Study Design

📊 Outcome Data

• ICD-10 codes: Define outcome clearly (e.g., cardiovascular: I00-I99, respiratory: J00-J99).
• Age stratification: Common groups: 0-64, 65-74, 75+.

🌡 Exposure Assessment

• Fixed monitors: Daily PM2.5, O3, temperature from regulatory networks.
• Spatial interpolation: Kriging, IDW to assign exposure to populations between monitors.
• Satellite data: AOD-derived PM2.5 estimates (e.g., van Donkelaar datasets).
• Lag alignment: Ensure exposure date aligns with outcome date (same day = lag 0).

⚠️ Key Considerations

• Minimum series length: ≥ 3 years recommended for seasonal control.
• Population stability: Assume fixed population over study period.
• Harvesting debate: Short-term displacement vs. true excess mortality.
• Multiple comparisons: Pre-specify primary analyses; label exploratory analyses.

📚 References

• Study Design Templates