An In-Depth Look at the CRA Challenge in 2016

RBM and changes in quality management are impacting study teams, CRAs and sites.

While at eXl’s 7th Proactive GCP Compliance summit, Sharon Reinhard, President of Clinical Compliance-who has more than 18 years’ experience in biopharmaceutical compliance-led a panel of experts on the impact of RBM on clinical trial quality. In this interview, Reinhard elaborates on how RBM and changes in quality management are impacting study teams, CRAs and sites.  

Moe Alsumidaie: How are centralized and risk-based monitoring impacting study sites and CRAs?

Sharon Reinhard: In theory, centralized and risk-based monitoring should have a positive effect on study sites. If sponsors are leveraging all the electronic sources of data such as central labs, ECGs, eDiary/ePRO, central pathology, less data is being collected at the trial sites, thus less data collection and entry for site personnel, and less data to be monitored on-site by CRAs. This should allow the CRA to focus on key GCP activities such as evaluating the quality of the site’s overall performance and source documents, ensuring the correct patients were enrolled according to inclusion/exclusion criteria, evaluating the informed consent and drug accountability processes and ensuring all adverse events are being identified and reported in a timely manner.

However, in conducting several dozen audits in the last 18 months, sites are indicating that they are frustrated with these monitoring techniques because numerous sponsors/CROs have elected to stop or greatly reduce on-site visits, asking study personnel to scan volumes and volumes of source data, consent forms and regulatory documents so that they can review the information “remotely.” In order to scan the source documents, the source must first be de-identified to adhere to HIPAA/privacy regulations; remote monitoring is not RBM, and the process not only places a significant budgetary and resource strain on sites, but also increases the risk of errors with incomplete redaction.  

I do not believe this was the spirit or intent of regulatory agencies encouraging and in some cases requesting a risk-based approach. I believe the regulators are trying to emphasize that not all data and errors are equal, thus, the industry needs to think more carefully about the best way to collect, access, review and analyze that data in a way that ensures patient safety and data integrity.

MA: How is the role of the CRA changing? How do you envision the role of the CRA in five and 10 years from now?

SR: The role of the CRA has evolved a lot over the last two decades and the changes have not been for the better. When I joined industry 18 years ago, CRAs had steady employment, received a lot of training and seemed to enjoy their jobs. They felt as though they were part of the study team and took a bit of ownership in the outcome of the trial. Site personnel viewed CRAs as an extension of the sponsor and trusting, long-term relationships were developed between CRAs and sites.

About 10 years ago, mainly due to financial and headcount constraints, the industry began to eliminate internal monitoring divisions and moved to an outsourcing model. Although I’ve seen many successful trials run in an outsourcing model and the model is scalable, there are also pitfalls. For example, CRAs are assigned to multiple sponsors across multiple therapeutic areas, which has resulted in no continuity with therapeutic experience or relationships with trial sites.

Similar to large biopharmaceutical enterprises, smaller companies are also heavily relying on the outsourcing model, however, with a unique characteristic: lean sponsor teams. In an informal survey we conducted at the eXL Proactive GCP conference, more than 80% of companies are outsourcing monitoring or utilizing a functional service provider to source monitors.

In the outsourcing model, it seems to me that the emphasis changed from a CRA having a holistic understanding of what was going on at their research sites to operating in a rather automated environment where they really only scratch the surface of understanding what is going on. For example, CRAs tend to focus efforts on non-important study risks, such as an outdated CV, rather than critical study risks, such as safety and efficacy measures.

RBM requires the ability for CRAs to prioritize tasks and evaluate the overall performance of a site, which involves taking an auditing approach to monitoring. Audits continue to be successful tools for identifying errors and risks in trials, as auditors go into a trial site and find many issues that were missed by the monitors. This is because auditors are focused on identifying errors with critical data, and not preoccupied with less impactful tasks, such as 100% SDV.   

In five to 10 years, I believe that monitors will be trained to think and perform like auditors and relationship managers. Monitoring will still remain a necessary part of the industry, as important GCP concepts such as delegation of PI responsibilities and drug accountability cannot be performed remotely. However, monitors will be trained to evaluate the operations of a site, thus providing a far more reliable and valuable input into risk-based systems.

MA: What current training aspects are lacking with CRAs, and how are they impacting clinical trial quality?

SR: Many years ago, in order to be a monitor, you had to go through several types of training. There was therapeutic training that provided baseline knowledge of the anatomy and physiology one should know in order to read medical charts. Training was provided on the investigational product so that the mechanism of action, anticipated side effects or other scientifically relevant information was shared. This training also included a review of non-clinical and clinical data known to date.

Monitors were educated about the scientific measures used in a study, such as a rating scale, a clinical lab or a pathology outcome. Aside from the more therapeutically focused training, monitors were taught job skills such as how to review a medical chart so that the review not only ensured the data on the CRF was accurate but also required some logic checking. I do not sense that operations teams are conducting these types of trainings today.

To elaborate, I have recently audited trials where blood pressures of 180/110 were verified by the monitor, and the monitor didn’t ask the PI if that BP should be considered an adverse event. No clinical significance was recorded in the source. The monitor didn’t think to see if there was a history of hypertension in the medical history or to see if the blood pressure had been within normal limits initially but gradually increased. Could this be a drug effect? Today, monitors are being trained simply to verify that the data in the source matches the data in the CRF.

Additionally, monitors aren’t trained on time management. There has to be clear understanding that the six to eight hours on site cannot be spent strictly performing SDV. Monitors must be able to quickly evaluate if there have been changes in site personnel, ensure the associated training and regulatory documents have been updated, ensure the drug accountability is being performed, ensure patients were consented properly, assess how work is being delegated, ensure the source meets ALCOA standards and that adverse events are being evaluated and reported in a timely manner. Without this kind of training, the quality of monitoring diminishes significantly.

MA: How can the industry change as it relates to incorporating RBM with quality management systems?

SR: I think the industry has to move in this direction; people in quality management see it but I think something is lost in translation when it comes to the functional areas such as clinical operations, data management, programming and pharmacovigilance. Timelines are becoming more and more aggressive. Headcount and budgets are growing leaner, yet the regulatory standards are not being reduced. The system has to be more efficient.

The switch is going to require a proactive approach to understanding where the data is being collected and how to access it. Very little of the key primary and secondary endpoints are being collected at the site. With the development of electronic diaries to capture rating scales, patient reported drug compliance and other endpoints such as clinical labs and pathology, we can get at the data much sooner than an on-site monitoring visit. Through programming, we can pull out outliers and trends. However, nothing can replace a monitor going on-site, periodically, to ensure the overall performance of a site is acceptable with regards to GCP standards.

The industry needs to redefine the roles we have working on clinical trials. We need programmers who can pull this data out well before programming tables, listings and figure/graphs and we need medically qualified people reviewing data outliers and trends, weekly and monthly, not at database lock. We also need hybrid roles that are doing some cross logic checking and evaluating the data for any safety trends or errors in interpretation of the protocol, which could be detected through these types of efforts.

Before we can do any of that, though, we need to think about how we are capturing the data in order to fully leverage eClinical technologies. CRFs need to be designed to capture information, so that errors and outliers can be found. Through proper EDC programming and RBM, the sponsor can easily determine if the data support compliance. This approach allows study teams to detect and alert sites to issues early, such as a patient over-or under-dosing sooner than the site might realize or sooner than that eight to 12 week monitoring visit.

A Quality Management System can work much more effectively if monitoring findings infuse a rich source of data or perspective into the QMS. If monitors were trained to work like auditors, the input would be more robust. If we are having a failure on the front line, so to say, where the monitoring is failing to detect the risk, then the input into the QMS is not thorough or accurate and risk cannot be appropriately or adequately assessed. Without proper training and leveraging the technology, we will continue to struggle.

If you’re interested in assessing and optimizing quality management, ExL's 7th Annual Clinical Quality Oversight Forum is taking place October 17-19 in Philadelphia.