Identifying optimal characteristics of Cochrane systematic review summary formats: qualitative study and three randomized controlled trials

Article type
Authors
Buljan I1, Tokalić R1, Puljak L2, Zakarija Grković I3, Marušić A1
1Department of Research in Biomedicine in Health, University of Split School of Medicine, Split; and Cochrane Croatia
2Department for Development, Research and Health Technology Assessment, Agency for Quality and Accreditation in Health Care and Social Welfare, Zagreb and Cochrane Croatia
3Departments of Family Medicine and Clinical Skills, University of Split School of Medicine, Split; and Cochrane Croatia
Abstract
Background: Users have difficulties with understanding health information. As Cochrane is continually developing various materials for evidence translation, it is important to identify optimal formats for end-users.

Objectives: The aim of the study was to conduct three different parallel randomized controlled trials to assess how the framing of summary information (positive versus negative framing of numerical information; standard format versus one sentence summary; and summaries with compared to summaries without numbers) affect the perception of the efficacy and intention for use of the intervention.

Methods: We first conducted three focus groups with students, patients and doctors (N = 19) which identified the differences between desirable and undesirable formats. In the first trial, we tested whether positive or negative framing of numerical expressions in Plain language summaries (PLS) had the effect on perception of effectiveness and willingness to use the intervention in a group of first year medical students (N = 91) during an anatomy class. Participants were randomly allocated by computer software to either positive (n = 44) or negative frame (n = 47). They had to read three different Cochrane PLSs and answer the questions about the effectiveness of an intervention.

Results: Focus groups identified susceptibility to numerical framing as an important factor in understanding Cochrane summary formats. In the first trial, none of the participants was excluded from analysis and there was no difference in the perception of the intervention efficacy between groups who read positively or negatively framed information (theoretical range 1-10; mean difference 1.53, 95% CI -0.33 to 3.39).

Conclusions: Framing of health information in a positive or negative direction did not affect the understanding of the intervention described. Ongoing trials are focused on testing other elements of summary formats experimentally which might contribute to better understanding of findings.

Patient or healthcare consumer involvement: Patients have participated in focus groups and will be participating in the trials based on the results of qualitative study.