Improving the health status of ex-prisoners: Hard lessons learned in reviewing complex interventions with multiple outcomes

Article type
Authors
Gallagher C1, Kinner S2, vanDooren K3
1Criminology, Law and Society, George Mason University, Fairfax, Virginia,
2Justice Health, Burnett Insitute, Melbourme, Victoria, Australia
3School of Population Health, University of Queensland, Queensland, Australia
Abstract
Background: As elegant as the underlying principles of systematic reviews may be, concern among researchers increasingly points to the difficulties encountered when studied interventions are complex and the potential outcomes diverse, as is nearly always the case when social components arise. This is especially true at the design stage and, as covered in Singapore, in packaging meta-analytic results so that they are meaningful for day-to-day use in practice. Objectives: Using the work presented by panelists in the Complexity in Cochrane reviews plenary from the Singapore Colloquium, this paper reports on attempts at an a priori management approach of these issues for a systematic review on interventions aimed at improving the health status of ex-prisoners. Methods: The authors use standard Cochrane systematic review methods, as well as those offered by the equity and economic methods groups. The authors anticipate and delineate the inherent complexities of the review in an attempt to preserve detail and package findings for practitioners. Results: A wealth of information regarding evidence is archived and described, and is illustrative of the complexities that were unanticipated in the design states. Though few studies meet the necessary rigor for inclusion, results fall prey to those foreshadowed by Paul Glasziou in last year’s plenary – takeaway points are more difficult from meta-analyses than from individual Randomized Control Trials (RCTs). Conclusions: This study reemphasizes the need to address and capture the profound complexities inherent in multiple component interventions with complex populations and in complex settings. The authors provide multiple examples of pre-review attempts to describe and archive these complexities for narrative review and to preserve some of the original intent of the interventions so they may be repackaged with the review results so they are equally as usable as those from RCTs. And simple as results from meta-analyses seem, research in the medical field increasingly suggests that gleaning specific meaning from systematic reviews for incorporation into day-to-day practice has proven to be a challenge. This is especially true when the studied interventions are complex and the outcomes diverse, which is nearly always the case when interventions contain a social component.