Systematic Review Methods for Observational Studies: Challenges and Solutions

Article type
Authors
Lach L1, Saini M2, Bailey S1, Cimino T3
1School of Social Work, McGill University, Montreal, Quebec, Canada
2Faculty of Social Work, University of Toronto, Toronto, Ontario, Canada
3Centre For Research on Children and Families,McGill University,Montreal, Quebec, Canada
Abstract
Background: Systematic review methods have been well established for interventions studies. However, less is known about how to manage reviews of observational studies. Grant submissions for observational studies require researchers to articulate state of art knowledge to substantiate the rationale for the study. Although cherry picking studies that adequately generate this rationale are appreciated, reviewers have no way of knowing if these studies are representative of what is known. Therefore, a more systematic way of collating knowledge is required. Objectives: The purpose of this presentation is to identify a protocol for systematically reviewing observational studies, how this compares to protocols for reviewing intervention studies, and the challenges that researchers faced and solutions generated in developing this protocol. We were interested in parenting of children with neurodevelopmental disorders, how parenting differed from that of parenting children without neurodevelopmental disorders, as well as the correlates of parenting. Methods: Campbell Collaboration protocol for intervention studies was adapted to conduct a systematic review of observational studies related to parenting of children with chronic health conditions and disabilities. Over 10,000 abstracts were generated using a keyword search in 6 different databases. Abstracts were submitted to first and second level initial screening, strict screening, and data extraction. Results: Challenges associated with the systematic review of observational studies include management of an inordinate quantity of data, multiple outcome measures, measures that purport to evaluate the same phenomenon but are conceptually disparate, measures of varying quality, studies that report correlational data but have no comparison or control group, studies that report only within group analysis, management of multiple findings (correlations and regression analysis). Conclusions: Although conducting a systematic review of observational studies is challenging, the issues are shared across substantive areas and require consensus building that can only be achieved through collaboration.