Nursing Home Quality Ratings and COVID-19 Outbreaks: West Virginia, 2020

This page summarizes a CDC report on the first months of the COVID-19 pandemic in 2020.

Nursing homes are high-risk settings for COVID-19 outbreaks, and during the pandemic health departments did not have the resources to give every home rapid help with infection prevention and control. Researchers from CDC and the West Virginia Department of Health and Human Resources asked whether a rating that already existed could show which homes were most at risk.

Since 2008, the Centers for Medicare & Medicaid Services (CMS) has published inspection results and quality ratings for every certified nursing home under its Five-Star Quality Rating System. Trained inspectors make annual unannounced visits; the deficiencies they find are scored and summarized into a rating from 1 star (lowest quality) to 5 stars (highest), adjusted for staffing (such as nursing hours per resident) and quality measures (such as hospital readmissions).

West Virginia was a good place to test the idea: on April 17, 2020, it became the first state to require COVID-19 testing of all nursing home residents and staff, carried out from April 21 to May 8.

The study

The team identified every nursing home outbreak in the state from March to June 11, 2020, through routine surveillance and the universal testing. An outbreak meant two or more laboratory-confirmed cases within 14 days, at least one of them in a resident. The ratings used came from each home's latest inspection, between December 13, 2018, and February 26, 2020, about two weeks before the state's first reported case.

The homes

West Virginia had 123 CMS-certified nursing homes:

Nineteen of the 20 one-star homes were for-profit. They tended to have more residents and generally lower nurse staffing than higher-rated homes.

The outbreaks

By June 11, 14 homes (11%) had outbreaks, with 226 cases among residents (median 2.5 per home, range 1–71) and 140 among staff (median 4, range 0–39). Compared with homes without outbreaks, those with outbreaks had:

• more residents on an average day (92 versus 76)
• fewer nurse aide hours per resident per day (1.9 versus 2.2), though total nurse staffing was similar
• higher COVID-19 incidence in their counties (178 versus 105 per 100,000)
• more health deficiencies on inspection (a mean of 15 versus 11)

Half of the outbreak homes (7 of 14) had a one-star rating, compared with 13 of the 109 homes without outbreaks (12%). One outbreak home was a CMS Special Focus Facility, a designation for the state's lowest-rated homes with a history of serious deficiencies, which has no star rating and was left out of the modeling.

After accounting for county incidence and the number of residents, the odds of an outbreak were 87% lower in 2- to 3-star homes (adjusted odds ratio 0.13) and 94% lower in 4- to 5-star homes (adjusted odds ratio 0.06) than in 1-star homes. Put the other way, one-star homes had about seven times the odds of an outbreak of 2- to 3-star homes and about 17 times the odds of 4- to 5-star homes.

Why ratings might matter

Controlling the virus in nursing homes is hard, but early infection control, such as visitor restrictions, masks, staff education, symptom screening, outbreak plans and repeated facility-wide testing, may prevent or contain outbreaks. Lower-rated homes may struggle to put these in place. Earlier studies linked low star ratings to more health care–associated infections, worse outcomes after surgery and more hospital readmissions, possibly because of lower staffing. Here, outbreak homes had fewer nurse aide hours, which could hamper symptom monitoring and rapid detection and may signal homes without the money to hire enough staff or buy supplies. The authors suggest health departments use star ratings to decide which homes to help first, with tools such as CDC's COVID-19 Infection Control Assessment and Response tool.

Limitations

• The star rating is a composite, so the study cannot say which specific factors drive the link, and general quality improvement that does not strengthen infection control might not reduce outbreaks.
• Other factors, such as the infection control measures in place and residents' demographics, were not available and might explain part of the association, though that matters less if ratings are used only to rank risk.
• The results come from one state.
• Staffing and resident numbers were annual daily averages that may not match the study period.

A question of equity

Lower-rated homes are more likely to serve people facing social and economic disadvantage, including people enrolled in both Medicare and Medicaid, racial and ethnic minority groups and people with low incomes, who may already be at higher risk of severe COVID-19. The authors argue that protecting high-risk homes is needed to reduce COVID-19 deaths and disparities, and that it should go together with improving the overall quality of life and care for residents and staff. In response to the pandemic, CMS had directed its Quality Innovation Network–Quality Improvement Organizations to low-rated nursing homes to address quality problems and give COVID-19 infection control support.

Sources

• Bui DP, See I, Hesse EM, et al. "Association Between CMS Quality Ratings and COVID-19 Outbreaks in Nursing Homes — West Virginia, March 17–June 11, 2020." MMWR 69(37). https://www.cdc.gov/mmwr/volumes/69/wr/mm6937a5.htm
• The report's title gives the study period as beginning March 17, 2020, and its methods as March 14; this page gives neither start date. It also gives two different 95% confidence intervals for the 4- to 5-star estimate (0.006–0.39 and 0.003–0.39), so neither is repeated here.