Abstract
Background: A number of clinical trials have faced problems of subject enrollments after a trial is open. Purpose: The purpose of this research is to develop a predict index which estimate poor accrual in clinical trials before the trial is open using Electronic Medical Records. Methods: Towards clinical trials that are concerned by our center, we calculated the predictive index by dividing target sample size by the number of searched eligible patients and compared with accrual achievement of these trials. Results: Thirteen of 19 eligible trials have not achieved their original target sample size. The accrual achievement negatively correlated with the predictive index significantly. All the nine trials whose predictive index were more than constant values have not achieved their target sample size. In other words, these trials were able to predict that they could not achieve their original target sample size before the trial was open. Conclusion: Investigators should set attainable target sample size based on their actual clinical practice records at the protocol development stage to avoid poor accrual.
