runs, we conduct several policy experiments.
In the formulation of our policy intervention runs, the primary step entails procuring final state
variables from our calibrated cases, which then serve as initial conditions for our model
projections. Notably, we are not constrained to employ the initial conditions derived from the
calibration process’s final states. Alternatively, we can opt for distinct calibrated case run
parameters while initiating projections from the same initial condition. This approach
circumvents the use of diverse initial conditions generated from final states of selected calibrated
runs, which were determined based on historical data.
The policy experiments we conducted involve modifying specific input values that can be
controlled through policy interventions. We examine the following policy scenarios separately:
1. Reduced barriers: This scenario involved a 50 percent reduction in the impact of certain
barriers.
2. Increased recruitment rate: In this scenario, we amplified the value of the recruitment rate
ν by 25 percent, leading to a 1.25-fold annual increase in newly graduated nurses.
3. Elevated wage growth rate: Here, we boosted the wage growth rate α_d by 50 percent.
While the initial calibration assumed a 2 percent average annual real wage growth, this
scenario accelerates the growth to an average of 3 percent per year.
4. Multi-intervention: We concurrently implemented all three of the above interventions.
In conjunction with the three aforementioned policies, we execute the baseline scenario,
entailing no policy interventions and adhering to status quo conditions. To assess the potential
impacts of the specified policy adjustments on the qualified nurse pool and supply within the
context of Virginia over the forthcoming 15 years, we utilize projected trajectories as depicted
below. Projected trajectories serve as the basis to evaluate the ramifications of these policy
adjustments on Virginia’s nurse’s supply over the next 15 years.
Consolidating our outcome distribution analysis across all policies, we showcase a unified
visualization in Figure F.3. by selecting the most representative case run. This encompassing
figure comprises four plots, each representing an outcome metric, displaying trajectories under
different policies. This arrangement streamlines the assessment of policy impacts across output
metrics, fostering an intuitive grasp of their combined effects.
36
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure F.3. Comparative Trajectories of Policy Effects for Registered Nurses
Regional Analysis
Drawing from our calibrated case runs tailored to the state of Virginia, we adjusted the initial
conditions of state variables in alignment with the wage and population characteristics of each of
the state’s eight unique AHEC regions. Figure F.4. is a map of Virginia that delineates these
37
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
eight AHEC regions. This was achieved through proportional adjustments to parameters
associated with nurse counts, leveraging demographic data to create a model that effectively
encapsulates the distinct attributes of these AHEC regions.
We employed distinct initial conditions for each of the eight regions and conducted
independent policy model runs for each region. Focusing solely on the most representative case
runs, we constructed analogous plots for each of the eight regions. These plots, akin to the one
showcased in Figure F.5. for the entire state of Virginia, illustrate the ramifications of policy
interventions across various outputs. These region-specific plots are showcased in Figures F.6.-
F.7, each centered on one of the four outputs. Importantly, each of these figures encompasses
eight individual plots, each corresponding to a unique region under analysis. It is worth noting
that these simulations for the eight regions were executed independently, with outcomes in one
region holding no influence over the outcomes in the others. Because of this independence and
because the primary differences between the regions are related to demographics and wages, the
output plots for each region are essentially scaled versions of one another.
Figure F.4. Map of Virginia and How it is Divided into Eight Regions of Analysis
38
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure F.5. Results of Policy Interventions on Employment by AHEC Region for Registered Nurses
Figure F.6. Results of Policy Interventions on the Gap Between Supply and Demand by AHEC
Region for Registered Nurses
39
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Appendix G. System Dynamics Model Results on Primary Care
Providers
Model Experimental Design
This appendix contains our dynamic model’s results for Virginia’s primary care providers. It
is structured similarly to Appendix F, which covered nurse-related outputs, but in a more
streamlined manner. Like before, we complement our findings with enlightening plots derived
from these outcomes. However, we’ve chosen to exclude certain plot types that appeared in
Appendix F, prioritizing those that provide the most insightful information. Sensitivity analysis
plots are omitted as well. Our primary focus is on policy analysis plots for the representative case
runs, excluding plots displaying outcome bands and distributions across all case runs.
Following a similar methodology, our process starts with model calibration and experimental
design. We utilize parameter uncertainty ranges to generate 40,000 distinct case runs, each
representing a unique parameter combination. These cases simulate trajectories for real wages,
physician supply, qualified physician pool, and work reluctance. Calibration spans from pre-
pandemic conditions on April 5, 2020, to June 4, 2023. Output trajectories are compared with
calibration targets, aligning wage and employed primary care provider numbers with predefined
benchmarks. This ensures thorough alignment with calibration time-series targets.
Calibration Phase Trajectories
Figure G.1. illustrates the calibrated trajectories for the four outputs: number employed,
qualified pool, wage, and health care work reluctance. These trajectories are labeled “calibrated”
as they closely match the wage and employed primary care providers calibration targets. The
calibration targets are depicted as black points on the plot, with black lines above and below
indicating the calibration window. This window signifies the allowed tolerance for trajectories in
relation to the calibration target point’s distance.
A color scale ranging from yellow to orange is utilized, showcasing the distribution of
calibrated model runs over time. Orange bands represent distributions closer to the median, while
yellow bands signify distributions further away from the median.
40
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure G.1. Final Calibrated Model Trajectories for Primary Care Providers
Upon examining Figure G.1, a notable observation arises—our model struggles to effectively
track the four designated calibration points, two pertaining to primary care providers employed
and two relating to their wages. This disparity stems from the fact that our calibration targets
vividly illustrate a decline in physicians’ real wages during the pandemic, occurring between
2021 and 2022, while the count of employed primary care providers remained relatively constant
during the same period. However, our model doesn’t account for the scenario where real wages
experience a gradual decrease over time.
Furthermore, based on our initial conditions and calibration targets, it becomes evident that
the count of employed primary care providers exhibited a swift surge between 2020 and 2021,
followed by a consistent level between 2021 and 2022. Regrettably, our model fails to replicate
this rapid upswing in employed primary care providers between 2020 and 2021, instead fitting a
more gradual growth curve. Despite these challenges, we are not overly concerned about the
perceived noncalibration of our model because of this issue.
This perspective is founded on the understanding that the pandemic-induced conditions of
negative wage growth and the dynamics of primary care provider employment are unlikely to be
41
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
sustained or replicated in the medium to long term. Consequently, we proceed with confidence
by selecting the showcased set of case runs in Figure G.1. for our policy projections.
Policy Experiments
Repeating our approach for nurses described in Appendix F, we build upon our calibrated
case runs selection and perform policy experiments for primary care providers.
Projected trajectories serve as the basis to evaluate the ramifications of these policy
adjustments on Virginia’s primary care provider pool and supply over the next 15 years.
Consolidating our outcome distribution analysis across all policies, we showcase a unified
visualization in Figure G.2. by selecting the most representative case run. This encompassing
figure comprises four plots, each representing an outcome metric, displaying trajectories under
different policies. This arrangement streamlines the assessment of policy impacts across output
metrics, fostering an intuitive grasp of their combined effects.
42
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Commonwealth Analysis
Figure G.2. Comparative Trajectories of Policy Effects for Primary Care Providers
43
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Regional Analysis
Applying the method outlined in Appendix F, Figures G.3.–G.4. display region-specific plots
for primary care providers, each dedicated to one of the four outputs. Within each figure, you’ll
find eight distinct plots, each representing a specific region under analysis.
Figure G.3. Results of Policy Interventions on Employment by AHEC Region for Primary Care
Providers
44
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure G.4. Results of Policy Interventions on the Gap Between Supply and Demand by AHEC
Region for Primary Care Providers
45
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Appendix H. System Dynamics Model Results on Behavioral
Health Specialists
Model Experimental Design
This appendix examines outcomes generated by our dynamic model for behavioral specialists
in Virginia. It follows a streamlined structure similar to Appendix F, which detailed nurse-related
outputs. As in Appendix F, we utilize insightful plots derived from these outcomes, focusing on
informative plot types and excluding sensitivity analysis plots. Our focus is on presenting policy
analysis plots for representative case runs, excluding plots showing bands and distributions
across all case runs for each output.
We begin with model calibration and experimental design, using a parameter uncertainty
range to create 40,000 distinct case runs. Each run captures a unique parameter combination.
Within each case, we simulate trajectories for real wages, behavioral specialist supply, qualified
pool, and health care work reluctance. The calibration phase spans from pre-pandemic conditions
on April 5, 2020, to June 4, 2023.
During calibration, we compare output trajectories (e.g., supply and wages) against
designated targets. We specifically assess wage outcomes and the number of employed
behavioral specialists against predefined calibration benchmarks. This ensures alignment with
our calibration time-series targets comprehensively.
Calibration Phase Trajectories
Figure H.1. illustrates the calibrated trajectories for the four outputs: number employed,
qualified pool, wage, and health care work reluctance. These trajectories are labeled “calibrated”
as they closely match the wage and employed behavioral health specialists calibration targets.
The calibration targets are depicted as black points on the plot, with black lines above and below
indicating the calibration window. This window signifies the allowed tolerance for trajectories in
relation to the calibration target point’s distance.
A color scale ranging from yellow to orange is utilized, showcasing the distribution of
calibrated model runs over time. Orange bands represent distributions closer to the median, while
yellow bands signify distributions further away from the median.
46
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure H.1. Final Calibrated Model Trajectories for Health Behavioral Specialists
Policy Experiments
Repeating our approach for nurses described in Appendix F, we build upon our calibrated
case runs selection and perform policy experiments for behavioral specialists.
Projected trajectories serve as the basis to evaluate the ramifications of these policy
adjustments on Virginia’s behavioral specialist pool and supply over the next 15 years.
Consolidating our outcome distribution analysis across all policies, we showcase a unified
visualization in Figure H.2. by selecting the most representative case run. This encompassing
figure comprises four plots, each representing an outcome metric, displaying trajectories under
different policies. This arrangement streamlines the assessment of policy impacts across output
metrics, fostering an intuitive grasp of their combined effects.
47
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Commonwealth Analysis
Figure H.2. Comparative Trajectories of Policy Effects for Behavioral Health Specialists
48
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Regional Analysis
Applying the method outlined in Appendix F, Figures H.3.–H.4. display region-specific plots
for behavioral specialists, each dedicated to one of the four outputs. Within each figure, you’ll
find eight distinct plots, each representing a specific region under analysis.
Figure H.3. Results of Policy Interventions on Employment by AHEC Region for Behavioral Health
Specialists
49
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Figure H.4. Results of Policy Interventions on the Gap Between Supply and Demand by AHEC
Region for Behavioral Specialists
50
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
Abbreviations
AHEC Area Heath Education Centers
AI artificial intelligence
BLS U.S. Bureau of Labor Statistics
COVID-19 coronavirus disease 2019
EWMA exponential weighted moving average
LHS Latin Hypercube Sampling
PDF probability density function
RN registered nurse
51
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
References
Abate, Semagn Mekonnen, Siraj Ahmed Ali, Bahiru Mantfardo, and Bivash Basu, “Rate of
Intensive Care Unit Admission and Outcomes Among Patients with Coronavirus: A
Systematic Review and Meta-Analysis,” PLoS One, Vol. 15, No. 7, 2020.
Alimohamadi, Yousef, Elahe Mansouri Yekta, Mojtaba Sepandi, Maedeh Sharafoddin, Maedeh
Arshadi, and Elahe Hesari, “Hospital Length of Stay for COVID-19 Patients: A Systematic
Review and Meta-Analysis,” Multidisciplinary Respiratory Medicine, Vol. 17, No. 1, 2022.
American Association of Colleges of Nursing, “Nursing Shortage Fact Sheet,” July 2023a. As of
August 17, 2023:
https://www.aacnnursing.org/news-data/fact-sheets/nursing-shortage
American Association of Colleges of Nursing, “Nursing Workforce Fact Sheet,” July 2023b. As
of September 15, 2023:
https://www.aacnnursing.org/news-data/fact-sheets/nursing-workforce-fact-sheet
American Hospital Association, “Fast Facts on U.S. Hospitals, 2023,” 2023. As of August 17,
2023:
https://www.aha.org/statistics/fast-facts-us-hospitals
Berlin, Gretchen, Meredith Lapointe, and Mhoire Murphy, “Surveyed Nurses Consider Leaving
Direct Patient Care at Elevated Rates,” McKinsey & Company, February 17, 2022.
Haines, Julia, “The State of the Nation’s Nursing Shortage,” U.S. News & World Report,
November 1, 2022.
Hindmarsh, Alan C., and Linda R. Petzold, “Algorithms and Software for Ordinary Differential
Equations and Differential‐Algebraic Equations, Part II: Higher‐Order Methods and Software
Packages,” Computers in Physics, Vol. 9, No. 2, March 1995.
Institute of Medicine, Finding What Works in Health Care: Standards for Systematic Reviews,
National Academies Press, 2011.
Michas, Frédéric, “Hospital Occupancy Rate U.S. 1960–2019,” Statista, October 6, 2022.
Salsberg, Edward S., “Changes in the Pipeline of New NPs and RNs: Implications for Health
Care Delivery and Educational Capacity,” Health Affairs Forefront, June 5, 2018.
Soetaert, Karline, Thomas Petzoldt, and R. Woodrow Setzer, “Solving Differential Equations in
R: Package Desolve,” Journal of Statistical Software, Vol. 33, No. 9, 2010.
52
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
U.S. Department of Health and Human Services, Health Resources and Services Administration,
Bureau of Health Workforce, and National Center for Health Workforce Analysis, Supply
and Demand Projections of the Nursing Workforce: 2014–2030, July 21, 2017.
53
PREPUBLICATION COPY. [DATE: SEPTEMBER 29, 2023]
File and source
- File
- 2023-RD481.pdf
- Size
- 4,400,809 bytes
- SHA-256
- c2fb4564b90e2515a4f72eaa1c7a404b018df232e166172220dfea852871c920
- Our copy
- 2023-RD481.pdf
- Original
- No public link identified.