Protocol Development

It is important to be proactive in considering how each trial will prepare for and manage participation changes. A main component of this preparedness is in the trial design and clear, comprehensive protocol content (see principle D2).

It is unlikely to be possible to foresee and prepare for all eventualities. Prospective risk assessment enables assessment of uncertainties and monitoring allows trial teams to see how well their processes are working).

Trial monitoring processes can ultimately lead to protocol amendments if it is agreed that existing processes could be improved.

When designing the trial follow-up procedures, consider how to reduce the burden for participants, trial sites and all others running the trial.

Consider also allowing flexibility in the timing, mode and frequency of follow-up so that participants can continue contributing with less commitment, if that suits them.

These considerations need to be balanced with the need for the trial to adequately meet its scientific objectives, and with feasibility. For example, trial follow-up cannot be so flexible that it is difficult for trial staff to manage. The potential benefits of more flexible approaches need to outweigh the costs and risks.

Consider documenting the decisions made in designing the trial follow-up methods (e.g. in the trial protocol), including where it is agreed best not to allow much or any flexibility in follow-up methods. This way, these decisions can be revisited during the trial as required (see Risk Assessment and Monitoring).

Other process documents, such as data management plans or work instructions, can go into more detail about how CTU staff should respond to notifications about participation changes as they occur in the trial.

This can include checks that the protocol instructions about managing participation changes are being followed correctly, and what advice to give trial site staff if they contact the CTU for more support in managing any specific cases. These process documents should be informed by trial risk assessments and can be important in staff training.

Points to consider including:

• The nature and extent of participation changes should be an informed and freely-given decision by the participant (see principle O2).
• If any elements of participation stop because of decisions made by others to protect participants (e.g. stopping trial intervention), these decisions should only apply to aspects of participation relevant to the decision, and not beyond these (O2). When decisions like these are made on behalf of the participant, they should be communicated and explained to the participant. The participant should have the chance to ask questions of an appropriate person so that they can understand why aspects of their participation had to stop.
  • These same points may apply if the trial design means that sometimes aspects of participation cannot continue even where the participant is willing (for example, in some early phase trials where stopping trial intervention means there will be no further trial follow-up).
  • If the protocol gives specific rules about when trial intervention should stop, it may be useful to be explicit about what follow-up activity would be expected after that, in the absence of a participant’s decision that they want that other trial activity to stop as well.
• Collecting as much as possible of a trial’s planned data can help the trial reach a reliable conclusion. This should be kept in mind by all trial staff during the trial (see principle O3).
• Loss of contact between trial staff and participants is not the same as an explicit wish to stop or reduce participation (see principle O4).
• Data collection (and any other aspects of participation where a ‘presumed ongoing consent’ approach applies*) will continue until a participant says they want it to stop (principle O5). The protocol could explain the conditions that need to apply when using this approach. It could give expectations around reviewing each specific case to check the appropriateness of the presumed ongoing consent, considering aspects like changes in the participant’s health or capacity to consent.
• Data generated until a participant says they want data collection to stop can still be collected for the trial (see principle O5). This may not apply in the same way if the data collection itself has confidentiality considerations, for example if it involves people outside the care team accessing medical notes to transfer data to trial forms. In this case, a ‘presumed ongoing consent’ approach could theoretically apply, but may not easily be done fairly and transparently.
• Data collected until such a point a participant says they want data collection to stop is retained until the end of the trial archiving period and used in trial analysis.
  • PeRSEVERE principle O6 states that this data can be made available to others outside the trial team for secondary research purposes; the protocol can state the trial position on this, including how it interacts with participants’ consent (for example, a ‘presumed ongoing consent’ approach may be appropriate).
  • The protocol could give a general instruction to follow in case of any requests (e.g. from participants) for trial data to be deleted. This may involve contacting the CTU as soon as possible to review and discuss details of the case.
• Information about how trial results will be shared with participants should make clear that results will also be made available to participants who stopped or reduced their participation, if they would like to receive those results (see principle O7).
  • Instructions can be given on how these participants’ contact details and preferences will be elicited and maintained, including to check whether these may have changed at the time participants stop or reduce their participation.
  • Participants could be offered a way, such as an online portal, to maintain their own records about their contact details and preferences. Alternatively, centralised services such as NHS Digitrials could be used to help manage participant communication.

* This means participants’ consent to a particular trial activity is presumed to be ongoing if the participant has not said they want that activity to stop, as long as certain conditions are met. See PeRSEVERE principle O5.

This may include:

• How to establish exactly what a participant wants to do when they say they want their participation to change, including their preferences for further contact about the trial.
• Expectations around how the participant’s wishes should be documented for trial purposes (e.g. in the medical notes and on trial-specific forms).
• If it would help the trial’s objectives, consider collecting trial outcome data at the time of a participation change, particularly where participants say they want no further data collected after this time. Care should be taken when communicating with participants about this.
• While participants have the right to stop taking part in the trial without having to give a reason, it is acceptable for trial staff to try to establish the participant’s motivations. The protocol can give advice about how to broach this with participants, gauging when it may not be appropriate to ask, and what sort of information is and is not useful to the trial.
• Any activity required to check participants’ safety during the time when they stop or reduce their participation, particularly in relation to stopping trial intervention. This could include a final safety assessment with the trial doctor, additional data collection by trial site staff or from routinely collected healthcare data, or specific advice to pass to the participant’s usual care teams.
• If there are plans to share any particular information or communications with participants when they stop or reduce their participation (or while they are deciding what to do), this could be explained in the protocol.

See Resources for template checklists to use when managing individual participation changes.

These instructions should be suitable to the trial and appropriate to the patient population. It may be sensible not to be too prescriptive, instead allowing trial site staff to use their own judgement for each specific participant. However, general guidance could include:

• A minimum (or maximum) number of attempts to contact the participant before stopping and considering them to be uncontactable (or uncontactable for now).
• Recommendations about how to attempt to get back in contact. This may include use of more than one contact method across the attempts where possible, if participants may be less likely to respond to certain methods (for example, some people may not answer the phone if they do not recognise the number). Each participant’s circumstances should be taken into account when planning these recontact attempts. For example, if it was known that a participant’s health was deteriorating at the last trial contact, or if they had already spoken about potentially leaving the country.
• Actions to take on establishing that a participant is now out of contact with the trial (following unsuccessful attempts to get in contact according to the protocol instructions). This could include notifying the CTU, or sending the participant a written confirmation of their change in status, if their contact details are still considered likely to be correct.
• Expectations for trying to regain contact nearer the end of the trial (only where this could be helpful to the trial’s objectives). Again, information known about each participant’s circumstances should be taken into account when deciding what to do.
• Expectations for action to take in the situation where a participant gets back in contact after a while out of contact. This could include whether the participant needs to be asked to formally restate their consent to participate (e.g. in writing) or whether their consent can be implied. This may depend on the specific circumstances in each case, so guidance could instead cover how to decide what to do. In any case, trial staff must be satisfied that the participant is happy to restart trial activities and understands what doing that would mean for them. Staff should document the interaction they had with the participant about this.
• Any processes for contacting other people or sources to regain contact and/or collect further trial data. This might include contact with GPs, family or carers, or accessing routine data sources, either centralised (e.g. NHS England) or localised (e.g. individual NHS organisations that are not participating in the trial).
  • The protocol should be clear about how participants’ consent impacts on these activities. This includes if there are any requirements for specific, prior consent to allow these activities where they might otherwise be considered intrusive. It also includes where ‘presumed ongoing consent’ approaches may apply to allow the activities to continue if certain conditions are met (see above).
  • Take care in any case to reduce the chance that these processes could be considered intrusive. Patient contributors may be able to advise on this aspect in particular.
  • Successfully obtaining data from non-trial sites (e.g. where data would be provided to trial sites, which then provide it to the CTU) can be challenging. Those other sites are not specifically resourced to do any of the work required to provide the data. Some may be unsure of the rules around collecting and sharing data in this way, or about whether it might constitute trial activity that should only happen at trial sites. Further work to explore and resolve these challenges could be useful.

See Resources for template checklists to use when managing situations where researchers and participants have lost contact.

• This may include organising elements of participation into ‘tiers’. While participation should change in line with participants’ wishes (see principle O2), in some trials where participation comprises multiple elements, there may be many ways that participation could change, in theory. Using participation tiers or layers can avoid overwhelming participants with choices and make things practically easier for those running the trial. Deviations from the tiers may still be possible if participants want their participation to change in a different way. An example might be to specify levels of participation such as:
  • 1) Participating in all aspects.
  • 2) Stopped trial intervention only, continuing other aspects.
  • 3) Stopped trial-specific clinic visits (with or without continuing questionnaire completion) but happy for data to be used from routine healthcare visits.
  • 4) Stopped trial-specific clinical visits and no longer wants data to be used from routine healthcare visits, but other ‘passive’ elements of participation continue (in line with a ‘presumed ongoing consent’ approach – see Standard Operating Procedures and PeRSEVERE principle O5).
  • 5) All aspects of participation stop but data collected already is retained.
  • 6) All aspects of participation stop and data collected already is not retained (only in very unusual circumstances; see principle O6).
  • Preferences for further contact about the trial, including receiving trial results, might not need to be tied to any particular levels of participation and could be managed independently.

• It may not be helpful to mention any of the reduced or alternative options in the initial pre-consent patient information (see Informing Potential Participants).
• The ‘tiering’ approach may include defining a minimum dataset that would still inform the main trial objective(s) for any participants whose participation changes, with a particular focus on the primary trial outcome measure and other key outcome measures.
• Another example of possible ‘flexibility’ in trial follow-up is to plan to allow participants to skip a planned trial visit where this might make it easier for them to keep contributing, and where otherwise they might want or need to stop attending all further trial visits. This could only be suitable if it still resulted in collection of adequate trial outcome data, and if it was feasible for all trial staff to manage.

This includes any specific action required by trial staff in response to those change types and what data collection would continue (in the absence of a participant’s decision that they want data collection to stop).

These instructions should be written with any defined estimands in mind (see principle D2). Statistical planning (in the protocol or Statistical Analysis Plan) should make clear if any types of participation change have trial-specific implications for the planned analysis, for example in terms of defining the outcome measures.