Full text
Department of Health and Human Services
OFFICE OF
INSPECTOR GENERAL
CMS’S COVID-19 DATA INCLUDED
REQUIRED INFORMATION FROM
THE VAST MAJORITY OF NURSING
HOMES, BUT CMS COULD TAKE
ACTIONS TO IMPROVE COMPLETENESS
AND ACCURACY OF THE DATA
Christi A. Grimm
Principal
Deputy Inspector General
September 2021
A-09-20-02005
Inquiries about this report may be addressed to the Office of Public Affairs at
Public.Affairs@oig.hhs.gov.
Office of Inspector General
https://oig.hhs.gov
The mission of the Office of Inspector General (OIG), as mandated by Public Law 95-452, as amended, is
to protect the integrity of the Department of Health and Human Services (HHS) programs, as well as the
health and welfare of beneficiaries served by those programs. This statutory mission is carried out
through a nationwide network of audits, investigations, and inspections conducted by the following
operating components:
Office of Audit Services
The Office of Audit Services (OAS) provides auditing services for HHS, either by conducting audits with
its own audit resources or by overseeing audit work done by others. Audits examine the performance of
HHS programs and/or its grantees and contractors in carrying out their respective responsibilities and are
intended to provide independent assessments of HHS programs and operations. These audits help reduce
waste, abuse, and mismanagement and promote economy and efficiency throughout HHS.
Office of Evaluation and Inspections
The Office of Evaluation and Inspections (OEI) conducts national evaluations to provide HHS, Congress,
and the public with timely, useful, and reliable information on significant issues. These evaluations focus
on preventing fraud, waste, or abuse and promoting economy, efficiency, and effectiveness of
departmental programs. To promote impact, OEI reports also present practical recommendations for
improving program operations.
Office of Investigations
The Office of Investigations (OI) conducts criminal, civil, and administrative investigations of fraud and
misconduct related to HHS programs, operations, and beneficiaries. With investigators working in all 50
States and the District of Columbia, OI utilizes its resources by actively coordinating with the Department
of Justice and other Federal, State, and local law enforcement authorities. The investigative efforts of OI
often lead to criminal convictions, administrative sanctions, and/or civil monetary penalties.
Office of Counsel to the Inspector General
The Office of Counsel to the Inspector General (OCIG) provides general legal services to OIG, rendering
advice and opinions on HHS programs and operations and providing all legal support for OIG’s internal
operations. OCIG represents OIG in all civil and administrative fraud and abuse cases involving HHS
programs, including False Claims Act, program exclusion, and civil monetary penalty cases. In
connection with these cases, OCIG also negotiates and monitors corporate integrity agreements. OCIG
renders advisory opinions, issues compliance program guidance, publishes fraud alerts, and provides
other guidance to the health care industry concerning the anti-kickback statute and other OIG enforcement
authorities.
Notices
THIS REPORT IS AVAILABLE TO THE PUBLIC
at https://oig.hhs.gov
Section 8M of the Inspector General Act, 5 U.S.C. App., requires
that OIG post its publicly available reports on the OIG website.
OFFICE OF AUDIT SERVICES FINDINGS AND OPINIONS
The designation of financial or management practices as
questionable, a recommendation for the disallowance of costs
incurred or claimed, and any other conclusions and
recommendations in this report represent the findings and
opinions of OAS. Authorized officials of the HHS operating
divisions will make final determination on these matters.
Report in Brief
Date: September 2021
Report No. A-09-20-02005
CMS’s COVID-19 Data Included Required Information
From the Vast Majority of Nursing Homes, but CMS
Could Take Actions To Improve Completeness and
Accuracy of the Data
What OIG Found
CMS’s COVID-19 data for nursing homes included the required data from the
vast majority of nursing homes (e.g., the number of confirmed COVID-19 cases
among residents); however, the data were not complete or accurate for some
nursing homes. Specifically, for 775 of the 15,388 nursing homes (about
5 percent), CMS’s COVID-19 data: (1) did not include all of the COVID-19 data
that nursing homes were required to report and (2) were not complete or
accurate after CMS had performed its quality assurance checks (e.g., the
number of confirmed COVID-19 cases among residents may have been under-
or overreported). These conditions occurred because, in part, CMS’s quality
assurance checks were not always effective in ensuring the accuracy and
completeness of the COVID-19 data for nursing homes.
In addition, we identified two areas in which CMS could take additional actions
to help ensure that its COVID-19 data are complete and accurate. First, CMS
could provide technical assistance to all nursing homes that fail its quality
assurance checks. Second, CMS could make additional efforts to ensure that:
(1) CMS’s and States’ COVID-19 data elements (e.g., confirmed COVID-19 cases
among residents) are comparable (i.e., CMS and States could use the same
data elements) and (2) the reported data are not substantially different.
When CMS’s COVID-19 data are complete and accurate, Federal and State
officials and other stakeholders may be able to more effectively monitor trends
in infection rates and develop public health policies when making decisions
about how to ensure the health and safety of nursing home residents and staff.
What OIG Recommends and CMS Comments
We recommend that CMS assess the costs and benefits of implementing the
six recommendations listed in our report (e.g., our recommendations that it
revise its quality assurance checks and contact nursing homes that fail quality
assurance checks to verify the accuracy of reported data or to correct
inaccurate data), and if CMS determines that the benefits outweigh the costs,
take action to implement the recommendations.
CMS concurred with three of our recommendations but did not concur with
the other three recommendations. After reviewing CMS’s comments, we
maintain that our recommendations are valid.
Why OIG Did This Audit
The United States currently faces a
nationwide public health emergency
because of the COVID-19 pandemic.
Federal regulations, effective May 8,
2020, required nursing homes to
report COVID-19 information, such as
the number of confirmed COVID-19
cases among residents, at least
weekly to the Centers for Disease
Control and Prevention’s (CDC’s)
National Healthcare Safety Network.
Each week, CDC aggregates the
reported information and sends the
data to the Centers for Medicare
& Medicaid Services (CMS) for
posting to the CMS website. These
data are used to assist with national
surveillance of COVID-19 in nursing
homes and to support actions to
protect the health and safety of
nursing home residents.
Our objective was to determine
whether CMS’s COVID-19 data for
nursing homes were complete and
accurate.
How OIG Did This Audit
Our audit covered CMS’s COVID-19
data for 15,388 nursing homes that
reported data to CMS for the 4-week
period from June 22 through July 19,
2020. We interviewed CMS officials
to understand the reporting process
for COVID-19 data, reviewed CMS’s
COVID-19 data and quality assurance
process, selected a statistical sample
of 120 nursing homes and sent
questionnaires to each sampled
nursing home, and sent
questionnaires to State health
departments in 50 States and the
District of Columbia.
The full report can be found at https://oig.hhs.gov/oas/reports/region9/92002005.asp.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
TABLE OF CONTENTS
INTRODUCTION ............................................................................................................................... 1
Why We Did This Audit ....................................................................................................... 1
Objective ............................................................................................................................. 1
Background ......................................................................................................................... 1
COVID-19 Public Health Emergency ....................................................................... 1
CMS’s Oversight of Nursing Homes ........................................................................ 2
Federal Requirements for Nursing Homes To Report COVID-19 Data ................... 2
Process for Nursing Homes’ Reporting of COVID-19 Data and CMS’s Analysis
and Posting of the Data ........................................................................................ 3
How We Conducted This Audit ........................................................................................... 7
FINDINGS ......................................................................................................................................... 9
CMS’s COVID-19 Data Did Not Include All of the Data That Nursing Homes
Were Required to Report ............................................................................................... 10
CMS’s COVID-19 Data Were Not Complete or Accurate After CMS Had Performed
Its Quality Assurance Checks .......................................................................................... 11
CMS Excluded Data for Some Nursing Homes From the COVID-19 Data Each
Week During Our Audit Period .......................................................................... 12
CMS’s COVID-19 Data Included Data for Nursing Homes That Reported More
Total COVID-19 Deaths Than Total COVID-19 Cases and Admissions ............... 14
CMS’s COVID-19 Data Included Data for Nursing Homes That Reported More
Total COVID-19 Deaths Than Total Deaths From Any Cause ............................. 16
CMS Could Take Additional Actions To Help Ensure That the COVID-19 Data for
Nursing Homes Are Complete and Accurate ................................................................. 17
Technical Assistance Is Needed for All Nursing Homes That Fail CMS’s
Quality Assurance Checks .................................................................................. 17
Additional Efforts Are Needed To Ensure That CMS’s and States’ COVID-19 Data
Elements Are Comparable and That Data Are Not Substantially Different ....... 18
CONCLUSION ................................................................................................................................. 21
RECOMMENDATIONS ................................................................................................................... 22
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
CMS COMMENTS AND OFFICE OF INSPECTOR GENERAL RESPONSE ........................................... 23
Third Recommendation: Revise Quality Assurance Checks for Ratios 2 and 4 ................ 23
CMS Comments ..................................................................................................... 23
Office of Inspector General Response .................................................................. 23
Fourth Recommendation: Identify Nursing Homes That Reported More COVID-19
Deaths Than Deaths From Any Cause ............................................................................ 24
CMS Comments ..................................................................................................... 24
Office of Inspector General Response .................................................................. 24
Sixth Recommendation: Determine the Feasibility of Using Comparable Data Elements
To Collect COVID-19 Data............................................................................................... 24
CMS Comments ..................................................................................................... 24
Office of Inspector General Response .................................................................. 24
APPENDICES
A: Audit Scope and Methodology ..................................................................................... 25
B: State Health Departments’ COVID-19 Data Related to Cases and Deaths Among
Nursing Home Residents and Staff ............................................................................. 28
C: CMS Comments ............................................................................................................ 34
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
1
INTRODUCTION
WHY WE DID THIS AUDIT
The United States currently faces a nationwide public health emergency because of the
COVID-19 pandemic. Federal regulations, effective May 8, 2020, required nursing homes to
report COVID-19 information, such as the number of confirmed COVID-19 cases among
residents, at least weekly to the Centers for Disease Control and Prevention’s (CDC’s) National
Healthcare Safety Network (NHSN).1 Each week, CDC aggregates the reported information and
sends the data to the Centers for Medicare & Medicaid Services (CMS) for posting to the CMS
website.2 These publicly available data are used to assist with national surveillance of COVID-19
in nursing homes and to support actions to protect the health and safety of nursing home
residents. As of February 28, 2021, CMS’s COVID-19 data showed that 640,271 nursing home
residents had had confirmed cases of COVID-19 and 130,174 residents had died from
COVID-19.3
Because Federal and State officials use COVID-19 data to monitor trends in infection rates and
develop public health policies when making decisions to ensure the health and safety of nursing
home residents and staff, we assessed the risk that CMS’s COVID-19 data for nursing homes
could be incomplete or inaccurate. We conducted this audit to provide a snapshot of the data’s
completeness and accuracy for the 4-week period from June 22 through July 19, 2020 (audit
period).
OBJECTIVE
Our objective was to determine whether CMS’s COVID-19 data for nursing homes were
complete and accurate.
BACKGROUND
COVID-19 Public Health Emergency
COVID-19 is a disease caused by a highly contagious coronavirus, called SARS-CoV-2.
COVID-19’s symptoms include fever, fatigue, cough, and shortness of breath. The disease is
1 In the memo “Interim Final Rule Updating Requirements for Notification of Confirmed and Suspected COVID-19
Cases Among Residents and Staff in Nursing Homes” (QSO-20-29-NH), issued May 6, 2020, CMS required nursing
homes to report COVID-19 data starting May 11, 2020, but made reporting of the data before May 11, 2020,
optional.
2 COVID-19 Nursing Home Data, available at https://data.cms.gov/stories/s/COVID-19-Nursing-Home-Data/bkwz-
xpvg. Accessed on Mar. 16, 2021. In this report, we refer to these data as “CMS’s COVID-19 data for nursing
homes.”
3 These data may not include the numbers of nursing home residents who had confirmed cases of COVID-19 and
had died from COVID-19 before May 11, 2020.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
2
fatal in some cases. Older adults and people who have severe underlying medical conditions,
such as heart or lung disease or diabetes, are at higher risk for developing more serious
complications from COVID-19.
On March 11, 2020, the World Health Organization characterized COVID-19 as a pandemic,
which is an epidemic that has spread over several countries or continents, usually affecting
many people.4 On March 13, 2020, then President Trump declared the COVID-19 outbreak a
national emergency. As of August 17, 2021, CDC had reported 37 million confirmed cases of
COVID-19 and 620,493 deaths from COVID-19 in the United States.
CMS’s Oversight of Nursing Homes
The Medicare and Medicaid programs cover care in skilled nursing and nursing facilities (i.e.,
nursing homes), respectively, for eligible beneficiaries in need of nursing services, specialized
rehabilitation services, medically related social services, pharmaceutical services, and dietary
services. CMS is responsible for overseeing nursing homes’ compliance with Medicare and
Medicaid standards for ensuring the health and safety of nursing home residents by enforcing
the standards required to help each resident attain or maintain the highest level of well-being.5
CMS delegates many of its responsibilities to State agencies (i.e., State health departments) and
issues guidance, such as State Survey Agency Directors memos, to State agencies.6
As of January 1, 2021, there were approximately 15,300 nursing homes in the United States.
These nursing homes had 1.6 million licensed beds, of which 1.2 million were occupied by
residents.
Federal Requirements for Nursing Homes To Report COVID-19 Data
To ensure that America’s health care facilities (e.g., nursing homes) are prepared to respond to
the COVID-19 public health emergency, on May 6, 2020, CMS issued to State survey agency
directors the memo “Interim Final Rule Updating Requirements for Notification of Confirmed
and Suspected COVID-19 Cases Among Residents and Staff in Nursing Homes” (QSO-20-29-NH).
The memo stated that: (1) nursing homes must submit COVID-19 data to CDC by May 17, 2020,
beginning with the week of May 11 through May 17, 2020, and continue to submit the data
weekly; (2) CMS would grant a 2-week grace period to submit the data, followed by a warning
letter in the third week; and (3) CMS would begin imposing civil monetary penalties (CMPs) on
4 An epidemic refers to an increase, often sudden, in the number of cases of a disease above what is normally
expected in a population in a specific area.
5 Sections 1819 and 1919 of the Social Security Act provide that nursing homes participating in the Medicare and
Medicaid programs, respectively, must meet Federal participation requirements, such as those related to quality of
care, nursing services, and infection control.
6 A State survey agency is the entity responsible for conducting most surveys of nursing homes to certify their
compliance with CMS’s Medicare and Medicaid participation requirements.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
3
nursing homes that had not reported data by June 7, which was the end of the fourth reporting
week after May 17, 2020.7
On May 8, 2020, CMS published an interim final rule that added Federal regulations requiring
nursing homes to report information (i.e., data) on confirmed and suspected COVID-19 cases
and other data, such as the supply of personal protective equipment (PPE), to CDC and to
nursing home residents, their representatives, and their families.8, 9 The interim final rule
stated that these data would be used to monitor trends in infection rates and develop public
health policies.
On September 2, 2020, CMS’s interim final rule codified the use of CMPs for each week that a
nursing home failed to report COVID-19 data. In the Federal Register, CMS stated:
We believe that a completely transparent CMP structure will help deter
noncompliance, encourage timely reporting, and eliminate possible gaps in
reporting that could hinder the government’s response to the [public health
emergency] for COVID-19 in specific geographic areas. For example, depending
on the circumstances, the failure of one facility to report COVID-19 cases on a
timely basis could delay our ability to detect and respond to an emerging
COVID-19 hot spot.10
The interim final rule stated that the amount of the CMP begins at a minimum of $1,000 for the
first occurrence of noncompliance with reporting requirements and increases by $500 for each
subsequent time that a nursing home fails to report the required COVID-19 data. (The
maximum allowable CMP amount is $6,500 per citation.) The interim final rule also stated that
the reporting requirements would be assessed weekly and that the regulation would continue
to be in effect for up to 1 year after the end of the public health emergency.
Process for Nursing Homes’ Reporting of COVID-19 Data and CMS’s Analysis and
Posting of the Data
Nursing homes report COVID-19 data to CDC, CDC performs quality assurance checks to reduce
misrepresentation and inaccuracies in the data, and CDC sends the aggregated data to CMS.
CMS performs quality assurance checks on the data and posts the data to its website. CMS
7 A CMP is a monetary penalty that CMS may impose against a nursing home for either the number of days or for
each instance that the nursing home is not in substantial compliance with one or more Medicare and Medicaid
participation requirements for long-term care facilities.
8 PPE is used by health care personnel to protect themselves, patients, and others when providing care. PPE
includes face masks, gloves, and gowns.
9 85 Fed. Reg. 27550, 27601 (May 8, 2020) and 42 CFR § 483.80(g).
10 85 Fed. Reg. 54820, 54825 (Sept. 2, 2020) and 42 CFR § 488.447.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
4
provides technical assistance to nursing homes that failed quality assurance checks so that they
can verify or correct the reported data.11 The following sections describe the process in detail.
Nursing Homes’ Reporting of COVID-19 Data to CDC
Nursing homes are required to report COVID-19 data to CDC at least weekly through CDC’s
NHSN system.12, 13 To support the Nation’s response to COVID-19, CDC introduced a new
COVID-19 module (a reporting section in the NHSN system) for long-term care facilities,
including nursing homes.14 This module allows nursing homes to electronically report COVID-19
data in four areas: (1) resident impact and facility capacity (e.g., the number of residents with
confirmed or suspected cases of COVID-19), (2) staff and personnel impact (e.g., the number of
staff with confirmed or suspected cases of COVID-19), (3) supplies and PPE (e.g., whether the
nursing home has at least a 1-week supply of face masks and gloves), and (4) ventilator capacity
and supplies.15 To respond to a question in the COVID-19 module, a nursing home provides
either a count (i.e., a specific number) or a Yes or No response. In this report, we refer to each
question as a “data element.”
Figure 1 on the following page shows the resident impact and facility capacity area in the
COVID-19 module for our audit period and the various data elements for this area.
11 CDC also provides technical assistance to nursing homes that failed quality assurance checks. See footnote 19.
12 The NHSN is a secure, internet-based system that provides health care facilities (such as nursing homes), States,
and the Nation with data needed to identify problem areas, measure the progress of prevention efforts, and
ultimately eliminate health-care-associated infections.
13 Individual nursing homes or groups (e.g., State and local health departments) may report COVID-19 information.
During our audit period, five State health departments (for California, Idaho, Maryland, Michigan, and Washington)
stated that they reported COVID-19 information through the NHSN system on behalf of nursing homes.
14 Long-term care facilities provide a variety of medical and personal care services to people who are unable to live
independently.
15 During our audit period, a resident could have been included in the number of residents with suspected cases of
COVID-19 in one week and included in the number of residents with confirmed cases of COVID-19 in another week.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
5
Figure 1: Resident Impact and Facility Capacity Area in the COVID-19 Module and
the Various Data Elements
Each week, CDC extracts the reported data from the NHSN system, performs quality assurance
checks to reduce misrepresentation and inaccuracies in the public reporting of these data, and
compiles and sends the aggregated data to CMS.16
CMS’s Quality Assurance Checks Related to Nursing Homes’ Reported COVID-19 Data
After CMS receives the CDC data each week, CMS analyzes the data (e.g., by checking whether
all Medicare-certified nursing homes reported the required data through the NHSN system) and
performs quality assurance checks to reduce misrepresentation and inaccuracies in the public
reporting of these data.
For each nursing home, CMS aggregates the data reported by the nursing home each week for
each data element. CMS also adds a new data element representing the cumulative total to
date for certain existing data elements, such as the number of residents with confirmed cases
of COVID-19.17 CMS uses these cumulative totals in performing its quality assurance checks.
16 During our audit period, CDC had three quality assurance checks. For example, if a nursing home entered the
same number (greater than zero) of new weekly confirmed COVID-19 cases for 3 or more consecutive weeks, CDC
considered the nursing home to have failed the quality assurance checks and excluded the data for weekly
confirmed COVID-19 cases before submitting the aggregated data to CMS.
17 According to CMS, it determines the cumulative total for a data element by using a nursing home’s data going
back to January 1, 2020, if those data are available (not May 11, 2020, the date when nursing homes were required
to report the data).
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
6
These checks involve comparing certain ratios (described below) related to resident and staff
COVID-19 data with specific ratio thresholds to determine whether the nursing home may have
reported inaccurate data.
Figure 2 shows the six quality assurance ratios that CMS used for its quality assurance checks.
For example, for ratio 1, CMS compares the number of total confirmed resident cases of
COVID-19 plus the number of total admissions with COVID-19 against the number of all beds in
the facility. If the ratio is greater than or equal to the ratio’s threshold (i.e., 1.5 for ratio 1), CMS
updates its COVID-19 data by excluding all of the nursing home’s data for the week from the
aggregated data it posts to its website. Based on CMS’s analysis of data and feedback from
subject matter experts (e.g., CDC representatives), CMS determined that it was improbable for
a nursing home to have 1.5 times or more the number of total confirmed resident cases of
COVID-19 plus total COVID-19 admissions than the number of all beds in the facility. Except for
the number of all beds in the denominators of ratios 1, 3, 5, and 6, all the totals in the
numerators and denominators of the six quality assurance ratios are cumulative totals.
Figure 2: CMS’s Six Quality Assurance Ratios18
18 As of January 24, 2021, CMS had revised its quality assurance checks. The threshold for ratio 1 was increased to
greater than or equal to 1.75. The threshold for ratio 6 was increased to greater than 2.0 if the number of all beds
was 50 or fewer or greater than 1.25 if the number of all beds was greater than 50. CMS also changed the
exception to ratios 1 through 3 to by replacing the word “and” with “or” in the exception description: “or fewer
than 25 total confirmed COVID-19 cases among residents.”
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
7
CMS’s Posting of COVID-19 Data to Its Website and CMS’s Technical Assistance to
Nursing Homes
After performing quality assurance checks, CMS compiles updated data for the nursing homes
and posts these data to its website. CMS also provides technical assistance by performing
periodic outreach activities by phone and email to nursing homes that failed CMS’s quality
assurance checks and had their data excluded from CMS’s COVID-19 data.19 If a nursing home
verifies that its data were accurate or if it corrects its data, CMS adds the nursing home’s data
to the COVID-19 data for all nursing homes.
Figure 3 summarizes the reporting process for COVID-19 data for nursing homes, from when
they report to CDC their COVID-19 data to when CMS posts the data to its website.
Figure 3: Reporting Process for COVID-19 Data for Nursing Homes
HOW WE CONDUCTED THIS AUDIT
Our audit covered CMS’s COVID-19 data for 15,388 nursing homes that reported data to CMS
through CDC’s NHSN system for the 4-week period from June 22 through July 19, 2020.20 The
19 According to CDC, it also performed periodic outreach activities to nursing homes that failed CDC’s quality
assurance checks. After CDC confirmed that a nursing home had corrected its data, CDC marked the corrected
data for the nursing home to show that it had passed the quality assurance checks and later provided the
corrected data to CMS. When CMS received the corrected data, it added the data to its COVID-19 data for all
nursing homes.
20 We obtained the COVID-19 data from CMS’s website at https://data.cms.gov/Special-Programs-Initiatives-
COVID-19-Nursing-Home/COVID-19-Nursing-Home-Dataset/s2uc-8wxp. Accessed on July 30, 2020.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
8
data consisted of 61,546 records; each record represented the COVID-19 data submitted by a
nursing home for 1 week of the 4-week audit period.
We interviewed CMS officials to understand the reporting process for COVID-19 data, the
quality assurance process for verifying the accuracy and completeness of the data, and how the
data were used. We also obtained information from CDC officials regarding nursing homes’
reporting of COVID-19 data through the NHSN system and CDC’s process for sending the data to
CMS. To determine whether CMS’s COVID-19 data for nursing homes were complete and
accurate, we reviewed those data for our audit period and other data and reports, including
raw (i.e., unmodified) data that nursing homes reported through CDC’s NHSN system and
weekly reports that contained the results of CMS’s quality assurance checks.
To understand how nursing homes reported COVID-19 data through the NHSN system (e.g.,
whether nursing homes reported data for all data elements for each week during our audit
period), we selected a statistical sample of 120 nursing homes across the Nation and sent a
questionnaire to each sampled nursing home. In addition, to understand nursing homes’
reporting processes at the State level, we sent questionnaires to State health departments in
50 States and the District of Columbia to identify those that collected COVID-19 data from
nursing homes and posted the data publicly on their websites.
After learning that 34 of these State health departments required nursing homes to report
COVID-19 data to them, we requested that these health departments provide information for
the same data elements that CMS obtained through CDC’s NHSN system (CMS’s data elements),
i.e., the numbers of COVID-19 cases and deaths among residents and staff for each week of our
audit period.21 We compiled the information provided by the State health departments.22 We
then compared the State health departments’ information with CMS’s COVID-19 data to
determine: (1) whether the States’ data elements, such as the number of confirmed COVID-19
cases among residents, were comparable to (i.e., the same as) CMS’s data elements23 and
(2) for those data elements that were comparable, whether the data that States reported (e.g.,
21 Of the 34 State health departments, 33 responded to our request for COVID-19 information. However, not all
33 State health departments responded regarding all of the data elements, which was caused, in part, by our
request that they not provide information on any data elements that reflected data collected from long-term care
facilities other than nursing homes.
22 We did not verify the information that State health departments provided (i.e., by reviewing supporting data or
documentation).
23 Because many of the data elements in CMS’s COVID-19 data differed from what the States used, CMS’s data and
the States’ data could not always be directly compared. For example, CMS’s COVID-19 data included a data
element for the number of resident deaths from both confirmed and suspected cases of COVID-19, but some
States required nursing homes to report the number of resident deaths from only confirmed cases of COVID-19.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
9
the specific count for the number of confirmed COVID-19 cases among residents) were
substantially different from CMS’s data.24
We conducted this performance audit in accordance with generally accepted government
auditing standards. Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions
based on our audit objectives. We believe that the evidence obtained provides a reasonable
basis for our findings and conclusions based on our audit objectives.
Appendix A contains the details of our audit scope and methodology. Appendix B contains
tables showing State health departments’ COVID-19 data related to cases and deaths among
nursing home residents and staff.
FINDINGS
CMS’s COVID-19 data for nursing homes included the required data from the vast majority of
nursing homes (e.g., the number of confirmed COVID-19 cases among residents); however, the
data were not complete or accurate for some nursing homes.25 Specifically, for 775 of the
15,388 nursing homes (about 5 percent), the data: (1) did not include all of the COVID-19 data
that nursing homes were required to report26 and (2) were not complete or accurate after CMS
had performed its quality assurance checks (e.g., the number of confirmed COVID-19 cases
among residents may have been under- or overreported).27 These conditions occurred
because: (1) CMS had limited resources to ensure that every nursing home reported all of the
required COVID-19 data and (2) CMS’s quality assurance checks were not always effective in
ensuring the accuracy and completeness of the COVID-19 data for nursing homes.
In addition, we identified two areas in which CMS could take additional actions to help ensure
that its COVID-19 data are complete and accurate. First, CMS could provide technical
24 If the difference between a State’s data and CMS’s data for a given data element was 20 percent or greater, we
determined that the State’s data were substantially different from CMS’s data. For example, if the total number of
confirmed COVID-19 cases among residents for our audit period was 1,000 in a State’s data and 790 in CMS’s data,
we determined that these data were substantially different because 1,000 is 27 percent greater than 790.
25 Of the 15,388 nursing homes, 13,890 nursing homes (about 90 percent) reported all of the required data (i.e.,
these nursing homes reported data for all of the data elements) each week during our audit period. The remaining
1,498 nursing homes (about 10 percent) did not report the required data for 1 or more of the data elements each
week; however, we did not include all of these nursing homes in our findings. We did not include a nursing home’s
data in our findings if we determined that CMS took action to ensure that the COVID-19 data were complete and
accurate, such as imposing a CMP for a nursing home that did not report resident and staff data.
26 For the purpose of this report, we determined that a nursing home did not report all of the required information
if it did not provide data for the data elements related to: (1) the number of COVID-19 cases and deaths among
residents and staff and (2) PPE and supplies.
27 The total number of nursing homes identified in our findings is greater than 775 because some nursing homes
were included in more than 1 finding. These nursing homes could have corrected the data after our audit period.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
10
assistance to all nursing homes that fail its quality assurance checks. Second, CMS could make
additional efforts to ensure that: (1) CMS’s and States’ COVID-19 data elements are comparable
(i.e., CMS and States could use the same data elements) and (2) the reported data are not
substantially different.
When CMS’s COVID-19 data are complete and accurate, Federal and State officials and other
stakeholders may be able to more effectively monitor trends in infection rates and develop
public health policies when making decisions about how to ensure the health and safety of
nursing home residents and staff. Incomplete or inaccurate data could delay CMS’s ability to
detect and respond to an emerging COVID-19 hotspot, such as a surge or resurgence of
COVID-19 cases in a community.
CMS’S COVID-19 DATA DID NOT INCLUDE ALL OF THE DATA THAT NURSING HOMES
WERE REQUIRED TO REPORT
Nursing homes must electronically report information about COVID-19 to CDC’s NHSN no less
than weekly, including data on the number of suspected and confirmed COVID-19 infections
(i.e., cases); the number of total deaths and COVID-19 deaths among residents and staff; PPE
and hand-hygiene supplies on hand; and ventilator capacity and supplies on hand in the facility
(42 CFR § 483.80(g)).28 If a nursing home fails to comply with reporting requirements, CMS may
impose a CMP for each week that the nursing home is noncompliant.29 The COVID-19 module
within the NHSN system instructs a nursing home to enter either a count or a “Yes” or “No”
response for each data element.
For some nursing homes, CMS’s COVID-19 data did not include all of the data that they were
required to report. Specifically, for the 15,388 nursing homes, we found that:
• 123 nursing homes did not report required data for at least 1 of the data elements
related to COVID-19 cases and deaths among residents and staff for at least 1 week of
our audit period and
• 83 nursing homes did not report required data for at least 1 of the data elements
related to PPE and supplies for at least 1 week of our audit period.
28 The number of total deaths and COVID-19 deaths among residents consists of residents who died in a nursing
home or another location (e.g., a hospital) as reported by the nursing home.
29 CMS’s memo “Interim Final Rule Updating Requirements for Notification of Confirmed and Suspected COVID-19
Cases Among Residents and Staff in Nursing Homes” (QSO-20-29-NH), May 6, 2020. On September 2, 2020, CMPs
were codified in 42 CFR § 488.447, which stated that CMS may impose a CMP for noncompliance with the
requirements in sections 483.80(g)(1) and (2) and would assess compliance with those requirements weekly.
Although this CMP authority was not in effect during our audit period (because it did not become effective until
September 2, 2020), CMS could have instead used its CMP authority under 42 CFR § 488.408 for failure to comply
with the requirements in sections 483.80(g)(1) and (2) during the audit period.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
11
Although these nursing homes failed to report required data for some of the data elements,
CMS determined that they had met reporting requirements. CMS did not have a process to:
(1) identify all nursing homes that submit only partial data related to COVID-19 cases and
deaths among residents and staff and related to PPE and supplies and (2) request that they
submit the required data for all of the data elements.
CMS stated that it did not take any actions (e.g., performing outreach) for nursing homes that
submitted partial COVID-19 data because it determined whether nursing homes had met
reporting requirements based on reporting in general, not on whether required data for all data
elements had been reported. For example, CMS determined that a nursing home had met
reporting requirements if it had reported data for any of the data elements related to resident
and staff impact. CMS also stated that it would be difficult to determine which data elements
would have to be reported to constitute meeting reporting requirements and that it would
require significant resources to make operational changes to impose CMPs based on nursing
homes not reporting required data for certain data elements.30 Finally, CMS stated that it
wanted to be aware of issues that nursing homes may have experienced in adjusting to the new
reporting requirements and did not want to penalize nursing homes for not reporting data on
PPE and ventilators.
CMS’S COVID-19 DATA WERE NOT COMPLETE OR ACCURATE AFTER CMS HAD PERFORMED
ITS QUALITY ASSURANCE CHECKS
CMS’s COVID-19 data were not complete or accurate after CMS had performed its quality
assurance checks on nursing homes’ reported data. Specifically, CMS’s quality assurance checks
resulted in CMS: (1) excluding some nursing homes’ data from the COVID-19 data each week
during our audit period when at least one of the quality assurance ratios exceeded the
established threshold, (2) including data for some nursing homes that reported more total
COVID-19 deaths than COVID-19 cases and admissions, and (3) including data for some nursing
homes that reported more total COVID-19 deaths than total deaths from any cause. If CMS’s
30 CMS stated that it could use its existing authority under 42 CFR § 488.408 to impose CMPs on nursing homes
that did not meet reporting requirements.
Example of a Nursing Home That Submitted Partial Data
A nursing home reported data through CDC’s NHSN system for only the data elements related
to the number of all beds, whether it had access to laboratory testing for COVID-19 while a
resident was in the facility, and whether it had staffing shortages. The nursing home did not
report any data for the data elements related to the number of COVID-19 cases and deaths
among residents and staff. After receiving the CDC data, CMS determined that the nursing
home had met reporting requirements even though the nursing home had submitted partial
data. CMS did not impose a CMP on this nursing home.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
12
quality assurance check process has weaknesses, there is an increased risk that nursing homes’
COVID-19 data may not be complete or accurate (i.e., the data may be over- or underreported).
CMS Excluded Data for Some Nursing Homes From the COVID-19 Data Each Week
During Our Audit Period
Generally, if one of the six quality assurance ratios for a nursing home exceeds the threshold
that CMS established for that ratio, CMS considers the nursing home to have failed the quality
assurance check and excludes all of that nursing home’s reported data for the week from the
COVID-19 data that CMS posts on its website. Four of CMS’s six quality assurance ratios (i.e.,
ratios 1, 3, 5, and 6) are dependent on the accuracy of the number of all beds and the total
number of COVID-19 cases and admissions or deaths among residents or staff.
CMS excluded nursing homes’ data from its COVID-19 data each week during our audit period
when at least one of the quality assurance ratios exceeded the established threshold.
Specifically, for the 15,388 nursing homes, CMS excluded 498 records for 150 nursing homes
that exceeded the threshold for at least 1 of the ratios during our audit period.31
See the following page for an example of a nursing home that exceeded CMS’s established
threshold for one of the quality assurance ratios each week during our audit period.
31 Each record represented the COVID-19 data submitted by a nursing home for 1 week of the 4-week audit period.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
13
CMS’s use of ratios 1, 3, 5, and 6 resulted in certain nursing homes’ data being excluded from
the COVID-19 data each week, regardless of whether the reported data were accurate.
Because each of these ratios had a constant number of all beds in the denominator but a
cumulative total in the numerator, this resulted in the ratio increasing every week, which
meant that the established threshold for the ratio was exceeded every week.
If CMS had used noncumulative totals in the numerator of its quality assurance ratios (e.g., the
weekly data reported by each nursing home) and had adjusted the threshold for each ratio to a
level that may have indicated erroneous data, CMS would not have excluded some nursing
Example of a Nursing Home That Exceeded CMS’s Established Threshold for One of the
Quality Assurance Ratios Each Week During Our Audit Period
One nursing home’s ratio of the total number of residents’ confirmed COVID-19 cases and
admissions to the number of all beds (i.e., ratio 1) increased each week during our audit
period, from 1.57 in the first week to 1.81 in the fourth week. The cumulative total in the
numerator increased each week, but the number of all beds stayed constant. All of the data
for this nursing home were excluded from CMS’s COVID-19 data each week. Figure 4 shows
how the nursing home failed CMS’s quality assurance check each week during our audit period
for exceeding ratio 1’s threshold of 1.5. (This nursing home reported a total of 25 confirmed
COVID-19 cases and admissions among residents before June 22, 2020, the start of our audit
period; this total is included in the cumulative totals shown for each week.)
Figure 4: The Nursing Home Failed CMS’s Quality Assurance Check Each Week
During Our Audit Period
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
14
homes’ data from the COVID-19 data each week. Figure 5 shows how the nursing home in
Figure 4 would not have failed CMS’s quality assurance check if CMS had used noncumulative
totals in the numerator of ratio 1 and a hypothetical threshold of 1 (for illustrative purposes
only) instead of CMS’s threshold of 1.5, using the actual weekly data reported by that nursing
home.
Figure 5: The Nursing Home in Figure 4 Would Not Have Failed CMS’s Quality Assurance
Check Each Week if CMS Had Used Noncumulative Totals and a Different Ratio Threshold
CMS stated that it was not aware of any ratio that would more accurately identify potentially
erroneous COVID-19 data for nursing homes and that revising the existing ratios may result in
excluding more data than should be excluded. CMS also stated that it did not have the
resources to work individually with every nursing home whose data may be inaccurate and that
nursing homes can contact CMS to verify their data for future periods so that the data would
not continue to be excluded each week. Finally, CMS stated that data would continue to be
excluded unless nursing homes contact CMS or CMS contacts nursing homes during its periodic
outreach.
CMS’s COVID-19 Data Included Data for Nursing Homes That Reported More Total COVID-19
Deaths Than Total COVID-19 Cases and Admissions
CMS’s quality assurance ratio 2 uses the number of total COVID-19 deaths from both suspected
and confirmed COVID-19 cases in the numerator and the number of total confirmed COVID-19
cases and admissions among residents in the denominator. Ratio 4 uses the number of total
COVID-19 deaths from both suspected and confirmed COVID-19 cases in the numerator and the
number of total confirmed COVID-19 cases among staff in the denominator. (These ratios use
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
15
cumulative, not weekly, totals.) In addition to these data, nursing homes are required to report
the number of suspected COVID-19 cases among residents and staff (42 CFR § 483.80(g)).
However, these data are not included in the denominator in CMS’s ratios 2 and 4.
CMS’s COVID-19 data included data for nursing homes that reported: (1) more total COVID-19
deaths among residents than total confirmed COVID-19 cases and admissions among residents
and (2) more total COVID-19 deaths among staff than total confirmed COVID-19 cases among
staff. Specifically, for the 15,388 nursing homes, we found that:
• 279 nursing homes reported a total number of COVID-19 deaths among residents that
was greater than the total number of confirmed COVID-19 cases and admissions among
residents for at least 1 week during our audit period but passed the quality assurance
checks32 and
• 30 nursing homes reported a total number of COVID-19 deaths among staff that was
greater than the total number of confirmed COVID-19 cases among staff for at least
1 week during our audit period but passed the quality assurance checks.33
32 For the 279 nursing homes, 946 records showed more total COVID-19 deaths among residents than total
confirmed COVID-19 cases and admissions among residents.
33 For the 30 nursing homes, 85 records showed more total COVID-19 deaths among staff than total confirmed
COVID-19 cases among staff.
Example of a Nursing Home That Reported More Total COVID-19 Deaths Among Residents
Than Total Confirmed COVID-19 Cases and Admissions Among Residents
CMS’s COVID-19 data included data from one nursing home that had reported a total of seven
COVID-19 deaths among residents as of the fourth week of our audit period but only one
confirmed COVID-19 case and four COVID-19 admissions. This nursing home passed CMS’s
quality assurance check because the nursing home’s ratio of 1.4 was less than the ratio
threshold of 1.5. As a result, the nursing home’s data were included in the COVID-19 data.
(See Figure 6.)
Figure 6: The Nursing Home Passed CMS’s Quality Assurance Check
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
16
Quality assurance ratios 2 and 4, which both used ratio thresholds of 1.5, did not identify a
nursing home’s data as potentially inaccurate unless the nursing home reported 1.5 times or
more the number of total COVID-19 deaths among residents or staff than total confirmed
COVID-19 cases and admissions among residents or total confirmed COVID-19 cases among
staff.
CMS stated that a nursing home may have more total COVID-19 deaths among residents and
staff than total confirmed COVID-19 cases and admissions because the nursing home may have
also reported COVID-19 deaths of residents and staff who had suspected cases of COVID-19.34
CMS’s COVID-19 Data Included Data for Nursing Homes That Reported More Total COVID-19
Deaths Than Total Deaths From Any Cause
Nursing homes must report data that include COVID-19 deaths and total deaths among
residents to CDC’s NHSN (42 CFR § 483.80(g)). CDC’s COVID-19 module instructs nursing homes
to provide at least once a week: (1) the number of COVID-19 deaths for residents with
suspected or laboratory-positive COVID-19 who died in the facility or another location and
(2) the number of all deaths for residents who died in the facility or another location.
CMS’s COVID-19 data included data for nursing homes that reported more total COVID-19
deaths than total deaths from any cause. Specifically, of 15,388 nursing homes, 208 reported
more total COVID-19 deaths (2,879) than total deaths from any cause (1,520) for at least
1 week during our audit period. For example, 1 nursing home’s data showed that it had
111 total COVID-19 deaths but only 22 total deaths from any cause.
CMS stated that it did not have a quality assurance check to identify nursing homes that
reported more total COVID-19 deaths than total deaths because it focused on identifying errors
that would have an impact on the accuracy of COVID-19 reporting and the COVID-19 response
in nursing homes. CMS also stated that such a check would likely not have an impact on the
Federal COVID-19 response. Furthermore, CMS stated that it focused on COVID-19 deaths, not
the number of all deaths. (However, we believe that more COVID-19 deaths than all deaths
could indicate that the reported number of COVID-19 deaths was inaccurate.) Finally, CMS
stated that as of November 2020, CDC’s COVID-19 module did not allow a nursing home to
report more COVID-19 deaths than all deaths.35
34 As of November 23, 2020, nursing homes were no longer required to report the number of suspected COVID-19
cases among residents and staff. However, the COVID-19 data still included total suspected COVID-19 cases
previously reported by nursing homes, so CMS could include suspected cases in the ratios to make the ratios more
accurate.
35 Although nursing homes were no longer allowed to report more COVID-19 deaths than deaths from any cause,
the COVID-19 data still included data that nursing homes reported before the system change was implemented.
CMS could reach out to these nursing homes and request that they verify or correct the reported data.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
17
CMS COULD TAKE ADDITIONAL ACTIONS TO HELP ENSURE THAT THE COVID-19 DATA FOR
NURSING HOMES ARE COMPLETE AND ACCURATE
We identified two areas in which CMS could take additional actions to help ensure that its
COVID-19 data are complete and accurate. First, CMS could provide technical assistance to all
nursing homes that fail its quality assurance checks. Second, CMS could make additional efforts
to ensure that: (1) CMS’s and States’ COVID-19 data elements are comparable (i.e., CMS and
States could use the same data elements) and (2) the reported data are not substantially
different.
Technical Assistance Is Needed for All Nursing Homes That Fail CMS’s
Quality Assurance Checks
According to the CMS document Nursing Home COVID-19 Data Quality – Frequently Asked
Questions, CMS provides technical assistance to nursing homes to help them submit COVID-19
data accurately.36 CMS provides assistance through help desk support (by email and phone)
and through periodic calls to assist nursing homes in complying with reporting requirements.
CMS could provide technical assistance to all nursing homes that fail its quality assurance
checks to ensure that its COVID-19 data are accurate. CMS performed outreach to provide
technical assistance to only 417 of 771 nursing homes (about 54 percent) that failed CMS’s
quality assurance checks for at least 1 week during our audit period. Further, of the 32 nursing
homes that responded to our questionnaire and failed quality assurance checks for at least
1 week during our audit period, only 2 stated that they had been contacted by CMS (or CDC) to
correct the data. The remaining 30 nursing homes stated that they did not know that they had
failed quality assurance checks.
CMS stated that it attempted to email and call nursing homes that failed quality assurance
checks. When making phone calls, CMS made three attempts and if there was no response, it
did not follow up because of limited resources. CMS also stated that it would continue to
monitor weekly data to verify whether nursing homes had corrected their data or had
continued to fail quality assurance checks.
If CMS does not perform outreach to all nursing homes that have failed quality assurance
checks, nursing homes may not be aware that they failed the checks and may need to verify
and revise their reported data if the data are incorrect. Nursing homes’ lack of awareness may
lead to inaccuracies in CMS’s COVID-19 data (i.e., nursing homes may be under- or
overreporting data).
36 This document is available at https://data.cms.gov/stories/s/COVID-19-Nursing-Home-Data/bkwz-xpvg/.
Accessed on Jan. 20, 2021.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
18
Additional Efforts Are Needed To Ensure That CMS’s and States’ COVID-19 Data Elements Are
Comparable and That Data Are Not Substantially Different
To improve reliability of CMS’s COVID-19 data, CMS could make additional efforts to ensure
that: (1) its COVID-19 data elements are comparable to the data elements that nursing homes
report to State health departments37 and (2) CMS’s and States’ data are not substantially
different.
Of the 51 State health departments, 34 collected COVID-19 data from nursing homes and
posted the data publicly on their websites.38 However, in response to our questionnaire, not all
of the State health departments provided to us information for the same data elements that
CMS obtained through the NHSN system for the number of COVID-19 cases and deaths among
residents and staff.39
For example, CDC’s NHSN collected data on the number of COVID-19 deaths among residents,
which consisted of deaths from both confirmed and suspected cases of COVID-19. Of the
17 State health departments that provided us with the number of COVID-19 deaths among
residents, 13 did not include the number of deaths from suspected cases of COVID-19.
Therefore, the number of COVID-19 deaths for each State in CMS’s COVID-19 data was not
comparable to the number of COVID-19 deaths reported by each of the 13 State health
departments.
Table 1 on the following page shows the number of State health departments that provided
information on specific data elements in response to our request and the number of State
health departments with data elements that were not comparable to the data elements in
CMS’s COVID-19 data.
37 There is no legal requirement that CMS and State health departments use the same data elements for the
purposes of reporting COVID-19 data for nursing homes.
38 We requested that the 34 State health departments provide us with information for the following data
elements: the number of residents who had suspected and confirmed cases of COVID-19, the number of residents
who had died from COVID-19, the number of staff who had suspected and confirmed cases of COVID-19, and the
number of staff who had died from COVID-19. Of the 34 State health departments, 33 responded to our request
for information. However, not all 33 State health departments responded regarding all of these data elements,
which was caused, in part, by our request that they not provide information on any data elements that reflected
data collected from long-term care facilities other than nursing homes.
39 Table 3 in Appendix B summarizes the responses from the 51 State health departments, identifying those that
collected COVID-19 data from nursing homes and posted the data publicly on their websites and whether they
provided information on specific data elements (i.e., the number of COVID-19 cases and deaths among residents
and staff) in response to our questionnaire.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
19
Table 1: Not All State Health Departments Had Data Elements That Were Comparable to
CMS’s Data Elements
Data Element That Office
of Inspector General
(OIG) Reviewed
No. of State Health
Departments That
Provided Information on
the Data Element to OIG
No. of State Health
Departments That Did
Not Have a Data
Element Comparable
to CMS’s
No. of State Health
Departments That
Had a Data Element
Comparable to CMS’s
Number of residents with
suspected and confirmed
cases of COVID-19
17
0
17
Number of residents who
died from COVID-19
17
13
4
Number of staff with
suspected and confirmed
cases of COVID-19
15
0
15
Number of staff who died
from COVID-19
13
12
1
Furthermore, for the data elements that were comparable, we identified substantial differences
between the data provided by State health departments and CMS’s COVID-19 data. If the
difference between a State’s data and CMS’s data for a given data element was 20 percent or
greater, we determined that the State’s data were substantially different from CMS’s data. For
example, if the total number of confirmed COVID-19 cases among residents was 1,000 in a
State’s data and 790 in CMS’s data, we determined that these data were substantially different
because 1,000 is 27 percent greater than 790.
Table 2 on the following page shows the number of State health departments that provided
information on data elements comparable to CMS’s and the number of these State health
departments with data that were substantially different from CMS’s COVID-19 data.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
20
Table 2: Some State Health Departments Provided Data That Were Substantially Different
From CMS’s COVID-19 Data
Data Element That OIG
Reviewed
No. of State Health
Departments That Had a
Data Element
Comparable to CMS’s
No. of State Health
Departments That Had
Substantially Different Data
From CMS’s Data
Number of residents with
suspected and confirmed cases
of COVID-19
17
11
Number of residents who died
from COVID-19
4
2
Number of staff with suspected
and confirmed cases of COVID-19
15
6
Number of staff who died from
COVID-19
1
0
Tables 4 through 7 in Appendix B show the results of our analysis for each of the four data
elements shown in Tables 1 and 2.
The CMS document Nursing Home COVID-19 Data Quality – Frequently Asked Questions40 listed
several possible reasons for differences between CMS’s COVID-19 data for nursing homes and
the data that nursing homes reported to State health departments:
• Nursing homes may have reported the number of COVID-19 cases or deaths to their
State health departments before these data were required to be reported to CDC
through the NHSN system.41
• Nursing homes may have reported different data to their States from what they
reported to CDC through the NHSN system.42
• States may have reported different numbers of COVID-19 cases or deaths for a facility
because their definitions and requirements for reporting may have differed from those
in the NHSN system.43
40 This document is available at https://data.cms.gov/stories/s/COVID-19-Nursing-Home-Data/bkwz-xpvg/.
Accessed on Jan. 20, 2021.
41 We requested that State health departments provide information for each week of our audit period.
42 We requested that State health departments provide data for the same data elements for which nursing homes
reported data through the NHSN system.
43 We noted this issue as not all State health departments had data elements that were comparable to CMS’s data
elements.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
21
• Nursing homes’ COVID-19 case or death data were identified as having not passed
CMS’s (or CDC’s) quality assurance checks.
• Some States collected information from facilities that were not required to submit
COVID-19 data through the NHSN system, such as assisted living facilities.44
CMS stated that although differences in CMS’s and States’ COVID-19 data may be confusing for
stakeholders, CMS does not have the authority to require State health departments to collect
data for the same data elements included in CMS’s COVID-19 data. CMS also stated that it
encourages States to “leverage the Federal reporting requirements and use the same data that
[CMS] collects.”
By using standardized data elements for reporting, nursing homes could ensure that they are
reporting the same data elements to different entities, which may improve the ability of
stakeholders to rely on the data to monitor trends in infection rates and develop public health
policies to protect the health and safety of nursing home residents and staff. In addition, when
CMS’s COVID-19 data and States’ data are substantially different, users of the data may not
know which set of data they can rely on.
CONCLUSION
CMS’s COVID-19 data for nursing homes included the required data from the vast majority of
nursing homes; however, the data were not complete or accurate for some nursing homes.
Specifically, for about 5 percent of nursing homes, the data: (1) did not include all of the
COVID-19 data that nursing homes were required to report and (2) were not complete or
accurate after CMS had performed its quality assurance checks. In addition, we identified two
areas in which CMS could take additional actions to help ensure that its COVID-19 data are
complete and accurate. First, CMS could provide technical assistance to all nursing homes that
fail its quality assurance checks. Second, CMS could make additional efforts to ensure that:
(1) CMS’s and States’ COVID-19 data elements are comparable (i.e., CMS and States could use
the same data elements) and (2) the reported data are not substantially different.
When CMS’s COVID-19 data are complete and accurate, Federal and State officials and other
stakeholders may be able to more effectively monitor trends in infection rates and develop
public health policies when making decisions about how to ensure the health and safety of
nursing home residents and staff. Incomplete or inaccurate data could delay CMS’s ability to
detect and respond to an emerging COVID-19 hotspot, such as a surge or resurgence of
COVID-19 cases in a community. Furthermore, if CMS does not contact all nursing homes that
have failed quality assurance checks, nursing homes may not be aware that they failed the
checks and may need to verify and revise their reported data if the data are incorrect. Nursing
homes’ lack of awareness may lead to inaccuracies in CMS’s COVID-19 data. In addition, by
using standardized data elements for reporting, nursing homes could ensure that they are
44 We requested that State health departments provide the data only for nursing homes.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
22
reporting the same data elements to different entities, which may improve the ability of
stakeholders to rely on COVID-19 data. Finally, when CMS’s COVID-19 data and States’ data are
substantially different, users of the data may not know which set of data they can rely on.
Therefore, although we acknowledge that CMS’s COVID-19 data for nursing homes included the
required data from the vast majority of nursing homes, we believe that CMS could take actions
to improve the data’s completeness and accuracy.
RECOMMENDATIONS
We recommend that the Centers for Medicare & Medicaid Services assess the costs and
benefits of implementing the following recommendations and if CMS determines that the
benefits outweigh the costs, take action to:
• develop a process to identify nursing homes that do not report required data for all of
the data elements related to COVID-19 cases and deaths among residents and staff and
PPE and supplies and request that they submit the required data;
• revise its quality assurance checks for ratios 1, 3, 5, and 6, which use the number of all
beds in the denominator, to ensure that it does not exclude from its COVID-19 data
nursing homes’ data that should be included;
• revise its quality assurance checks for ratios 2 and 4, which use the numbers of
confirmed cases and admissions among residents and confirmed cases among staff in
the denominator, by adding the number of suspected COVID-19 cases to the
denominator in both ratios and adjust the ratios’ thresholds to reflect the change;
• identify the nursing homes that reported more total COVID-19 deaths than total deaths
from any cause and request that they verify or correct the reported data;
• contact nursing homes that failed quality assurance checks to verify the accuracy of
reported data or to correct inaccurate data; and
• work with CDC and with State health departments to determine the feasibility of using
comparable data elements to collect COVID-19 data and the feasibility of monitoring
substantial differences in the data and, if determined feasible, take actions to provide
the public with more complete and accurate COVID-19 data for nursing homes.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
23
CMS COMMENTS AND OFFICE OF INSPECTOR GENERAL RESPONSE
In written comments on our draft report, CMS concurred with our first, second, and fifth
recommendations and provided information on actions that it had taken or planned to take to
address these recommendations. However, CMS did not concur with our third, fourth, and
sixth recommendations.
Regarding our first, second, and fifth recommendations, CMS stated that it will assess the costs
and benefits of each recommendation. Regarding our first recommendation, CMS stated that it
performs ongoing data quality checks to identify instances in which nursing homes may have
entered incorrect data. CMS also stated that, as part of this process, it employs flags that alert
nursing homes if their data are not correct so that they have an opportunity to correct their
data and attempts to contact nursing homes that fail quality assurance checks by email and
phone. Regarding our fifth recommendation, CMS stated that it provides technical assistance
by performing periodic outreach activities by email and phone to nursing homes that failed
CMS’s quality assurance checks and had their data excluded from CMS’s COVID-19 data.
CMS also provided technical comments on our draft report, which we addressed as
appropriate. CMS comments, excluding the technical comments, are included as Appendix C.
After reviewing CMS’s comments, we maintain that our recommendations are valid. Our
responses to CMS’s specific comments are described in the sections below.
THIRD RECOMMENDATION: REVISE QUALITY ASSURANCE CHECKS FOR RATIOS 2 AND 4
CMS Comments
Regarding our third recommendation, CMS stated that because CDC’s NHSN COVID-19 module
no longer collects suspected COVID-19 cases, “it appears that this recommendation is no longer
relevant.” CMS suggested that we remove this recommendation.
Office of Inspector General Response
We understand that nursing homes were no longer required to report the number of suspected
COVID-19 cases among residents and staff each week in the NHSN COVID-19 module as of
November 23, 2020. However, CMS’s COVID-19 data still included the number of suspected
COVID-19 cases previously reported by nursing homes.45 Ratios 2 and 4 in CMS’s quality
assurance checks used the number of total COVID-19 deaths from both suspected and
confirmed COVID-19 cases in the numerator, but did not use suspected COVID-19 cases in the
denominator. CMS could include the number of previously reported suspected COVID-19 cases
45 As of November 2020, the number of suspected COVID-19 cases for residents and staff was 388,225, which was
about 31 percent of the total number of COVID-19 cases (i.e., confirmed and suspected COVID-19 cases for
residents and staff and confirmed COVID-19 admissions for residents).
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
24
in the denominator to make the ratios more accurate. If inaccurate ratios are used, the COVID-
19 data reported by nursing homes may be excluded from CMS’s COVID-19 data when they
should have been included, or vice versa. Further, CMS may not be able to provide technical
assistance to nursing homes to verify the accuracy of the data reported by nursing homes.
FOURTH RECOMMENDATION: IDENTIFY NURSING HOMES THAT REPORTED MORE
COVID-19 DEATHS THAN DEATHS FROM ANY CAUSE
CMS Comments
Regarding our fourth recommendation, CMS stated that because CDC has confirmed that
updates to the NHSN interface no longer allow nursing homes to enter a number of COVID-19
deaths that is greater than total deaths from any cause, “it appears that this recommendation is
no longer relevant.”46 CMS suggested that we remove this recommendation.
Office of Inspector General Response
We understand that the NHSN no longer allows nursing homes to report more COVID-19 deaths
than deaths from any cause. However, we identified nursing homes that had previously
reported more COVID-19 deaths than deaths from any cause before the NHSN interface was
updated. If CMS does not request that those nursing homes verify or correct their previously
reported data, CMS cannot ensure that the total number of COVID-19 deaths is accurate.
SIXTH RECOMMENDATION: DETERMINE THE FEASIBILITY OF USING COMPARABLE DATA
ELEMENTS TO COLLECT COVID-19 DATA
CMS Comments
Regarding our sixth recommendation, CMS agreed that, to the extent possible, data reporting
elements should be aligned among States and the Federal Government. However, CMS stated
that it does not have oversight of or authority over States’ reporting efforts for COVID-19 data.
Office of Inspector General Response
We understand that CMS does not have the authority to mandate that States align COVID-19
data elements in their own reporting systems with Federal reporting requirements. However,
CMS, CDC, and State health departments may be able to work together, to the extent possible,
to use comparable data elements and monitor substantial differences in CMS’s COVID-19 data
and States’ data. These actions would improve the ability of stakeholders to rely on the data
and to protect the health and safety of nursing home residents and staff.
46 Nursing homes report to CDC through the NHSN system the weekly totals of COVID-19 deaths and deaths from
any cause.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
25
APPENDIX A: AUDIT SCOPE AND METHODOLOGY
SCOPE
Our audit covered CMS’s COVID-19 data for 15,388 nursing homes that reported data to CMS
through CDC’s NHSN system for the 4-week period from June 22 through July 19, 2020.47 The
data consisted of 61,546 records; each record represented the COVID-19 data submitted by a
nursing home for 1 week of the 4-week audit period.
We did not perform an overall assessment of the internal control structure of CMS. Rather, we
limited our review to those internal controls related to CMS’s COVID-19 data for nursing
homes. We reviewed all five components of internal controls: control environment, risk
assessment, control activities, information and communication, and monitoring.48 Because our
audit was designed to provide only reasonable assurance that the internal controls we
reviewed were effective, it would not necessarily have detected all internal control deficiencies.
We conducted our audit from June 2020 to May 2021.
METHODOLOGY
To accomplish our objective, we:
• reviewed applicable Federal laws, regulations, and guidance;
• interviewed officials from CMS to obtain an understanding of the reporting process for
COVID-19 data, the quality assurance process for verifying the accuracy and
completeness of the data, and how the data were used;
• obtained information from CDC officials regarding nursing homes’ reporting of data
through the NHSN system, CDC’s quality assurance process, and CDC’s process for
sending the data to CMS;
• obtained COVID-19 nursing home data from CMS’s website on July 30, 2020;
47 We obtained the COVID-19 data from CMS’s website at https://data.cms.gov/Special-Programs-Initiatives-
COVID-19-Nursing-Home/COVID-19-Nursing-Home-Dataset/s2uc-8wxp. Accessed on July 30, 2020.
48 The Government Accountability Office’s (GAO’s) Standards for Internal Control in the Federal Government:
September 2014 (GAO-14-704G), known as the Green Book, sets the internal control standards for Federal entities.
The Green Book defines internal control as the plans, methods, policies, and procedures used by management to
fulfill the mission, strategic plan, goals, and objectives of the entity. The Green Book approaches internal control
through a hierarchal structure made up of five components: (1) control environment, (2) risk assessment,
(3) control activities, (4) information and communication, and (5) monitoring.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
26
• obtained other data and reports from CMS and CDC, including raw (i.e., unmodified)
data that nursing homes reported through CDC’s NHSN system and weekly reports that
contained the results of CMS’s quality assurance checks;
• analyzed CMS’s COVID-19 data and the other data and reports obtained from CMS and
CDC to determine whether CMS’s data were complete and accurate (e.g., whether
nursing homes reported more COVID-19 deaths than COVID-19 cases);
• selected a statistical sample of 120 nursing homes across the Nation and sent a
questionnaire to each sampled nursing home to obtain an understanding of how nursing
homes reported COVID-19 data through the NHSN system (e.g., whether nursing homes
reported data for all data elements for each week during our audit period);49
• sent questionnaires to State health departments in 50 States and the District of
Columbia to identify those that collected COVID-19 data from nursing homes and posted
the data publicly on their websites and to obtain an understanding of nursing homes’
reporting processes at the State level;
• requested that State health departments that required nursing homes to report
COVID-19 data provide us with information for the same data elements that CMS
obtained through the NHSN system for our audit period: the number of residents with
suspected and confirmed cases of COVID-19, the number of residents who had died
from COVID-19, the number of staff who had suspected and confirmed cases of
COVID-19, and the number of staff who had died from COVID-19;
• compiled information provided by State health departments and compared it with
CMS’s COVID-19 data to determine: (1) whether the States’ data elements were
comparable to CMS’s data elements and (2) for those data elements that were
comparable, whether the data that States reported were substantially different from
CMS’s data;50 and
49 We selected a stratified random sample of nursing homes based on whether, for our audit period, they had:
(1) reported data for all 4 weeks, (2) reported data for all 4 weeks but failed quality assurance checks for 1 to 3
weeks, (3) reported data for all 4 weeks but failed quality assurance checks for all 4 weeks, and (4) not submitted
data for 1 or more weeks. Of the 120 nursing homes, 103 responded to our questionnaire.
50 We did not verify the information that State health departments provided (i.e., by reviewing supporting data or
documentation). Because the COVID-19 data elements for nursing homes used by CMS and States differed, CMS’s
data and the States’ data could not be directly compared. For example, CMS required nursing homes to report the
number of resident deaths from both confirmed and suspected cases of COVID-19, but some States required
nursing homes to report the number of resident deaths from only confirmed cases of COVID-19. If the difference
between a State’s data and CMS’s data for a given data element was 20 percent or greater, we determined that
the State’s data were substantially different from CMS’s data. For example, if the total number of confirmed
COVID-19 cases among residents for our audit period was 1,000 in a State’s data and 790 in CMS’s data, we
determined that these data were substantially different because 1,000 is 27 percent greater than 790.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
27
• discussed the results of our audit with CMS officials.
We conducted this performance audit in accordance with generally accepted government
auditing standards. Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions
based on our audit objectives. We believe that the evidence obtained provides a reasonable
basis for our findings and conclusions based on our audit objectives.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
28
APPENDIX B: STATE HEALTH DEPARTMENTS’ COVID-19 DATA RELATED TO CASES AND DEATHS
AMONG NURSING HOME RESIDENTS AND STAFF
Table 3 summarizes the 51 State health departments’ responses to our questionnaire,
identifying those States that collected COVID-19 data from nursing homes and posted the data
publicly on their websites. The table also identifies whether the States provided information on
specific data elements (i.e., the numbers of COVID-19 cases and deaths among residents and
staff).
Table 3: State Health Departments That Collected and Posted COVID-19 Data and Provided
Requested Information (i.e., Count) for Cases and Deaths Among Residents and Staff
State
Did State Health
Department
Collect and Post
COVID-19 Data
Publicly?
Did State Health Department Provide Information for
the Following Data Elements to OIG?
Resident
COVID-19
Cases
Resident
COVID-19
Deaths
Staff
COVID-19
Cases
Staff
COVID-19
Deaths
Alabama
Yes
No
No
No
No
Alaska
No
-
-
-
-
Arizona
No
-
-
-
-
Arkansas
No
-
-
-
-
California
Yes
Yes
Yes
Yes
Yes
Colorado
Yes
No
No
No
No
Connecticut
Yes
Yes
Yes
Yes
Yes
Delaware
Yes
No
No
No
No
District of
Columbia51
-
-
-
-
-
Florida
Yes
No
No
No
No
Georgia
Yes
Yes
Yes
Yes
No
Hawaii
No
-
-
-
-
Idaho
No
-
-
-
-
Illinois52
Yes
No
No
No
No
Indiana
Yes
No
No
No
No
Iowa
Yes
Yes
Yes
No
No
Kansas
Yes
No
No
No
No
Kentucky
Yes
Yes
Yes
Yes
Yes
Louisiana
Yes
Yes
Yes
Yes
No
Maine
No
-
-
-
-
51 The District of Columbia’s health department did not respond to our questionnaire.
52 In the information that Illinois’ State health department provided to us, the numbers of COVID-19 cases and
deaths among residents and staff were combined into one total. Because we could not separate the data
elements, we excluded Illinois from our analysis.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
29
State
Did State Health
Department
Collect and Post
COVID-19 Data
Publicly?
Did State Health Department Provide Information for
the Following Data Elements to OIG?
Resident
COVID-19
Cases
Resident
COVID-19
Deaths
Staff
COVID-19
Cases
Staff
COVID-19
Deaths
Maryland
Yes
No
No
No
No
Massachusetts
Yes
Yes
Yes
Yes
Yes
Michigan
Yes
Yes
Yes
Yes
Yes
Minnesota
No
-
-
-
-
Mississippi
Yes
Yes
Yes
Yes
Yes
Missouri
No
-
-
-
-
Montana
No
-
-
-
-
Nebraska
No
-
-
-
-
Nevada
Yes
No
No
No
No
New Hampshire
Yes
Yes
Yes
Yes
Yes
New Jersey
No
-
-
-
-
New Mexico
No
-
-
-
-
New York
Yes
No
Yes
No
No
North Carolina
Yes
No
No
No
No
North Dakota
Yes
No
No
No
No
Ohio
Yes
Yes
No
No
No
Oklahoma
No
-
-
-
-
Oregon
Yes
No
No
No
No
Pennsylvania
Yes
No
No
No
No
Rhode Island
Yes
Yes
Yes
Yes
Yes
South Carolina
Yes
Yes
Yes
Yes
Yes
South Dakota
No
-
-
-
-
Tennessee
Yes
Yes
Yes
Yes
Yes
Texas
Yes
Yes
Yes
Yes
Yes
Utah
Yes
Yes
Yes
Yes
Yes
Vermont
No
-
-
-
-
Virginia
Yes
No
No
No
No
Washington
Yes
No
No
No
No
West Virginia
Yes
Yes
Yes
Yes
Yes
Wisconsin
Yes
No
No
No
No
Wyoming
No
-
-
-
-
Tables 4 through 7 on the following pages summarize the information that State health
departments provided us for each of the four data elements, whether each data element was
comparable to CMS’s data element, and the results of our analysis to determine whether there
were substantial differences between the States’ and CMS’s COVID-19 data. For each State
shown in these tables, we calculated the difference in percentage between the State’s data and
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
30
CMS’s COVID-19 data by: (1) subtracting CMS’s count for the data element from the State
health department’s count for the data element and (2) dividing the result by CMS’s count. If
the difference in percentage was 20 percent or greater, we determined that the State’s data
were substantially different from CMS’s data.
Table 4: Comparability of Data Element and Difference of State Health Departments’ Data
and CMS’s COVID-19 Data (i.e., Count) for Resident COVID-19 Cases
State Health
Department
That Provided
Data To OIG
Data
Element:
Resident
COVID-19
Cases
Comparable to
Data Element
From CMS’s
COVID-19
Data?
Count From
State Health
Department
Count
From CMS
Difference
in
Percentage
(%)
Substantially
Different
From CMS’s
COVID-19
Data?
California
Confirmed
Yes
4,495
3,943
14%
No
Connecticut
Confirmed
Yes
127
180
–29%
Yes
Georgia
Confirmed
Yes
1,688
1,529
10%
No
Iowa
Confirmed
Yes
401
295
36%
Yes
Kentucky
Confirmed
Yes
547
590
–7%
No
Louisiana
Confirmed
Yes
1,425
1,354
5%
No
Massachusetts
Confirmed
Yes
365
228
60%
Yes
Michigan
Confirmed
Yes
429
298
44%
Yes
Mississippi
Confirmed
Yes
1,382
968
43%
Yes
New Hampshire
Confirmed
Yes
61
91
–33%
Yes
Ohio
Confirmed
and
Suspected
Yes
1,290
1,894
–32%
Yes
Rhode Island
Confirmed
Yes
55
42
31%
Yes
South Carolina
Confirmed
Yes
781
1,059
–26%
Yes
Tennessee
Confirmed
Yes
1,193
357
234%
Yes
Texas
Confirmed
Yes
5,835
5,476
7%
No
Utah
Confirmed
Yes
301
352
–14%
No
West Virginia
Confirmed
Yes
85
14
507%
Yes
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
31
Table 5: Comparability of Data Element and Difference of State Health Departments’ Data and
CMS’s COVID-19 Data (i.e., Count) for Resident COVID-19 Deaths53
State Health
Department
That Provided
Data To OIG
Data Element:
Resident
COVID-19
Deaths
Comparable to
Data Element
From CMS’s
COVID-19
Data?
Count From
State Health
Department
Count
From CMS
Difference
in
Percentage
(%)
Substantially
Different
From CMS’s
COVID-19
Data?
California
Confirmed and
Suspected
Yes
661
537
23%
Yes
Connecticut
Confirmed and
Suspected
Yes
60
44
36%
Yes
Georgia
Confirmed and
Suspected
Yes
221
243
–9%
No
Iowa
Confirmed
No
52
-
-
-
Kentucky
Confirmed
No
99
-
-
-
Louisiana
Confirmed
No
158
-
-
-
Massachusetts
Confirmed
No
95
-
-
-
Michigan
Confirmed
No
64
-
-
-
Mississippi
Confirmed
No
259
-
-
-
New
Hampshire
Confirmed
No
10
-
-
-
New York
Confirmed and
Suspected
Yes
141
131
8%
No
Rhode Island
Confirmed
No
39
-
-
-
South Carolina
Confirmed
No
223
-
-
-
Tennessee
Confirmed
No
162
-
-
-
Texas
Confirmed
No
556
-
-
-
Utah
Confirmed
No
43
-
-
-
West Virginia
Confirmed
No
26
-
-
-
53 CMS’s COVID-19 data for resident COVID-19 deaths consisted of the number of deaths from confirmed and
suspected cases. If a State health department provided the number of deaths for only confirmed cases, the data
were not comparable, and we did not include the count from CMS.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
32
Table 6: Comparability of Data Element and Difference of State Health Departments’ Data and CMS’s
COVID-19 Data (i.e., Count) for Staff COVID-19 Cases
State Health
Department
That Provided
Data To OIG
Data
Element:
Staff
COVID-19
Cases
Comparable to
Data Element
From CMS’s
COVID-19
Data?
Count From
State Health
Department
Count
From CMS
Difference
in
Percentage
(%)
Substantially
Different From
CMS’s
COVID-19
Data?
California
Confirmed
Yes
4,226
3,718
14%
No
Connecticut
Confirmed
Yes
204
172
19%
No
Georgia
Confirmed
Yes
973
1,199
–19%
No
Kentucky
Confirmed
Yes
376
441
–15%
No
Louisiana
Confirmed
Yes
1,195
1,340
–11%
No
Massachusetts
Confirmed
Yes
320
200
60%
Yes
Michigan
Confirmed
Yes
613
637
–4%
No
Mississippi
Confirmed
Yes
633
720
–12%
No
New Hampshire
Confirmed
Yes
62
63
–2%
No
Rhode Island
Confirmed
Yes
63
97
–35%
Yes
South Carolina
Confirmed
Yes
335
811
–59%
Yes
Tennessee
Confirmed
Yes
1,365
795
72%
Yes
Texas
Confirmed
Yes
3,825
4,838
–21%
Yes
Utah
Confirmed
Yes
182
202
–10%
No
West Virginia
Confirmed
Yes
92
27
241%
Yes
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
33
Table 7: Comparability of Data Element and Difference of State Health Departments’ Data and
CMS’s COVID-19 Data (i.e., Count) for Staff COVID-19 Deaths54
State Health
Department
That Provided
Data To OIG
Data
Element:
Staff
COVID-19
Deaths
Comparable to
Data Element
From CMS’s
COVID-19
Data?
Count From
State Health
Department
Count
from CMS
Difference
in
Percentage
(%)
Substantially
Different From
CMS’s
COVID-19
Data?
California
Confirmed
and
Suspected
Yes
14
15
–7%
No
Connecticut
Confirmed
No
3
-
-
-
Kentucky
Confirmed
No
0
-
-
-
Massachusetts
Confirmed
No
0
-
-
-
Michigan
Confirmed
No
2
-
-
-
Mississippi
Confirmed
No
4
-
-
-
New Hampshire
Confirmed
No
0
-
-
-
Rhode Island
Confirmed
No
1
-
-
-
South Carolina
Confirmed
No
5
-
-
-
Tennessee
Confirmed
No
0
-
-
-
Texas
Confirmed
No
16
-
-
-
Utah
Confirmed
No
0
-
-
-
West Virginia
Confirmed
No
0
-
-
-
54 CMS’s COVID-19 data for staff COVID-19 deaths consisted of the number of deaths from confirmed and
suspected cases. If a State health department provided the number of COVID-19 deaths for only confirmed cases,
the data were not comparable, and we did not include the count from CMS.
TMENT OF HEALTH & HUMAN SERVICES
Centers for Medicare & Medicaid Services
DATE:
TO:
FROM:
SUBJECT:
June 21, 2021
Christi A. Grimm
P1incipal Deputy Inspector General
Chiquita Brooks-LaSme ~
f;J_ ";( a.S}
Administrator
O
Administrator
Washington, DC 20201
Office of Inspector General (OIG) Draft Repo1t: CMS's COVID-19 Data
Included Required Infonnation From Most Nursing Homes, but CMS Could Take
Actions To In1prove Completeness and Accuracy of the Data A-09-20-02005
The Centers for Medicare & Medicaid Se1vices (CMS) appreciates the opp01tunity to review and
coll11llent on the Office of Inspector General's (OIG) draft repo1t.
CMS is charged with developing and enforcing quality and safety standards across the nation's
health care system, a responsibility that the Agency takes se1iously. This duty is especially
important when it comes to the care provided for some of the most vulnerable in our society,
beneficiaries residing in nursing homes, and is especially critical now as we respond to the
coronavirns disease 2019 (COVID-19) pandeinic.
As part of our response to the pandeinic, on May 8, 2020, CMS published an inte1im final mle
with comment period (85 FR 27550) requiring nursing homes to report data on a weekly basis to
the Centers for Disease Control and Prevention (CDC) about COVID-19 cases, deaths, and
supply levels, among other metrics (42 C.F.R. 483 .80(g), 85 FR 27550, 27627, May 8, 2020).
Consistent with CMS' collllnitment to transparency, CMS began posting the reported COVID-19
data for viewing by nursing home residents and families, facilities, stakeholders, and the general
public beginning in early June 2020. The data posted by CMS on the COVID-19 Nursing Home
Data Website is rep01ted by nursing homes and collected at the federal level by the CDC through
the National Healthcare Safety Network (NHSN) Long-Term Care Facility COVID-19 Module.
The CDC NHSN system is used to ensure a nationwide, standardized process of collecting
COVID-19 data from nursing homes in each state. While the CDC is the owner and is
responsible for the NHSN system, CMS and CDC use this data as part of a coordinated eff01t to
provide detailed info1mation to state and local health depaitments, as well as nursing homes, and
to inf01m national infection prevention and control policies and strategies to fwther support
nursing home residents. It is impo1tant to note, however, that this system and requirement is
separate from state-level COVID-19 data collection eff01ts, and CMS does not have the authority
to mandate that state agencies ( other than the state Medicaid and CHIP agencies) comply with
federal repo1ting requirements. Although some states may have adopted the NHSN system after
utilizing a state-based system, CMS expects that federally-repo1ted data and state-repmted may
differ for individual facilities.
APPENDIX C: CMS COMMENTS
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
34
OIG's findings show that the vast majority of nursing homes have entered data conectly,
with less than five percent of nursing homes not reporting all required data. As such, the data
posted provides an informative and actionable representation of how COVID-19 has impacted
nursing homes across the U.S. Nursing homes that are not in compliance with reporting
requirements will receive a deficiency citation, and CMS will impose a civil money penalty
(CMP). As such, CMS has established an automated process to track facilities that are not in
compliance with the requirement to report data to CDC, and to impose CMPs on these facilities.
CMPs become due and payable after the facility has been afforded an opportunity to dispute the
findings through Independent Info1mal Dispute Resolution, or an administrative hearing is final,
if requested. For example, if facilities provided evidence of a good faith effo1t to report, but were
unable to report due to technical issues, CMS may rescind the CMPs.
As with any new reporting program, CMS provides technical assistance to facilities to help them
submit data accurately. This is accomplished through help desk suppo1t and periodic stakeholder
calls, with the goal of assisting nursing homes in how they can best comply with the repo1ting
requirements. CMS 's outreach focuses on the data variables that have the greatest impact on
nursing home quality and safety, such as the number ofresident and staff COVID-19 hospital
admissions, cases, and deaths. In addition, the CDC offers training and technical suppo1t on
repo1ting through the NHSN Module, including web-based trainings and live office hours with
CDC subject matter experts.
CMS and the CDC are perfonning ongoing data quality checks to identify instances where
facilities may have entered inconect data, such as entering cumulative counts over time instead
of new cases, and other data entry enors. As pmt of its quality assurance process, CMS has flags
that ale1t facilities if their data are not conect so that the facilities have the opportunity to make
conections. Additionally, CMS attempts to contact nursing homes that fail quality assurance
checks by email and phone calls. CMS encourages state and local health departments that have
access to these data to reach out to their facilities to work with them on amending their data. A
common reason for not repo1ting in the beginning has been difficulty using NHSN, and both
CMS and CDC have reached out to these facilities to assist them. CMS will also, on occasion,
directly contact facilities reporting inaccurate data, such as when there is a spike in the number
of facilities that have triggered quality assurance flags, in order to help resolve any reporting
issues.
In addition to the data collected through the CDC's NHSN, CMS uses eve1y tool at its disposal
to effectively monitor trends in infection rates, and develop public health policies when making
decisions about how to ensure the health and safety of nursing home residents and staff. CMS
has remained focused on holding Medicare and Medicaid certified nursing homes accountable
for safety and quality of care. For example, state smvey agencies (SSAs) prioritized their work to
focus on facilities whose conditions pose irmnediate jeopardy to facility residents. CMS shares
data with the SSAs each week to help ir1fo1m which nursing homes may have potential problems
preventing or controlling the spread of COVID-19, and therefore can target swift onsite facility
smveys. To fmther assist nursing homes in respondir1g to the ongoir1g pandemic, the Department
of Health and Human Se1vices deployed federal Task Force Strike Teams to coordinate with
states and provide onsite technical assistance and education to nursing homes experiencing
outbreaks. Additionally, CMS strategically refocused the approach of Quality Innovation
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
35
- Quality Improvement Organizations (QIN - QIOs) to work with providers,
community partners, beneficiaries and caregivers on data-driven quality improvement initiatives
designed to improve the quality of care for beneficiaries across the United States. QIN - QIOs
also reach out to nursing homes across the country to provide infection control training and
virtual technical assistance.
CMS thanks OIG for its effo1ts on this important issue and looks fo1ward to working with OIG
on this and other issues in the funire. OIG's recommendations and CMS's responses are below.
OIG Recommendation
CMS should assess the cost and benefits of implementing the following reco1mnendations and if
CMS dete1mines that the benefits outweigh the costs, take action to:
•
Develop a process to identify nursing homes that do not repo1t required data for all of the
data elements related to COVID-19 cases and deaths among residents and staff and PPE
and supplies and request that they submit the required data.
CMS Response
CMS concurs to assess the costs and benefits of the OIG's suggested recommendation. CMS
perfo1ms ongoing data quality checks to identify instances where facilities may have entered
inconect data. As pa1t of this process, CMS employs flags that ale1t facilities if their data is not
conect so facilities have the opportunity to conect their data. CMS also attempts to contact
nursing homes that fail quality assurance checks by email and phones calls.
OIG Recommendation
CMS should assess the cost and benefits of implementing the following reco1mnendations and if
CMS dete1mines that the benefits outweigh the costs, take action to:
•
Revise its quality assurance checks for ratios I, 3, 5, and 6, which use the number of all
beds in the denominator, to ensure that it does not exclude from its COVID-19 data
nursing home's data that should be included;
CMS Response
CMS concurs to assess the costs and benefits of the OIG's suggested recommendation.
OIG Recommendation
CMS should assess the cost and benefits of implementing the following reco1mnendations and if
CMS dete1mines that the benefits outweigh the costs, take action to:
•
Revise its quality assurance checks for ratios 2 and 4, which use the numbers of
confirmed cases and admissions among residents and confirmed cases among staff in the
denominator, by adding the number of suspected COVID-19 cases to the denominator in
both ratios and adjust the ratios' thresholds to reflect the change;
CMS Response
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
36
the CDC's NHSN Long Tenn Care Facility COVID-19 Module no longer collects
suspected cases, it appears that this recormnendation is no longer relevant, therefore CMS non
concurs and suggests the OIG remove this recormnendation.
OIG Recommendation
CMS should assess the cost and benefits of implementing the following recormnendations and if
CMS determines that the benefits outweigh the costs, take action to:
•
Identify the nursing homes that reported more COVID-19 deaths than total deaths from
any cause and request that they velify or conect the reported data;
CMS Response
The CDC has confinned that updates to the NHSN interface no longer allow facilities to enter a
number of COVID-19 deaths that is greater than total deaths from any cause. It appears that this
recormnendation is no longer relevant, and therefore, CMS non-concurs and suggests the OIG
remove this recommendation.
OIG Recommendation
CMS should assess the cost and benefits of implementing the following recormnendations and if
CMS determines that the benefits outweigh the costs, take action to:
•
Contact nursing homes that failed quality assurance checks to verify the accuracy of
reported data or to conect inaccurate data.
CMS Response
CMS concurs to assess the costs and benefits of the OIG's suggested recommendation. CMS
provides technical assistance by performing periodic outreach activities by phone and email to
nursing homes that failed CMS 's quality assurance checks and had their data excluded from
CMS's COVID-19 data.
OIG Recommendation
CMS should assess the cost and benefits of implementing the following recormnendations and if
CMS determines that the benefits outweigh the costs, take action to:
•
Work with CDC and with State health departments to detennine the feasibility of using
comparable data elements to collect COVID-19 data and the feasibility of monitoring
substantial differences in the data and, if determined feasible, take actions to provide the
public with more complete and accurate COVID-19 data for nursing homes.
CMS Response
CMS agrees that - to the extent possible - data reporting elements should be aligned among
States and the Federal Government. As stated above, CMS does not have oversight of or
authority over States' reporting efforts. Because of this lack of authority, CMS non-concurs with
this recormnendation.
CMS’s COVID-19 Data for Nursing Homes (A-09-20-02005)
37