Risk assessments about participation changes
Conducting a risk assessment prior to starting a trial is considered best practice and is included in recent revisions of ICH Good Clinical Practice guidance.
In the context of planning for participation changes, risk assessment means considering the likelihood that the number or type of participation changes could impact negatively on the trial’s objectives or on participants’ rights or wellbeing.
It might be that aspects of a trial’s population, design or setup make certain sorts of participation change more likely. Factors to consider could include the length of follow-up, possible changes in participants’ health, media coverage about the trial intervention, or availability of treatments outside of the trial.
Understanding the risks (which may be high, low or unknown in terms of likelihood and impact) means trial conduct and analysis can be risk-proportionate. It also gives an opportunity to review aspects of the trial design in light of the assessed risks, such as whether potential burden on participants can be further reduced.
Using prompts to guide risk assessments
From a practical point of view, it may be useful to provide those performing each trial’s risk assessment with a list of prompts. This can help them think through risks applying to their trial, though it is important for the prompt list not to discourage trial staff from considering risks not present in the list.
Prompts might include the likelihood of many participants stopping intervention early (for any reason), many participants saying they want other aspects of their participation to stop, many participants losing contact with the trial staff or many losing capacity to consent during the trial.
Trial staff might be prompted to consider risks associated with the participation change processes themselves, and how these could be mitigated.
For example, if there are plans to try to regain contact with, or collect further data about, participants who are no longer in contact with trial staff, might these activities be seen as intrusive or unexpected?
Where intervention delivery and outcome data collection are closely linked (e.g. the intervention delivery and outcome measure collection take place at the same clinic visits), could efforts to improve visit attendance affect the generalisability of the trial results?