Contribution of AOD-PM2.5 surfaces to respiratory-cardiovascular hospital events in urban and rural areas in Baltimore, Maryland, USA: New analytical method correctly identified true positive cases and true negative controls

Epidemiologic studies have used aerosol optical depth (AOD)-PM2.5 as a proxy for ambient PM2.5 in urban and rural areas, even though its validation with air monitors has only occurred in urban areas. The contribution of elevated AOD-PM2.5 on respiratory-cardiovascular true positive (TP) cases, exposed to high PM2.5, and true negative (TN) controls, not exposed to elevated PM2.5, was evaluated in 72 Community Multiscale Air Quality (CMAQ) grids with (urban) and without (rural) air monitors. The odds ratio (OR) algorithm and the newly developed beta (ẞ) algorithm were used to evaluate the reliability and validity of TP cases, and TN controls in grids with and without air monitors. Four experimental AOD-PM2.5 fused surfaces and four health outcomes were evaluated. Only the linear predictor (ẞ) algorithm reliably and correctly identified TP cases and TN controls, with probabilities ~1.00. The OR algorithm only identified TN controls, with probabilities ~1.00, and significantly overestimated the percentage of TP cases. Regression analyses demonstrated that the OR algorithm's accuracy could be improved if the number of cases for all health outcomes was increased 50.8% in all grids and 73.9% in grids without monitors. Since the number and percentage of TP cases and TN controls were similar in grids with and without air monitors, this outcome suggests that the AOD-PM2.5 and health outcome concentration-response function evaluated in grids with monitors also holds in grids without air monitors. The possible use of AOD-PM2.5 fused surfaces, as another epidemiologic tool, to assess elevated ambient PM2.5 concentration levels to respiratory-cardiovascular hospital events in rural areas is discussed. Contribution of AOD-PM2.5 Bayesian Surfaces to Respiratory-Cardiovascular Hospital Events in Urban and Rural Areas in Baltimore, Maryland, USA: New Analytical Method Correctly Identified True Positive Cases and True Negative Controls Figure Caption: New analytical method correctly detected case-control status of participants exposed to PM2.5 in the case-crossover study. [Display omitted] •Accuracy of respiratory-cardiovascular case-control evaluated in AOD-PM2.5 study.•OR method overestimated true positive cases and correctly detected true negative controls.•The New Beta algorithm accurately detected true positive cases and true negative controls.•The new method confirmed the correct classification cases and controls in urban and rural areas.