COVID-19 Nursing Home Resident and Staff Deaths: AARP Nursing Home Dashboard

All nursing home data are from the Centers for Medicare & Medicaid Services (CMS) Nursing Home COVID-19 Public File (downloaded most recently on 5/17/2021). These data are self-reported by facilities to the Centers for Disease Control and Prevention (CDC) at least weekly. The five dashboard measures use this CMS data source.

The PPE measures previously used in the dashboard have been discontinued as of the week ending 3/14/2021. As of last month, we continue to track PPE shortages with a different measure that is not comparable to previous months.

Several data points in the state fact sheets include general population state data (that is, not limited to nursing homes) as a denominator or stand-alone measure. These data are from the COVID Tracking Project (statewide positivity rate; downloaded most recently on 2/25/2021) and USAFacts (total deaths and cases in the state; downloaded most recently on 5/27/2021).

Data were analyzed by Scripps Gerontology Center at Miami University in Ohio; additional analysis and preparation of the dashboard by the AARP Public Policy Institute.

Key Definitions

• COVID deaths (residents): Total number of residents with suspected COVID-19 or a positive COVID-19 test result who died in the facility or another location as a result of COVID-19 related complications.
• Total deaths (residents): This count includes new COVID-19 related deaths and non-COVID-19 related deaths. Includes residents who died in another location, such as a hospital.
• COVID deaths (staff): The number of deaths for staff and facility personnel with suspected COVID-19 or a positive COVID-19 test result.
• COVID cases: Number of residents, or staff and facility personnel, with new laboratory positive COVID-19 test results, as reported by the facility.
• PPE shortage: Having an “urgent need” for each type of PPE, meaning that the nursing home will run out within seven days if additional supply is not found. For the “all PPE” measure, a shortage refers to having a shortage of one or more of the five categories: N95 masks, facemasks, eye protection, gowns, and gloves. 
• Staffing shortage: Identified staffing shortage for each personnel category based on facility needs and internal policies for staffing ratios for at least one week out of the four weeks in the reporting period.

CDC has issued detailed instructions to nursing homes for reporting these data:

Inclusion Criteria

For the four-week measures, nursing facilities were included only if the facility reported to CDC for all four weeks (nationally, 93% of facilities for the most recent four-week period, for states ranging from 85% to 99%).. If a nursing facility reported, but had missing data for a specific measure (this is rare), that facility is excluded from the calculation of that measure for the dashboard.

Most nursing facilities with missing data are only missing the most recent week (ending 5/16/2021).  That is, most missing data are due to late responses, not skipped entirely.  In order to have the most current data possible, we must exclude those facilities that were late in reporting the most recent week of data as well as those with one or more weeks of non-response in earlier weeks. 

Aggregate counts of deaths and cases may be an undercount if there are facilities that are not reporting. Percentages or rates might be slightly biased if the average of non-reporting facilities differs significantly from the average of reporting facilities.

For the “since 6/1/2020” and “since January 2020” measures, all nursing homes reporting at least one week of data are included. The national response rate is greater than 99% for both measures.

Comparability to Other Data Sources

The first reporting date for the CMS Nursing Home COVID-19 data was May 24, 2020, and includes all cases and deaths since the beginning of the year that were reported; however, retroactive reporting is not mandatory, and the accuracy of reporting at the state level is unknown.

Data points that go back prior to the first reporting date, including the “since January” counts of resident cases and deaths in the state fact sheets, may significantly undercount the total number of cases and deaths. At the national level, the CMS data source gives a significant undercount of the number of cases and deaths before June, compared to other sources that were reporting in real time.

Since June, the CMS data are much more reliable and at the national level track well against data reported by the states (comparisons to individual states are difficult because each state categorizes and reports the data differently).

The state fact sheets include several measures of the percentage of total state deaths and cases that occurred among nursing home residents and nursing home staff. Because numerator and denominator data are from different sources, the reported data may result in a percentage greater than 100%, which is impossible. The value of each such measure is capped at 100%. These measures should not be used to compute the number of cases or deaths occurring outside of nursing homes.