Observational studies of traditional Chinese medicine may provide evidence nearly consistent with the randomized controlled trials: A meta-epidemiological study

BACKGROUND: In traditional Chinese medicine (TCM) field, the benefits of observational studies was more significant. Whether the evidence from observational studies agreed with RCTs in the field of TCM was still unclear. METHODS: A meta-epidemiological study was conducted. Meta-analyses and systematic reviews including cohort studies and case-control studies of TCM were included. Ratio of odds ratio (ROR) of randomized controlled trials and observational studies were calculated individually and intercomparisons were conducted by pool analysis. RESULTS: A total of 11 studies and 30 outcome pairs were included in the pool analysis. Using results from the observational studies as the reference group, the polled ROR comparing randomized controlled trials with observational studies was 1.23 (95% confidence interval 1.05 to 1.44, and 95% prediction interval 0.90 to 1.68). The ROR by subgroup analysis were 1.15 (95% confidence interval 0.96 to 1.38; 95% prediction interval 0.95 to 1.39) and 1.12 (95% confidence interval 0.86 to 1.46; 95% prediction interval 0.51 to 2.47) for cohort studies and case-control studies, respectively. CONCLUSIONS: There is difference in pooled results between randomized controlled studies and observational studies on TCM. However, the prediction interval shows the difference is small, which suggests observational studies of TCM can be included in data analysis to provide evidence for TCM. Future studies are needed to verify the above conclusion.