Introduction: Vaccines are one of the most widely used health technologies. It is estimated that 1479 million doses were used worldwide in 1990. Despite their wide acceptance, few population-based evaluations of their effectiveness have been carried out and evidence of their cost-effectiveness is equivocal.

Objective: This paper reports the outcome of a handsearch of the journal Vaccine from 1983 (the first year of publication) to 1994. The handsearch was meant as a first step to establish a register of both trials and economic evaluations of vaccines.

Methods/Results: 93 issues were handsearched to retrieve randomised and clinical controlled trials, and to examine the characteristics of the population of retrieved trials. Additionally, all economic evaluations were retrieved and classified according to study design. 228 trials were retrieved of which 114 (50%) were RCTs. The three most frequent trial topics were those of vaccines against hepatitis B (34 trials or 29% of RCTs), hepatitis A (14 or 12% of RCTs) and influenza (10 trials or 9% of RCTs). A Cochrane-strategy Medline search yielded 60 trials (26% yield of com-bined CCTs and RCTs). 10 papers with a primarily economic aim were identified. Of these, 5 (50%) were descriptive, 4 (40%) were analytical and 1 (10%) methodological. The most popular topic was vaccination against hepatitis B (3 papers - 30%). 5 (50%) of these papers were correctly identified by Medline as primarily economic.

Discussion: The poor yield of Medline searches is confirmed by the result of this exercise. Additionally, this handsearch indicates that there may be a discrepancy between world incidence of a disease (and presumably its global importance) and the frequency of published trials on its vaccine. This situation appears to mirror that in economic literature. Further systematic searches for trials would help to explore the size and effects of possible selection and publication biases on and by world vaccine literature.