Article Text

other Versions

PPI, paradoxes and Plato: who's sailing the ship?
  1. Jonathan Ives,
  2. Sarah Damery,
  3. Sabi Redwod
  1. Department of Primary Care Clinical Sciences, University of Birmingham, Birmingham, UK
  1. Correspondence to Dr Jonathan Ives, Medicine, Ethics, Society & History (MESH), School of Health and Population Sciences, The University of Birmingham, 90 Vincent Drive, Edgbaston, Birmingham, B15 2TT, UK; j.c.ives{at}


Over the last decade, patient and public involvement (PPI) has become a requisite in applied health research. Some funding bodies demand explicit evidence of PPI, while others have made a commitment to developing PPI in the projects they fund. Despite being commonplace, there remains a dearth of engagement with the ethical and theoretical underpinnings of PPI processes and practices. More specifically, while there is a small (but growing) body of literature examining the effectiveness and impact of PPI, there has been relatively little reflection on whether the concept/practice of PPI is internally coherent. Here, the authors unpick a ‘paradox’ within PPI, which highlights a tension between its moral and pragmatic motivations and its implementation. The authors argue that this ‘professionalisation paradox’ means we need to rethink the practice, and purpose, of PPI in research.

  • Medical ethics
  • philosophy of the family
  • research ethics
  • fatherhood
  • ethics in medical education
  • epidemiology
  • primary care
  • research on special populations
  • sociology
  • applied and professional ethics
  • ethics committees/consultation

Statistics from


Over the last decade, patient and public involvement (PPI) has become a requisite in applied health research. Some funding bodies demand explicit evidence of PPI, while others have made a commitment to developing PPI in the projects they fund. The Research Governance Framework,1 which regulates research in the UK National Health Service, stipulates ‘[r]elevant service users and carers or their representative groups should be involved wherever possible in the design, conduct, analysis and reporting of research’. Yet, despite being commonplace, there remains a dearth of engagement with the ethical and theoretical underpinnings of PPI processes and practices. More specifically, while there is a small (but growing) body of literature examining the effectiveness and impact of PPI, there has been relatively little reflection on whether the concept/practice of PPI is internally coherent. Here, we unpick a ‘paradox’ that we have identified within PPI, which highlights a tension between its aims and implementation.

Motivation for PPI

The move towards PPI in research appears to have two broad motivations, which are illustrated in table 1. We do not intend these motivations to be considered settled accounts of the justification, purpose and process of PPI, but rather broad illustrations of two overarching rationales that encompass a range of broadly similar concerns. One set of concerns are pragmatic and outcome orientated, and the other set are process orientated. The first, outcome orientated, set of concerns is pragmatic insofar as patients and the public are said to bring unique perspectives to research, making it more appropriate to their needs and the needs of healthcare services.2–6 This rationale draws on the contention that PPI has a number of beneficial effects on the quality of research processes and outputs, and promotes more reliable, relevant research7 and is pragmatic insofar as the justification for PPI is its ‘usefulness’, where ‘usefulness’ is defined in terms of producing better quality research. This suggests an approach to PPI similar in nature to a transaction, in which the services of one party are contracted by another in order to bring necessary skills to the project. While unequivocal evidence of these pragmatic benefits is scarce, there is some indication that PPI can improve the relevance of participant information sheets and the quality of informed consent.8 Similarly, there is some evidence that PPI can lead to improvements in the research process, such as recruitment, data collection, analysis, interpretation and dissemination.9

Table 1

Rationales for patient and public involvement (PPI)

The second rationale can be characterised as ideological or process orientated, and draws on broader social and ethical narratives around democratic representation, transparency, accountability, responsibility and the redressing of power imbalances.10–13 This rationale is reflected by INVOLVE, a publicly funded national advisory group, which promotes and supports greater public involvement in health research. According to INVOLVE, PPI in research is underpinned by the notion that people who are affected by research should have a right to have a say in what, and how, research is undertaken.6 While one aspect of this ideological motivation may be a commitment to a democratic ideal, this seems underpinned by a, perhaps more fundamental, commitment to the respect for autonomy, such that it is considered a breach of autonomy for research be conducted ‘on’ people as opposed to ‘with’ them. Indeed, this form of argument is rooted in the human and civil rights movement of the 1960s and 1970s which was characterised by ‘profound suspicion and distrust of constituted authority’ (p 256),14 and a commitment to defining one's own interests and speaking for oneself. Certainly, within mental health and disability studies research, PPI has evolved as a potent challenge to the production of knowledge from an exclusionary and elitist research community. This kind of rationale is explicitly political insofar as it challenges institutions, agencies and organisations to promote people's rights to self-determination and inclusion.15 With regard to involvement in research, this approach is deeply suspicious of mainstream research methodologies for their propensity to reinforce existing power imbalances and to constitute a source for exploitation. Instead, it emphasises social change and the redistribution of power at a macro level, while challenging the micro politics of research processes and arguing for an equalisation of research relationships.

There are, of course, immediate problems with the exploitation and exclusionary lines of argument, and we might invoke a Kantian response to point out that so long as participants enter into research of their own free will, knowing what is involved, they are not being used merely as a means to an end and their autonomy is respected. This reply is problematic for two reasons. First, and perhaps obviously, we might note that the very concept of informed consent is problematic, and highlight the fact that PPI is often regarded as a way of enhancing the process of informed consent. As such, informed consent does not mitigate the need for PPI, as PPI itself might be seen as a requirement for obtaining appropriate consent. Second, and we think more importantly, it appears to ignore the wider point that taking informed consent from an individual participant does not mitigate the sense in which specific communities or populations might feel they are targeted for ‘use’ by researchers, may feel they take on the larger share of the research burden, and notwithstanding the taking of consent may not feel respected as persons by the research community, or empowered. PPI seeks to remedy this by inviting members of those communities into the research team, and treating them as valued equals who can make a significant contribution to the process beyond being a research ‘subject’ or consultant. It is noteworthy that to bring about this kind of ideological/normative benefit, much more than tokenistic inclusion or consultation is required. It requires, instead, creating a model of PPI that draws on the PPI agent as a lay expert (in relation to their experience of illness or service use) but also as a representative of the public, who can act as a conduit to mitigate concerns about exploitation and elitism and open up the research process. For this to occur, genuine and significant partnership is necessary, with no party having a greater share of the power in the relationship than the other. This approach to PPI is analogous to a cooperative venture, in which one or more parties form a reciprocal partnership, where all parties develop and manage the venture from the start, and all are involved on an equal footing. This kind of involvement is not akin to the involvement a research fellow or project manager might have. These people are employed by the principal investigator, and while the principal investigator might listen to, and value, their views, the relationship they have to the research is pragmatic and transactional. They are paid to bring their specific skills and knowledge to the project on the understanding that those skills will bring about better quality research. There is no suggestion that those people are being used merely as a means to an end in a morally repugnant way. We appreciate that principal investigators do have obligations to research staff that go beyond payment, notably for career development, and particularly for those institutions that have signed up to the CONCORDAT. Even so, we do not think this employee–employer relationship can be described as a cooperative venture in the way required to address the concerns outlined above. Despite benefits such as joint authorship (where justified) and curriculum vitae building (which are both arguably available to the PPI agent in any case), there is still a considerable differentiation of power that results in an unequal relationship.

Staley and Minogue have argued that the active inclusion of patients and the public leads to more ethically sound research.16 This follows, they argue, because PPI: (a) ensures that the acceptable limits of controversial research are defined; (b) ensures the research is useful, relevant and of high quality; and (c) improves the ethical conduct of the research itself (by improving processes of informed consent, ensuring research is respectful, etc). While we are broadly, and theoretically, in agreement (although we note that (a) and (c) are significantly more complex than the authors seem to suppose given the difficulties inherent in simplistically deriving normative moral claims from public opinion),17–19 we will now go on to outline and justify our scepticism about the extent to which these benefits can be combined in practice.

The importance of being ‘lay’…

Accessing the benefits of PPI depends on the extent to which the PPI ‘agent’ is able to bring a ‘lay’ perspective to research. A PPI agent may have a wide range of expertise generated through their experience of being a service user or member of the public, and it is this expertise that justifies their involvement in the first place. However, this expertise can be accessed though a simple consultative model. Similarly, a professional researcher who also happens to be a cancer survivor would be equally well placed to offer a service user perspective on the research s/he is conducting. The fact that a simple consultation model is not usually considered sufficient, and that it is not really considered acceptable for a professional researcher to take on the dual role of PPI agent, suggests that the real value of PPI does not lie in its ability to access lay expertise. Rather, the value of PPI lies in the capacity of the agent to be an ‘outsider’ who can reflect and comment, with relative objectivity, on the research process. Crucially, they do this in a way that is informed by their own experience of illness, disability or having accessed particular services. What is most valuable to the research is the agent's lay and outsider perspective, leading to, inter alia, the following benefits: making the language on research documentation more accessible; offering analytic/interpretive insights that may not be available to the professional researcher; and acting as a critical observer who can speak out against assumptions, prejudices or practices that are not in line with the public or patient interest. By virtue of being a lay outsider, the agent is free from concerns arising from cultural and institutional pressures (eg, to publish, to generate research income, to build a curriculum vitae, to demonstrate scholarship), which the professional researcher must always take into account.

The importance of being expert…

The notion of expertise is particularly salient when considering the implications of the cooperative model of PPI advocated by INVOLVE, in which lay people are significantly involved, and responsible for, the ‘design, conduct, analysis and reporting of research’. We contend that, in making this kind of move, we risk being drawn into a paradox that will render PPI at best ineffective and at worst tokenistic. To elucidate the argument, we first draw on Plato's20 classic argument against democratic government. Plato creates an analogy between the state and a ship, and challenges us to consider the following:

Imagine you are on a ship. The captain is in charge. He is bigger and stronger than anyone else, and can impose his orders by sheer force of will. However, he is a little deaf and has little knowledge of navigation. The crew are all individually expert in the mechanics of sailing, though ambitious to dominate the captain. Their legitimate presence on this ship, however, is justified solely insofar as they act to carry out the orders of the captain. Last, there is an expert pilot-navigator, who is a skilled seaman, and who knows how to navigate by the stars.

In Plato's analogy to a democratic state, the captain represents the public, which is able to impose its will by force of numbers, despite having no expertise in statesmanship. The crew represents the politicians, who are expert statesmen, but are obliged to execute the will of the public. The pilot-navigator represents the philosopher who, according to Plato, has the knowledge, moral expertise and wisdom to steer the ‘ship of state’ along the right course.

Plato's challenge to the democratic ideal is to ask who we would really want sailing the ship: the unskilled but forceful captain, the ambitious crew or the skilled and experienced pilot-navigator? The assumption is that we should favour the pilot-navigator. The challenge to PPI is similar: why would we want to put unskilled lay people in a position where they can direct a process about which they know and understand little? An intuitive response is that we should not allow untrained people to play a significant role in carrying out research, in a way that gives them some power and control over the research process. Rather, the public (the captain) should have a say in where the ship is going, but the expert (the pilot-navigator) is then given the rudder.

Empowerment and training as a means of facilitating PPI in research

We find intuitive the claim that people who have a significant role in the design, conduct, analysis and reporting of research must have some level of expertise, and at the very least must be conversant in the language of research and its methodologies. This much we can take as a given, for if we place unskilled and untrained people in positions of power in the research process, we take unjustifiable risks with public funds and with the safety of participants. Additionally, placing unskilled and untrained people in positions of power may lead to undue caution that might obstruct valuable and necessary research. For this reason, an unskilled and untrained member of a research team must always play ‘second fiddle’ to the professionals (similar to the trained builder and the unskilled labourer, the professional golfer and the caddy, or the cook and the kitchen hand). This does not mean that PPI should be abandoned, only that PPI agents need to be trained in research in order to be effective and equal members of the research team. Further, they must be given sufficient training so that they are empowered, and have the confidence and wherewithal to challenge and question professional researchers where necessary. Following research into the PPI experience, this kind of training and empowerment has been favoured as an appropriate response to concerns about tokenism and lack of understanding of the research process.5 This kind of empowerment comes from an understanding of the research process, its technicalities and its culture, such that the trained PPI agent becomes an insider, with insider knowledge.

Once adequately trained and empowered, there is no reason why a lay person cannot play a significant role in the design, conduct, analysis and reporting of research, and legitimately obtain a position of power in the research process such that they can bring about both the practical (outcome orientated) and ideological (process orientated) benefits of PPI.

The professionalisation paradox

This line of argument is not unproblematic. Through the process of formal training and ‘on the job’ familiarisation with research, the PPI agent will necessarily undergo some degree of professional socialisation. As a result, their ability to act as a lay person, representing the public interest, is undermined, and arguably can no longer bring the benefits of ‘layness’ to research.

The paradox thus created can be stated simply:

  • Unless PPI agents obtain adequate training, they cannot contribute substantially to the conduct of research.

  • But…

  • Once a PPI agent undergoes training, and becomes familiar enough with research to be substantially involved, their ‘lay’ status is compromised.

Once ‘trained’, a PPI agent does not lose their experience of illness, disability or accessing services, and may still bring that expertise to bear. However, having undergone something of the process of professional socialisation, and become ‘more expert’, they may be less able to authentically reflect the lay perspective. This is not a question of the lay person losing their identity as ‘lay’, but of their lay perspective being ‘tamed’ to make theirs more congruous with that of the professional researcher. Where this occurs, we might ask what the point of ‘lay’ representation is.

What next?

The solution to avoiding this paradox seems to be to accept that a truly cooperative model of PPI in research may not be achievable. If our thesis is correct, then the fully democratic, cooperative model describes an ideal that can never truly flourish, simply because the concept is not internally coherent; we need to train lay people to enable them to participate in research cooperatively, but through training they become professionalised and therefore unable to be ‘lay’.

The appropriate response is not to abandon PPI, but to recognise that the best place for a cooperative model is not during the conduct of research, but at the funding, reviewing and dissemination stages, for which specialist knowledge and training are not necessary and in which the democratic/cooperative ideal can be fully expressed, because at these stages genuine ‘layness’ is not a barrier to inclusion in the way that it is in the research process. It may be argued that even at these stages some level of expertise is required. We are, however, hesitant to extend our argument to all phases of the research process simply because the further we move away from the carrying out of research the fewer specific and unique skills are required. In fact, we are confident that PPI has a significant role to play in some stages of the research process, and our argument only really applies to the process of carrying out the research itself, because of the very specific skills that are required. We are not kicking the captain off the ship, we are merely suggesting s/he occupies a different position.

Through involvement in other stages (eg, commissioning; planning; ethical review; dissemination; evaluation; interpretation of evidence for policy and practice guidance development), the public (the captain) can tell researchers (the pilot-navigator) where they want the ‘ship of research’ to go, and then leave it to the ‘experts’ to make sure it arrives safely.

The pragmatic benefits of PPI can still be achieved during the research process, but through a transactional/consultative model. This would not generate the professionalisation paradox because the PPI agent would not require training in order to act as a lay consultant who provides a lay perspective at key stages, where evidence suggests it can bring tangible benefits. The role of the lay consultant would not be to challenge or represent the public interest (as this is achieved at the funding, review, commissioning, etc, stages, where detailed knowledge of the research process is not necessary and where public and patient values have clear relevance), but to provide a service. Such work will undoubtedly require effort on the part of the researchers to ensure that they are explicit about the input they require. The onus would be on the professional researchers to communicate effectively with the lay consultant, rather than make that person less ‘lay’.

A few objections

Before concluding, we must briefly deal with two anticipated objections. The first might comprise the claim that a PPI agent might want to be trained and might benefit personally from training. PPI agents have reported that they want training and that they can find the process challenging, rewarding and in some cases therapeutic.13 To this, we would reply that the personal benefits to the PPI agent are incidental. Unless the intention is to use PPI as a therapeutic technique, which is formally adopted and funded as such, PPI presence during the carrying out of research is legitimated by the pragmatic benefits they bring. If these benefits are undermined by training, then the need for training should be reconsidered. It is similarly important that PPI is not seen as a route to a research career. The person who sees PPI as an apprenticeship/work experience (though we only have anecdotal evidence that this does occur) has no incentive to act as a lay person, and every incentive to act as a professional and to become more expert, which compromises the integrity of PPI.

The second objection might comprise the claim that our suggested consultative model does not account for one of the ideological motivations for PPI described above—that being the need to prevent communities from feeling used and to redress power imbalances at the macrolevel. To this, we would acknowledge that our model cannot deal with that problem, but point out that neither can the cooperative model in anything other than a merely tokenistic way. This is because it only deals with a small number of individuals who are unlikely to be connected to the community in a way that is sufficient to mitigate any sense of community disenchantment. While a cooperative model may make one or two people feel valued, that does not seem enough to address the wider problem. Further than that, the way that PPI agents are selected does not seem consistent with the aims of community engagement and democratic representation. In order to address the ideological concerns here, the cooperative PPI agent would have to be democratically elected from the community in question, and would have to report to that community (and not merely the PPI community). While selection/recruitment and reporting processes could be changed to address our latter point, it seems unlikely that any model of PPI could incorporate the kind of wide-scale community inclusiveness necessary to ensure that whole communities of research participants feel engaged and valued, rather than a few (probably unrepresentative) individuals.


Although we support PPI as a strategy to make research more accountable and relevant to local communities, we are sceptical about the coherence of a fully democratic cooperative model in the context of current arrangements for the conduct of research. Despite its ideological appeal, we are concerned that the concept is not internally coherent. The paradox that lies at its heart leads us to conclude that its espoused benefits can never be fully realised. We suggest that to avoid this paradox, and to ensure that research gains in relevance and usefulness through PPI, we should not seek to train lay people in research. Rather, we should seek to involve lay people in funding decisions, prioritisation of research agendas, research governance and ethical review, etc. However, during the carrying out of research, we should acknowledge the value and honesty of a more transactional/consultative model, and reap the benefits of genuine ‘layness’.


View Abstract


  • Funding Sabi Redwood is funded by the National Institute for Health Research (NIHR) through the Birmingham and Black Country Collaborations in Leadership in Health Research and Care (CLAHRC-BBC). The views expressed in this publication are not necessarily those of the NIHR, the Department of Health, NHS Partner Trusts, or the CLAHRC-BBC.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles