The use of 'PICO for synthesis' and methods for synthesis without meta-analysis: protocol for a survey of current practice in systematic reviews of health interventions.

ABSTRACT

Introduction: Systematic reviews are used to synthesise research and inform decision making by clinicians, consumers and policy makers. The synthesis component of systematic reviews is often narrowly considered as the use of statistical methods to combine the results of studies, primarily meta-analysis. However, synthesis can be considered more broadly as a process beginning with (i) defining the groupings of populations, interventions and outcomes to be compared (the 'PICO for each synthesis'); (ii) examining the characteristics of the available studies; and (iii) applying synthesis methods from among multiple options. To date, there has been limited examination of approaches used in reviews to define and group PICO characteristics and synthesis methods other than meta-analysis. Objectives: To identify and describe current practice in systematic reviews in relation to structuring the PICO for each synthesis and methods for synthesis when meta-analysis is not used. Methods: We will randomly sample 100 systematic reviews of the effects of public health and health systems interventions published in 2018 and indexed in the Health Evidence and Health Systems Evidence databases. Two authors will independently screen studies for eligibility. One author will extract data on approaches to grouping and defining populations, interventions and outcomes, and the rationale for the chosen groups; and the presentation and synthesis methods used (e.g. tabulation, visual displays, statistical synthesis methods such as combining P values, vote counting based on direction of effect). A second author will undertake independent data extraction for a subsample of reviews. Descriptive statistics will be used to summarise the findings. Specifically, we will compare approaches to grouping in reviews that primarily use meta-analysis versus those that do not. Conclusion: This study will provide an understanding of current practice in two important aspects of the synthesis process, enabling future research to test the feasibility and impact of different methodological approaches.