Table 1 Effect of (wildfire and non-wildfire) PM2.5 on respiratory hospital admissions.

From: Wildfire smoke impacts respiratory health more than fine particles from other sources: observational evidence from Southern California

Fire upwind + strong SAW (1999–2012)

Regression model for respiratory admissions (rate per 100,000 people)

Aggregated sources (smoke and non-smoke)

Approach used to isolate wildfire-specific PM2.5

Instrumental Variable

Imputation

Interaction

Seasonal Interpolation

Wildfire-specific

Non-smoke

Wildfire-specific

Non-smoke

Wildfire-specific

Non-smoke

Wildfire-specific

PM2.5 coefficient

0.0014

0.0071

0.0013

0.018

1.00068

1.00061

0.0024

0.0055

(95% CI)

(0.00077–0.0021)

(−0.0022 to 0.017)

(0.00068–0.0020)

(0.0064–0.030)

(1.00049–1.00087)

(0.10–1.0015)

(0.0018–0.0030)

(−0.00068 to 0.012)

% change with 10 µg m−3 PM2.5

0.76

3.8

0.72

10

0.67

1.28

1.3

3.0

(95% CI)

(0.42–1.1)

(−1.2 to 8.9)

(0.36–1.1)

(3.5–16.5)

(0.48–0.86)

(0.37–2.19)

(0.97–1.7)

(−0.37 to 6.3)

  1. All regressions include controls:flu admissions, weather covariates, day-of-week effects, month-of-year effects, zip code fixed effects, and a time trend.
  2. Summer months (June, July, August) are excluded.
  3. Mean PM2.5 = 15.6 μg m−3 (IQR = 9.2 μg m−3).
  4. Mean rate of respiratory admissions per 100,000 people = 1.85.
Back to article page