Ambient air pollution epidemiology systematic review and meta-analysis: A review of reporting and methods practice

• Published in: Environment international Vol. 92-93; pp. 647 - 656
• Main Authors: Sheehan, Mary C., Lam, Juleen, Navas-Acien, Ana, Chang, Howard H.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier Ltd 01-07-2016
• Subjects: Air Pollutants - toxicity; Air pollution; Climate change; Environmental Exposure; Environmental health; Environmental Health - standards; Epidemiologic Studies; Guidelines; Meta analysis; Meta-Analysis as Topic; Review Literature as Topic; Systematic review; Climate change; Systematic review; Air pollution; Meta analysis; Environmental health; Guidelines
• ISSN: 0160-4120; 1873-6750
• DOI: 10.1016/j.envint.2016.02.016

Systematic review and meta-analysis (SRMA) are increasingly employed in environmental health (EH) epidemiology and, provided methods and reporting are sound, contribute to translating science evidence to policy. Ambient air pollution (AAP) is both among the leading environmental causes of mortality and morbidity worldwide, and of growing policy relevance due to health co-benefits associated with greenhouse gas emissions reductions.
We reviewed the published AAP SRMA literature (2009 to mid-2015), and evaluated the consistency of methods, reporting and evidence evaluation using a 22-point questionnaire developed from available best-practice consensus guidelines and emerging recommendations for EH. Our goal was to contribute to enhancing the utility of AAP SRMAs to EH policy.
We identified 43 studies that used both SR and MA techniques to examine associations between the AAPs PM2.5, PM10, NO2, SO2, CO and O3, and various health outcomes. On average AAP SRMAs partially or thoroughly addressed 16 of 22 questions (range 10–21), and thoroughly addressed 13 of 22 (range 5–19). We found evidence of an improving trend over the period. However, we observed some weaknesses, particularly infrequent formal reviews of underlying study quality and risk-of-bias that correlated with lower frequency of thorough evaluation for key study quality parameters. Several other areas for enhanced reporting are highlighted.
The AAP SRMA literature, in particular more recent studies, indicate broad concordance with current and emerging best practice guidance. Development of an EH-specific SRMA consensus statement including a risk-of-bias evaluation tool, would be a contribution to enhanced reliability and robustness as well as policy utility.
•Of 43 SRMAs (2009 to mid-2015) reviewed, >60% were published in the last 2years.•SRMAs met at least partially 16 of 22 criteria based on consensus guidelines.•<30% performed study quality reviews and those were more likely to comprehensively evaluate bias.•Development of EH-specific SRMA guidelines, including for quality review, is needed.