In the industry there is a big confusion in terminology used to describe new methods of study monitoring. At least partially this is caused by the fact that at least 3 major organizations (CTTI Project, FDA, EMA) are producing guidance documents in the area of quality in clinical trials and each touches on monitoring practices in a different way as part of an overall Quality and Risk Management approach to clinical trials. TransCelerate have delivered a Position Paper based on these guidance documents called 'Risk Based Monitoring Methodology'.
The intent of this article is to have a more practical look at the definitions. We recommend using Risk Based Monitoring term only when a monitoring approach has the three features discussed below.
Risk-based monitoring definitions in guidance documents can be perceived as vague:
The FDA writes that "[RBM] directs sponsor oversight activities on preventing or mitigating important and likely risks to data quality and to processes critical to human subject protection and trial integrity by appropriate use of centralized monitoring and reliance on technological advances."
TransCelerate, however, defines RBM as "an adaptive approach that directs monitoring focus to the evolving areas of greatest need which have the most potential to impact patient safety and data quality."
With a deeper analysis of the guidance documents while merging that with knowledge on industry practices, a clearer picture starts to develop. What becomes apparent is that RBM consists of different components:
- Quality planning where specific study risks are assessed and management plans created (in TransCelerate documents this would include RACT tool, IQRMP, critical data and processes definition, in FDA guidance this is Identify critical data and processes, Risk assessment, factors to consider when developing monitoring plan)
- Study specific monitoring plan defining including appropriate mix of central, remote and on-site monitoring activities to address the risks defined in quality planning stage (adjusting SDV to actual need belongs over here)
- Central review of data modifies execution of monitoring activities and triggering interventions on site level (FDA: definition of results that would trigger changes in planned monitoring activities)
The most common misunderstanding of RBM is taking one of the elements presented above without others as RBM implementation. For example, both monitors and leaders of big clinical projects fear that RBM is only a reduction in SDV. On the other hand, some industry RBM implementations are very focused on centralized monitoring implementation.
The RBM concept can be illustrated by a simple example:
Quality planning: Quality of local measurements [analyte X] determined as critical to quality as done locally using a desktop device and used to determine investigational product dosing in IVRS
Appropriate mix of central, remote and on-site monitoring activities: Sites' [analyte X] measurements are analysed centrally by a Centralized Monitoring analyst by comparing local and central lab results. Sites needing intervention are selected and monitors informed
Review of data modifies execution of monitoring activities: In centrally selected sites monitors perform remote and on-site interventions (retraining, equipment change)
This example also illustrates some prerequisites for RBM implementation:
- Technology - is needed to integrate data from various systems (in this case IVRS and C-lab data) and effectively communicate the findings to monitors as well as track management of the identified issues
- Real-time data entry at site and real time data management - with focus on timely data entry and cleaning enables timely central data review for trend analysis
- New monitors' capabilities - these are needed to handle centralized monitoring signals of new nature (based on probability rather than evident errors) via remote communication channels (in contrast to traditional on-site visits)
Based on the RBM definition above Centralized Monitoring (CM, also called Centralized Data Monitoring, Central Monitoring) is an activity that is a component of RBM. It is a process whereby a remote evaluation and monitoring of study data and site performance is carried out by experts at a separate location to where the clinical investigation is being conducted. The activity requires appropriate technology, process and the necessary skillset. Some companies decided to create a separate function to perform centralized monitoring.
Centralized Monitoring implementation in studies is possible without implementation of other elements of RBM. In that instance, the central study team gets a good overview of emerging data but site monitoring practices are unchanged. Such implementations should not be called RBM in our opinion.
Within the centralized monitoring area, a term that creates a lot of confusion is "statistical monitoring" or "central statistical monitoring". Classical central monitoring utilizes simple statistical models to find outliers in defined parts of data (e.g. low endpoint reporting sites). In contrast, in statistical monitoring a whole dataset is analysed using sophisticated stats engines to find outlying data patterns. A significant volume of data is needed making this type of analysis more reasonable in big studies.
Quality by Design
The term QbD is most often used by CTTI. Most commonly it is used to describe the planned approach to quality in all aspects of the study using the plan-do-check-act methodology. By partially simplifying we can say that CTTI considers RBM (or tailored monitoring) as a component of the QbD approach. On the other hand TransCelerate considers the QbD approach as a component of RBM.
Based on our experiences, using the terms QbD, RBM, CM interchangeably creates confusion within clinical organizations. Our suggestion to those on a journey to RBM implementation is to have an organizational agreement to use one term: RBM to describe the new monitoring approach. However, we suggest calling RBM only the strategies where all 3 elements (quality planning, appropriate monitoring mix, and centrally guided site interventions) are present.