All Times EDT
Keywords: clinical trial, predictive model, machine learning, analytics, risk, database lock
Lack of insight to proactively predict, in a data-driven and objective manner, which clinical trials are at risk of experiencing the delay in clinical database lock is a problem in clinical development. We develop a predictive model that predicts the risk of experiencing a possible delay in clinical database lock for a given clinical trial based on information typically available between First Subject First Visit (FSFV) and Last Subject Last Visit (LSLV) to address this problem. The model is based on historical data from clinical trials and utilizes machine learning techniques to achieve high accuracy at approximately 85%. It has been demonstrated to reflect the potential risk of delay in clinical database lock. Providing this capability increases awareness of the downstream performance of the clinical database lock associated with the upstream performance of the enrollment. It also helps to drive risk-based planning and oversight. In this poster, we will share our approach to solve this problem and the methodology that we used.