Research Priority Setting in Plastic and Reconstructive Surgery: A Systematic Review

Stakeholders from lower-income countries are underrepresented in priority setting initiatives for plastic surgery, despite the global burden of disease. Future studies should recruit more patient and multidisciplinary stakeholders, to achieve meaningful consensus. Clear implementation strategies are needed to maximise impact.


Summary Background
The health research agenda has historically been led by researchers; however, their priorities may not necessarily align with those of patients, caregivers, and clinicians. Research priority setting initiatives identify and prioritise topics which lack evidence. This is particularly important in plastic surgery, a specialty lacking high-quality evidence to definitively answer many common clinical questions. Research priorities direct research activity and funding, so their selection process must be representative and transparent. This review appraised all priority setting initiatives in plastic surgery using the reporting guideline for priority setting of health research (REPRISE).

Methods
OVID Medline, EMBASE, CINAHL and the James Lind Alliance (JLA) repository were searched (inception -11/06/21) using search terms for 'research priority setting' and 'plastic and reconstructive surgery'. Dual-author screening and data extraction was conducted, according to PRISMA.

Conclusions Introduction
The health research agenda has historically been led by researchers; however, their research priorities may not necessarily align with those of patients, caregivers and clinicians delivering patient care [1][2][3] . This mismatch may result in funding for research questions which do not address the shared interests of all relevant stakeholders. This is a frequent cause of avoidable research waste, alongside duplication of research [4][5][6] . Research waste can be defined as research which is not prioritised or warranted, and does not lead to worthwhile achievements 5,6 . Research priority setting initiatives are designed to identify and prioritise research questions without existing evidence which are important to stakeholders in discrete areas of healthcare. Numerous prioritisation approaches are available to gather and rank evidence uncertainties, using primary and/or secondary research methods 4,7 .
Research prioritisation is particularly important in surgical research, given the disproportionately lower funding received compared with non-surgical research 8,9 . For example, in the United Kingdom (UK), surgical research receives less than 5% of Government health research funding despite one third of admissions requiring surgical care 9,10 . Compared with other surgical specialties, plastic and reconstructive surgery particularly lacks high-quality evidence to definitively answer many common clinical questions 11,12 . This may be due to the specialty's small size 11 and difficulty measuring subjective and aesthetic outcomes 13 . This has generated many clinical uncertainties for research prioritisation and to date, various priority setting initiatives have been delivered in areas such as hand and upper limb surgery [14][15][16][17] , burns [18][19][20][21][22][23][24] , congenital defects 25 , skin surgery 26,27 , breast surgery 28 and aesthetic surgery 29 .
Research priorities have a significant impact on research activity and funding 14,30 , so it is important that their selection process is transparent, representative and adequately reported 7,31,32 . Previous systematic reviews of priority setting initiatives in various fields have demonstrated suboptimal 5 reporting 33-39 and lack of patient involvement 36,38,39 . The aim of this systematic review is to describe the scope, methodology and reporting quality of priority setting initiatives in plastic and reconstructive surgery, using the reporting guideline for priority setting of health research (REPRISE) checklist 31 . The results will inform future researchers conducting research prioritisation in plastic surgery by summarising previous work, identifying areas still in need of prioritisation and highlighting both common deficiencies and examples of best practice in the methodology of previous initiatives.

Methods
This systematic review adheres to a pre-specified protocol (see supplemental digital content 1) and is reported in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement 40 . It was ineligible for PROSPERO registration as there are no direct health outcomes.

Study eligibility
Inclusion and exclusion criteria are shown in Table 1. Included studies were full-length peerreviewed research articles describing research priority setting initiatives in plastic and reconstructive surgery, from database inception to 11 th June 2021. Protocol studies were excluded because they lacked information on outcomes. If a research publication describing the priority setting initiative was not available, formal reports from the James Lind Alliance (JLA) website (or equivalent organisation) were used. Multiple publications describing the same research priority setting initiative were merged as one study.

Types of participants
6 All eligible research priority setting initiatives were included, regardless of type of participant.
Stakeholders were defined as patients, carers, healthcare providers, researchers, policy makers and industry representatives 31 .

Types of interventions
Included studies elicited research priorities from stakeholders in areas related to plastic and reconstructive surgery, as defined by the Royal College of Surgeons 41 .

Types of outcome
For all included studies, the primary outcome was a final prioritised list of research topics or questions related to plastic and reconstructive surgery.

Search strategy
The following electronic databases were searched from inception to 11 th June 2021: OVID Medline, OVID EMBASE and CINAHL. The JLA website was also searched for reports of relevant priority setting partnerships. A search string was developed to identify relevant papers, which included key search terms and medical subject headings for 'research priority setting' and 'plastic and reconstructive surgery' 36,42 . The database search strings can be found in the supplemental digital content.

Study selection process
Search results were combined and de-duplicated using Covidence (Veritas Health Innovation Ltd, Melbourne, Australia). Articles were screened independently by two authors (AL, GH) in two stages (by title and abstract, and full text), according to prespecified inclusion criteria (Figure 1). A third author (AY) was consulted if discrepancies in article screening could not be resolved.

7
There is currently no risk of bias assessment tool for studies of research priority setting and tools designed for trials and observational studies are not applicable. The REPRISE reporting guideline was used for data extraction and to assesses reporting quality 31 . This checklist included items covering the context and scope of the priority setting initiative, governance and team members, stakeholder recruitment and characteristics, identification and prioritisation of research topics, dissemination of results, implementation and evaluation, and disclosures of funding and conflicts of interest.

Data extraction
Data were extracted using a piloted data extraction form (Microsoft Excel) developed for the purposes of this review. Data were extracted independently by two authors (AL, GH). A third author (AY) was consulted if discrepancies in data extraction could not be resolved. Studies were categorised into eight subspecialty areas, adapted from the Royal College of Surgeons classification 41 , including congenital conditions, breast surgery, skin surgery, trauma, burns, hand and upper limb surgery, aesthetic surgery and other cancer-related reconstruction (e.g., head and neck cancer, sarcoma or perineal malignancy). Studies were defined as international if their stakeholder groups were multinational. Income status of the research setting (as stated in the text or affiliated with the corresponding author) was described according to the World Bank 43 . The World Bank categorises income status into high, upper-middle, lower-middle and low income countries, based on Gross National Income per capita 43 . Studies were aggregated by year of publication, pre-or post-2004, when the JLA was established 44 . Stakeholders were regarded multidisciplinary if more than one health profession (e.g., surgeon, nursing, therapy, etc) was included.

Data synthesis
Information relating to priority setting context and scope, stakeholder characteristics and study methodology was tabulated (Tables 2-3). Reporting quality (compliance with the REPRISE checklist) is summarised in Table 4 and Figure 2. Table S1 shows a detailed breakdown of reporting 8 compliance with REPRISE per study. A list of included studies can be found in supplemental digital content 2. Based on compliance with the REPRISE checklist and author consensus, a summary of recommendations for future priority setting exercises in plastic surgery is shown in Table 5.

Identification of studies
The database search returned 3899 de-duplicated citations, of which 17 were included in the final analysis ( Figure 1). Five publications were merged as they represented one priority setting exercise [45][46][47][48][49] . Two studies identified from the JLA website were not published in peer-reviewed journals; data were therefore extracted from available JLA reports 25,50 .

Context and scope
Most priority setting initiatives were conducted in single country (14/17) Table 2). The three international studies included only high-income countries (United States, Canada, UK, Australia and Norway) 17,21,29 . Only one study was conducted in a lower-middle income setting (India) 20 , and none in low-income settings. More priority setting initiatives focussed on burns (6/17) 20,21,2318,19,22 and hand and upper limb surgery (4/17) [14][15][16]51 , than other subspecialty groups. Some burns priority setting initiatives had a specific focus, e.g., paediatric burn care 23 , burn nursing 22 , burn recovery 20 and rehabilitation 21 , which may account for some repetition of the topic. Similarly, the British Association of Hand Therapists regularly repeat their priority setting exercise 15,16 , hence the duplication of UK hand therapy priority setting. No priority setting initiatives were identified in (non-breast) cancer surgery and trauma subspecialty groups. Most priority setting initiatives (13/17) were published after 2004 11,14,[27][28][29]15,16,[18][19][20][21]25,26 . All studies included priorities relevant to clinical practice; some additionally looked at delivery of health services 17,18,21,22 and basic science 11,19 . For example, in their priority setting process for general plastic surgery, Henderson et al 11 included both clinical topics (clinical trials, lymphoedema surgery) and basic science topics (tissue engineering, gene therapy) in the final priorities.

Quality of reporting
The quality of reporting varied across studies. Out of a possible 32 items included in the modified REPRISE checklist (Table 4), the number of items met ranged from 9 (28%) to 27 (84%). The overall completeness of reporting, as defined by the number of checklist items met, did not correlate significantly with year of publication (r = 0.43, p = 0.084). All 17 studies defined the geographical scope, health area or focus, and research focus of the priority setting exercise. Basic characteristics of stakeholders were described in most studies (16/17) 11,14,[24][25][26]28,29,51,15,16,[18][19][20][21][22][23] , but these were often limited to occupation and did not specify age, gender, or income status. Few studies identified the project leaders or steering group members (6/17) [14][15][16]19,26,28 , described their characteristics (a minimum of total number, stakeholder type, and occupation; 4/17), or described how the individuals within these groups were selected (7/17)  No studies reported if reimbursement was provided for patient participation, and no studies outlined the budget for the project.

Discussion
This systematic review has comprehensively summarised and appraised global research priority setting initiatives in plastic surgery. Overall, the findings demonstrate underrepresentation of stakeholders (particularly patients) from lower-income countries, as well as lack of multidisciplinary initiatives. Compliance with reporting standards for research priority setting was variable and especially deficient in the areas of governance and team members, evaluation and feedback, translation and implementation and funding and conflicts of interest.
There were several priority setting initiatives for burns 20,21,2318,19,22 and hand conditions [14][15][16]51 ; these tended to have more non-multidisciplinary steering groups and focussed on the care delivered by specific professions (e.g., nursing 22 or therapy 15,16,21,51 ). Arguably, a better approach would be to work cohesively with multidisciplinary input to maximise clinical relevance of the final priorities and reduce repetition, particularly with respect to subsequent funding applications. No priority setting initiatives were identified in the trauma and non-breast cancer categories, although major trauma and skin cancer JLA priority setting partnerships are in progress. Some studies e.g., Henderson et al 11 had a broad scope traversing clinical practice and basic science, however the distinct funding streams for clinical and translational research would favour a more clearly defined scope.
With regards to stakeholder characteristics, there was only one priority setting exercise from a lower middle-income country 20 , and none from low-income countries, despite the burden of disease (especially burns and trauma) in lower income nations 52,53 . Increasing the participation of stakeholders from lower-income countries will increase generalisability of the final priorities but raises questions for study design. For example, how to define adequate international representation and how to manage international data, particularly when local factors (e.g., health infrastructure and sociocultural beliefs) may significantly affect prioritisation behaviours 20,54 . Patients were underrepresented as both steering group members and stakeholders, despite clear differences in their prioritisation behaviours compared with healthcare professionals 6 . Research has shown that patients are less likely to prioritise pharmacological and surgical interventions (the focus of most registered trials), favouring education and training, service delivery, and physical and psychological interventions 6 . Neglecting patients' needs is a common cause of research waste and should be avoided through adequate representation at steering group and stakeholder levels. Although basic stakeholder characteristics were described in most studies, the level of detail varied, and often did not include age, gender, or ethnicity, as recommended by the World Health Organisation 55 . setting exercise regularly, allowing them to identify persistently highly ranked research topics which require ongoing investment (e.g., complex regional pain syndrome), as well as emerging topics (e.g., treatment modalities) 15,16 . Regularly repeating priority setting exercises could be a useful method for assessing the implementation and impact of priorities, but further consideration is needed to determine appropriate time intervals on a case-by-case basis. Future researchers could also consider conducting an 'impact survey' to determine how much competitive funding was awarded to proposals referencing the research priority exercise 14 .
Strengths of this review include the comprehensive search of three electronic databases and handsearching the JLA repository, with dual-author screening and data extraction. Limitations include lack of non-English language articles and formal risk of bias assessment, although no specific tool exists for research priority setting exercises, and reporting quality was assessed with the REPRISE checklist. Study compliance with some of the REPRISE checklist items was subjective. For example, most studies described stakeholder characteristics, but the level of detail (not specified in REPRISE) varied substantially, from only reporting stakeholder type (i.e., patient versus healthcare professional) to detailed demographics (age, gender, and ethnicity).

Conclusion
Research priority setting exercises have been conducted in most subspecialty areas of plastic and reconstructive surgery, except trauma and non-breast cancer. Reporting quality was variable and generally poor when describing project team and stakeholder characteristics, reasoning for exclusion of research uncertainties, authors' conflicts of interest and funding sources. A core aim of establishing research priorities is to influence funding allocation, yet strategies to assess their implementation were rarely described. Future initiatives should also consider recruitment of multidisciplinary steering groups, including an advisor with experience in research priority setting, and more patients and stakeholders from lower-income countries.

23.
Anonymous. Outcomes measurement in pediatric burn care: An agenda for research: