Practical Steps for Centralized Monitoring Success

The successful implementation of Centralized Monitoring requires meticulous planning, process reform and alignment, cross-functional expertise, and most importantly the right technology in place.

With regulators advocating greater reliance on Centralized Monitoring practices for effective study oversight, the right understanding and implementation of a Centralized Monitoring approach is essential for maximizing the benefits it can bring.

In our latest blog on the topic, Jennifer Campbell, Data Analyst at CluePoints discusses why an individualised approach is often required and how connecting all parts of the Centralized Monitoring process is crucial for success, especially when it comes to turning insights and data into actionable metrics.

Even though many studies have some risks in common, just as often there will be a few that are unique to a particular therapeutic area or a specific protocol. Variety is inherent in trial design and so approaches must be considered beyond “one size fits all”. The first step on this path is getting to know the protocol. What is the design? It is vital to listen to the study team and understand the protocol to establish study priorities.

The protocol and risk assessment are crucial for identifying the critical processes and data. This can then inform significant questions and priorities to help define what risks are key for detecting, such as critical endpoints. Where is the most impact going to be achieved? There is an ocean of data in each trial, so initial review and discussion is vital to focus on what is important and where the greatest impact will be in order to establish the most productive monitoring.

Study specific metrics can come in many flavours. Most center around an adherence to the protocol conducts of critical assessments or actions that have been taken in the study. For instance, let us take the example of critical endpoint data that has been collected in a patient assessment such as a clinician reported outcome scale. If the clinician’s evaluation of an assessment can be subjective then there may be a requirement to limit the number of changes in raters, aiming to reduce variations. The risk would be that a lack of rater consistency would degrade the reliability of the scale, undermining the quality of that endpoint. The goal is then to identify sites that are experiencing a higher level of rater changes. The metric in response to that goal would be to measure the frequency that raters are changing across subsequent patient evaluations. Other considerations for different trials may include looking for an abnormal fluctuation in evaluation scores, which might point to a different kind of risk. This would lead to a different mitigation approach and investigation by the study team.

Different steps and approaches require a variety of skills and mindsets; however, it is important to remember that not everyone within the team requires all the pieces of the puzzle. A shared goal is vital here to make a clear picture. It is crucial to understand what risks are being detected and understanding the concepts of risk assessment. Asking practical questions is a very important skill set. Everyone can contribute here.

Technical skills are another key element, including knowledge around the specifications of the database and CRF forms – the nitty gritty of how the data is represented is crucial to gathering the right ingredients for metrics. General data literacy is a good foundation, but a fine understanding of what the data looks like, how it is collected, stored, and retrievable is needed.

Lastly, some of the most important skills that are often overlooked because they are hard to measure and quantify, but are key for making a challenging process run more smoothly: communication and critical thinking. The ability to process a lot of different information and messages from multiple levels is essential to ensure clarity and that nothing is lost in translation. Everyone needs to be mindful of each other’s perspective, the jargon used and how different skills and mindsets can all work in harmony.

One challenge that can often hinder the success of Risk-Based Monitoring is a failure to understand the needs of the study. This can result in metrics that are not useful. Communication and critical thinking upfront are key to focus and prioritise on important data. It may be tempting to measure everything – all possible risks – all possible critical data of interest, however this logical instinct is often not needed. Less is more, as sometimes too many metrics can result in “risk-fatigue” – it is only possible to review so much. What is critical? What is standard? What is practical to measure? Be specific and clear. This stems from thoughtful preparation and really understanding the goals of the study to identify where the most impact can be achieved. There is often a learning cycle to be aware of and focus may change. This is OK – revisions are sometimes needed and being adaptable and flexible is key to successful metrics. This is where a continual loop of communication and feedback is vital to the process of risk monitoring.

Having metrics and data is all well and good, but the recipient of the analysis needs to be able to grasp what the potential risk situation is and what the appropriate follow-up action is. A clear and meaningful picture is necessary to arm the team with the information that they need to conduct a root-cause analysis and to determine what – if any – mitigations are needed.

Once actions are taken and outcomes are found this needs to be documented. This may take the form of a signal and action tracker. Taking a broad view of what is going on can be helpful. This comes from taking both comprehensive data quality analysis and targeted metrics hand in hand. Feedback will inform future analyses and the cycle of monitoring, for instance if something is flagged that is not important or meaningful it can be eliminated or adjusted in future statistical analyses, helping focus on what matters most for actionable insights.

Often, we need to focus big, then small, and then big again. This involves listening to the study team and understanding the protocol, critical processes, and risks for a study. Next, it is about delving into the technical details of a database, developing metrics, and applying statistical analyses. Only then can the results be translated into the context of the bigger picture of the study. For real success to happen this stage, teams may require the expertise of analysts who specialize in that middle step and are adept at connecting all three parts of the process together.