Discrepancias entre las frecuencias de los desenlaces primarios propuestos y obtenidos en ensayos clínicos aleatorizados

Objective: To determine the differences between the frequencies of the outcomes proposed in the initial protocol of randomized clinical trials for the outcomes obtained at endpoint. Methods: An observational, cross-sectional that included cardiology clinical trials published in The New England Journal of Medicine from 2002-2013. Results: We identified 28 trials. In 60.7 % of the estimated outcome studies incidence was higher than that found, showing discrepancy (difference greater than 10 %) in 13 (46.4%) studies. 10.7 % of the studies showed a discrepancy between the size estimated by the authors and the sample calculated by us. Conclusion: We found that although the authors of clinical trials have a good estimate of the sample size, its assumptions are not good, because the expected incidence in the control group is often overestimated in 60 % of cases.