English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Background More than 360,000 people infected with COVID-19 in New York State (NYS) by the end of May 2020. Although expanded testing could effectively control statewide COVID-19 outbreak, the county-level factors predicting the number of testing are unknown. Accurately identifying the county-level predictors of testing may contribute to more effective testing allocation across counties in NYS. This study leveraged multiple public datasets and machine learning algorithms to construct and compare county-level prediction models of COVID-19 testing in NYS. Methods Testing data by May 15th was extracted from the Department of Health in NYS. A total of 28 county-level predictors derived from multiple public datasets (e.g., American Community Survey and US Health Data) were used to construct the prediction models. Three machine learning algorithms, including generalized linear regression with the least absolute shrinkage and selection operator(LASSO), ridge regression, and regression tree, were used to identify the most important county-level predictors, adjusting for prevalence and incidence. Model performances were assessed using the mean square error (MSE), with smaller MSE indicating a better model performance. Results The testing rate was 70.3 per 1,000 people in NYS. Counties (Rockland and Westchester) closed to the epicenter had high testing rates while counties (Chautauqua and Clinton) located at the boundary of NYS and were far away from the epicenter had low testing rates. The MSEs of linear regression with the LASSO penalty, ridge regression, and regression tree was 123.60, 40.59, and 298.0, respectively. Ridge regression was selected as the final model and revealed that the mental health provider rate was positively associated with testing (β=5.11, p=.04) while the proportion of religious adherents (β=-3.91, p=.05) was inversely related to the variation of testing rate across counties. Conclusion This study identified healthcare resources and religious environment as the strongest predictor of spatial variations of COVID-19 testing across NYS. Structural or policy efforts should address the spatial variations and target the relevant county-level predictors to promote statewide testing. Disclosures All Authors: No reported disclosures

429. County-level predictors of COVID-19 testing across the 62 counties in New York State: A comparison across machine learning algorithms