[Randomized clinical trials in the Nederlands Tijdschrift voor Geneeskunde].

OBJECTIVE: To collect all randomized clinical trials (RCT) published in the Nederlands Tijdschrift voor Geneeskunde (Ned Tijdschr Geneeskd) in order to present them to the Cochrane Centre, Oxford, United Kingdom, to be included in the international database of RCT's, which will be used to perform meta-analyses. DESIGN: Retrospective bibliometric study. SETTING: Editorial office of Ned Tijdschr Geneeskd. METHOD: All original papers (OP), published between the May 1st, 1948 and May 1st, 1993 were studied. Papers were selected that met the criteria for RCT: randomization or the similar assignment of patients to different treatment groups. RESULTS: Of 6820 OP 259 (3.8%) were selected. In 69 (26.6%) randomization seemed to be performed correctly and in 20 (7.7%) randomization was not correct; in 170 (65.6%) there was no randomization. In the period May 1st, 1948 till May 1st, 1968 24 RCT's were published, in the period May 1st, 1973 till May 1st, 1993 60. Most RCT's (78%) describe the effects of pharmaceutical treatment. CONCLUSION: RCT's as such were difficult to recognize in Ned Tijdschr Geneeskd. Over the years there was an increase in the number of published RCT's. The problems we found with the design of RCT's and randomization are still current. An international register of published and/or planned RCT's, as envisaged by the Cochrane Centre, will make it easier to carry out meta-analyses.