关键词: 临床试验, 严重不良事件一致性核对, 人工核查, SAS程序

Abstract:

AIM: To realize the semi-automation of SAE (serious adverse events) reconciliation between clinical database and safety database through SAS (statistic analysis system) program, reduce the manual work and improve the accuracy of reconciliation.  METHODS: A CRO company's SAE reconciliation was taken as example, the SAE of safety database was imported into SAS through Proc Import procedure, which was combined with SAE of clinical database and was sorted by subject, AE number or AE name; the same variable one by one was compared, located and marked the inconsistent data point in EXCEL by SAS DDE language, the consistency check was more targeted and more intuitive.RESULTS:The inconsistent data point was accurately located and the workload of manual check was reduced; meanwhile, screening and matching through SAS program reduced the random errors which manual operation may produce. CONCLUSION: This method can assist traditional manual check to a certain extent; also, it can reduce the human dependence, guarantee the quality of check and process a high value of practical application.

Key words: clinical trial, SAE reconciliation, manual check, SAS program