Figure 2.
Reported statistically significant results across medical specialties. (A) Proportion (with 95% CI) of reported statistically significant effects in published studies for efficacy and safety outcomes. (B) Relationship between the proportion of statistically significant effects and the median sample size. ORs (with 95% CI) for a statistically significant effect in a primary study (C) and in a combined effect from a meta-analysis (D). RCT, randomised controlled trial. ENT, ear, nose and throat. SJR, Scimago journal rank. n.s., not statistically significant.