In a systematic review, is consequence of exposure the same as effect of intervention? Reflections on the challenges in conducting a systematic review on the sexual consequences of Female Genital Mutilation/Cutting

Article type
Authors
Berg R1, Denison E1, Fretheim A1
1Norwegian Knowledge Centre for the Health Services, Oslo, Norway
Abstract
Background: A plethora of guidance for systematic reviews on the effectiveness of interventions exists. Organizations such as the Cochrane Collaboration have contributed to an emerging consensus of best practices and the development of tools to answer questions of effectiveness and harm of controlled interventions. Much less research synthesis work has been developed for questions of consequences of natural harm/exposure. Objective: To present some methodological challenges in conducting a systematic review of the consequences of a socially prescribed practice through a recently completed systematic review about the sexual consequences of female genital mutilation/ cutting (FGM/C). Methods: The search retrieved 3700 records, which were screened for inclusion. Included studies were subject to quality assessment, data extraction, and synthesis including meta-analysis. Results: Across the 17 samples a total of 12,755 women between the ages 15 77 from nine different countries were included. All studies compared women with FGM/C to women without FGM/C. Among several challenges in our systematic review, three were particularly complex: 1) Identifying appropriate quality assessment check-lists for observational studies investigating not effectiveness but the relation between exposure (FGM/C) and consequences (sexual outcomes). Most check-lists are design specific (e.g. cross-sectional, cohort), not research question specific. 2) Some health issues do not lend themselves to manipulation thus, to what extent, if at all, is it reasonable to apply GRADE for studies not assessing effect but rather relationship? 3) Communicating the increased likelihood of an adverse sexual outcome from FGM/C rather than causal effect of the practice was linguistically demanding. Conclusions: Questions of consequences of practices do not always lend themselves to RCTs, but for observational study designs guides and tools available to systematic reviewers are scarce. There is a need to organize the relevant methodological challenges and develop resources to advance such research.