A real-world example demonstrating the application of Cochrane guidance to ensure transparent synthesis of evidence addressing a broad policy question
2Methods in Evidence Synthesis Unit, School of Public Health and Preventive Medicine, Monash University, Australia
3Indigenous Health Equity Unit, Melbourne School of Population and Global Health, The University of Melbourne, Australia
Background:
Specifying questions and criteria at the level of each synthesis can enhance the transparency of reviews and help ensure that reviews address questions of importance to decision-makers. Version 6 of the Cochrane Handbook for Systematic Reviews of Interventions introduced the concept of ‘PICO for each synthesis’ to bring greater focus on the need to plan and report details of the synthesis questions addressed within a systematic review. The results of a systematic review are ultimately determined by these synthesis questions and the decisions authors take in deciding which studies are eligible to answer each question. Without changing the review eligibility criteria, a synthesis can be structured to address different questions (e.g., broader or narrower) simply by grouping interventions, outcomes, or populations differently. Yet this level of specification is uncommon and can be challenging in broad systematic reviews such as those commissioned to address policy questions.
Objectives:
To describe our approach to question specification in a broad systematic review commissioned to inform government decisions about health insurance rebates, including:
• Use of an analytic framework and staged consultation to pre-specify a proposed structure for synthesis, agree on the final PICO for each synthesis, and triage studies for analysis.
• The benefits of this approach for decision-making, workflow, and relevance of the synthesis.
Methods:
This was 1 of 16 reviews each examining the effects of a natural therapy in any population (condition, injury, risk factor) on any health outcome. Following guidance in Chapter 3 of the Cochrane handbook, we prespecified a framework with proposed population and outcome groupings for synthesis (https://doi.org/10.1186/s13643-022-02015-1). We compiled an aggregate list of populations and outcomes from included studies organised by the framework. Populations and outcomes were prioritised by the commissioners, enabling us to finalise the synthesis PICOs and triage studies for meta-analysis.
Results:
The synthesis examined effects on six outcomes (pain, fatigue, sleep, mental health, quality of life, physical function) from 234 studies categorised into 20 population groups. For patients, the approach underpins transparent synthesis of evidence commissioned to inform decisions that affect their health.
Patient, public, and/or healthcare consumer involvement: through the commissioner (www.nhmrc.gov.au/health-advice/all-topics/complementary-medicines/natural-therapies-review).