Deviations Demystified
Clinical studies take place in the real world and for this reason departures from the protocol are inevitable. Misinterpretation, imponderables, human error, and now a pandemic, all contribute to a growing list of protocol deviations which, depending on the study, can easily number in the thousands. Clinical teams spend a lot of time collecting, reviewing and analyzing protocol deviations over the course of the study. At Acumen we write numerous clinical study reports (CSRs), and for nearly every report there seems to be some confusion around summarizing these findings.
What is an important protocol deviation? Most clinical teams we work with have a robust and adequately reviewed list of all protocol deviations. However, for the CSR, ICH guidance confirms that all important protocol deviations should be presented. This means that some additional curation of your master deviations list is needed. In its Q&A to the ICH E3 guideline, the FDA states that important protocol deviations are:
“a subset of protocol deviations that might significantly affect the completeness, accuracy, and/or reliability of the study data or that might significantly affect a subject's rights, safety, or well-being.”
By this definition, it follows that important deviations should be exceptional. In contrast, we have seen analyses where more than a third of all protocol deviations are deemed important, giving the erroneous impression that the study was not properly conducted or monitored. In our experience, important protocol deviations are few and most of them are associated with informed consent (patient safety), enrollment criteria (patient safety, reliability of data), and prohibited medications (patient safety, reliability of data). Out of window visits, or missed routine lab draws are not typically important deviations.
When should the analysis be complete? Too often, we see protocol deviations becoming a rate-limiting step for the CSR. We recommend that you perform an early analysis of all protocol deviations at database lock. Determine which ones are important, per the ICH definition, and then assess what impact these may have had on the analyses or the conduct of the study. Completing this process quickly will also expedite the generation of the programmable outputs needed for the CSR.
Lastly, deviations are not violations. A violation is an infringement; a deviation is doing something that’s different. The former implies punishment, the latter calls for examination.
Fernando Venegas
Head of Operations and Quality