Cardiac Data Integrity Score (CDIS): Development of a Composite Metric to Quantify Data Quality in Cardiovascular Clinical Trials

Abstract

This paper assesses the Clinical Data Integrity Score for three completed cardiovascular trials by integrating datasets from mobile-based behavioural therapy, acute heart failure, and diabetes-cardiology coordination trials. The use of CDIS entails consolidating completeness, timeliness, consistency, error burden, and anomaly load into a weighted composite score that assesses data quality on a multidimensional scale. Using 46 sites and 2,738 participants, CDIS identified sites that were performing poorly, even when using traditional indicators like Query Rate, Edit-Check Pass Rate, and Audit Finding Density. The measure indicated cross-trial coherence in the accumulation of AE and SAE reporting gaps, temporal delays, and inter-form inconsistencies in the human-machine workflows. CDIS is valuable because it is a scalable, interpretable tool that can enhance oversight and improve the reliability of cardiovascular clinical trial data.