Figure 2.

Percentage difference and 95% confidence intervals in hospitalizations during 2008–2010, among U.S. Medicare recipients $\ge 65$ y of age per $10{\text{-}\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$ increase in ${\mathrm{PM}}_{2.5}$, lag day 0 to lag day 6. Smoke days are defined as having a wildfire-specific contribution $>5{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$, and non-smoke days as wildfire-specific contribution $\le 5{\mathrm{\mu }\mathrm{g}/\mathrm{m}}^3$. Associations are estimated using a single lag model for the interaction between ${\mathrm{PM}}_{2.5}$ (${{\mathrm{PM}}_{2.5}}^{\text{Tot}\text{CMAQ}}$, ${{\mathrm{PM}}_{2.5}}^{\text{Tot}\text{CMAQ}\text{-}\mathrm{M}}$, or ${{\mathrm{PM}}_{2.5}}^{\text{Tot}}$) and SmokeDay adjusting for day of the week, day [natural spline with 6 degrees of freedom (df) per year], temperature (natural spline with 3 df), and relative humidity (natural spline with 3 df) for each county, followed by a meta-analysis. Using ${{\mathrm{PM}}_{2.5}}^{\text{Tot}\text{CMAQ}}$ 595 counties, 341 counties, and 607 counties were used in the meta-analyses for (A) respiratory; (B) asthma, bronchitis, and wheezing; (C) and cardiovascular hospitalizations, respectively. Using the other metrics (${{\mathrm{PM}}_{2.5}}^{\text{Tot}\text{CMAQ}\text{-}\mathrm{M}}$ or ${{\mathrm{PM}}_{2.5}}^{\text{Tot}}$) 134 counties, 92 counties, and 137 counties were used in the meta-analyses for (A) respiratory; (B) asthma, bronchitis, and wheezing; (C) and cardiovascular hospitalizations, respectively. The y-axes limits differ between hospitalization types. (See Table S1 for corresponding numeric data.)