Applying the GRADE (Grading of Recommendations Assessment, Development and Evaluation) methodology to evaluate the evidence in environmental health: the SENTIERI Project case study

Abstract

Introduction: SENTIERI Project (Epidemiological Study of Residents in Italian Contaminated Sites) had examined the epidemiological literature (1998-2009) to a priori assess the strength of the causal association between environmental exposures and health outcomes.The a priori evaluation increases the confidence in the causal interpretation of the observed associations, containing the post hoc observations and multiple comparisons problems. We updated the evidence of SENTIERI Project adopting the GRADE approach for summarising and rating the overall quality of evidence in a transparent and systematic way. Methods: We conducted a systematic review of studies on the association between residence near a petrochemical plant/refinery and risk of lung cancer. We searched Pubmed, Embase and other databases to identify relevant studies (2010-2015). Three reviewers extracted data independently and assessed risk of bias-RoB using a modified version of the Cochrane risk of bias tool. For each study 9 domains (e.g. population selection, exposure assessment, confounding) were classified in terms of RoB (Low, Probably low, Probably high, High) and then, applying the GRADE, the quality of the overall evidence and the strength of evidence were rated. Results: We identified 1690 records, 7 studies met inclusion criteria. Out of the 9 items, exposure and confounding assessments were rated Probably High/High for risk of bias in 3 studies; in most domains the RoB was Probably Low/Low. The quality of overall evidence was rated as moderate and strength of evidence across all studies was classified as limited. Conclusions: SENTIERI Project adopted the GRADE methodology to update the a priori evaluation of the strength of the epidemiological evidence to be used in the interpretation of study results. This example responds to the demand in environmental health for the adoption of a structured, transparent and reproducible synthesis of the evidence, it also documents its feasibility.