Integrating findings from a Cochrane Systematic Review of effectiveness and a Cochrane qualitative evidence synthesis: methods and lessons learnt

Article type
Authors
Glenton C1, Lewin S2, Colvin C3, Carlsen B4, Noyes J5, Schwartz A6, Rashidian A7
1Norwegian Branch of the Nordic Cochrane Centre, Norwegian Knowledge Centre for the Health Services, Norway
2Norwegian Satellite of the Cochrane EPOC Group, Norwegian Knowledge Centre for the Health Services and MRC South Africa, Norway
3Centre for Infectious Disease Epidemiology and Research (CIDER), University of Cape Town, South Africa
4Uni Rokkan Centre, Norway
5University of Bangor, UK
6University of Cape Town, South Africa
7National Institute of Health Research and School of Public Health, Tehran University of Medical Sciences, Iran
Abstract
Background: There is a growing interest globally and within the Cochrane Collaboration in conducting qualitative evidence syntheses to complement reviews of intervention effectiveness. Qualitative data synthesis can be used, for instance, to explore diversity in populations and interventions included in reviews of effectiveness and to shed light on processes behind the results of these reviews. However, there is very little knowledge or experience on how to integrate these two sources of evidence.

Objectives: To describe an approach used to integrate evidence from a Cochrane Review of the effectiveness of lay health worker programmes with evidence from a Cochrane qualitative evidence synthesis exploring barriers and facilitators to the implementation of these programmes.

Methods: Based on the draft qualitative evidence synthesis, we produced a summary of qualitative findings table of factors that appear to affect lay health worker programme implementation. We are currently using this summary of findings to build a logic model that indicates how these factors might relate to each other and how they might lead to specific lay health worker programme outcomes, such as changes in health care utilization or health outcomes. Our next step is to examine the effectiveness review to see whether this includes outcomes that correspond to the various stages of the model, and to identify where we have gaps in both outcome assessment and evidence regarding specific explanatory pathways within the model.

Results: In our presentation we will describe our experiences when moving from the summary of qualitative findings table to the logic model, including the challenges encountered in linking together specific qualitative findings in a relational model; in the development of methods for indicating the certainty of each finding or factor, as well as the certainty of the relationship between each factor; and in bringing together the logic model and the effectiveness review data.