The right to stop taking part
All guidelines, policies and laws about clinical trials and other research with human participants are clear that, before taking part in research, individuals* must voluntarily give informed consent, and that they can withdraw that consent at any time after having given it, without providing a reason and without negative consequences for them.
In real terms, study participants can therefore stop participating whenever they like.
Participation is often made of multiple elements
Participation in a study may often be made up of different elements, for example undergoing the medical treatment that the study is evaluating, attending hospital visits for study-specific medical assessments, completing study-specific questionnaires, and so on.
It is possible for some of these to stop while others continue (with some limitations – for example it would not be safe for a study participant to continue receiving study treatment while stopping medical assessments intended to monitor their safety).
Participants who decide
We argue that the spirit of the right to withdraw consent suggests that individual study participants must stay in charge of their own destiny. It should therefore be they who decide which elements of study participation they are happy to continue and which they would like to stop, rather than being given an all-or-nothing choice.
For example, if it is possible in a given study for participants to stop study-specific visits but continue providing relevant data from routine healthcare visits, participants should be given the choice to accept or refuse this arrangement, and they should be given adequate information to inform this choice.
The only exception to this is where others need to take action to protect study participants, for example if those responsible for the participant’s clinical care feel it is in their best interests for them to stop receiving study treatment.
Balancing conflicting priorities
The third principle of the ICH Guideline for Good Clinical Practice says: “The rights, safety, and well-being of the [study participants] are the most important considerations and should prevail over interests of science and society.”
This points to the potential for conflict between the needs of individual study participants wanting to stop participating, and the needs of the study and the robustness of its results (“science and society”), which will be negatively affected by participants stopping participation earlier than was planned in the study protocol.
Despite this, we argue there is at least theoretically a way to prepare and run studies so that participants’ rights are not compromised but also the negative impact on the study results is kept to a minimum (while not being completely eradicated).
We must aim to achieve this for at least the following reasons:
- a) the results of clinical studies are used to inform and improve healthcare for the rest of society, so they must be as useful and reliable as possible, and
- b) when a study does not produce useful results (this means we don’t learn anything new from the results, regardless of whether what we learn is ‘positive’ or ‘negative’**), the study’s participants have given their time and been exposed to any risks associated with study participation for no benefit to society.
The PeRSEVERE principles are intended to guide how study participants’ right to withdraw their informed consent should be applied in practice so that we can achieve the following aims:
- Do the best by individual study participants: individual study participants should get the best possible information to inform their choices about ongoing study participation, and be as involved as possible in any decisions that are made about their participation. All decisions to change (or not change) the nature of their study participation must be informed and freely-made, in the same way as initial consent.
- Do the best by the study, and by research and society in general: the actual or perceived robustness of a study’s results should not be impaired by individual study participants’ decisions to stop participating unless it is completely unavoidable; this protects those patients in future whose treatment might be affected by the study’s results. Studies should be transparently reported so that it is clear what changes in participation occurred, and how these changes were handled in the study analyses.
How we developed the PeRSEVERE principles
Our principles were initially developed through extensive discussion and debate within our collaborative group (see full list of contributors [add link]) which was formed through the UKCRC Registered Clinical Trials Unit Network and which includes patients, statisticians, methodologists, clinical trial managers, data managers, and specialists in quality assurance, research regulation and bioethics.
Our draft principles were then tested through a public consultation run in 2021, aiming in particular to gather additional feedback on the principles’ clarity, feasibility, novelty and acceptability.
Throughout the process of developing our principles, we have been guided by the high-level principles of ethical research conduct (including good clinical practice), our knowledge and understanding of existing clinical research regulations and guidance, and our collective experience of designing, running, analysing, reporting on and participating in clinical trials and other research.
The need to be proactive
In general, we suggest it is best to be proactive in ensuring the above aims are achieved. This means designing studies and preparing to run them in the right way, rather than only reacting to problems as they occur. We aimed to keep this in mind as we have developed the principles.
A team effort
Everyone within research teams can play a part in ensuring that studies are designed, prepared and run in the right way, but we suggest there may be particularly important roles for patients involved in running and overseeing a study (to ensure all actions taken and processes implemented are fair, appropriate and uphold study participants’ rights) and statisticians, methodologists and trialists (to ensure that all actions and processes support the study objectives and uphold study integrity).
Where PeRSEVERE fits with existing practice
We acknowledge that some of our principles are already reasonably well-established in clinical research (at least in the UK). However, we consider it important to present them all as a complete, coherent set. We are not aware of any previous attempts to comprehensively define, as we have done here, how the ethical right of study participants to withdraw their consent should be put into practice.
Other relevant background information guiding our principles has included:
- The SPIRIT statement, which says what should be included in clinical trial protocols.
- The CONSORT statement, which says how clinical trials should be reported clearly and transparently.
- Revision 1 (R1) of the ICH guidance document E9: “addendum on estimands and sensitivity analysis in clinical trials to the guideline on statistical principles for clinical trials”, and other sources on the use of ‘estimands’ in clinical trials. The estimand concept is important in handling participation changes in clinical trials, as it requires researchers to describe a highly specific research question up front, and then design, conduct and analyse their study with this research question in mind.
- Current data protection laws in the European Union and UK (the EU and UK General Data Protection Regulation [GDPR] and the UK 2018 Data Protection Act). These give special protection to data used in research, including to allow researchers to keep data after a study participant decides to stop participating in a study, as long as certain conditions are met. These laws also mean that potential study participants must be clearly informed, before they join a study, about how their data will be used in the study, including about any limitations on their usual data protection rights and about any possible international data transfers.
- Laws governing how to run studies involving medicines mandate that all data collected in a study must be kept for a long time after the end of the study (in some cases, as long as 25 years). This includes data about study participants who stopped participating early. This is so that it is possible for the regulatory authorities to check the results of the study are accurate at any time during this period. This is particularly important when a study led to changes in the standard of care for a particular health condition.
Other things to note when using the PeRSEVERE principles
The PeRSEVERE collaborator group is mostly based in the UK and we do not claim to be experts in laws and policies applying to research in other areas of the world.
We are also conscious of the breadth of research activity globally, including settings and study types that none of us have experience of.
We aimed for our principles to be general enough to be able to apply, at least in theory, to any setting.
However, we certainly encourage others to apply our principles with some flexibility to account for each specific study, considering factors such as the patient group taking part in the study, the study’s characteristics (e.g. the number of participants taking part, the study design and so on), and the study-specific risks to patient safety and study integrity.
* Here, this means individuals with capacity to give informed consent – see ‘Scope and limitations’ [add link].
** A trial could reliably show that a new treatment is not better than an existing treatment. This might be disappointing, but this would still be something new we had learned, and would be a useful result. The worst outcome from a trial is that the results are not reliable and so we have learned nothing new.