Background

The EQUATOR network has developed high standard reporting guidelines such as the CONSORT (Consolidated standards of Reporting Trials) statement and the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement enhancing the quality of research reporting, but no guidance has been developed specifically for the reporting of patient and public involvement (PPI). This prompted the development of the original Guidance for Reporting Involvement of Patients and the Public (GRIPP), which tackled inconsistent reporting by helping researchers, patients, carers, and the public to improve the quality, consistency, and transparency of PPI reporting, to strengthen the quality of the international PPI evidence base [1]. While the original GRIPP checklist represented an important starting point in creating high quality PPI reporting, its development drew on systematic review evidence, without broader input from the international PPI research community [2,3,4]. Achieving consensus is now acknowledged as a crucial step in producing a reporting guideline [5]. GRIPP2 tackled this gap by developing consensus in the international PPI community.

INVOLVE defines public involvement in research as being carried out with or by members of the public rather than to, about, or for them. PPI in research can improve the relevance and overall quality of research, by ensuring that it focuses on the issues of importance to patients [1]. This includes, for example, working with research funders to prioritise research; the development of more patient relevant research questions, study designs, and outcomes; offering the patient perspective as members of a project steering group; commenting on and developing research materials to improve readability; assisting with recruitment to studies; lay write up of the studies; and advocacy of study results [2,3,4, 6,7,8,9]. In the UK, the National Institute for Health Research has provided vital strategic and infrastructure support to embed PPI across publicly funded research, creating a context where PPI is seen as a key element in research. Internationally PPI is also developing, with similar initiatives in Canada, United States, Australia, and Europe [10, 11]. Networks such as the citizen and patient involvement group of Health Technology International have evolved, enabling international collaboration in relation to involvement and engagement [12].

While the PPI evidence base has expanded significantly over the past decade, the reporting of PPI in papers has often been inconsistent and partial, with little information about the context, process, and impact of public involvement and with limited reporting of conceptualisation or theorisation [1,2,3,4]. Inadequate reporting can create problems for systematic reviews that attempt to synthesise PPI evidence [2,3,4]. Appraisal, interpretation, and synthesis of results are difficult, aside from the ethical imperatives of reporting research in a way that others understand and can use [13,14,15,16]. Inconsistent reporting creates a fragmented evidence base making it difficult to draw together our collective understanding of what works, for whom, why, and in what context. Furthermore, researchers, patients, carers, or clinicians cannot learn from previous experience, and precious resources devoted to involving patients and the public are wasted. Omitting descriptions of PPI activities from a study can represent a form of misreporting and might misrepresent the initial intentions of a study.

This article introduces the two versions of the GRIPP2 reporting checklist: GRIPP2-LF, a longer checklist for studies where PPI forms the primary focus of a study (Table 1) and GRIPP2-SF, a short checklist for studies where PPI is a secondary or tertiary focus (Table 2). We also describe the development of GRIPP2 and outline how it can be used.

Table 1 GRIPP2 long form
Full size table
Table 2 GRIPP2 short form
Full size table

GRIPP2 reporting checklist development methods

The study used the EQUATOR method for developing reporting guidelines [5], which included: systematic review evidence; a three stage Delphi survey including key stakeholders in the field of PPI; and a face-to-face collaborative meeting to develop consensus on items outstanding from the Delphi survey. A summary of methods is presented, with a companion paper reporting the rationale for GRIPP2 and the full methods [17]. For the purposes of this paper, we have therefore reported only a summary of key steps in Appendix 1.

The systematic reviews that underpinned the original GRIPP checklist had already identified the need for the guidance [2,3,4]. The PIRICOM systematic review, which included the conceptualisation, definition, measurement, impact, and outcomes of PPI on research, researchers, service users, participants, funders, and policy makers, was updated for GRIPP2 to ensure no additional concepts were omitted from the Delphi survey. In addition, searches were conducted to identify any other reporting guidelines for PPI.

Three rounds of the Delphi survey were conducted to gain consensus (see Appendix 2). This included 143 international participants in round one, with an 86% (123/143) response for round two and a 78% (112/143) response for round three reflecting the standard number of participants used in the development of previous EQUATOR guidance [5]. Participants of the Delphi survey included researchers, funders, patients, carers, editors, and individuals from international research agencies from countries including Australia, the United States, Canada, and Europe. Collectively, participants represented a wide range of expertise relevant to the development of consensus in PPI reporting.

Participants were asked to rate each item in the checklist on a scale of 1–10, with 1 considered unimportant and 10 considered very important, and medians and interquartile ranges were calculated for each item in the Delphi survey. Space next to each item was used for free text comments with suggested refinements, reiterations, and additional items. If items reached a median score of ≥8 in round one and round two they were considered to have reached positive consensus and included. Items that reached a median ≤ 5 in rounds one and two were excluded from the checklist. Items that reached a medium score of 6 or 7 in one round and a median score of ≥8 in the other round were voted on again in round three. Positive consensus was gained if the items scored a median score of ≥8 in two of the rounds. An important finding from the first round was that participants thought GRIPP items were most relevant when the main focus of a study was on PPI, and many felt there should be a shorter version for papers that included some element of PPI. As a result participants were asked to identify and score “core” items in round two which could be included in a shortened version of the guideline, suitable for studies that have included PPI as a secondary focus. The five core items that comprise the GRIPP2-SF all gained a median score of 9 in round two. Thus all five were included in round three and again gained median scores of 9, indicating that consensus was reached on the short form version.

Qualitative comments were analysed thematically to identify common themes, points of feedback, challenges to the items, and queries about wording [18]. Qualitative comments suggested the need to reword some items to simplify them and ensure clarity of meaning. Two sections from the original GRIPP checklist, section 5, which focused on measurement, and section 6, which was focused on capture of impact, were combined as it was recognised that they were conceptually overlapping. The original section 8 was deleted as participants thought it duplicated existing items.

Appendices 3 and 4 report the results of the Delphi survey. Following the Delphi survey, a collaborative consensus meeting was held with 25 key experts with knowledge, experience, or both, of PPI, including patient partners and carers (n = 8), researchers (n = 9), clinicians (n = 6), and healthcare journal editors (n = 2). The aim of this meeting was to finalise consensus on the seven items on the threshold of consensus following the Delphi survey (Appendix 2) and ensure clarity of the items.

Patient partners were collaboratively involved at key stages of the study. Three patient partners were recruited to the research team and were involved in refining the focus of the research questions, in development of the search strategy and interpretation of results of the systematic review, in discussions identifying the need for development of guidelines, and in selecting the items for the original GRIPP checklist. Furthermore, the patient partners assisted in developing the electronic survey for the first phase of the Delphi survey consensus process and were instrumental in assisting in recruitment to the Delphi study and in collation of comments from each Delphi survey round, and contributed to adapting items for GRIPP2. The consensus meeting involved eight patient partners in total, and the three patient partners recruited to the research team were involved in the write-up of the study and are coauthors in papers. More detailed information of their contribution to the development of GRIPP is described using GRIPP2-SF in Table 3 and used to populate the BMJ PPI guidance in Table 4.

Table 3 PPI in the development of GRIPP2 using GRIPP 2-SFa
Full size table
Table 4 Patient and public involvement in GRIPP2 according to BMJ guidance
Full size table

Scope and illustration of use

GRIPP2-LF (Table 1) and GRIPP2-SF (Table 2) are the first international, evidence based, community consensus informed guidelines for the reporting of PPI in research. The checklists provide key PPI concepts that authors should report in papers, to enhance the overall quality and transparency of the PPI evidence base. GRIPP2-LF and GRIPP2-SF ultimately aspire to guide PPI reporting in different types of studies, from reporting on PPI in trials (GRIPP2-SF) to reporting of PPI focused studies (GRIPP2-LF). Researchers can use the reporting guideline prospectively to plan PPI in studies and retrospectively as a quality assurance step in the writing up of PPI in publications and reports. Health and social care research funders and research institutions could promote adherence to the GRIPP2 reporting checklist as a means to optimise the creation of transparent, consistent, and high quality PPI evidence. Journal editors could use GRIPP2 reporting checklists to set their reporting expectations for submitted manuscripts. Higher quality reporting will gradually lead to the development of a stronger PPI evidence base that will facilitate more effective synthesis of PPI studies.

GRIPP2 can be used in different ways within a paper. For GRIPP 2-LF the entire paper can be shaped by the guidance, with researchers selecting the items of relevance. With GRIPP2-SF researchers could present all the information in the body of the paper under the relevant reporting titles or in a separate box. Table 3 provides an illustration of GRIPP2 –SF using this study as an example. This table is an illustration of the potential of GRIPP2 reporting. It is purposefully long to demonstrate the type of information it could include. A more specific, shorter form of reporting would also be acceptable, as long as it contained the key information.

Availability

GRIPP2-SF and GRIPP2-LF are available on the EQUATOR webpage (www.equator-network.org/), or at http://www2.warwick.ac.uk/fac/med/research/hscience/wrn/research/themea.

Discussion and limitations

GRIPP2-LF and GRIPP2-SF are the first international, evidence based, community consensus informed guidelines for the reporting of patient and public involvement in research. Although consensus was achieved in the development of GRIPP2, further refinements are expected over time as the evidence base underpinning PPI evolves, reflecting the iterative EQUATOR method of guideline development. In addition, it has not yet been possible to conduct any usability testing to understand how GRIPP2 works in practice with different types of study designs. The final consensus meeting did not include international experts because of a restricted budget, which might have limited the discussion from an international perspective. Thus, the next phase of development for GRIPP2-LF and GRIPP2-SF should include wider international application and piloting to test conceptual equivalence in different country contexts. Feedback from researchers using GRIPP2 will help refine it. We have created a comment box on the Warwick Medical School website to facilitate this http://www2.warwick.ac.uk/fac/med/research/hscience/wrn/research/themea.

Guidelines such as the CONSORT statement for randomised controlled trials (RCTs) are regularly updated to reflect changes in health research more widely [19]. Such evolution is particularly important for GRIPP2 because PPI is at a pre-paradigm stage in its development and recognition, reflecting Kuhn’s conceptualisation of how science changes over time with significant paradigm shifts that generate new ways of thinking [20].

While GRIPP2-LF and GRIPP2-SF aim to guide consistent reporting, it is not possible to be prescriptive about the exact content of each item, as the current evidence base is not advanced enough to make this possible [2,3,4, 21]. Authors should carefully consider the relevance of each GRIPP2 item but recognise that it is sometimes not necessary, or even possible, to include each item in a particular manuscript. With future development of the evidence base, it will be possible to refine GRIPP items, and some may become mandatory.

The success of the PPI in this study may relate to several factors. Firstly, the patient partners had received training around research methods in previous studies and were actively involved in a patient and public involvement group attached to the University of Warwick Medical School. Furthermore, the researchers were experienced at involving patient partners in their research [22]. Finally, good relationships and ways of working were established, which are known as key factors for facilitating high quality PPI [4, 21].

We recognise that GRIPP2 was developed with experts familiar with PPI and that there are still significant challenges in academic culture in enacting the behaviour changes that public involvement requires. PPI needs to become embedded practice within research rather than an option, and both researchers and patients need to recognise their own training and development needs, drawing on the evidence base to guide effective practice.

A further limitation is that GRIPP2-LF and GRIPP2-SF are conceptualised within the culture and language of research. Bearing in mind that the ultimate intention of high quality reporting in PPI is to develop best practice, there is a need to develop a patient or service user version of GRIPP2 to ensure comprehensibility and usefulness and to ensure that patient important concepts indicative of high quality research are included, although these are yet to be identified. This would reflect important changes in academic publishing, where patients are regularly writing and peer reviewing academic papers and require ways of understanding reporting quality [22]. Used alongside other EQUATOR guidance, the intention is to guide the development of a transparent, consistent, and high quality PPI evidence base. More effective synthesis of the PPI evidence base will help to identify best practice, avoid poor practice, and contribute to research that is acceptable, relevant, appropriate, and high quality and that has the potential to generate benefit for all.