P109: Association Between Operational Complexity of Clinical Trials and Average Database Lock Time During Mid-Study Updates
Biomedical Research Program Manager
IBM Watson Health United States
The objective of this study is to understand the relationship between operational and protocol complexity of clinical trials and the number of mid-study updates and database lock time and to understand that database lock duration is predicated on clinical trial variables.
The outcomes of interest were average lock duration and number of database changes. Groups of interest were defined based on phase of clinical trial, number of sites, number of participants, number of data points, therapeutic area, and whether the trial was a device trial or not.
In this study we evaluated 586 clinical trials that used IBM Clinical Development (ICD) as the clinical data management system and analyzed variables associated with complexity of clinical trials and their impact on database lock time and the number of database changes.
The Kruskal-Wallis (or Mann-Whitney U) test was performed to determine differences between groups for non-normally distributed continuous data. If overall test was statistically significant, post hoc comparisons were performed with Bonferroni correction.
We found statistically significant differences in average lock duration between the different groups (trial phase, number of sites, number of participants, number of data points, and therapeutic area), but no significant differences were found between device and non-device trials.
Increasing volume and diversity of data, sophisticated protocol design elements associated with trial execution and adaptive nature of protocols have rendered clinical trials more complex. Increasing complexity of trials is associated with increased average database lock time. Planned and unplanned mid-study updates have emerged as a major pain point in clinical trial data management with on average 30 days required for implementing each change and site downtime within that 30-day period as a major concern. As such, flexible clinical data management systems (CDMS) with system-aided design validation can mitigate the challenges associated with mid-study updates, reduce site downtime, and expedite execution. In this study, we found however that a a clinical data management system with automated validation substantially reduced average database lock time to just 30 minutes per trial per database change.