Pharmaceutical companies and contract research organizations (CROs) are increasingly trying to leverage technology to optimize risk-based monitoring. While technology is a critical component, roles also need to be looked at again. Recently we interviewed Dr. Nimita Limaye, an expert in risk-based monitoring, about the future of the monitoring role. In her opinion, we are now witnessing the emergence of a new role in clinical research—the “risk monitor”. Nimita sees the role of the risk monitor as resulting from the merging of roles across data management, clinical analytics and clinical monitoring into a new and vitally important role for the success of a clinical trial.
What do you really mean by the term “risk monitor”?
Nimita: In my opinion, this is a new and an evolving role—the risk monitoring lead has the capabilities to analyze site performance, perform subject review for critical risk indicators and review data visualizations to determine study risk. The role demands a new set of skills and knowledge and is a far more enriched role. Whether you manage risk for a clinical trial or you manage risk when you drive, the logic is fundamentally the same. While driving, you can get hit by someone from behind, but you also have to watch out for someone who may hit you head-on; so, you have to look out for both. Similarly, you cannot have different people monitoring different aspects of risk in a stand-alone manner. One needs to have a holistic view of the risk. Further, as people are acquiring new skills, it is extremely important that they undergo an assessment to demonstrate proficiency in the newly acquired skill sets. Change management also plays a key role here. Lastly, the new role also helps provide a distinct career path for young, aspiring, data managers, CRAs and clinical analysts.
How should the new role be upskilled?
Nimita: While the new role can be trained technically to monitor risk across sites and subjects, what is needed in addition is the development of risk acumen to ensure that a risk monitor knows how to assess risk, how to select key risk indicators (KRIs), which visualizations to look at and how to interpret them, how to assess the impact, probability, detectability (IPD) and derive a risk probability number (RPN), and how to look at process and audit compliance. The new risk monitors need to be aware of the ICHE6 R2 guidelines and their implications for RBM, what the leading bodies such as Transcelerate Biopharma and regulators such as the FDA and EMA are recommending, and what failure mode effect analysis (FMEA) is all about. Every effort should be taken by organizations to raise the bar and move team players from transactional to insightful players.
Can the training of a data manager lead to a better understanding of what is happening on the site?
Nimita: Data managers, if they want to step up as risk monitors, need to learn to look across data in its totality, understanding the global stature of a clinical trial and the multiple components that impact each other. The industry lacks experienced people who have the skills to assess study-related risks, by blending in risk management aspects, across data management and site monitoring. Many times silos and structures in organizations result in different team players assessing different aspects of risk. That can cause serious damage, with some key data points being missed out totally. In addition, while assessing risk, one needs to look at the flow of data across the duration of the trial as it is not meaningful to carve out specific segments of data to be reviewed by different team members playing different roles at different time points. This actually adds to study risk. It is also important to detect risk real-time and while it may be cost effective to have junior team members review different pieces of information, with the assumption that someone else will perform the holistic review at a later stage, this may well be too late.
The challenge today lies in the fact that a CRA does not understand the subject data aspects, and hence may not be able to see the bigger picture; and the situation applies vice versa to the data manager as well.
This is an emerging opportunity for data management to understand and guide monitoring on how to leverage risk analytics. The role of the risk monitor involves the assessment of trends, outliers, and the evaluation of risk parameters across countries, sites and patients. Predictive analytics need to be leveraged to forecast future risk.
What role does historical data play in this process?
Nimita: Actually, it is criminal not to use past data for future analysis. Leveraging past data helps in the assessment of risk probabilities for the future, especially when one looks across a particular therapeutic area, a molecule, or a site. There is so much data sitting out there—how often do we mine that data? We are missing out on a significant amount of intelligence if we ignore historic data.
Nowadays, centralized risk monitoring teams try to generate an integrated view of risks, by pulling data from various applications, such as EDC, CTMS, labs, etc. Different tools are also developed for planning the SDV strategy, the site risk tiering strategy, the generation of an integrated quality risk management plan, a subject data review plan and so on and so forth. All of these typically leverage data only of the particular trial and do not leverage historic data from prior studies. While this needs to be done with caution and one cannot randomly draw parallels, I think the full potential of RBM has been far from tapped.
Do all central monitors understand what charts show?
Nimita: Not always. However, education and mentoring play a key role in addressing this. Many times resources are not trained adequately.RBM is not a transactional activity and team members need to understand the study protocol, the therapeutic area, why certain parameters have been identified as KRIs and what is the scientific basis upon which thresholds have been defined. They need to know how to interpret visualizations and how to look across different data points to draw meaningful insights. As younger and newer players come into the picture, it is important that processes and guidelines are well delineated to eliminate personal variance in view points and to establish sound and consistent outcomes. Adequate training and mentoring is essential, otherwise we are actually putting studies at risk.
How do you deal with an overflow of action items, too many escalations, if that happens?
Nimita: Sometimes there are so many KRIs defined and so many actions that the study team starts ignoring them. This may result in an important escalation of a KRI not being actioned. Thus, we need to be more selective regarding which KRIs we include. Here, the rule of ‘less is more’, works. Of course, the basic ones are needed anyway, such as rates of reporting of AEs, SAEs, etc. However, unnecessary triggers should be avoided.
One can create composite KRIs to minimize multiple triggers from firing. However, what is often not done is the assignment of weightages to individual parameters in composite KRIs. Giving every parameter the same weightage is not advisable, as the risk implications of different parameters do vary.
How do you define thresholds for KRIs?
Nimita: Thresholds are usually defined by a consensus of several stakeholders; they are a combination of individual experience and past industry metrics.
Are risk monitors allowed to change thresholds for KRIs during a trial?
Nimita: Yes, thresholds for some key data points do need to be dynamic. If a steering panel decides to change thresholds for certain KRIs based on some justification (for example a protocol amendment, new regulations, some changes at the site), then this must happen.
Should risk monitors apply different KRIs/thresholds for different countries?
Nimita: This would vary on a case-by-case basis. For example, if there is a reason to believe that there are certain local politico-regulatory changes that may impact the risk of the trial, then these need to be factored in. For example, if a regulatory agency demands that 100% SDV should be performed for all sites, then one can effectively address this by assigning the highest risk for all the sites in that location, thus ensuring that 100% SDV is carried out. Country-specific and state-specific triggers can be implemented to identify debarred investigators as well.
A new and a critical ‘risk monitor’ role is evolving. Industry needs to invest to ensure that people are trained and capabilities are assessed before these resources are assigned to projects. Career paths need to be created to ensure retention of such skilled resources. At all times it should be kept in mind that RBM is not a transactional activity and that an understanding of critical concepts, recent guidance and industry trends is important. Finally, this is about patient safety and study risk—a lot is at stake.
About Nimita Limaye:
Dr. Nimita Limaye, Principal Consultant, Nymro Clinical Consulting Services, has over twenty years of experience working across the Pharma CRO industry, playing key leadership roles at companies such as Quintiles, TCS, SIRO, Altana Pharma, Cadila Healthcare, etc. She has led global operations, managed strategic relationships, and has led risk-based monitoring, medical writing, CDM, clinical analytics, and healthcare offerings. She has led teams of domain experts and has played a key role in contributing to technology roadmaps, data visualization strategies, helping to define key risk indicators and to providing key insights for platform development. She has chaired multiple conferences, presented at diverse global forums and has multiple publications to her credit. She was on the editorial board of the Journal of Applied and Translational Genomics, has given a keynote talk in London on ‘Disruptive Innovation in Clinical Trials’ and a chapter has been authored by her in the book ‘How India Found Its Feet: The Story of India Business Leadership and Value Creation 1991–2010’. She is also the past chair of the Society of Clinical Data Management (SCDM), and the first person from India to chair this association. She recently presented at the SCDM Leadership Forum and Conference in San Diego on RBM strategy and implementation challenges.
MyRBQM® Portal's new Machine Learning predictive analytics model is far less sensitive to noise, it better generalizes data, and provides narrower CIs in long time series. Singular Spectrum Analysis is particularly valuable for long time series, as mostly observed in clinical trials.
We've created this "ask the experts" Q&A to provide clear responses to the questions that were most often asked by the forum's participants.
Frontiers Health Ecosystem Companies Fighting Back the COVID-19 Outbreak As the COVID-19 outbreak continues, it has become even more challenging for patients to seek in- person medical care and for healthcare providers (HCP) to [...]
Cyntegrity developed a special COVID-19 RACT that guides those involved in clinical trials on specific risks which may arise as a result of COVID-19, and how they can mitigate them. This special RACT is part of the COVID-19 rapid deployment pack for ad-hoc centralized monitoring.
We can help establish near real-time centralized monitoring, enable the use of "ad-hoc crisis management" Key Risk Indicators (KRIs), and mitigate patient safety and study quality risks.
This week Applied Clinical Trials published an interesting case study of the risk-based quality management learning curve. The provision of a set of training modules helped Merz Pharma to initiate a smooth RBQM rollout.