LAUSR

In this issue of Acta Anaesthesiologica Scandinavica, Karlsen et al. present an interesting meta-epidemiological study in which they assessed temporal trends in risk of bias and sample size in randomised clinical trials (RCTs) of post-operative pain management after hip and knee arthroplasty between 1990 and 2016. The authors included all trials from two recently published systematic reviews of analgesic interventions for post-operative pain treatment following total hip or knee arthroplasty. Temporal trends in sample size and in the seven domains in the Cochrane risk of bias tool in the 171 included trials were assessed. The authors used statistical process control, that is run and control charts, to assess whether any variation in risk of bias and sample size over time was random or non-random. They found that overall risk of bias decreased over time, mainly due to better sequence generation and allocation concealment, whereas sample size did not change significantly.