Single-center trials show larger treatment effects than multicenter trials: evidence from a meta-epidemiological study

Tags: Oral
Dechartres A1, Boutron I2, Trinquart L3, Charles P4, Ravaud P2
1INSERM, Universite Paris Descartes, Hotel-Dieu Hospital (APHP), France, 2French Cochrane Centre, INSERM, Universite Paris Descartes, Hotel-Dieu Hospital (APHP), France, 3French Cochrane Centre, France, 4Cochin Hospital (APHP), Universite Paris Descartes, France

Background: For many physicians, multicenter randomized controlled trials (RCTs) are considered more valid than single-center RCTs but there is no formal evidence about this fact.

Objectives: We performed a meta-epidemiological study to evaluate whether treatment effect estimates differ between single-center and multicenter RCTs.

Methods: PubMed was searched for meta-analyses of RCTs with binary outcomes published between August 2008 and January 2009 and in the first 6 months of 2010 in the 10 leading journals of each medical specialty. One issue of the Cochrane Database of Systematic Reviews was also searched. All individual RCTs included in the meta-analyses were selected and searched. Data were extracted and quality assessed by use of the Risk of Bias (RoB) tool of The Cochrane Collaboration. The primary outcome was the ratio of odds ratio (ROR) quantifying the difference in estimated intervention effect between single-center and multicenter RCTs. A ({ROR}<1) would indicate larger estimates of the intervention effect in single-center trials. We performed sensitivity analyses with adjustment for sample size, risk of bias within RCTs and variance of the log OR to take into account publication bias.

Results: We selected 48 meta-analyses including 421 RCTs: 223 were single-center and 198 multicenter RCTs. Single-center RCTs showed larger intervention effect than multicenter RCTs (combined ROR = 0.73, 95% CI: 0.64 to 0.83), with low heterogeneity across individual meta-analyses (I2=12.0 %, {p}=0.24). Adjustment for sample size yielded consistent results (ROR = 0.85, 95% CI: 0.74 to 0.97), as did adjustment for ROB within RCTs, such as allocation concealment (ROR = 0.76, 95% CI: 0.67 to 0.86) and variance of log OR (ROR = 0.83, 95% CI: 0.72 to 0.96).

Conclusions: Single-center RCTs showed larger treatment effects than multicenter RCTs, which was consistent in all sensitivity analyses. Our results suggest taking this item into consideration when interpreting the results of RCTs and meta-analyses.