The implementation of risk-based monitoring is moving forward, but it is not without its challenges. Many sponsors still view the practice as extra or optional in a trial, added at their discretion, so it is difficult to justify the costs up front. However, the new guidance is clear that the sponsor is required to develop a risk-based approach to monitoring. This includes any and all trials that fall under ICH E6t, i.e. every industry-sponsored clinical trial.
The first lines from the guideline introduction read, "Good Clinical Practice (GCP) is an international ethical and scientific quality standard for designing, conducting, recording and reporting trials that involve the participation of human subjects."
Yet another objection we hear is that ICH E6 does not apply to US-only clinical trials. Now, there is a chance that the argument is now moot since FDA has officially published ICH E6 as their industry guidelines; but even before that, the argument did not stand up to scrutiny. In the vast majority, 99.99% of clinical trial protocols, including US-only, sponsors include verbiage that the trial will be run in accordance with GCP. This automatically means that, regardless of the geography of a particular study, compliance is built directly into the protocol design.
That said, since the new guide demands risk-based monitoring implementation in all clinical trials, we find that it’s the scope of central monitoring activities that require discussion in each particular trial.
Currently, our practice is to have our Key Risk Indicators planned and monitored in every study while reserving more complex statistical data monitoring to pivotal, large-scale trials and central safety reviews for those trials with specific safety concerns.
And yes, implementing the monitoring KRIs seems relatively simple at first; the problem is, as usual, in the details.
Trial design is the first of many challenges, especially regarding its connection to the planned enrollment rate. The complexity arises when we have to deal with a relatively small and slowly enrolling study (think a rare indication with a handful of sites and even fewer subjects enrolled over the course of multiple years) while also planning the efficient monitoring of a clinical trial in infectious diseases (with dozens of sites enrolling hundreds of subjects over several months).
Establishing appropriate KRI thresholds poses a problem in the slower, smaller study. One simply cannot apply traditional site performance comparisons because these slower, smaller enrolling studies do not produce the data in the volume that lends itself to standard statistics. If one patient was screened out at a single site, they would create a 100% screen out rate which, when compared to the study average, will trigger red flags everywhere. The solution is increasing collaboration with statistical teams to ensure that the established controls work properly.
Ultimately, the success of any CROs' risk-based monitoring model comes down to the knowledge of their own methodologies and an increased attention to detail. There is no "one-size fits all" in this new world, and it’s up to CROs to keep a close look – and to collaborate, adapt and improve – their own processes to remain in compliance with the new guidance.