This 2013 study by Resnick et al., published by the New Mexico Department of Health, evaluated the cardiorespiratory health impacts of wildfire smoke exposure in Albuquerque during the 2011 Wallow Fire. The fire burned over 538,000 acres in Arizona, causing fine particulate matter (PM2.5) levels to spike in Albuquerque, over 150 miles away.

1. Core Methodological Design

Study Design: A retrospective observational cohort study utilizing Poisson regression to calculate emergency department (ED) visit rates, standard errors, and 95% confidence intervals across three distinct temporal blocks. Rate Ratios (RRs) were calculated by comparing ED visit rates during acute and post-acute fire periods directly to the pre-wildfire reference period.
Study Population: Pooled individual-level ED visit records from 7 major non-federal licensed hospitals in the metropolitan area of Albuquerque, New Mexico, between May 1 and July 8, 2011. Analyses were stratified across three age cohorts:
- Youth: Ages 0–19
- Working-Age Adults: Ages 20–64
- Older Adults: Ages 65+

2. Exposure Measurement & Categorization

Instead of using satellite computer simulations, they used real-world ground monitoring data [7–9]:

Ground monitors (operated by the City of Albuquerque, Los Alamos National Laboratory, New Mexico Environment Department, and USFS mobile units) recorded hourly \text{PM}_{2.5} levels, which were averaged into daily 24-hour \text{PM}_{2.5} levels [7–9].
Rather than analyzing daily pollution as a continuous number, they divided the timeline into three distinct calendar blocks68:
1. Pre-Wildfire Period (May 1 – May 31): Clean air baseline (mean \text{PM}_{2.5} = 6.8\,\mu\text{g/m}^3)89.
2. Acute Exposure Period (June 1 – June 13): Heavy smoke days when \text{PM}_{2.5} spiked (mean \text{PM}_{2.5} = 31.3\,\mu\text{g/m}^3, exceeding 15\,\mu\text{g/m}^3 on 12 out of 13 days)3more_horiz.
3. Post-Acute Period (June 14 – July 8): Recovery window after the main smoke plume cleared (mean \text{PM}_{2.5} = 14.5\,\mu\text{g/m}^3)68.

3. How the Model Worked

Statistical Method: They used Poisson regression to calculate emergency room visit rates6.
Comparison: They compared the rate of hospital visits during the Acute Smoke Period (June 1–13) directly against the Pre-Wildfire Period (May 1–31), which served as the clean baseline reference67.
Output: The model calculated Rate Ratios (RR)6. An 1.00" data-display="false">\text{RR} > 1.00 meant visits went up during the smoke event compared to normal clean-air days67.

3. Outcomes Studied

Clinical visits were grouped based on ICD-9 codes appearing in the first three diagnostic fields of a patient's ED record:

Respiratory Outcomes: Asthma (ICD-9 493) and 'Other diseases of the respiratory system' (ICD-9 510–519) [27–28].
Cardiovascular Outcomes: All cardiovascular conditions (ICD-9 390–459), Diseases of pulmonary circulation (ICD-9 415–417), Cerebrovascular disease (ICD-9 430–438), and Diseases of veins, lymphatics, and the circulatory system (ICD-9 451–459) [28–30].

4. Key Results

Overall Population (All Ages)

Respiratory Visits: During the acute exposure period, asthma ED visits increased by 8% (RR = 1.08, 95% CI: 0.91-1.29) and 'other respiratory diseases' increased by 23% (RR = 1.23, 95% CI: 0.96-1.57) relative to baseline [14, 27–28].
Cardiovascular Visits: All cardiovascular visits combined showed an 8% increase (RR = 1.08, 95% CI: 1.00-1.16, p = 0.065) during the acute period.

Age-Stratified Findings

The study revealed that age strongly mediated the specific health risks:

Older Adults (Ages 65+): Showed the highest vulnerability for asthma and venous disorders:
- Asthma: 73% significant increase (RR = 1.73, 95% CI: 1.03-2.93, p < 0.05).
- Diseases of Veins, Lymphatics & Circulatory System: 56% significant increase (RR = 1.56, 95% CI: 1.00-2.43, p = 0.05).
Working-Age Adults (Ages 20–64): Faced heightened cardiovascular and cerebrovascular risks:
- Diseases of Pulmonary Circulation: 164% significant increase (RR = 2.64, 95% CI: 1.42-4.90, p < 0.01).
- Cerebrovascular Disease (Stroke/TIA): 69% significant increase (RR = 1.69, 95% CI: 1.03-2.77, p < 0.05).
Youth (Ages 0–19): Case counts were too low to detect statistically significant changes despite small observed increases [16, 27–29].

Post-Acute Recovery Window

Following the acute exposure period, average daily ED visits for nearly all conditions returned to lower levels than during the active smoke window.

5. Study Conclusion & Limitations

Conclusion: Wildfire smoke exposure from distant fires can significantly increase emergency department visits for acute respiratory and cardiovascular conditions hundreds of miles away. Age is a critical modifying factor, with older adults exhibiting heightened asthma risks and middle-aged adults experiencing sharp increases in vascular and cerebrovascular complications.
Documented Limitations: The study relied on administrative ED hospital databases with varied reporting capacities; PM2.5 exposure was averaged across two primary monitors rather than modeled at individual home addresses; and the models did not adjust for daily weather variables (like humidity or ambient temperature).

robin: ROBINS-E Domain Judgment Summary

ROBINS-E Bias Domain	Judgment	Key Risk Driver / Methodological Justification
Domain 1: Risk of bias due to confounding	🔴 HIGH RISK	The study used a simple pre-post comparison without controlling for critical time-varying meteorological confounders (ambient temperature, heat index, humidity) or temporal seasonality and day-of-week trends12.
Domain 2: Risk of bias arising from measurement of exposure	🟡 SOME CONCERNS / HIGH RISK	Exposure was measured by averaging daily readings from two static citywide ambient air monitors across all of Albuquerque, using a single citywide average as a proxy for individual short-term personal exposure [7–10, 20].
Domain 3: Risk of bias in selection of participants	🟢 LOW RISK	High population coverage pooling complete state-mandated ED visit records from all 7 major non-federal hospitals in Albuquerque [6, 11–12].
Domain 4: Risk of bias due to post-exposure interventions	🟢 LOW RISK	Short acute timeframe (2-week exposure period); no post-exposure medical co-interventions altered exposure status before ED arrival13.
Domain 5: Risk of bias due to missing data	🟢 LOW RISK	Complete health administrative database maintained by the New Mexico Department of Health (NMDOH)4.
Domain 6: Risk of bias arising from measurement of outcomes	🟢 LOW RISK	Outcomes were objectively extracted from hospital records based on standardized ICD-9 codes in any of the first three diagnostic fields [6–7, 26].
Domain 7: Risk of bias in selection of reported result	🟢 LOW RISK	Transparent reporting across all pre-specified respiratory and cardiovascular diagnostic codes, age brackets, and temporal windows [10, 14–16, 25].
OVERALL RISK OF BIAS	🔴 HIGH RISK	Driven by Domain 1 (uncontrolled meteorological/seasonal confounding in Poisson regression) and Domain 2 (coarse citywide monitor exposure proxy) [10, 20–21, 80].