ABSTRACT
Study Objectives: Due to their ubiquity, smartphone applications are becoming increasingly important for emergency response, including providing a means of mobilizing volunteer responders. Data from these applications may be useful for identifying potential disparities in emergency response by revealing geographic gaps and racial and income-based inequity in the availability of volunteers. This could in turn be used to create targeted interventions to increase equitable emergency response coverage. The purpose of our study was to examine associations between race, SES factors, and access to emergency resources using data from PulsePoint (PP), a smartphone-based emergency response application for public cardiac arrest. We sought to contextualize this investigation to the COVID-19 pandemic, to further understand how pandemic conditions may intersect with existing inequities. Methods: The PP responder position data from the Allegheny County PP deployment was aggregated into zip code-level totals from data samplings taken from August 2019 to May 2020 using geospatial informatics software (QGIS). These totals were stratified into pre- and intra-pandemic periods, as well as by racial and demographic characteristics obtained from the US Census Bureau. The change in available responders at the zip-code level, as well as the association between number of available responders and racial and demographic characteristics, were examined using Mann-Whitney U Tests due to non-normal distribution of responder counts. Results: The median (IQR) of available PP responders before and after the stay at home order were 67.4125 (116.9375) and 73.05 (127.95), respectively. Fifteen percent of zip codes in the Pittsburgh area have > 30% of African Americans with a median (IQR) of 280 (1488). This compared to 95.6% of zip codes in the Pittsburgh area that have > 30% of Caucasian-Americans with a median (IQR) of 8582 (12538). The median (IQR) for the percent below the poverty level for all zip codes was 9% (10.8%). The p-value of available PP responders before the shutdown for high-income vs. low-income zip codes was 0.493. The p-value of available PP responders after the shutdown for high-income vs. low-income zip codes was 0.197. Lastly, the p-values of available PP after the shutdown to zip codes with > 30% vs. <30% Caucasian-Americans and > 30% vs. <30% African Americans were -0.443 and 1.095, respectively. Conclusion: In summary, SES was associated with the number of PP responders at the zip code level in Allegheny county. Interestingly, the pandemic shifted the distribution of responders to a net increase in available responders which did not entirely differ by race, but by income.