Detection of Fraud in a Clinical Trial Using Unsupervised Statistical Monitoring

Abstract

Background A central statistical assessment of the quality of data collected in clinical trials can improve the quality and efficiency of sponsor oversight of clinical investigations. Material and Methods The database of a large randomized clinical trial with known fraud was reanalyzed with a view to identifying, using only statistical monitoring techniques, the center where fraud had been confirmed. The analysis was conducted with an unsupervised statistical monitoring software using mixed-effects statistical models. The statistical analyst was unaware of the location, nature, and extent of the fraud. Results Five centers were detected as atypical, including the center with known fraud (which was ranked 2). An incremental analysis showed that the center with known fraud could have been detected after only 25% of its data had been reported. Conclusion An unsupervised approach to central monitoring, using mixed-effects statistical models, is effective at detecting centers with fraud or other data anomalies in clinical trials.

Does Central Monitoring Lead to Higher Quality? An Analysis of Key Risk Indicator Outcomes

“…For years, regulatory agencies such as FDA and EMA have required that the conduct and the progress of clinical trials be monitored to ensure patient protection and high-quality studies [ 1 , 2 ]. Until recently, the primary approach to meeting this requirement included frequent visits to each investigative site by designated site monitors who manually reviewed all of the patient source data to ensure it was reliably reported to the trial sponsor—a practice known as 100% source data verification (SDV) [ 3 – 6 ]. However, a major revision to the ICH GCP guidance was published in 2016 which strongly encouraged the use of central monitoring to more effectively and efficiently monitor trial conduct across all sites [ 7 ].…”

“… Statistical data monitoring—The execution of a number of statistical tests against some or all of the patient data in a study, which are designed to identify highly atypical data patterns at sites that may represent various forms of study misconduct. The types of misconduct identified may include fraud, inaccurate recording, training issues and study equipment malfunction or miscalibration [ 1 , 3 , 9 – 13 ]. …”

Advanced analytics for clinical trial quality: Commentary on ‘Can quality management drive evidence generation?’

“…As quality risks are identified, real-time and holistic approaches can then be applied 2–5 to detect anomalies (e.g. identification of safety underreporting).…”

“…In recent years, there have been significant efforts to build advanced analytics for clinical trials quality. [2][3][4] There is a good rationale as to why it has become a necessity. In brief, traditional Quality Assurance (QA) processes that involved on-site and manual source data validation and verification, together with heavy travels for QA staff, were not sustainable.…”

“…Statistical models can be used to assess risks throughout the execution of the trial, but also for Quality by Design. 3 As quality risks are identified, real-time and holistic approaches can then be applied [2][3][4][5] to detect anomalies (e.g. identification of safety underreporting).…”