The growth of clinical research is extraordinary and the number of clinical trials being approved to run is increasing rapidly in the world. Therefore, there is an increase in the complexity of newer trials as well as the cost of monitoring them. This increasing complexity leads to clinical sites generating more errors during data collection, thus causing delays in the data-cleaning process. Furthermore, there is a higher chance for deviations to occur and be unnoticed at a clinical site. These deviations may directly affect patient safety and data integrity, and is in violation of Good Clinical Practice (GCP). Considering this development, most pharmaceutical companies (or sponsors) are moving towards a new monitoring approach called the Risk-Based Monitoring (RBM) model. For RBM to be defined, a comparative analysis needs to be made between RBM and the traditional monitoring model.
One of the benefits of RBM is better resource management. This includes both costs and labor. Traditional monitoring involves 100% source document verification (SDV) whereby case report forms are compared directly with source documents on site. A clinical research associate (CRA) acting on behalf of the sponsor is required to carry out the SDV. Despite 100% SDV being the oldest way of verifying data collected during a clinical trial at a site, the scope of this activity gives very little insight into a clinical site’s daily processes and day-to-day tasks. Clinical sites tend to prioritize case report form completion (paper or electronic entry) rather than maintaining source documents that fulfill Good Documentation Practices (GDP). This approach has proven to be more time-intensive and resource-dependent. On the other hand, RBM allocates time and resources to each clinical site, based on a calculated individual site risk assessment, whereby sites given a higher risk assessment will require larger resources. RBM is structured in such a way that different risk management approaches are employed to tackle the varying complexities of different trials. One fundamental basis of RBM is that during the progress of a study, as more data is collected and analyzed during the life of a trial, the monitoring strategy for the trial may change to focus on more pivotal elements of that study. Not all data-points are SDV-ed during an RBM visit. This allows a CRA to focus only on study-specific critical data that may affect safety evaluations and endpoints. However, this does not mean data that is not required to be SDV is unimportant or does not need to be collected. This just means that based on the RBM model, a higher emphasis is given to the critical data, hence the need to SDV these data points.
Most sites are familiar with the term SDV but are unaccustomed to the term Source Document Review (SDR). Via SDR, source documents are reviewed thoroughly to assess its quality while ensuring adequacy and adoption of appropriate critical processes. For example, the documentation of a complete informed consent process. The RBM model emphasizes an integrated combination of SDR and targeted SDV to reveal a bigger picture of what happens at a site. Even though SDV is targeted in RBM (<100%), all source data should still be reviewed. If documentation appears to be inadequate, this prompts a CRA to verify it against what is recorded in CRF (even if it is not a targeted SDV-able data point). A discrepancy would question whether a finding is a recurring issue at that site. The risk of potential fraud cannot be ignored (for example, identical vital signs across different patients or across different visits of the same patient). Therefore, RBM allows more focus on protocol-specific processes and significant data points while identifying potential risks effectively as compared to the traditional monitoring model.
Aside from targeted SDV, RBM employs a less frequent monitoring schedule by a CRA (visit frequencies can range from 12-16 weeks) as opposed to the traditional model (average frequency ranges from 4-10 weeks). This would be dependent on the therapeutic area and risk profile of the site. Interestingly, the frequency and duration of a visit are also affected by the number of enrolled patients at the site. Some RBM trials may even require more frequent CRA monitoring visits, triggered by significant events that affect study end-points (adverse events or trial milestones such as the recruitment phase). The traditional model’s visit frequency is more labor-intensive and requires a higher cost. Contrariwise, RBM results in a higher benefit-cost ratio, while allowing more flexibility and better time management when monitoring sites.
Since its inception, investigators have raised many concerns with RBM. One of these is the less frequent on-site visits by CRAs. Investigators feel more comfortable with the presence and guidance of a sponsor CRA and the RBM’s visit frequency may not be suitable to their needs. However, RBM also implements an activity called remote monitoring, allowing a CRA to remain in contact with their sites regularly, in between scheduled on-site visits. Previous on-site visit activities and follow-ups can now be discussed during remote monitoring contacts (for example, CRF data discrepancies and queries, and providing documents for signing or filing via email or fax). These remote visits are official and conducting them would require adherence to relevant SOPs (for example, writing reports following such visits or providing visit follow-up letters to the site). Another concern from site is the increased labor from site staff in complying with remote monitoring contacts, such as scanning and filing of documents and more administrative tasks to cover the absence of a CRA. Through years of implementing the RBM approach in phases, it was discovered that the workload of site staff has remained roughly the same.
The RBM methodology encourages CRAs to employ subtle ways of investigating a site’s clinical process. This is particularly useful when the CRA has no access to hard-copy source documents, such as during a remote monitoring visit. For example, instead of asking the site, “Was the informed consent process carried out for this patient?”, which is a simple yes or no question, a CRA may ask, “How was the informed consent carried out and documented for this patient?”, allowing insight into the site’s understanding of the informed consent process. From asking open-ended questions, CRAs can identify any deficiencies in the site’s process in obtaining informed consent.
Part of the RBM model, Sponsors may also implement a centralized monitoring approach. Like remote monitoring, centralized monitoring evaluates study data and trends remotely at a location other than the site. Usually carried out centrally by a team of medical monitors, data managers and statisticians, they detect data trends across CRF pages as well as issues and findings identified at a site level (for example, repeated late re-consenting of a subject using amended ICFs). They also assess the quality and consistency of data collected, and identify missing data and outliers. Lastly, they monitor key performance indicators of a site (for example, the screen failure rate or number of deviations noted), which eventually contribute to its risk assessment profile. Centralized monitoring is integral in driving the success of RBM trials as it complements local monitoring done by CRAs.
RBM is an ever-changing method of managing clinical trials. By leveraging on a wide array of tools and techniques, the quality of clinical trial data and subject safety can be secured for the better. As trial data becomes progressively more electronic, more and more trials are routinely managed via RBM to alleviate high costs and mitigate risk. The favorable cost-effectiveness ratio of an RBM trial eventually contributes to the market and retail price of the approved drug in the future. Therefore, it is in the best interest of sponsors to adopt RBM in their trial management strategies.
By Jeremy-Marc Morais, Clinical Research Associate 1, PRAHS Strategic Solutions, PRA Health Sciences, Performing Services as Site Manager on behalf of Janssen-Cilag Pty Ltd, Global Clinical Operations