How Biometrics can support a Risk Based Monitoring Approach for ICH GCP E6 (R2) Guidelines?
The clinical phase is the most complex part in a drug development process and requires efficient planning, conduct and monitoring to achieve the desired quality and obtain reliable study data for regulatory submission.
As complexities in clinical trials have increased significantly in the last few years, the clinical monitoring cost and, in turn, the trial management cost have risen significantly in order to meet strict regulatory requirements that demand higher data quality and better monitoring of patient safety.
The FDA guidance for industry and the EMA reflection paper encourages greater reliance on centralized monitoring practices, highlighting the fact that Sponsors should be “targeting” their on-site monitoring activities promoting risk mitigation and early issue detection to improve data quality and patient safety in a cost‐efficient manner.
More importantly, the ICH GCP E6 (R2) guidelines were approved in December and are set to take effect in June 2017. These guidelines represent the most important revision in the past 20 years requiring Sponsors to implement a more risk-based approach and find a way of being more targeted in their on-site monitoring activities.
The new ICH GCP guidelines have amended requirements to Quality Management, the responsibility of the CRO, trial management, data handling and record keeping, monitoring and non-compliance.
Why Centralized Monitoring
Over the years the obligation on sponsors to ensure that the study participants are adequately protected and the integrity of the study data is maintained have commonly been translated as a need for frequent on-site visits and 100% Source Data Verification (SDV). On-site monitoring is costly, it accounts for 30% of the overall cost of clinical trial management and furthermore 100% of SDV does not necessarily result in higher data quality because the CRA usually has a very narrow perspective on its own sites. They have limited ability to provide insight for data trends across time, patients, and clinical sites.
There is growing evidence to show that centralized monitoring may be more effective in detecting non-compliance and data fabrication than even 100% SDV because there are certain critical deviations that appear only when viewing the summary data with a broader perspective.
Risk based monitoring relies on a systematic process to identify, assess, control, share and review the risks associated with the clinical trial during its lifecycle. CROS NT has identified two solutions that support this approach:
- SUPERVISED method: based on RISK BASED METRICS (RBM). This approach is more data specific, and trial specific, based on established risk indicators and thresholds as this is often based on subjective clinical experience.
- UNSUPERVISED method: based on CENTRALIZED STATISTICAL MONITORING. CSM is more holistic, unsupervised. It is an approach free from defined hypothesis and uses statistical tests to ensure that the quality and integrity of data.
What is Changing in RBM
Adopting RBM is challenging. It requires business process transformations, and even organizational restructuring.
Sponsors and their research partners should begin planning for RBM at the outset of the research project to ensure processes are aligned across the organizations and determine who will handle various monitoring activities and corrective actions. According to the new ICH GCP guidelines, Sponsors ultimately hold the responsibility for non-compliance to the protocol, GCP and written procedures. In addition Sponsors and third-party providers must consider the degree of organizational change necessary to implement this methodology. This is a dynamic and proactive approach where activities of some parties change respect to the traditional workflow. For example Data Manager gains an active role in the:
- Risk assessment process (set up phase)
- Central monitoring of data for risk detection (ongoing phase)
- Sharing of study progression with the team (ongoing phase)
Also, the CRA will be able to focus its activity specifically on high risk sites identified by outcomes of RBM and CSM instead of equally distributing efforts over all the sites.
RISK BASED METRICS
In our supervised approach, designed taking into account the consideration of the TransCelerate Paper, we identified 5 risk categories that include different risk indicators associated with 28 metrics integrated in an EDC platform. Behind each risk indicators there is a rationale and the study team should evaluate which are relevant for the study.
For example the category of Patient Disposition considers:
1. Enrollment status and Enrolled patients by site: because recruitment of participants below the expected or predicted recruitment rates can jeopardize the power of the study, impose time delays and financial penalties on the trial, and can lead to premature trial closure.
2. Reason for discontinuation and dropout patient by site: because high discontinuation could be a possible safety signal and can lead to insufficient data for statistical analysis
Based on the results of the metrics, if the outcome exceeds the predefined threshold, an action is triggered with the aim to mitigate the risk. The type of actions can range from re-training the site, to calling the site with demonstrating that successful solutions do not always necessarily require an on-site visit.
CENTRALIZED STATISTICAL MONITORING
It is able to identify a lack of variability or implausible values, which would go undetected using other approaches.
Examples of analysis that can be implemented are Time-Trend Analysis, Univariate Analysis, Weekdays Analysis, Visit Scheduling Analysis, Rounding to Integers Analysis, Digit Preference Analysis, Multivariate Analysis. Once tests are performed on a large number of variables available in a clinical trial database, they become part of a high-dimensional matrix which is analyzed by statistical methods to identify extreme centers. Sites with higher risk need further actions to ensure data integrity.
Implementing RBM means start working with a proactive approach to data quality and safety to address shortcomings while the study is ongoing.
Though Risk-based metrics and CSM are different in their methodology and focus, the two strategies complement each other when used effectively and the combination of both is therefore best suited for effective identification and management of risks. Both solutions also take a proactive approach to satisfying the ICH GCP E6 (R2) guidelines by focusing on a risk-based approach to data quality and monitoring using innovative technology solutions.