Seeing the forests and the trees--innovative approaches to exploring heterogeneity in systematic reviews of complex interventions to enhance health system decision-making: a protocol

Syst Rev. 2014 Aug 12:3:88. doi: 10.1186/2046-4053-3-88.

Abstract

Background: To improve quality of care and patient outcomes, health system decision-makers need to identify and implement effective interventions. An increasing number of systematic reviews document the effects of quality improvement programs to assist decision-makers in developing new initiatives. However, limitations in the reporting of primary studies and current meta-analysis methods (including approaches for exploring heterogeneity) reduce the utility of existing syntheses for health system decision-makers. This study will explore the role of innovative meta-analysis approaches and the added value of enriched and updated data for increasing the utility of systematic reviews of complex interventions.

Methods/design: We will use the dataset from our recent systematic review of 142 randomized trials of diabetes quality improvement programs to evaluate novel approaches for exploring heterogeneity. These will include exploratory methods, such as multivariate meta-regression analyses and all-subsets combinatorial meta-analysis. We will then update our systematic review to include new trials and enrich the dataset by surveying authors of all included trials. In doing so, we will explore the impact of variables not, reported in previous publications, such as details of study context, on the effectiveness of the intervention. We will use innovative analytical methods on the enriched and updated dataset to identify key success factors in the implementation of quality improvement interventions for diabetes. Decision-makers will be involved throughout to help identify and prioritize variables to be explored and to aid in the interpretation and dissemination of results.

Discussion: This study will inform future systematic reviews of complex interventions and describe the value of enriching and updating data for exploring heterogeneity in meta-analysis. It will also result in an updated comprehensive systematic review of diabetes quality improvement interventions that will be useful to health system decision-makers in developing interventions to improve outcomes for people with diabetes.

Systematic review registration: PROSPERO registration no. CRD42013005165.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Biomedical Research
  • Decision Making, Organizational
  • Diabetes Mellitus / therapy*
  • Humans
  • Meta-Analysis as Topic*
  • Quality Improvement*
  • Research Design / standards*
  • Review Literature as Topic*
  • Systematic Reviews as Topic