Is There Too Much Risk in Clinical Trials?

by Tom Groves

Mention the word ‘risk’ to any patient participating in a clinical trial and it will bring up ideas of alarming side effects, the issues experienced with TGN1412 at Northwick Park in 2006 or the thalidomide experience of the 1950s and 1960s.  Mention the word ‘risk’ to monitors and CRAs working in clinical trials and that will bring up the idea of ‘Risk Based Monitoring’. Mention the word ‘risk’ to a Clinical Project Manager and this will bring up the ideas of the trial going over budget, the trial not delivering on time, the investigational drug being placed on clinical hold, resource limitations etc. The simple fact is that the word ‘risk’ means different things to different people. This can have far reaching consequences if different groups of people have a different interpretation of the same word.

So what actually is risk?

Well that is exactly the question that needs to be addressed. The definition of risk is “The possibility that something unpleasant or unwelcome will happen”.  The term “possibility” means that it may or may not happen, it certainly does not mean it WILL happen or has already happened.  If something has already happened then this is no longer a risk but instead it becomes an “issue”.  The terms “unpleasant” or “unwelcome” highlight the fact that with risks we are looking at have a negative effect. The opposite of the term “risk” is opportunity. An opportunity is the possibility of a better than expected event, for example the possibility of recruitment being completed well ahead of schedule.

Types of “Risk” in Clinical Trials

There are many classification systems for risk.  Many of these have been related to specific classifications of trials, e.g. commercial vs non-commercial.  I am not attempting here to provide a summary or overview of these and equally not looking to critically appraise these. Instead, I am attempting to provide a framework for the considerations of risks in clinical trials and to try to move away from the use of a single word of “risk”

Risks to Patient

• Safety related to investigational product, medical device and comparators used in the trial
• Safety related to specific procedures and assessments used in the clinical trial
• Failure to protect personal data

Risks to Sponsor

• Failure to obtain desired outcome from the trial (i.e. inconclusive/negative outcome)
• Exceed/run-out-of funding
• Late delivery of results of trial
• Integrity of data and/or results

Risks to the Healthcare Market

• Incorrect/unreliable results from trial
• Patients being enrolled in sub-standard clinical trials rather than being included in trials of a greater scientific value
• Resource usage reviewing trial results that are of little value

There are obviously a multitude of risk reduction techniques that can be applied to the above risks, many of which are the result of an increased financial investment in the trial, e.g. more time and scientific input to the design of the trial, larger patient population, 100% SDV and data validation.

The many people the term “Risk Based Monitoring” is the buzz-word of the moment.  Whether this is actually a risk reduction technique is questionable. Perhaps this is more of a ‘focus on the important’ approach to monitoring to increase the efficiency of resources without

compromising the quality of the trial and the trial results.  In the same sense as “Risk Based Monitoring” should we also be seeing the terms “Risk Based Data Management”, “Risk Based Statistical Analysis”, “Risk Based Protocol Design” being used more often?

In 2011 the Department of Health, Medical Research Council (MRC) and the Medicines and Healthcare Products Regulatory Agency (MHRA) published a paper entitled “Risk-adapted Approaches to the Management of Clinical Trials of Investigational Products”. This is a framework for the managing trials based on different levels of risk and these risks cover many of the areas discussed above.  Risk-adapted monitoring is just a part of this framework.  In essence the framework looks at determining the level of “risk” there is to the patient and the importance and impact of the trial result. Once these have been determined the design and management of the trial can be done to the required levels so as to address the risk. FDA also recently finalised their guidance on risk based monitoring which focuses mainly on monitoring but also mentions other aspects of trial management.  Neither document however, looks specifically at the delivery aspect of the trial, e.g. time and money, which are the some of the risks to the sponsor.

So what are the risks in clinical trials?

Patient Safety & Reliability of Results Risks

These are the risks that can be addressed through Risk-Based and Risk-Adapted approaches to clinical trials. Determine the level of risk and then design and conduct a trial to an acceptable level.  I would refer to these as Risk associated and addressed through Design, Conduct & Analysis

Delivery of the Trial to Cost & Time

These are the risks to the sponsor of time and budget over-runs. In some cases these can have major consequences (time to market, first to market, providing safety updates following approval). I would refer to these as Delivery Risks.

This therefore leaves one question. Who manages which category of risk? This is not a simple question to answer. Unless the different kinds of risk are understood, categorised and assessed from the beginning then the risks will not be managed. Patient Safety and to some extent the reliability of results have always been risk assessed as Ethics, Regulatory Authorities and Protocol Review Committees will review clinical trial protocols to ensure that these aspects are met.  The Delivery to cost and time are often a more challenging set of risks to a clinical project manager for many reasons including lack of PM experience, focus on the scientific and monitoring aspects of trials through to lack of visibility and management tools.  Failure to address risks prospectively frequently leads to the project manager dealing with “managing issues” relating to delivery, whilst ironically at the same time probably leading a project deploying “risk based monitoring”.