The quest for trials on the efficacy of human vaccines. Results of the handsearch of Vaccine

We report the results of the handsearch of all the issues of Vaccine from 1983 to 1994 to identify randomized and controlled clinical trials (RCTs and CCTs) with the aim of setting up a Cochrane data base of vaccine trials. We read 1119 original papers and located 231 trials of which 117 (50.6%) were RCTs and 114 (49.4%) were CCTs. This compared with a MEDLINE search yield of only 60 trials (25.9% of reports identified by handsearching). Hepatitis B vaccine was the most popular topic (25.5% of trial reports), followed by influenza (13.4% of trial reports) and Hepatitis A (10.4% of trial reports). Seventy per cent of trials had been carried out in Europe or North America and mean overall trial size was 1136patients (median 110). Overall only 20 trial reports mentioned placebo explicitly, while 22 reports contained data on both overall and arm size which could not be reconciled. Reviewers of vaccine trials should not base searches exclusively on MEDLINE as they are likely to miss up to 75% of trial reports. Vaccine should adopt a standardized format of reporting of trials and contributors should conform to these rules.