Risk based monitoring in clinical studies – improving data quality

Abstract: Study quality is of paramount importance in clinical studies to ensure patient safety and reliable evaluation of the treatment, where the latter also entails the safety of future patients. Therefore, for example monitoring is required to minimise any risk of quality loss. Traditionally, the method for monitoring has been 100% source data verification which is costly and not sufficient. Today, authoritiesworldwide recommend risk based monitoring (RBM), which is a tool to monitor study site activities and it signals for unexpected deviations in processes or in data. RBM is to a large extent based on central statistical monitoring (CSM) using statistical analysis, and organisations involved in clinical studies are in the process of implementing these recommendations. However, since the methods are still in a stage of development further knowledge on the subject is needed. In this thesis, RBM was reviewed and selected methods were used to monitor a subsection of data from a clinical study conducted by TFS (Trial Form Support, Lund, Sweden). The key risk indicators for adverse events and serious adverse events were analysed using supervised and unsupervised statistical analysis which resulted in 19% of the sites being flagged for further investigation. Further adjustments of the methods are needed. The major difficulties in implementing RBM lie in the set-up and especially in that of supervised analysis. Further studies that share technical details and hands-on experience of CSM are needed to drive the development of RBM in clinical studies globally for general study quality improvement