A 30 year perspective of the quality of evidence published in 25 clinical journals: Signs of change?

Methods: The quality of clinical studies published in five different specialties, over three decades was evaluated. Computerised search of the Medline database was undertaken to evaluate the articles published in 25 clinical journals in 1983, 1993, and 2003 from five different specialties ( medicine, surgery, paediatrics, anaesthesia, and psychiatry). The number of randomised controlled trials (RCTs), meta-analyses, and other clinical trials (non-RCT) were noted. Results: From the 27 030 articles evaluated, there were 2283 (8.4%) RCTs, 166 ( 0.6%) meta-analyses, and 4153 ( 15.4%) other clinical trials. For the proportion of RCTs, the rank order of the specialties was; anaesthesia ( 503; 18%), psychiatry ( 294; 9.6%), medicine ( 899; 8.1%), paediatrics ( 326; 6.4%), and surgery ( 261; 5.3%) ( p< 0.001). For the proportion of meta-analysis, the rank order of the specialties was; psychiatry ( 36; 1.2%), medicine ( 105; 0.9%), paediatrics ( 15; 0.3%), anaesthesia ( 6; 0.2%), and surgery ( 4; 0.1%) ( p< 0.001). Overall, from 1983 to 2003, there were increases in the proportion of RCTs ( 449, 5.9% to 1027, 9.6%), meta-analysis ( 0, 0% to 127, 1.2%), and other clinical trials ( 897, 12% to 1983, 19%) ( p< 0.001). This trend was apparent in each clinical specialty ( p< 0.001). Conclusions: Over the three decades evaluated, clinical trials, notably RCTs and meta-analysis form only a small proportion of articles published in prominent journals from five clinical specialties. This is notwithstanding the modest increases in the proportions of RCTs and meta-analysis over the same period.