The impact of China’s clinical trials on estimates of intervention effects in meta-analyses of randomised-controlled trials: A meta-epidemiological study

Tags: Poster
Ge L1, Pan B2, Niu Y1, Xie D1, Chen W3, Yao Y1, Shi X4, Tian J1, Yang K1
1Evidence-Based Medicine Center, School of Basic Medical Sciences, Lanzhou University; Key Laboratory of Evidence-Based Medicine and Knowledge Translation of Gansu Province, 2School of Public Health of Lanzhou University, 3Philosophy and Sociology School, Lanzhou University, 4Gansu Rehabilitation Center Hospital

Background: Trials from less-developed countries in a few cases show significantly more favourable treatment effects than trials in more-developed countries and, on average, treatment effects are more favourable in less-developed countries. China, as a rapidly developing country with the largest population in the world, may become a leading force in biomedical research in the future. However, the role of China’s clinical trials in meta-analyses is not clear.

Objective: To explore the impact of China’s clinical trials on estimates of intervention effects in meta-analyses of randomised-controlled trials.

Methods: We hand searched the full-text of Cochrane reviews published in 2015 from Cochrane Database of Systematic Reviews for meta-analyses, which included at least one trial published by China, and which involved dichotomous outcomes or continue outcomes. Information on Cochrane reviews and included randomised-controlled trials were extracted respectively. Data items included general characteristics, results of risk of bias, country name of trials included, primary outcome of each trial and effect size. We will estimate intervention effects as risk ratio (RR) for dichotomous outcome and standard mean difference (SMD) for continuous outcome in random effects models. We will use sensitivity analysis to explore the impact of China’s clinical trials on estimates of intervention effects by excluding China’s clinical trials, and relative risk ratio (RRR) or relative standard mean difference (RSMD) with 95% confident interval (CI) are calculated. And pooled RRR or RSMD value will be calculated using random-effects models. For dichotomous outcome, when pooled RRR is not 1, it would indicate that the China’s clinical trials may alter the result of meta-analysis. For continuous outcome, when difference of pooled RSMD is not 0, it would indicate that China’s clinical trials may alter the result of meta-analysis. In addition, subgroup analysis and meta-meta-regression will be conducted to explore the robust of results. Analysis will be performed using STATA 12.0.

Results and conclusions: This study is ongoing and results will be presented at the Summit.