[Randomized clinical trials in the Ugeskrift for Laeger 1948-1995].

The Danish Medical Journal (Ugeskrift for Laeger) was handsearched for randomised and pseudorandomised clinical trials over the years 1948-1995. We found 665 trials of which 539 (81%) were classified as randomised; the randomisation method was not stated for 422 of them (78%) and was insufficiently described in the majority of the rest. This problem became worse with time (p = 0.01). In contrast to the international literature where the increase was more gradual and occurred several years later, the number of randomised trials increased markedly around 1973. This change is associated with the introduction of courses in clinical research methodology during the previous years. Since randomisation is the most important safeguard against bias, the method should be described in detail.