Our results demonstrate a varied picture of involvement intent across different organisations. No single document addressed all five domains of 4Pi. Document readers are required to work hard to search for and disentangle the ‘why’ (strategic aim) and ‘how’ (plan of action) of involvement. Dense language has to be unpacked to gain an understanding of aspiration and direction. The documents differed in accessibility of language and content. Some were closer to Plain English and clearly explained NHS or organisational structures. Managerial and technical language with jargon and acronyms dominated the sample. Very few documents stated any meaningful detail about who was to be involved, for example by offering data about the local community or demographic information about the patient population across services. This was especially concerning when considering issues of inclusivity, equity and equality.
Lack of clarity in rationale and presentation of content
In our introduction we identified that there is limited actionable evidence to support those tasked with developing, implementing and evaluating meaningful patient and public involvement. The identified policy tensions are reflected in the documents we investigated [8, 10, 12]. There are limited, absent, and confusing aims that are rarely, if ever, logically linked to the proposed process of involvement. There are unclear rationales for involvement and no attempts to reflect on the influence of structural, positional, and political power [11, 14, 24,25,26]. Recent findings about the availability of frameworks for use in PPI in the research context revealed 65 were available, although all of them had very little transferability beyond the groups that developed them [27]. Practitioners, researchers and public contributors who share responsibility for the operationalisation and delivery of strategic aims for PPI are faced with a complex challenge.
In drawing out the key results, the purpose domain of 4Pi is the one which was fully met in eight out 15 documents. This leaves seven documents without a clearly defined purpose. If we look at how impact is addressed, only two out of 15 documents fully met our criteria. Six documents were unable to meet the criteria for impact in any way. This raises questions about if and how purpose and impact can be logically linked. These domains are essential features of contemporary involvement practice. Each may be considered separately and can be valued differently. For example, the demonstration of impact continues to invite debate [28,29,30,31,32]. There are concerns about the emphasis on impact. Those who favour an emancipatory approach to involvement may be less willing to see this as the endpoint when participation in the process itself may be as or more valuable. Expressions of purpose may reflect aspects of both instrumental and emancipatory rationales. For example, some purpose statements in the sample used language that we interpreted as framed around the need to meet policy, legislative or regulatory incentives rather than an expression of future ambition underpinned by theory and articulated through a shared vision and values.
The relationship between the theory and practice of involvement has parallels with other fields, for example QI [33]. Whilst difficulties with reporting, impact and evaluation are documented [34,35,36,37], the generation of actionable theory to evidence why we should involve, and in what way, to achieve a desired outcome requires further exploration. If we pursue the comparison with QI, we note similar gaps between academic perspectives on the theory of involvement and the intuitive knowledge and experience of those who are engaged in its practice, especially those considered to have non-expert knowledge i.e. patients, carers and service users. The responsibility for the generation of theory is perceived to be in the academic domain. The role of patients and public contributors in generating insight about theory and practice is not well understood. Jones et al. [18] invited public contributor support when analysing qualitative data in a study to investigate the reporting of QI, although this group were excluded as potential interviewees. Barber et al. [38] report the nuanced and complementary roles patients and public contributors can play alongside professionals to achieve strategic goals, support practice and spread new ideas.
Alternative to strategic documents
A gap between the theory and practice of involvement remains. This explains a commonly expressed frustration in the field when attempting to answer ‘why’ we should involve and difficulty in describing ‘how’ to do it with greater certainty that we will achieve meaningful results to improve care or strengthen research. We therefore open the debate by posing the question: ‘is there an alternative approach to construct PPI strategy to generate context specific theory, capture learning and evaluate effectiveness at the same time?’
We offer perspective from our applied health research programme where PPI is considered an integral component in supporting the translation of research evidence into practice using a QI approach [39]. We suggest the generation of programme theory. For example, the action effect method, or similar, could offer a way forward [40].
Programme theory and the action effect method
There is growing interest in the relationship between PPI and QI. Bergerum and colleagues [3] report a literature review and realist synthesis to generate programme theory for active patient involvement in QI efforts. In our own programme we have resisted the construction of a ‘traditional’ strategy in favour of testing this approach. In QI practice the identification and articulation of programme theory can support effective initiatives if carried out with rigour [40, 41]. Programme theory is described as ‘including an agreed aim, anticipated cause/effect relationships between the interventions, and the aim and measures to monitor improvement’. In the context of programme theory, the action effect method and diagram specifically ‘provides a framework to guide the execution and evaluation of a QI initiative, a focal point for other QI methods and a communication tool to engage stakeholders’ [40]. If we repeat this description replacing PPI for QI, it offers a different prospect for those required to act who wish to deliberatively study its effect and for those who see an effect (desired or undesired) and need to understand the steps that produced it. The notion of communication can include the explicit opportunity to learn together, for example through small tests of change or Plan-Do-Study-Act (PDSA) cycles. If conducted with discipline and support, PDSAs offer a way to capture both predictable and unpredictable outcomes and events that are context specific and lead to valuable emergent learning [32, 42]. Small tests of change were used to successfully build a shared learning network hosted by NIHR CLAHRC NWL [43].
The action effect diagram (Fig. 1) demonstrates, with horizontal arrows, the directions by which the ‘why’ and ‘how’ of an initiative works. While the notion of PPI as an intervention is contentious [32], we suggest that the discipline offered by articulating programme theory, for example through an action effect method, is worth considering. The action effect diagram explicitly states that all components are not within direct influence, indicating the roles of context and complexity. This frees practitioners to focus scarce resources where they are likely to have the most influence. The use of associated PDSA cycles has the potential to capture learning better. In our analysis we found very little evidence in the documents to explain ‘why’ PPI was being conducted and similarly scant identification of confident actions that might show others ‘how’. The action effect method or similar could ensure that these aspects are better identified and articulated and realistically tested rather than assumed to deliver results.
Limitations of this study
There are limitations to our research. The study is localised, however the use of documentation in PPI is widespread and will be relevant internationally and for readers from different backgrounds. Further work could be conducted to test transferability to other settings. The final sample documents were not constructed with knowledge of or alignment with 4Pi. Our intention was to bring discipline and order to the subjective interpretation of qualitative information. We cannot rule out unconscious bias in our appraisal of the sample.