Impact of correcting peak flow for nonlinear errors on air pollutant effect estimates from a panel study

Area of Science:

• Environmental Health
• Respiratory Medicine
• Biostatistics

Background:

• Panel studies frequently assess air pollutant effects using daily peak expiratory flow (PEF) measurements.
• Variable orifice PEF meters exhibit inaccuracies and a nonlinear response, potentially biasing study outcomes.

Purpose of the Study:

• To evaluate the impact of correcting nonlinear errors in PEF meter readings on air pollutant effect estimates within a pediatric panel study.
• To determine if PEF meter error correction influences the interpretation of environmental health findings.

Main Methods:

• 147 children aged 9 years recorded PEF twice daily for 8 weeks.
• PEF readings were corrected using a validated equation derived from Vitalograph meter performance.
• Regression analysis, incorporating confounders, compared pollutant effect estimates from corrected and uncorrected data.

Main Results:

• Correction minimally affected mean PEF values but significantly altered effect sizes for air pollutants.
• Nitrogen dioxide effect estimates decreased substantially, with a significant result lost for symptomatic/atopic children.
• Particulate matter (PM2.5) effect estimates shifted direction and magnitude across different lag times and averaging periods.

Conclusions:

• Nonlinearity correction in PEF measurements critically influences panel study outcomes and interpretation of air pollution impacts.
• The observed changes in effect estimates underscore the necessity of accurate PEF data calibration in environmental exposure research.
• Findings suggest that larger PEF corrections may be needed for adults, potentially leading to greater alterations in effect estimates.