Advancing the methods of conducting a systematic review alongside a realist review: an example investigating 'what' interventions are effective and 'why' for improving outcomes for seniors with multi-morbidity

Article type
Authors
Kastner M1
1University of Toronto; North York General Hospital
Abstract
Objectives: Systematic reviews tell us 'what' interventions can work, but to be useful for clinicians and policy makers, we need to know 'why/how' they work (via realist reviews). Few systematic reviews conducted alongside realist reviews exist. To address the unmet needs of seniors with multimorbidity, we conducted a systematic review of multi-chronic disease interventions, and a realist review alongside to understand why/how these interventions influence outcomes.

Methods: We used the PRISMA, Cochrane (systematic review) and RAMESES (realist review) criteria to guide our methods. After testing screening reliability, reviewer pairs independently screened citations/full-text articles. We performed meta-analyses if appropriate. For the realist review, we created a three-step synthesis process (adapted from meta-ethnography) to help us identify context-mechanism-outcome configurations that explained systematic review findings and to derive our final programme theories.

Results: 25 articles (n = 12,131) were included in the systematic review; 106 articles contributed to the realist review. Care co-ordination interventions showed greatest potential for improving a wide range of outcomes in seniors with depression plus other chronic diseases(s). We identified and refined three programme theories in our realist review:
1) efficacy of care co-ordination interventions can be explained by their structured approach to comprehensive care, particularly in primary care;
2) multimorbidity management requires both clinical management and patient self-management - see figure for conceptual model; and
3) in prioritizing multimorbidity management, patients focus on the condition with the most undesired symptoms, while providers focus on those that most threaten patients' morbidity and mortality.

Conclusions: Our findings can be used by clinicians to optimize multimorbidity care for seniors, and by policy makers to inform policy/funding decisions. We advanced the knowledge of how to conduct a combined systematic review plus realist review by developing an innovative process for their co-conduct.

Patient or healthcare consumer involvement: We used an integrated knowledge translation (iKT) strategy to design/execute our reviews (clinicians, researchers); and 10 seniors who helped interpret findings are currently co-designing a multimorbidity tool (KeepWell).