Extended Data Table 3 Comparison of ED visits, non-elective hospitalizations, and mortality in Medicare beneficiaries in affected versus control counties for post-disaster weeks 1-2 after onset of short-term U.S. NOAA NCEI billion-dollar weather disasters

From: Impact of extreme weather events on healthcare utilization and mortality in the United States

  1. A negative binomial regression model was used, with counties as the unit of analysis. The primary predictors were an indicator of whether a county was affected or control, time (pre- versus post-disaster), and the interaction between the two. All models included the disaster-specific match groups as fixed effects (matched on RUCA codes and census region) and were adjusted for the county characteristics: population of county, median household income, average age, percent male, percent non-Hispanic White, percent eligible for Medicaid (dual status), percent in poverty, and percent of persons over 25 with a high school diploma. U.S., United States; ED, emergency department; NOAA, National Oceanic and Atmospheric Administration; NCEI, National Centers for Environmental Information; CI, confidence interval; RR, risk ratio; Diff-in-Diff, difference-in-differences.
Back to article page