Is behaviour change communication guidance for general practice healthcare professionals evidence based? A systematic review

Abstract

Background

Although guidance for primary care professionals on how to communicate with patients about health behaviour change is prevalent, the evidential base for this guidance is unclear. We aimed to systematically review the current guidance for general practice healthcare professionals on how to communicate about health behaviour change with patients, and ascertain what evidence underpins these.

Methods

In this systematic review without meta-analysis (SWiM), we searched the electronic databases MEDLINE, Embase, PsycINFO, CINAHL, Scopus, and Web of Science on 23rd June 2023. We also searched national guidelines, training, and magazines aimed at providing behaviour change communication guidance for clinicians, on 6th August 2023. Eligible sources were resources providing health behaviour change communication guidance for healthcare professionals in general practice. Two reviewers screened sources for inclusion. Once included, one reviewer extracted relevant communication recommendations from each document. For each communication recommendation, one reviewer identified the presence or absence of citations to supporting evidence. Citations were then examined and classified by relevance. The key outcome measure was the number of communication recommendations supported by evidence and relevance of the supporting evidence cited.

Results

Across 1163 communication recommendations identified, 677/1163 (58.2%) included one or more citations, totalling 3640 citations. Of these, 1432/3640 (39.3%) were considered relevant. Of 1163 specific communication recommendations identified, 233/1163 (20.0%) used exclusively relevant evidence sources, whilst 61 (5.2%) used exclusively irrelevant evidence sources. Four hundred thirty-one (37.0%) recommendations included at least 1 relevant citation.

Conclusions

Guidance aimed at supporting healthcare professionals to communicate about behaviour change with their patients is rarely clearly substantiated with a citation to relevant supporting evidence. Where evidence is provided, it is rarely explicitly relevant to the claim in the guidance which cited it. Given that communication is fundamental to a range of crucial healthcare outcomes, guidance should draw on available evidence to better support healthcare professionals. Further research should examine the evidence underpinning other communication guidance.

Systematic review registration

Schwarze-Chintapatla A, Livingstone-Banks J & Albury C. What evidence underpins communication guidance regarding behaviour changes for healthcare professionals within general practice: A systematic review. PROTOCOL. Open Science Framework (2024).

Supplementary Information

The online version contains supplementary material available at 10.1186/s12916-026-04681-7.

Background

Communication between patients and professionals is crucial to healthcare encounters. Effective communication supports informed decision-making, treatment adherence, satisfaction, and is associated with improved patient outcomes [1–3]. Effective communication can also reduce the likelihood of malpractice claims [4]. Communication is recognised as an essential skill for healthcare professionals, and guidance and training consistently foreground its importance. Much communication between professionals and patients occurs in primary care, which is the most common first point of contact for patients.

A key role of primary care professionals is to communicate with patients about disease prevention and health promotion [5]. Doing so is crucial in reducing the burden of disease and improving healthy life expectancy, as many long-term conditions can be prevented or addressed early to limit progression. Whilst wider prevention and promotion strategies should focus on individuals and systems [5, 6], primary healthcare professionals are encouraged to act at the individual level, encouraging individuals to take action to change their health behaviours [7].

Guidance exhorts primary healthcare professionals to encourage and support patients to make changes to health behaviours (including dietary, physical activity, and smoking behaviours) which are major factors contributing to morbidity and chronic disease. For example, UK Public health guidance recommends primary care clinicians take ‘every opportunity’ to communicate about health behaviour change to support prevention.

Guidance for healthcare professionals on how to communicate with patients about health behaviour change is available through a range of sources, including clinical practice guidelines [8, 9], training, and information resources. However, despite the range of guidance available, clinicians report having difficulties communicating with patients about behaviour change [10] and fear these conversations will create discomfort and interactional difficulties between the clinician and patient [11]. This is important, because there are clear relationships between how behaviour change advice is communicated and its effectiveness. For example, a 2024 cohort study examined primary care behaviour change conversations which included the offer of referral to a weight management service. Conversations included the same content, but varied in how this content was communicated. Evidence showed that certain words, phrases, and tone were more likely to support service uptake, and weight loss in the longer term [12]. Healthcare professionals and patients also highlight that how behaviour change advice is communicated is important. Despite this, clinicians report a need for more support to have these conversations .

Although guidance for primary care professionals on communicating about behaviour change is prevalent, the evidential base for this guidance is unclear. It is possible that guidance is well supported by robust evidence, and yet current evidence is insufficient to support clinicians, or guidance is not influencing clinician behaviour [13]. However, there is also a possibility that communication guidance does not support clinicians to communicate effectively because it does not draw on the available evidence for how to do this. This is what we aim to find out. We aim to systematically review the current guidance for primary care professionals on how to communicate about health behaviour change with patients, and ascertain what, if any, evidence underpins these.

This review will enable understanding of how clinicians are currently supported to communicate with patients about health behaviour change, and where improvement may be needed. However, as communication is fundamental in underpinning all aspects of care, this review has potential for wider reaching implications. This review additionally acts as an important ‘test case’ to identify if review of communication guidance in other areas may be warranted. There is a very real possibility that communications guidance is a house of cards, destined to fail due to poorly evidenced foundations.

Methods

This systematic review and synthesis without meta-analysis (SWiM) followed Cochrane-recommended processes [14] and is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [15]. Protocol registration: Open Science Framework [16].

Eligibility criteria

We included sources of health behaviour change communication guidance for healthcare professionals in general practice (hereafter referred to as ‘guidance sources’, see Table 1). Eligible guidance sources comprised resources that provide recommendations targeted at healthcare professionals, and included national guidelines, articles in medical journals or magazines aimed at providing recommendations for practice, and continuing professional development training. Eligible guidance sources were applicable to a UK healthcare setting through either being (a) developed specifically for UK healthcare professionals, or (b) developed as generic recommendations that were neither healthcare system nor country specific. A focus on UK guidance can provide relevant evidence for national guidance, training, and policy stakeholders.

Table 1.

Terminology

Term	Definition	Examples
Guidance sources	Sources of health behaviour change communication guidance for healthcare professionals in general practice	National guidelines, articles in medical journals or magazines aimed at providing recommendations for practice, and continuing professional development training
Communication recommendations	A statement within a guidance source which provides guidance on how to verbally communicate with a patient regarding health behaviours	‘Listen to the patient without interrupting. Show your genuine interest in patient’s replies’
Evidence document	Documents cited in a guidance source to provide evidence for a communication recommendation	‘One BCT, “provide information on the consequences of behaviour in general”, was significantly associated with a positive change in intention. One BCT, “relapse prevention/coping planning”, was associated with a negative change in intention’ [17]

Eligible guidance sources focussed on verbal communication with adult patients about ‘health behaviour change’ in general, or included a specific focus on diet, physical activity (which we considered as one category because they are so often discussed together), smoking, unsafe sexual behaviour, and/or alcohol misuse. These specific behaviours were chosen as they are included in the National Institute for Health and Care Excellence (NICE) Guidelines for behaviour change [9].

We excluded guidance sources comprising teaching and training provided to medical students, and research studies discussing the recommendations/implications for practice of their results.

Search strategy

We undertook two approaches to identifying eligible guidance sources:

1. To identify guidance sources available through databases (such as articles in medical journals), we searched the electronic databases MEDLINE, Embase, PsycINFO, CINAHL, Scopus, and Web of Science from database inception to 23rd June 2023. Searches included relevant keywords and MeSH (Medical Subject Heading) terms. The full search strategy is available in Additional file 1: Table 1. Search results were imported to Covidence [18].

2. To identify guidance sources which were not in databases (such as national guidelines, websites, training, and magazines), we consulted general practice healthcare professionals on commonly used sources of guidance to form list of potential sources for guidance, which we then searched within. Handsearching was used to locate these guidance sources and the communication recommendations within them. The list is available in Additional file 1: Table 2. Search results were imported to excel.

Study selection

Two reviewers (ASC, a medical student with a special interest in behaviour change communication; and either CA, a Clinical Communication and Behaviour Change researcher; or JC, a psychology and linguistics student) independently screened all titles and abstracts of search results identified from electronic databases. Guidance sources identified which were not in databases did not always include abstracts, and so titles were screened only. Where reviewers disagreed, guidance was considered for full-text screening.

Two reviewers (ASC and JC) independently screened full texts. Disagreements were resolved through discussion or involvement of a third reviewer (CA).

Data extraction

Within each wider guidance source, we identified specific communication recommendations. ‘Communication recommendations’ were defined as a statement which provides guidance on how to verbally communicate with a patient regarding health behaviours. Examples of these included ‘Listen to the patient without interrupting. Show your genuine interest in patient’s replies’ [19] or ‘Adopting a persuasive and supportive rather than an argumentative and confrontational position’ [20].

To identify the content of current guidance, one reviewer (ASC) extracted each specific communication recommendation as well as key characteristics from each guidance source: including title, author, year of publication, publication source, health behaviour targeted, type of information source (e.g. training module, clinical practice guideline, or journal article), and the specific health behaviour change communication practices/recommendation that were included.

To ascertain the evidence-base behind current guidance, one reviewer (ASC) determined which specific communication recommendations were substantiated with a citation to supporting evidence. Where supporting evidence was cited, the same reviewer located and read the supporting evidence, extracting key characteristics: including title, source, document type, study type, and author.

Assessment

We next assessed the relevance of the supporting evidence to the specific communication recommendations they were cited to substantiate. One reviewer (ASC) assessed the relevance of the supporting evidence, discussing decisions with a second reviewer (CA). In addition, 100 randomly selected evidence documents were assessed in duplicate by two reviewers (ASC, and either CA or JC). Any disagreements were resolved through discussion.

We classed supporting evidence as relevant if both these criteria were met:

• Evidence about behaviour change communication is included (either in general practice or a non-specific clinical setting)

• Evidence is from an adult population

We classed supporting evidence as not relevant if:

• No evidence about behaviour change communication is included

• No evidence relating to verbal communication

• Evidence was not from an adult population

• Communication techniques are discussed, but no evidence is provided on their effectiveness

• Evidence is unrelated to the recommendation stated in the guidance

• Evidence was from a book

Data synthesis

We undertook a narrative synthesis of results and presented the quantitative findings into comparative tables, charts, and graphs, such that comparisons can be drawn between different health behaviours.

Patient and public involvement

Patients and members of the public were not involved in the design or conduct of this study.

Results

We identified 84 guidance sources. Figure 1 illustrates the screening process. Guidance sources were located across national clinical practice guidelines (n = 33), journals (n = 16), websites (n = 15), and professional magazines (n = 20). Sources of included guidance documents can be seen in Additional file 1: Table 3. Each guidance document contained multiple specific communication recommendations. Of these, 49.5% (576 recommendations across 33 sources) were in national guidelines; 18.3% (213 recommendations across 16 sources) were in journals; 16.3% (189 recommendations across 15 sources) were in websites; and 15.9% (185 recommendations across 20 sources) were in professional magazines. A single guidance source could contain communication recommendations relating to multiple specific health behaviours. There were 1163 specific communication recommendations across the 84 guidance sources. Together, these 1163 communication recommendations related to four specific health behaviours: (1) diet and physical activity (434 recommendations across 38 documents); alcohol use (116 recommendations across 11 documents); smoking cessation (102 recommendations across 11 documents); sexual health behaviours (40 recommendations across 4 documents); and ‘general’ unspecified health behaviours (471 recommendations across 38 documents). Some guidance sources contained communication recommendations relating to multiple specific health behaviours. Details on relevance of communication recommendations are shown in Fig. 2. Overview of communication recommendations and citation practices by health behaviour is shown in Table 2.

Fig. 1.

PRISMA flow diagram

Fig. 2.

Relevance of communication recommendations

Table 2.

Overview of communication recommendations and citation practices by health behaviour

Item	Alcohol, n (%)	Sexual health, n (%)	Smoking, n (%)	Diet and physical activity, n (%)	General behaviour change, n (%)
Guidance sources*	11/84 (13.1)	4/84 (4.8)	11/84 (13.1)	38/84 (45.2)	38/84 (45.2)
Communication recommendations*	116/1163 (10.0)	40/1163 (3.4)	102/1163 (8.8)	434/1163 (37.3)	471/1163 (40.5)
Communication recommendations with citations
Of total communication recommendations across all behaviours	96/1163 (8.3)	30/1163 (2.3)	74/1163 (6.4)	244/1163 (21.0)	233/1163 (20.0)
Of total communication recommendations for specific behaviour	96/116 (82.8)	30/40 (75)	74/102 (72.5)	244/434 (56.2)	233/471 (49.5)
Communication recommendations based on at least one relevant citation
Of total communication recommendations	79/1163 (6.8)	26/1163 (2.2)	29/1163 (2.5)	151/1163 (13.0)	146/1163 (12.6)
Of total communication recommendations for specific behaviour	79/116 (68.1)	26/40 (65)	29/102 (28.4)	151/434 (34.8)	146/471 (33.6)
Of total communication recommendations for specific behaviour including citations	79/96 (82.3)	26/30 (86.7)	29/74 (39.2)	151/244 (61.9)	146/233 (62.7)
Communication recommendations based entirely on relevant citations
Of total communication recommendations	32/1163 (2.8)	21/1163 (1.8)	11/1163 (0.95)	75/1163 (6.4)	94/1163 (8.1)
Of total communication recommendations for specific behaviour	32/116 (27.6)	21/40 (52.5)	11/102 (10.8)	75/434 (17.3)	94/471 (20)
Of total communication recommendations for specific behaviour including citations	32/96 (33.3)	21/30 (70)	11/74 (14.9)	75/244 (30.7)	94/233 (40.3)
Communication recommendations based entirely on irrelevant citations
Of total communication recommendations	5/1163 (0.4)	2/1163 (0.2)	8/1163 (0.7)	30/1163 (2.6)	16/1163 (1.4)
Of total communication recommendations for specific behaviour	5/116 (4.3)	2/40 (5.0)	8/102 (7.8)	30/434 (6.9)	16/471 (3.4)
Of total communication recommendations for specific behaviour including citations	5/96 (5.2)	2/30 (6.7)	8/74 (10.8)	30/244 (12.3)	16/233 (6.9)

*A single guidance source could contain communication recommendations relating to multiple specific health behaviours

Citations to supporting evidence documents

In total, evidence documents were cited 3640 times (Fig. 3). This included the same evidence document being cited multiple times within one guidance document.

Fig. 3.

Citations by health behaviour

Of the 1163 specific communication recommendations identified, 677 (58.2%) included citations to evidence the claim being made, whilst 486 (41.8%) did not include citations. The number and percentage of citations included varied between communication recommendations for different health behaviours. Specific communication recommendations about alcohol included the highest percentage of specific recommendations with citations (n = 96/116, 82.8%) followed by sexual health (n = 30/40, 75.0%), smoking (n = 74/102, 72.5%), diet and physical activity (n = 244/434, 56.2%), and ‘general’ behaviour change (n = 233/471, 49.5%). Diet and physical activity communication recommendations however included the highest overall number of recommendations with evidence citations (244), seen in Table 2.

Of the 677 specific communication recommendations which included citations, 272/677 were supported by more than one citation. For alcohol, there were 622 citations to support the 96 specific communication recommendations; for sexual health, there were 62 citations to support the 30 specific communication recommendations; for smoking, there were 121 citations to support 74 specific communication recommendations; for diet and physical activity, there were 2183 citations to support 244 specific communication recommendations; for general behaviour change, there were 652 citations to support 233 specific communication recommendations.

Eighteen out of 1163 (1.5%) specific communication recommendations included explicit claims that they were based on evidence, but did not include a citation to the evidence mentioned; this comprised diet and physical activity, 61% (n = 11/18); sexual health, 5.6% (n = 1/18); alcohol use, 16.7% (n = 3/18); and ‘general’ unspecified health behaviours, 16.7% (n = 3/18).

Thirty-nine out of 1163 (3.6%) specific communication recommendations included citations that were explicitly reported as ‘inference derived from evidence’. These did not make clear what evidence was used for such inferences. Communication theories or models were cited in 4/1163 (0.3%) specific communication recommendations, but no evidence was cited of their effectiveness.

Evidence use was different depending on where the guidance was published. Journals and national guidelines had the highest number of citations to number of recommendations (journals: 217 citations for 213 recommendations, national guidelines: 3268 citations for 576 recommendations), whilst websites and magazines had fewer citations than number of recommendations made (websites: 55 citations for 189 recommendations, magazines: 100 citations for 185 recommendations) (Additional file 1: Table 4).

Relevance of supporting evidence cited

Of the 3640 times evidence documents were cited, less than half (1432/3640, 39.3%) were relevant to the specific communication recommendations which cited them, and 1582/3640 (43.5%) were not relevant to behaviour change communication. Not all of the supporting evidence sources that were cited were accessible or assessable. Of 3640 citations, 86/3640 (2.4%) referred to books (in addition to 2 book chapters) whose relevance could not be categorised. In total, 62/3640 (17.2%) of evidence could not be assessed (due to unclear citations, inaccessible documents, or by their nature being books).

Of the 677/1163 communication recommendations identified which included citations, 431/677 (63.6%) included at least 1 relevant citation. However, of these recommendations, 198/677 (29.0%) also relied on irrelevant citations. Overall, only 233/677 (34.0%) communication recommendations which included citations were based entirely on relevant citations and 61/1163 (5.2%) of communication recommendations were based entirely on irrelevant citations.

The remaining 671/1163 (57.7%) recommendations either did not include a citation (486/1163, 41.8%); or included unclear citations, cited supporting documents that were inaccessible, or included citations to documents that could not be assessed (185/1163, 16%) (e.g. books). Further details of these according to source are available in Additional file 1: Table 5. Figure 4 illustrates citation practices across communication recommendations.

Fig. 4.

Citation practices across communication recommendations

Specific communication recommendations relating to alcohol included the highest percentage of citations to relevant evidence (n = 374/622, 60.1%), followed by general behaviour change (n = 381/652, 58.4%); smoking cessation (68/121, 56.2%); sexual health (30/64, 48.4%); and diet and physical activity (587/2183 26.9%).

The most cited evidence documents were systematic reviews (total 875; 379 relevant, 356 not relevant, and 140 inaccessible), followed by qualitative studies (total 634; 316 relevant, 306 not relevant, and 12 inaccessible) and quantitative studies (total 210; 57 relevant, 117 not relevant, and 36 inaccessible). Additional file 1: Table 6 details the nature of the citations as referred to by the guidance documents (for example, systematic review, randomised controlled trial, or qualitative study) and their relevance.

The highest percentage of relevant citations were in magazines (62/100, 62%), followed by journals (120/217, 55.3%) and websites (28/55, 50.9%). The lowest percentage of relevant citations were in national Guidelines (1225/3268, 37.5%).

Discussion

Statement of principal findings

In this systematic review without meta-analysis, we aimed to review the current guidance for primary care professionals on how to communicate about health behaviour change with patients, and ascertain what, if any, evidence underpins these.

We identified 84 guidance sources containing 1163 specific communication recommendations. Of these, 677 included citations to underpinning evidence, and 486 did not include citations. The 677 specific communication recommendations which included citations contained citations to 3640 supporting evidence documents. Less than half of these supporting evidence documents (1432/3640, 39.3%) were obviously relevant to the specific communication recommendation they were cited to support. Overall across the 1163 specific communication recommendations identified, only 233 (20.0%) were based entirely on relevant evidence to the communication recommendation being made.

Whilst some guidance sources were well cited with a clear chain of evidence, other citations (whilst documented) were unclear. Examples include studies whose results were described in appendices, but lacked key identifying features (e.g. titles) and thus could not be located. This is important in regard to accessibility and transparency of evidence for a reader, who should be able to access and read the evidence for themselves. Regardless of the quality of evidence, lack of citation clarity hinders a reader’s ability to understand what evidence relates to what guidance. Some guidance used evidence statements to claim an evidence-base of x studies, however provided citations for fewer studies than stated [21]. For example, a statement claiming an evidence-base of 29 randomised controlled trials (RCTs) only cited 24 RCTs .

Comparison with other studies

Evidence-based practice and decision-making is associated with improved patient outcomes [22], and extant research shows that grounding communication training and guidance in evidence can lead to changes in healthcare professionals’ communication behaviours [23–25]. In our study, we identified that behaviour change communication guidance is rarely substantiated by including citations to relevant evidence, meaning that guidance may not advise evidence-based practice, and clinicians and patients therefore may not realise the associated benefits. Our results complement results from a 2007 study assessing the quality of guidelines for doctor–patient communication used in communication programmes for general practitioner trainees across 8 Dutch university centres. This Dutch study also identified a lack of evidence underpinning these general communication guidelines [26].

Empirical research in the field of conversation analysis also highlights that there is often a gap between evidence of effective communication practices and the content of training and guidance. A series of empirical analysis of consultations between healthcare professionals and patients [12, 24, 27–29] show that evidence of effective communication practices exists across a range of healthcare settings, and yet this evidence is not always ‘written down or embedded in institutional training, policy, or related documentation’ [30].

The complexity of developing clinical guidelines is well reported; a 2000 review highlighted that evidence is only one part of guideline development, which should also incorporate clinical judgement and patient perspectives [31]; and a 2004 Lancet study highlights that, where evidence is considered, ‘translation of evidence into recommendations is not straightforward’ as data can be interpreted in different ways during guideline development [32]. However, although development can be complex, these studies report that evidence should be considered and incorporated, even if it is not the sole basis for recommendations. A strength of our study was the ability to quantify the extent to which relevant evidence was incorporated in final guidelines and training for clinicians.

Communication skills developed through medical training have been shown to be inadequate in early practice [33]. Our review shows that training and guidance available post-qualification are also likely to be inadequate as these rarely draws on evidence of effective practice. Taken together, this shows that primary care clinicians likely lack access to training in effective communication techniques for health behaviour change throughout their career.

Strengths and limitations

Some guidance presented a list of references at the end of the document; however, these references were not clearly linked to guidance in the document. Such unclear indications of evidence were labelled as ‘no cited evidence’ as there was no explicit citation. Whilst the content of such documents may indeed be based on evidence, such assessments were unfeasible within the constraints of this study. The lack of citing clarity itself, however, reflects a limitation of the guidance, through the absence of a transparent and explicit chain of evidence. Similarly, some documents [20, 34] alluded to evidence, but it was unclear what evidence related to which guidance. Such evidence was not included in the analysis. Whilst a possible limitation to the evidence assessment, the unclear citations themselves present a problem to the reader who wishes to review the evidence used.

Handsearching was used to locate additional guidance documents within websites. By the nature of handsearching, it is possible that some documents were missed; however, it is unlikely that such items have been systematically missed.

If more time were available, a full circle of the chain of citations could be followed of each document. This would produce results of interest to see what the evidence documents themselves are citing and if there were convergent documents. Another time permitting factor is an assessment of the quality and relevance of the books cited as evidence. Some guidance sources were inaccessible and so we could not assess if they included citations to supporting evidence.

Implications for research and practice

There are clear relationships between clinician communication and patient health outcomes. Effective ways to communicate about behaviour change are often counterintuitive and can conflict with communication recommendations in guidance and training [35]. The lack of evidence explicitly underpinning guidelines and training could explain this conflict, and why clinicians find behaviour change advice challenging, as the training and guidance available does not recommend effective approaches based on the best available evidence.

A lack of both citations and consistent citation practices in guidance sources presents a challenge to understanding the exact evidence upon which guidelines may be based. Policymakers and training developers should consider providing citations wherever specific claims are made, to increase transparency and facilitate assessment of the quality of the supporting evidence.

We have shown that behaviour change communication guidance for primary care professionals is seldom explicitly based on evidence. We were able to ascertain where evidence was cited and future research could build on this, to assess the quality of cited evidence. Future studies can also use the methods we have developed here to (a) assess the evidence underpinning other guidance sources, to understand if this issue is pervasive across other communication guidance and training for professionals, (b) assess the evidence used to underpin communication training and guidance provided in medical schools, and (c) review the evidence underpinning guidance sources in other countries to generate locally relevant knowledge, and assess evidence use and citation practices vary internationally.

Conclusions

Behaviour change communication guidance for primary care professionals is rarely substantiated with citations to relevant supporting evidence. Although effective communication is associated with improved patient outcomes, guidance and training are not clearly based on the best evidence available to achieve this.

Abbreviations

SWiM

Systematic review without meta-analysis

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Guidance Sources

Sources of health behaviour change communication guidance for healthcare professionals in general practice

NICE

National Institute for Health and Care Excellence

BCT

Behaviour change training

MeSH

Medical Subject Heading

RCTs

Randomised controlled trials

ASC

Anika Schwarze-Chintapatla

CA

Charlotte Albury

JC

Jamie Chua

JLB

Jonathan Livingstone-Banks

Authors’ contributions

CA and JLB conceived of this idea. ASC, CA and JLB undertook the design of this study. ASC collected the data, ASC and JC analysed the data with input from CA and JLB. ASC led the interpretation of the data, with input from CA and JLB. ASC led the drafting of this work, with CA carrying out substantive revisions. All authors edited drafts of this work. All authors read and approved the final manuscript.

Funding

This study was written by the authors and we request a fee waiver as the first author is a medical student and this work was unfunded. There are no prior publications from this study.

Data availability

Data is provided within the manuscript or additional information files. Any additional datasets used during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Ethical approval and consent to participate was not required for this systematic review.

Consent for publication

Not applicable.

Competing interests

Anika Schwarze-Chintapatla has no competing interests. Charlotte Albury has worked as an independent consultant for the Behavioural Insights team, Wildfowl Wetlands Trust, Adelphi Real World, Oxford Health BRC, and Linney Create for which she was paid personally. She was an academic advisor for NESTA, and did not receive personal payment. Jonathan Livingstone-Banks and Jamie Chua have no competing interests.