How to incorporate patient and public perspectives into the design and conduct of research

Understand what patient and public involvement is

At the outset, it is important to understand the theory underpinning PPI. In depth reviews and discussion of theory are available^1,4,10–13 and suggest that, depending on the circumstances, PPI will:

All of the above could reduce research waste^14–16 if PPI is put into practice in ways that ensure that research is meaningful, acceptable, ethical and useful.

How does patient involvement differ from patient participation in research?

Patient perspectives can be sought through patient involvement and through patients participating in surveys, interviews or focus groups to provide data for others to analyse, interpret and act on. The authors have observed that in grant applications and study protocols, PPI is often conflated with qualitative research or patient opinion surveys. Collecting data from patients can be important to gain diverse or representative views, but it is different from PPI and both are often needed (Table 2). Discussion with patients at a workshop can seem similar to collecting data in a focus group, because both involve listening to patients’ perspectives, but the context and outcomes from listening differ. PPI means that researchers are in a continuing and reciprocal relationship with patients and make decisions with them about the research. In qualitative research, researchers listen to patients in order to improve their understanding of a topic. Focus group discussions or qualitative interviews are audio-recorded and transcribed. Researchers collate, analyse and interpret text data from carefully sampled patients to produce valid new knowledge and generate hypotheses. Qualitative and survey research have systematic methodological quality standards. However, the researcher holds the power and patients may express strong views which may not be reported. In any research, the PPI and the data collection to gain wider patient perspectives can be separate, combined or overlap in some study phases, or they can be completely integrated throughout (Figure 1). Any combination is possible (Supplementary File 2, Example 1). They are often combined and integrated in equitable partnership research methods like action research, and ‘co-‘ prefixes to research terms, e.g. co-learning and co-production (Box 2).

Figure 1. The interface between Public Patient Involvement (PPI), qualitative and survey research across all stakeholders in research.

Action research, co-design, co-learning and co-production

Action research historically precedes co-production and gathered momentum in the 1940’s as a community-led action in research initiative^19,20,30. The UK National Institute of Health Research (NIHR) who fund research advocate co-production²⁷ as a method of involving patients meaningfully from start to finish of the research process. Differences in definitions (Box 2) are subtle, vary internationally and researchers may apply the approaches flexibly in practice. ‘Partnership approaches’ is used in this article as an umbrella term because it acknowledges the changing roles of patients beyond being ‘participants’ or ‘subjects’. Partnership research methods involve patients, clinicians, academics and other relevant stakeholders as equal mutually respected partners in the research team. Being a patient partner implies equal opportunity and equal voice. Equal power in decision-making is sometimes implied, however there are structural and economic power differentials between different types of partner in terms of pay, employment contracts, status and workplace environments. As language is evolving internationally it is more helpful to describe actual patient roles, tasks and responsibilities explicitly rather than use a label for an approach that is open to misinterpretation. For example, co-production²⁷ may mean consulting patients regularly or patients may actively collect and interpret research data. Terms like ‘Participatory Action Research’ confuse because the definition of ‘participation’ in a study means to contribute data, rather than active involvement in research decisions. Partnership research teams decide who has access to participant level data, how to share data securely and how decisions will be made collectively. Partnership approaches can be resource intensive require leadership skills to balance equity of decision-making with a strong scientific rationale. Negotiation skills are required to accommodate different perspectives in order to reach consensus in a timely manner. An important limitation to consider is how the partnership approach is interacting with the intervention: for example action research can become an active intervention component (Supplementary File 2, Example 2).

When starting to design a study about migraine, understand how PPI will add value to the research and which uncertainties about patient perspectives might benefit from additional analysis of patient data from a survey or qualitative interviews.

Find out what research questions are priorities for patients

Many funders require researchers to justify that their research question addresses what is important to patients^31–33. If a research question is of low priority to the people affected by the condition, or important outcomes are not considered, and/or the intervention in question is considered unacceptable to patients, then further research is wasteful¹¹.

A starting point for researchers is to find out if patients’ priorities already exist for their topic. Many national and international organisations involve patients to identify and publish research priorities specific to a healthcare condition. In the UK, the James Lind Alliance (JLA) specifically identifies and prioritises research questions for funders and there is a register^34,35. JLA establish Priority Setting Partnerships which involve collaborations between patients, carers and clinicians. The NIHR funds JLA advisors and the infrastructure, but a Priority Setting Partnership is responsible for its own funding. The JLA has a guidebook which provides step by step processes to identify research uncertainties and prioritise a top 10 list for different conditions³⁶. Researchers are advised to evaluate how priorities were established and the rigour of processes, as priorities can change with time and some groups may not have had an opportunity to be involved.

If patients’ priorities are unknown, and a Priority Setting Partnership is not available, contacting the potential funder to discuss options may be helpful. When researchers plan bespoke methods to prioritise research, it is important to find patients as soon as possible to identify the topic and refine the research question to ensure relevance. For example: work with a charity or a research organisation to conduct an on-line survey (Supplementary File 2, Example 3); advertise and run open public workshops with patients to rank research priorities; or ask participants in qualitative interviews what would make a difference, then construct research scenarios for them to ‘think aloud’ which one they would prioritise. Once patients have prioritised the research topic and questions, the next step is to prioritise the outcomes that matter, patient-centered outcome measures and identify acceptable research methods.

A first step for a researcher is to search the internet for key organisations and guidelines to see if patient research priorities for migraine are available. If not, a researcher can contact migraine charities and talk to a potential funder to seek their advice.

Identify patients and/or patient groups to involve as early as possible

Researchers are advised to find people to involve and to plan potential roles, responsibilities and tasks for their study as early as possible. Research teams may approach patients through formal patient groups, charities, community groups, University or Health and Social Care patient advisory panels, national directories such as ‘People in Research’³⁷, invoDIRECT³⁸, patients who are involved in producing guidelines like The National Institute of Health and Care Excellence³⁹, or through personal recommendation or advertisement. See Supplementary File 2, Example 4. It is usually not considered appropriate to involve patients that members of the research team are currently providing clinical care to⁴⁰. In the UK, InvoDIRECT³⁸ provides an A-Z on-line resource of organisations, networks and groups that support PPI in health and social care research (Box 1).

Lay or professional coordinators or link people may help and different sources of patients may be used for different purposes. For example, a head office of a patient charity may be invited to nominate a person to join a study steering committee, whereas a local patient group may help to make recruitment materials appealing and easily understood. Participants in a preparatory survey, focus group or qualitative interview may be invited to volunteer for patient involvement in future research. The qualitative research and PPI then become synergistic.

A researcher wanting to study migraine could contact a charity, their University or Health Service patient advisory panel or consult directories of patients who are interested in being involved in research. Invite a patient link worker to join the team who will co-ordinate wider patient involvement.

Decide who and how many patients to involve

As with any appointment, selection criteria for patients based on the research plan are useful to inform decisions. Deciding the number of patients to involve in a study requires careful consideration. Two is the minimum number recommended by INVOLVE³, however international guidance is less specific. The patient characteristics, skills and numbers will vary according to:

Consider equity of opportunity, unheard perspectives and health inequalities

Equity of opportunity for patients to be involved in research underpins UK guidance. The NIHR standards for PPI⁷ provide practical examples for how researchers can offer inclusive opportunities and sustain respectful, productive relationships. There is a danger that patient contributors are atypical, as the more confident and financially secure are more likely to volunteer. It can be easier to involve older, white and educated people, which can marginalise other perspectives. Health inequalities and equity are important when making research decisions⁴². Aim to find patients who represent the demographic of those affected by the condition. It can be challenging to access ‘typical’ members of the target population for the specific research question^42–44. See Supplementary File 2, Example 5. An adult or child may be selected to represent their own views⁴⁵ or, when the research involves children, vulnerable patients or patients with cognitive impairment, then a guardian, relative or carer may represent the patient’s views. A lack of resources can hinder recruiting some patients, such as those from ethnic minorities, the less privileged and less literate. Yet this is important because they tend to experience lower health status and poorer access to services. For these patients it can feel intimidating to meet researchers and attend meetings in a University. Alternative strategies include researchers going out into the community in order to build rapport and trust with patients on their own turf, which can then lead to discussions about research (Supplementary File 2, Example 1)^46,47. An outreach model for patient involvement via a link coordinator (professional or lay) can help to access less heard perspectives (Supplementary File 2, Example 5)⁴⁸. A useful guide for getting started and arranging a meeting with patients is available on the INVOLVE website⁴⁹.

A charity partner might help a researcher to plan how patients on low-income or from ethnic minorities can contribute to a research study on migraine. Adverts, social media and attractive visual information in local newspapers and chemist shops may help.

Consider the potential for bias and conflicts of interest

PPI in research and political lobbying can co-occur and introduce conflicts of interest with the potential to influence research decisions in ways that have been under-researched⁵⁰. Researchers are advised to consider sources of funding and affiliations of patient contributors, and to re-assess any arising conflicts of interest during their study.

Patients can work with research teams over many years, attend training courses and become a ‘PPI methodologist’ or expert individual or group. This has advantages and risks. Experienced patients can have an overview of a particular health condition that is invaluable. However, becoming embedded in a research team or an organisation can risk losing the ‘eye of the public’⁵¹. Researchers are advised to consider whether bias due to ‘group think’⁵² is possible. This is a risk in any established team, for either researchers or patients to become so familiar with the group or clinical area that they lose sight of fresh perspectives. Selecting new untrained patients for a study can highlight researchers’ preconceptions and assumptions. However, this also has limitations, as it can be difficult for patients to understand, question and challenge researchers when the language and culture are unfamiliar. Patients who have benefited from or experienced adverse events from a particular treatment can introduce bias. Select patients to balance views, for example patients who have positive and negative outcomes from a new procedure or treatment. It may add rigour to include qualitative or survey research to gain diverse and/or representative patient perspectives.

Throughout all stages of a study, researchers and patients make decisions that need to balance and prioritise evidence, personal experiences and competing values held at the individual, family, organisational, political, cultural and environmental levels. Rigour and quality standards for PPI in research are important to counter critics, as there is still some resistance to implementing PPI⁵³.

A researcher is advised to consider conflicts of interest and sources of bias, for example links to industry or private companies. Seek to balance positive and negative patient experiences relevant to the study.

Negotiate and agree an approach, tasks and responsibilities at an early stage

Once patients are involved, it is advisable to agree clear boundaries about the scope of the role, specific tasks and responsibilities. Some flexibility is desirable to accommodate unexpected issues that can arise in research and there are grey areas. See Supplementary File 2, Example 6. The approach can be bespoke for each study or for each phase within a study^{12,17,44,48,54} and can vary in the level of patient engagement, responsibility and control. Patients can contribute to three key functions: research decision-making; enhancing understanding of patient experience; and advising how to capture knowledge from other patients. For each function, a question to ask is: which method for involving people will add value and rigour? Example 7 (Supplementary File 2) draws on the work of Gamble and Colleagues who have produced a useful list of tips for patient roles in clinical trials derived from a cohort study of 111 funded trials²⁸.

Be realistic about what will be possible to achieve and the resources required³. A template for Terms of Reference is available on the INVOLVE website⁴⁹. Terms of Reference acknowledge the importance of mutual respect, practical communication issues and can be reviewed as the research progresses. Researchers may invite patients to propose ground rules for the length of time required to read and respond to emails and comment on documents, for mutual agreement. It is important for researchers to remember that patients may be managing ongoing health conditions which can be unpredictable. Patients value individual constructive and honest feedback about their contributions in order to learn, gain confidence and maintain motivation⁷.

At an early stage a researcher is advised to discuss roles and tasks involved in the migraine study. For example: help to design an appealing patient leaflet, recruit patients, attend project management meetings, interpret findings and present them to lay audiences.

Agree appropriate funding for patient involvement

International arrangements for supporting patient involvement in research vary according to the funding opportunity. It is important for researchers to check current guidance for the funding call they are applying to and budgeting guidance is usually available (Table 1). Negotiate with patients the costs: payment for patient time, any special needs (e.g. childcare, hearing impairment, translation services), training, reimbursement of travel and subsistence expenses. In addition, include costs for staff time to co-ordinate, support, train and facilitate patient involvement. Researchers are advised to spell out to patients the best case and worst case scenarios (e.g. delays to study start and finish), and what contributing to the study would and could involve. Some patients prefer to volunteer, others prefer cash payment or vouchers. Consider patients who are less financially secure. Patients may rely on benefits, part time work or retirement pensions, therefore consider how difficult it is to pay upfront for travel, to scan travel tickets in order to claim research expenses or to have access to computers or printers to access documents for a meeting.

Preparatory PPI activity prior to submitting the grant application can pose a problem for researchers because funding for this is seldom available prior to a grant. Yet this is precisely when patients can have important impact on the study research question, design and plan. In England, the NIHR Research Design Service will provide small amounts of money to cover PPI at the design stage⁵⁵. Some Universities fund generic patient partnership panels (e.g.56) to work with researchers who are seeking funding and larger charities can often help⁵⁷.

When costing a study about migraine, negotiate sufficient funds to pay for the planned PPI activities, be realistic about the workload and the resources required and consider special needs.

Training for patients involved in research

Providing or offering training may or may not be appropriate depending on the patient role and the purpose of training. Training may be desirable in order to undertake highly skilled roles like reviewing grant applications or sitting on independent trial steering committees. In particular, training in the principles of evidence based medicine, with consideration of where and how patient stories fit in evidence hierarchies may be useful. Example 8 (see Supplementary File 2) provides some training programmes that support patient involvement in research. For patients new to a PPI role, support to develop their abilities and confidence to express their views and question researchers may be relevant. Many universities, research funders and charities provide learning and support activities.

There are many PPI tasks where training is not necessary, where a different perspective is what really matters and patient experience of a healthcare condition is the required expertise. For example, when helping to choose important outcomes or advising on patient information or recruitment strategies, ‘untrained’ patients may make particularly valuable contributions.

Patients doing research

Traditionally, academics with qualifications, experience and recognised research skills collect and analyse data. However, increasingly patients are helping to recruit participants, collect or analyse data and some UK grant application forms ask about this (Supplementary File 1, Section B). Such questions arguably prime researchers to think that all boxes should be ticked, without considering the implications. Only appropriately trained patients or lay people should undertake research. Shared experiences of a condition can build trust, empathy and a bond which may help to recruit difficult to engage groups, for example children in care⁴⁵. However, attention is required to individual expertise, training requirements, supervision and the scientific rigour necessary to execute high quality research. Patients may do research alongside researchers in partnership research methods⁵⁸ and a paradigm of patient-led research is emerging facilitated by social media and digital technologies⁵⁹. INVOLVE has a Patient-Led-Research-Hub to support patients who want to pursue their own research ideas³⁸.

In the UK, any researcher accessing study participants who are NHS patients or staff requires a letter of access, sometimes referred to as a ‘research passport’, obtained from the NHS Research and Development offices (Supplementary File 2, Example 9)⁶⁰. If patients or lay people help to recruit participants to research, gain informed consent or collect, share or analyse data from individual or group discussions, qualitative research or surveys, then they are ‘doing research' and there are potential governance implications for the sponsor of the research in terms of employment law, ethics, leave entitlement and indemnity. Researchers should not encourage patients to do research because it requires less resource, or because it obviates the need for relatively costly skilled researchers whilst simultaneously bypassing regulatory hurdles. Rather, researchers and patient partners can decide together whether patient researchers are appropriate and beneficial to specific research projects.

Researchers wanting to study migraine may consider the pros and cons of patients doing aspects of the research and the governance issues.

Working together ethically

Consider how to work with patients ethically. PPI can be empowering for individuals and communities, but there are tensions and risks, including exploitation²⁵, and the burden and resource implications can be considerable¹⁰. Some ethical principles for researchers to consider when involving patients in research include:

INVOLVE³ states that UK ethics committee approval is not required when patients advise research teams, prioritise research questions, make choices relating to design, share decision-making or disseminate research findings. However, there can be grey areas particularly in relation to defining ‘data collection’. NHS or University Ethics committee approval is required in the UK if personal information, i.e. data as defined in the Data Protection Act⁶¹, is collected, shared and stored for future analysis and reporting. For iterative partnership research approaches like co-production, the current ethics committee processes create many challenges⁶². Researchers can request informed consent from participants to share anonymised data with patient partners, so that they can be involved in analysis and interpretation as members of the study team.

There are international differences in requirements for research ethical and governance approvals, and particular challenges with digital health research²⁵ which are beyond the remit of this article. New EU General Data Protection Regulation⁶³ commenced in May 2018, and requires transparency about the source of personal data, the purpose and who data will be shared with.

Audio-recording of PPI meetings in order to write accurate but not verbatim minutes, does not require ethics committee approval. However, it does require at least verbal consent from all present at the start of the meeting and the recording should be destroyed as soon as the minutes are agreed. People should receive forewarning of the intention to audio-record, know the purpose, what will happen to the recording and to the content, and be able to object or withdraw. If audio-recordings are stored for longer than is necessary, transcribed verbatim or if there is an intention to report or publish potentially identifiable quotations or content arising from PPI activities, then ethics committee approval is required. Ethics committees have lay committee members, who consider the ethical issues relating to patient involvement.

A researcher wanting to study migraine should consider the ethical issues when involving patients in the design and conduct of their study. Consider patient burden, equity and power, fair and respectful arrangements, confidentiality and the purpose, processes and consequences of any data collected or stored.

Involve patients in writing a grant application

Patients sit on research prioritisation committees and funding panels, alongside clinicians and academics, to decide which research is commissioned and which grants are awarded. See Supplementary File 2, Example 10. Many UK funding panels expect to read convincing and meaningful accounts of how patients have had an impact at key stages: preparatory work to inform the planned research; writing the application form particularly the lay summary; and the proposed PPI activity during the study. Expect to be challenged if PPI appears tokenistic. It is important to consider the trade-offs between specifying a plan for PPI in a research protocol and building in some flexibility for change as the research progresses. This may be challenging in countries where regulatory approvals for amending protocols is time consuming.

Patients can help researchers to write the whole grant application in an engaging, easy to understand language. The lay summary is often one of the first sections in a grant application that funding committee members read to gain an overview of the study. Reviewers like to understand exactly what study participants will experience from start to finish. Describe PPI clearly so that the reader understands who, why, how many, how often, what methods and what impact patients have already had on the grant application and will have in contributing to future research decisions. For example, decisions about recruitment methods, intervention delivery or components, which outcomes will be primary or secondary and how to collect data. It helps to use language precisely and to understand how involving, participating, collaborating, consulting and engaging with patients in research differ (Box 2).

A patient helping to write and edit a grant application can make it clear what will happen to patients who participate and how patients will be involved from study conception to dissemination of findings.

Plan future dissemination of findings

Patients can advise on how research might have an impact on health and health care beyond an academic audience. They often have in-depth knowledge of their condition and of on-line sources of information beyond that of academics and clinicians. They can help to write reports, blogs or summaries of findings creatively. See Supplementary File 2, Example 11. Offering participants a lay summary of the research findings is good practice. ‘Patients Included Charters’ provide accreditation for involving patients in conferences and in journal publications⁶⁴ and GRIPP2 PPI reporting guidelines^29,65 are available. Involvement of patients and the public is a critical component in successful implementation of research findings into healthcare, although evidence for best practice is limited^3,66,67.

The grant application for a study about migraine may propose a public event with a charity to present the results of the study. Researchers and patient partners may give joint talks. Small group discussions with migraine patients can suggest ways to spread the news and change care.