Analyzing disparities in COVID-19 testing trends according to risk for COVID-19 severity across New York City

Given the interplay between race and comorbidities on COVID-19 morbidity and mortality, it is vital that testing be performed in areas of greatest need, where more severe cases are expected. The goal of this analysis is to evaluate COVID-19 testing data in NYC relative to risk factors for COVID-19 disease severity and demographic characteristics of NYC neighborhoods.

COVID-19 testing and the racial/ethnic composition of NYC Zip Code Tabulation Areas (ZCTA) were obtained from the NYC Coronavirus data repository and the American Community Survey, respectively. The prevalence of neighborhood-level risk factors for COVID-19 severity according to the Centers for Disease Control and Prevention criteria for risk of severe illness and complications from COVID-19 were used to create a ZCTA-level risk index. Poisson regressions were performed to study the ratio of total tests relative to the total ZCTA population and the proportion of positive tests relative to the total tests performed over time.

From March 2nd-April 6th, the total tests/population (%) was positively associated with the proportion of white residents (IRR_adj: 1.0003, 95% CI: 1.0003-1.0004) and the COVID risk index (IRR_adj: 1.038, 95% CI: 1.029-1.046). The risk index (IRR_adj: 1.017, 95% CI: 0.939-1.101) was not associated with total tests performed from April 6th-May 12th, and inversely associated from May 12th-July 6th (IRR_adj: 0.862, 95% CI: 0.814-0.913). From March 2nd-April 6th the COVID risk index was not statistically associated (IRR_adj: 1.010, 95% CI: 0.987-1.034) with positive tests/total tests. From April 6th-May 12th, the COVID risk index was positively associated (IRR_adj: 1.031, 95% CI: 1.002-1.060), while from May 12th-July 6th, the risk index was inversely associated (IRR_adj: 1.135, 95% CI: 1.042-1.237) with positivity.

Testing in NYC has suffered from the lack of availability in high-risk populations, and was initially limited as a diagnostic tool for those with severe symptoms, which were mostly concentrated in areas where vulnerable residents live. Subsequent time periods of testing were not targeted in areas according to COVID-19 disease risk, as these areas still experience more positive tests.