Generalizability of randomized controlled trials in primary health care: Applying the PRECIS-2 tool on published protocols.

Abstract

Pragmatic design may facilitate the generalizability of effectiveness of randomized controlled trials (RCTs) in primary health care (PHC). The aim of this study was to investigate whether published protocols in PHC were designed pragmatically and to explore whether specific trial characteristics may be related to a pragmatic design. Using the Pragmatic Explanatory Continuum Indicator Summary-2 (PRECIS-2), we assessed pragmatism for 123 published RCT protocols. For each domain, we calculated the mean score with the 95% confidence interval (95% CI). Interrater reliability was assessed by weighted κ-coefficient with 95% CI. We examined potential associations of published protocol characteristics with overall pragmatism by performing univariate and multivariate analyses. We observed the highest score for primary analysis (4.66, 95%CI: 4.51, 4.82). The eligibility score was intermediate (3.16, 95%CI: 3.01, 3.32). Domains with scores towards the explanatory side included organization (2.50, 95%CI: 2.36, 2.63), flexibility of delivery (2.74, 95%CI: 2.60, 2.88) and flexibility of adherence (3.00, 95%CI: 2.83, 3.17). Interrater agreement was good (κ = 0.61; 95%CI: 0.34, 0.80; p < 0.001). Higher sample sizes were correlated to a pragmatic design (odds ratio: 6.86, 95%CI: 1.64, 28.75; p = 0.04). Most RCT protocols were rated as intermediate in the pragmatic-explanatory continuum. Future research may guide all stakeholders on how best to incorporate the level of pragmatism in the interpretation of the resultsso that the trials are more likely to be applicable in real-world settings.