As on-site monitoring can account for as much as 30 per cent of a late-stage trial’s costs, CluePoints last week launched a new software system to help companies target on-site monitoring of clinical trials by identifying centers with anomalous and inconsistent data.
CROs in particular have shown an interest in this software because they recognize that in order to get ahead of competition, they will have to begin shifting away from the standard of on-site monitoring, CluePoints Chief Commercial Officer Patrick Hughes told Outsourcing-Pharma.com.
The inefficiencies and high costs of on-site monitoring have been documented in trials, although the FDA and EMA have left it largely up to the sponsors to decide how to improve on-site monitoring. Both agencies recommend using central statistical monitoring to keep track of data quality.
Unlike other monitoring options that evaluate pre-determined risk indicators, Hughes said the CluePoints tool can rank each of the clinical trial sites in terms of their risk profiles in relation to the other sites in a trial. The software can also be run multiple times during the course of a trial.
A typical run of the software will generate thousands of p-values and then CluePoints analysts use this data to compile a report on the severity of the risks. It’s a service-based model right now, Hughes added, but it may ultimately fall under the control of a sponsor.
In case reports, the software has helped companies to:
- Identify under-reported adverse events;
- Confirm fraud in a trial where patient diaries were completed by site staff;
- Identify a site that used different labs for the same patients over time, which resulted in variability;
- Mis-calibrated equipment used across clinical trial centers in the same country; and
- Replicated data.
The International Drug Development Institute developed the software and statistical algorithms over a decade through research endorsed by GlaxoSmithKline Vaccines, the Institute of Statistics at Université Catholique de Louvain, and the Artificial Intelligence Research Laboratory of Université Libre de Bruxelles.
The process used to detect anomalies in site data is currently pending patent protection with the United States Patent and Trademark Office.