Assessing the Quality of Randomization From Reports of Controlled Trials Published in Obstetrics and Gynecology Journals

摘要: Objective. —To assess the methodologic quality of approaches used to allocate participants comparison groups in randomized controlled trials from one medical specialty. Design. —Survey published, parallel group trials. Data Sources. —All 206 reports with allocation described as 1990 and 1991 volumes four journals obstetrics gynecology. Main Outcome Measures. —Direct indirect measures adequacy randomization baseline comparisons. Results. —Only 32% an adequate method for generating a sequence random numbers, only 23% contained information showing that steps had been taken conceal assignment until point treatment allocation. A mere 9% both generation concealment. In apparently unrestricted randomization, differences sample sizes between control were much smaller than would be expected due chance. which hypothesis tests compare characteristics, 2% reported test results statistically significant, lower rate 5%. Conclusions. —Proper is required generate unbiased trials, yet these usually provided inadequate or unacceptable on Additional analyses suggest nonrandom manipulation selective reporting comparisons may have occurred. ( JAMA . 1994;272:125-128)