Healthcare workers’ perspectives on barriers and facilitators to referral and recruitment to diabetes prevention programmes: a systematic review protocol

Background Diabetes is a growing global health problem. International guidelines recommend identification, screening, and referral to behavioural programmes for those at high risk of developing type 2 diabetes. Diabetes prevention programmes (DPPs) can prevent type 2 diabetes in those at high risk, however many eligible participants are not referred to these programmes. Healthcare workers (HCWs) are pivotal to the referral and recruitment processes. This study aims to identify, appraise and synthesise the evidence on barriers and facilitators to referral and recruitment to DPPs from the perspective of HCWs. Methods A “best fit” framework synthesis method will synthesise qualitative, quantitative, and mixed methods evidence on factors that affect HCWs referral and recruitment to DPPs, with the Theoretical Domains Framework (TDF) as the a priori framework. MEDLINE, EMBASE, CINAHL, PsychINFO, Web of Science and Scopus will be searched for primary studies published in English. Year of publication will be restricted to the last 26 years (1997–2023). Quality will be assessed using the Mixed Methods Appraisal Tool. A mix of deductive coding using the TDF and inductive coding of data that does not fit the TDF will be synthesised into themes representing the whole dataset. The relationships between the final set of themes will be explored to create a new model to understand HCWs’ perspectives on referral and recruitment to DPPs. Sensitivity analysis will be carried out on this conceptual model. Confidence in the synthesised findings will be assessed using the GRADE-CERQual approach. One author will screen, extract, appraise the literature while a second author will independently verify a 20% sample at each stage. Discussion Participation in DPPs is key for programme impact. HCWs typically identify those at risk and refer them to DPPs. Understanding HCWs’ perspectives on the barriers and facilitators to referral and recruitment will inform future implementation of DPPs.


Introduction
Diabetes is a growing global health problem that places considerable burden on individuals and health systems 1 .Type 2 diabetes accounts for over 90% of all cases of diabetes with 438 million prevalent cases worldwide in 2019 2,3 .The prevalence and the death rates from type 2 diabetes rose by 49% and 11% respectively from 1990 to 2019 2 .This rising prevalence of type 2 diabetes has been attributed to ageing populations, improving medical care (which leads to people living longer with diabetes) and the rising incidence of diabetes risk factors, especially obesity 4 .
The Corona virus disease 2019 (Covid 19) pandemic was particularly devastating for people with diabetes.The risk of severe disease or death from Covid 19 in those with type 2 diabetes was at least twice as high than for those without diabetes 5 .Diabetes increased the burden of Covid 19 on health systems and economies globally 4 with evidence that the risk of developing type 2 diabetes is increased after Covid 19 infection even in those who were at low risk or had a mild infection 6 .The predicted rise in type 2 diabetes prevalence combined with the potential further increase in the incidence of type 2 diabetes due to Covid 19 infection has given rise to calls for urgent action to tackle the problem 4 .
Similar to type 2 diabetes, the prevalence of prediabetes is also rising globally 7 .Prediabetes is a stage of raised blood glucose levels where people at high risk of developing diabetes can be identified through measurement of fasting blood glucose, glycated haemoglobin or 2-hour post prandial load 7 .The International Diabetes Federation estimated 7.5% of people (374 million) worldwide live with impaired glucose tolerance 8 and identifying people at high risk is essential in tackling the growing problem of type 2 diabetes.Landmark randomised controlled trials investigating the effect of lifestyle change on diabetes incidence in those at high risk have demonstrated that type 2 diabetes can be prevented [9][10][11][12] .A meta-analysis by Galaviz et al. in 2018  synthesising the global impact of 63 diabetes prevention programmes (DPPs) that were implemented under real-world conditions or translated from proven interventions found that people receiving a lifestyle intervention had a 29% lower risk of developing type 2 diabetes than people who did not receive one 1 .DPPs aim to prevent type 2 diabetes by changing dietary and physical activity behaviours by raising awareness of the change needed, setting dietary and physical activity goals, action planning and relapse prevention 13 .International guidelines recommend identification, screening, and referral to DPPs for those at high risk of developing type 2 diabetes 13,14 .However, many eligible participants are not referred to or do not participate in DPPs when recruited 15,16 .A study investigating reach and use of diabetes prevention services in the USA found that only 4.9% of people at high risk were referred to a DPP and less than half (39.6%) of those attended 17 .The national DPP in England reports uptake by the first 100,000 referred to the programme was 56%, with only 19% completing it 18 .In a scoping review of existing evidence of the National Health Service DPP in England, the lack of evidence on how to improve initial engagement with the programme was highlighted and efforts to improve recruitment of those of working age, those from deprived and ethnic minority backgrounds was recommended 19 .
The reasons for low rates of referral and recruitment to DPPs are complex and multifaceted.System level factors can include a lack of engagement by leadership and a poor fit of the programme within the existing organisation 20 .Healthcare workers play a critical role in the referral pathway 21 .In a systematic review examining the implementation of DPPs, Aziz et al. (2015) suggests that high risk people identified and referred by health professionals resulted in higher participation rates, highlighting the importance of the healthcare worker in the referral process 22 .Studies have highlighted how healthcare professionals influence a person's view of their diagnosis of prediabetes and their decision to participate in DPP programmes 17,23 .Furthermore, a 2022 meta-synthesis on the barriers and facilitators to lifestyle change from the perspective of those at risk of type 2 diabetes reported that the guidance and education given by healthcare professionals facilitated positive change 24 .In a 2017 systematic review that explored the factors affecting diabetes prevention in the primary care setting, healthcare professionals were concerned about the extra workload that identifying people at risk of type 2 diabetes would bring and the impact on resources 25 .Some identified their lack of knowledge about preventing diabetes and a lack of confidence in providing diabetes prevention advice as barriers to providing preventative services.While most referrals to DPPs come through health professionals, not all DPPs are delivered in primary care.The national DPP in England for example is delivered by providers outside the national health service 26 .In the United States DPPs are delivered by community-based organisations, pharmacies, as well as health clinics 27 and programmes are available online 28 .The perspective of this broader group on referral and recruitment to DPPs warrants further investigation.
Therefore, this review aims to identify, appraise and synthesise the evidence on barriers and facilitators to referral and recruitment to DPPs from the perspective of healthcare workers (HCWs).Referral and recruitment to DPPs are the target behaviours.This review will adopt a "best fit" framework synthesis, a pragmatic and flexible approach, which integrates theory into systematic reviews 29 .The framework was primarily applied in the analysis stage and it did not inform data extraction.The "best fit" framework synthesis uses an existing theoretical framework (in this case the Theoretical Domains Framework) to organise the extracted data thereby ensuring the researcher is engaging with theory throughout the review process 30 .The "best fit" framework has the flexibility to consider themes outside of the a priori framework which allows for the theoretical framework to be tested and refined if necessary to better encapsulate the evidence available 29 .

Methods
This protocol describes a systematic review of qualitative, quantitative and mixed methods research on HCWs' perspectives on factors that affect referral and recruitment to DPPs.The systematic review has been prospectively registered on 21/12/2022 with the International Prospective Register of Systematic Reviews (PROSPERO) (CRD42022383023).This protocol adheres to the Preferred Reporting Items for Systematic review and Meta-Analysis Protocols (PRISMA-P) reporting guidelines 31,32 .

Review design
A "best fit" framework approach will be used to synthesise the evidence 29 and the Preferred Reporting Items for Systematic review and Meta-Analysis (PRISMA) guidelines 33 will be followed for reporting.The "best fit" framework synthesis method is flexible, pragmatic and is suitable for describing or explaining decisions about health and health behaviours 29 .The selection of this approach was informed by the RETREAT framework (Review question, Epistemology, Time, Resources, Expertise, Audience and purpose, Type of data) 34 .This review follows the seven steps of the "best fit" framework 29 (Table 1).
Step 1: Identifying a research question An initial scoping search was conducted to inform the research question and aims.The question was refined using the SPIDER (Sample, Phenomenon of Interest, Design, Evaluation and Research) framework 35 .This framework was adapted from the PICO (Population, Intervention, Comparison, Outcome) tool to facilitate the inclusion of qualitative and mixed methods as well as quantitative studies in reviews 35 .

Research question:
What are HCWs' perspectives on the barriers and facilitators to referral and recruitment to DPPs? Eligibility criteria: The eligibility criteria for this study are informed by the SPIDER framework (Table 2) 35 .
Sample: Any HCW involved in identifying, screening, referring or recruiting people at high risk of diabetes to DPPs.This could include, but not limited to, professionals such as primary care physicians, practice nurses, pharmacists, physiotherapists, dietitians, public health nurses, hospital doctors and nurses and could also include educators on the DPPs such as community health workers and lifestyle coaches.In line with the WHO definition, which encompasses support personnel, HCWs could also include administrative staff  or management who may have knowledge of the referral or recruitment process 38 .Studies that include the participants' perspectives on DPPs will only be included if data from the perspective of the HCW can be extracted separately.

Phenomenon of Interest:
Studies that report HCWs' perspectives on referral or recruitment to DPPs will be included.The perspective could include barriers and facilitators to recruitment and referral.A barrier will be defined as any factor that obstructs or impedes referral or recruitment to DPPs.A facilitator will be defined as any factor that promotes or eases referral or recruitment to DPPs.The barriers and facilitators can also include HCWs' perspectives of broader social and environmental influences.The DPPs should be programmes aimed at decreasing body weight if appropriate, achieving dietary recommendations and achieving physical activity recommendations.The programmes should be aimed at adults at high risk of developing type 2 diabetes.As this is a study on HCWs' opinions and not on the effectiveness of the programme the definition of high risk will not be limited to those with non-diabetic hyperglycaemia but could include those identified as being at high risk on a diabetes risk score (e.g., FINDRISK) or being overweight or obese 39 .There will be no restriction based on the setting, intensity or delivery mode of the programme.Studies of HCWs' perspectives on programmes for type 2 diabetes or gestational diabetes will be excluded.
Design and Research Type: All primary qualitative, quantitative and mixed methods research that meet the eligibility criteria will be included (Table 2).These could include qualitative (e.g., case studies or grounded theory), quantitative (e.g., cross sectional studies) or mixed methods studies which combine qualitative and quantitative data collection and analysis.Data collection methods could include observations, interviews, focus groups, surveys and questionnaires.The year of publication will be restricted to studies published after 1997 as this was when the first landmark trial on diabetes prevention was published 9 .Only studies published in English will be included due to the limited time available to conduct the review and limited access to translation services.
Step 2a: Identify framework The "best fit" approach allows for either identifying an appropriate framework from the published literature to guide the evidence synthesis or for generating a new meta-framework by systematically searching for and synthesising published frameworks.This study will use the former approach, selecting the Theoretical Domains Framework (TDF) as the a priori framework 36,40 .This framework was chosen as it is a synthesis of systematically reviewed behaviour change theories, developed specifically to understand and inform the implementation of evidence-based interventions 36,40 .The TDF is a comprehensive and rich source of theory that has been used previously in primary studies to understand implementation in DPPs 41 and it has been used as the a priori framework in the "best fit" method in a systematic review of patients' perceptions of diabetes prevention and cardiovascular disease programmes 42 .
Step 2b: Information sources and search strategy Systematic searching of electronic databases will be combined with supplementary search methods (i.e., reference and citation searching) to increase the identification of relevant research 43 .The databases commonly used in diabetes prevention systematic reviews will be searched.These include MEDLINE, EMBASE, CINAHL, PsychINFO, Web of Science and Scopus.A university librarian was consulted to further refine the search strategy.The search will use database specific controlled vocabulary (e.g., medical subject headings-MeSH), as well as key words in title and abstract, spelling variants, truncation and synonyms.Search dates (January 1997-March 2023).An example of a MEDLINE search is provided in Table 3. Step 3: Screening All references will be imported to Covidence, a systematic review software to manage screening, data extraction and quality assessment.If this software were unavailable an open access free alternative such as Rayyan could be used.Duplicates will be removed.At the beginning of the screening process the lead author and a second author will screen a small number of studies and meet to iteratively discuss their screening decisions, reflect on any differences of opinion and resolve any disagreements.A third author will be consulted if consensus cannot be reached.The lead author will then screen the abstract and titles of the remaining studies and the second reviewer will screen a random sample of 20% of the articles.Studies that do not meet the eligibility criteria will be excluded.If there is any doubt about exclusion at the title and abstract stage the studies will be kept in until the full text stage.The full text of all the potentially eligible research papers will then be read by the lead author to see if they meet the criteria to be included in the review.A second author will review a random sample of 20% of the articles.The lead author and a second reviewer will meet to iteratively discuss decisions and a third author will be consulted if required to reach consensus.A PRISMA flow diagram will be used to illustrate the screening and inclusion process and will outline the reasons for exclusion at the full text stages 33 .
Step 3 (continued): Data extraction Verbatim quotes and authors' interpretations relating to HCWs' perceptions of barriers and facilitators to recruitment and referral will be extracted from the results and discussion sections using a data extraction form.The data extracted will note whether it is a direct quote or author interpretation.Analysis from quantitative and mixed methods studies, such as survey results and descriptive summaries will also be extracted.Two authors will pilot this form with three of the included studies (one qualitative, one quantitative and one mixed methods if available) and the form will be refined and modified as needed with input from the wider research team if necessary 29 .The lead author will then extract the data from all included studies while the second author will extract the data from a random sample of 20% of the studies.Both authors will meet to iteratively discuss their decisions and a third author will be consulted if agreement cannot be reached.
The data extraction form will contain the following sections: study information (e.g., study number, title, first author and year of publication), characteristics of the study (e.g., aims, research design, sampling and sample size), participant characteristics (e.g., age, gender, years qualified, profession), programme characteristics (e.g., frequency, mode of delivery, length of programme, programme participants, setting (e.g.clinic, hospital, community), socioeconomic context) data collection approach and method of analysis and reported barriers and facilitators or related themes.If multiple reports from the same study are found, then these will be linked to avoid extracting the same data more than once.
Step 3 (continued): Quality assessment The quality of the studies will be appraised using the Mixed Methods Appraisal Tool 44 .This tool contains five different criteria which allows for concomitant appraisal of qualitative research, randomised controlled trials, nonrandomised quantitative studies, descriptive quantitative studies and mixed methods studies.The tool has been piloted for reliability and efficiency 45,46 and was updated in 2018 37 .The 2018 version recommends three steps to screen, choose the appropriate category and to rate the criteria.Studies will not be excluded if rated as low quality as they may have important insights 47 .The lead author will perform the quality assessment on all included studies while a second author will assess a random 20% sample.This will be an iterative process whereby three studies (qualitative, quantitative and mixed methods if available) will be assessed by each author and notes on decisions will be compared.This iterative process will continue with further studies until both authors are mostly in agreement to ensure a similar approach to the task.A third author will be consulted if necessary to resolve any disputes.
Step 4 and step 5: Deductive and inductive coding The data extracted from the studies on the barriers and facilitators to referral and recruitment will be coded deductively against the TDF Version 2 36 .This version has 14 theoretical domains and 84 component constructs related to behaviour change.The lead author will develop a coding tree in Nvivo with guidance from the 14 TDF domains and the 84 constructs.If this software were unavailable, analysis could be conducted using Excel and Word or open access software such as Taguette.Data that do not fit into the 14 TDF domains will be coded to an 'other' code and secondary thematic analysis of these codes will be performed 29 .An example of determinants coded to the "other" domain could be HCWs' perspectives of broader social and environmental factors which limit their patient's ability to participate.To ensure consistency in analysis three studies (qualitative, quantitative and mixed methods if available) will be jointly coded by the lead author and a second author.Quantitative data on the perceptions of HCWs will be coded to the relevant TDF domains.All data will be thematically synthesised.Quantitative data may be maintained in the reporting of results.Both authors will discuss their coding decisions and any discrepancies that cannot be resolved by consensus will be discussed with a third author.The lead author will conduct the analysis on the remaining studies and a second author on 20% of the articles.
Step 6: Develop the "best fit" framework The new themes will be added to the TDF domains, and an adapted model will then be produced by synthesising the a priori themes from the TDF and the additional concepts identified 29 .
Step 7: Assess the potential for bias in the new model/framework Relationships between the themes will be explored to create a new model or framework to understand the barriers and facilitators to recruitment and referral to DPPs from the HCWs' perspective.Differences between the TDF and the new framework (i.e., the absence or addition of themes) will be explored to assess and explain the selection and reporting of the research 29 .This will highlight if the differences are justifiable or if the literature needs to be reviewed again.If there is no evidence of dissonance in the synthesis, then purposeful efforts will be made to seek out "negative" cases 48 .Any sensitivity of the new framework to the study design, setting, quality assessment, subgroups of HCWs, year of data collection, delivery mode, and location will be examined.This will be conducted by examining the effect of excluding studies (e.g., those assessed as being low quality) on the themes in the framework.

Confidence
Confidence in the synthesised findings will be assessed using the GRADE-CERQual (Grades of Recommendation, Assessment, Development and Evaluation-Confidence in Evidence from Reviews of Qualitative research) approach 49 .The outcome of interest will be confidence in the evidence for barriers and facilitators to referral and recruitment to DPPs from the HCWs' perspective.The assessment of confidence will be based on four components: any limitations in the methodology of the primary studies contributing to a finding, how relevant the review question is to the review finding, how coherent a review finding is, and how adequate the data supporting a review finding is 49 .Ratings of confidence (high, moderate, low or very low) will be assigned to each review finding.Two authors will be involved in the confidence assessment and the review team will be consulted if further input is required to reach consensus.

Outcomes
The outcomes will be barriers or facilitators to referral or recruitment to DPPs from the HCWs' perspective.

Ethics
Ethical approval is not required for systematic reviews

Study status
The study has not yet commenced.The planned start for this study is March 2023.

Conclusion/Discussion
The potential of diabetes prevention programmes to impact on type 2 diabetes has not been fully realised due to the low numbers of people referred to and participating in the programmes.HCWs are a vital part of the referral pathway and can influence how seriously a person views their risk and their decision to take part.Using theory to understand the HCWs perspectives can give insights into why referral to DPPs is not routine practice and can identify what needs to be changed to improve referral and recruitment.This review will develop a conceptual framework to provide a broader understanding of these behaviours and will move beyond the description of the data to try to understand the relationships between the barriers and enablers identified.Using the "best fit" framework with the TDF as the a priori framework ensures that important themes are not overlooked while also allowing for new themes to be developed.The "best fit" framework is flexible and pragmatic allowing for the integration of qualitative, quantitative and mixed methods data.The inclusion of different forms of evidence will strengthen this review as "complex questions demand complex forms of evidence" 50 .
The findings of this review will be relevant to health systems and policy makers who have developed or are developing diabetes prevention programmes and will support further implementation efforts to improve recruitment and referral to DPPs.
review and framework synthesis of the evidence on barriers and facilitators to referral and recruitment to diabetes prevention programmes from the perspective of healthcare workers.The protocol is clearly presented and addresses an important step in the process of diabetes prevention, with anticipated implications for future implementation of such prevention programmes.Minor suggestions are offered below for the authors' consideration.

Introduction:
While it is implied that DPPs involve behaviour change, the content of these programmes could be briefly described in the introduction section to provide more context on what individuals are being asked to do when referred/recruited.

○
It is stated that the "best fit" method integrates theory into all parts of the systematic review and it seems theory is most explicitly integrated at the analysis/synthesis stage.Did theory also inform the development of the research question and identification of relevant studies, and if so was this based on guidance for applying the TDF? ○

Methods:
The TDF will provide a comprehensive framework for coding data on barriers and facilitators to DPP referral.As a behaviour change framework has been selected to structure the analysis, further attention could be given to the behavioural perspective in the justification for its use.For example, is referral and recruitment to DPPs positioned as the target behaviour, and if so is the overarching aim of this review to identify potential determinants of the current and desired behaviour?Being more explicit about the behavioural aspect will strengthen the justification for using the TDF here and may also help to guide the coding process.

○
In keeping with the "best fit" method, the protocol is framed with the assumption that new themes will be generated to account for data that cannot be deductively coded within the domains of the TDF.While the secondary thematic analysis will be conducted inductively, can the authors provide some suggestions as to how the TDF might be limited for coding relevant barriers and facilitators, for example are determinants beyond behavioural determinants anticipated to be relevant to recruitment and referral to DPP?
○ Challenges with coding data to TDF domains have been reported previously, including challenges defining the boundaries of domains and instances where data points appear to fit within several domains 1,2 .Will the coding tree be developed at the level of the 14 domains, or will it also refer to the 84 component constructs?

○
The approach to coding mixed and quantitative data to the TDF domains could be described in further detail.Will quantitative data be maintained in the reporting of results and if so, how will this be synthesised with qualitative evidence in the TDF coding and/or secondary thematic analysis? ○

Is the study design appropriate for the research question? Yes
Are sufficient details of the methods provided to allow replication by others?Yes Are the datasets clearly presented in a useable and accessible format?

Not applicable
Competing Interests: No competing interests were disclosed.
Reviewer Expertise: health psychology, self-management, health behaviour change I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
In keeping with the "best fit" method, the protocol is framed with the assumption that new themes will be generated to account for data that cannot be deductively coded within the domains of the TDF. 4. While the secondary thematic analysis will be conducted inductively, can the authors provide some suggestions as to how the TDF might be limited for coding relevant barriers and facilitators, for example are determinants beyond behavioural determinants anticipated to be relevant to recruitment and referral to DPP? Response 4. While the TDF will provide a framework to understand the influences on the HCWs behaviour it may not capture the HCWs perceptions on the broader social and environmental issues as highlighted by reviewer 1.An example of this could be the HCWs perception of financial constraints which limit their patient's ability to participate in DPPs.An example has been added to Step 4 and 5. Reviewer 2 5. Challenges with coding data to TDF domains have been reported previously, including challenges defining the boundaries of domains and instances where data points appear to fit within several domains1,2.Will the coding tree be developed at the level of the 14 domains, or will it also refer to the 84 component constructs?
NYU School of Medicine, New York, USA Thank you for the opportunity to review this article.The protocol is well-motivated with evidence on the rising prevalence of type 2 diabetes, role of lifestyle modification in preventing diabetes, low referral rates to diabetes prevention programs (DPPs), and role of health care workers (HCWs) in facilitating referrals.Understanding the barriers and facilitators to DPP referral and recruitment from HCWs' perspectives is an important goal, and the research question is clearly stated.My comments mainly focus on whether and how the review process and findings will illuminate social and environmental aspects of barriers and facilitators to DPP referral and recruitment.I discuss and encourage the authors to consider some of the pathways of influence that may be implicated here, and I offer suggestions relevant to their phenomenon of interest and theoretical framework.
The authors rightly note that reasons for low rates of referral and recruitment are complex.I draw out particular insights from the evidence they cite to motivate attention to social and environmental influences.For instance, the literature points to the role of structural factors (e.g., "outer setting" from the Consolidated Framework for Implementation Research or CFIR) in influencing DPP implementation, and it highlights that such factors are not widely surfaced in the literature (Madrigal et al., 2022 1 ).Evidence shows how HCWs perceive and respond, or feel unable to respond, to the social determinants of patients' health in relation to prediabetes management (Barry and Greenhalgh, 2022 2 ).Relatedly, a 2017 systematic review (Messina et al., 2017 3 ), although focused on primary care settings, presents three themes broadly relevant for the delivery and uptake of diabetes prevention interventions.The themes cover the role of context and setting, including features of the healthcare setting and perceived responsibility for lifestyle change; the role of patient factors such as motivation, including social determinants such as financial or time constraints that place diabetes prevention resources out of patients' reach; and the role of professional factors, e.g., workload and time (including competing health priorities), self-efficacy, knowledge, and perceptions of patients' motivation.
Taken together, this background literature suggests complex direct and indirect effects of social and environmental factors on HCWs' perceptions and behavior.A potential illustrative pathway involving social and environmental factors could be the following: Structural, external "outer setting" factors and patients' social determinants of health can influence patient motivation, which can inform HCWs' perceptions about patients' behavior and likely receptiveness to DPPs, which in turn can influence (and reinforce) HCWs' decisions and actions in counseling and referrals for diabetes prevention."Outer setting" factors can also directly influence HCWs' behavior and perspectives on DPP referrals and recruitment.
While the Theoretical Domains Framework (TDF) is an appropriate choice for an a priori framework, it would likely provide a close understanding of reflective, psychological processes that inform HCWs' perspectives and behavior (Atkins et al., 2017 4 ).The TDF does include "social influences" and "environmental context and resources" as constructs (Atkins et al., 2017 4 ), yet as defined in the TDF these refer to domains of influence on HCWs' behavior.My concern is that analysis based on the TDF alone may not adequately capture and distinguish social and environmental influences on patient behavior, which in turn can affect HCWs' perceptions and behavior.Can the authors discuss whether and how their review will address this?To what extent and how will the TDF be used to analyze social and environmental influences on HCWs' behavior and perspectives?
Examining social and environmental influences is significant not only because the social determinants of health are associated with risk of diabetes (Hill-Briggs et al., 2020 5 ).Social factors are additionally salient here because the review will examine the perspectives of a wide range of HCWs.HCWs' awareness of patients' social circumstances and sense of responsibility to address patients' social needs and lifestyle change can vary by their role, professional identity, and service setting, which can shape HCWs' perspectives and actions with regard to DPP referrals.An outcome of low DPP referrals may be differently explained by different mechanisms depending on the type of HCW.For example, community health workers may understand their role as encompassing support and assistance to resolve patients' social needs (Islam et al., 2017 6 and Malcarney et al., 2017 7 ).They may expend effort on activities related to social needs and could delay or defer DPP referrals and counseling on health behavior.Physicians in resource-constrained, safety-net health care settings may have little time for health counseling within clinical encounters and may feel compelled to prioritize patients with diabetes rather than those at risk for diabetes (Gore et al., 2020 8 ).They may fail to direct patients at risk for diabetes to (limited) resources such as community health workers who could refer patients to diabetes prevention programs.Can the authors discuss how the TDF will be applied to analyze the views and experiences of diverse subgroups of HCWs within the target population?
The "best fit" approach would lend flexibility and capture themes that lie outside the TDF.However, an extension or a complementary theoretical framework (Atkins et al., 2017 4 ) -one option is the Consolidated Framework for Implementation Research (CFIR) --could be useful from the outset to guide analysis of and elaborate social-environmental factors that shape HCWs' perspectives.
Another reason to account for social and environmental factors is to be alert to gaps in the data.A relevant and notable finding from the 2017 systematic review examining the delivery and uptake of diabetes prevention interventions was that socio-demographic data -such as on patients' income, ethnicity, and education --were poorly reported in its reviewed literature (Messina et al., 2017 3 ).This made it difficult to fully understand the context behind study findings (Messina et al., 2017 3 ).
A final point.In describing their phenomenon of interest, the authors state that they will seek studies that report, from HCWs' perspective, "the experience of a barrier or facilitator to referral or recruitment to DPPs."This framing of the phenomenon of interest, namely the HCWs' experience of a barrier or facilitator, seems to imply that HCWs attempt to make referrals and are unsuccessful or successful based on barriers and facilitators.Yet evidence suggests that low referral rates are partly due to factors such as HCWs' workload, lack of awareness, and competing health priorities in providing care.These factors indicate in part HCWs' inability to initiate referrals in the first place; they may not be attempting to make DPP referrals at all.While the research question is clear --"What are HCWs perspectives on the barriers and facilitators to referral and recruitment to DPPs?" --the phenomenon of interest appears to comprise several elements: HCWs' behavior with respect to DPP referrals; HCWs' experiences of barriers and facilitators when making referrals; HCWs' perspectives on barriers and facilitators, whether or not they make or attempt to make DPP referrals.
A reference to HCWs' behavior is present in the ).However, the authors also note that using both frameworks had implications for project management and timelines.A 2017 systematic review examining the combined use of the TDF and CFIR found using both frameworks could introduce unnecessary complexity.(Birkin et al., 2017).In light of the changes made to capture the HCWs views of the social and environmental factors affecting their patients (response 1 above) we are optimistic that the "best fit" approach will be flexible enough to capture the themes outside of the TDF and acknowledge that the project timeline will not accommodate the inclusion of another framework at this stage.The strengths and limitations of the fit of the TDF and the factors coded to the "other" domain will be considered in the discussion.To identify any gaps the data extraction form will now collect information on the programme participants, the setting (e.g.clinic, hospital, community) and the socioeconomic context.Reviewer 1 6.Can the authors clarify the scope and content of the phenomenon of interest with regard to the question of perspectives and behavior?Response 6. Thank you for highlighting this.The phenomenon of interest has now been clarified and the word "experience" removed to allow for perspectives other than those that HCWs have directly experienced such as workload, lack of awareness and competing health priorities.In the conclusion the word "behaviour" has been removed and "perspective" has been added to clarify that the phenomenon of interest is the HCWs perspective.

Table 3 . Example of a MEDLINE search strategy.
*" OR doctor* OR physician* OR nurse* OR "general practitioner" or GP* OR clinician* OR "health* worker" OR pharmacist* OR physiotherapist* OR "physical therapist*" OR dietitian* OR nutrition* OR "lifestyle coach*" OR "community health worker*" OR "health coach*" OR educator* OR "practice manager" OR "practice staff" OR practitioner*) OR ((family OR health OR healthcare OR "health care" OR medical OR nursing OR nutrition OR pharmacy OR physiotherapy OR "physical therapy" OR "public health" OR trainee) N2 (assistant* OR director* OR manager* OR officer* OR personnel OR practice OR practitioner* OR professional* OR provider OR staff)) ti,ab

Is the rationale for, and objectives of, the study clearly described? Yes Is the study design appropriate for the research question? Partly Are sufficient details of the methods provided to allow replication by others? Yes Are the datasets clearly presented in a useable and accessible format?
The proposed review is likely to provide insights into both behavior and perspectives, i.e., what HCWs do in terms of DPP referrals and HCWs' explanations for and views on why they do or don't make DPP referrals.Can the authors clarify the scope and content of the phenomenon of interest with regard to the question of perspectives and behavior?(Suppl 1): 360-382 PubMed Abstract | Publisher Full Text 8. Gore R, Brown A, Wong G, Sherman S, et al.: Integrating Community Health Workers into Safety-Net Primary Care for Diabetes Prevention: Qualitative Analysis of Clinicians' Perspectives.J Gen Intern Med.2020; 35 (4): 1199-1210 PubMed Abstract | Publisher Full Text No competing interests were disclosed.Reviewer Expertise: I have expertise in implementation science research, qualitative methods, and the application of social science theory to health services and public health research questions.I found that the combination can assist with the interpretation of findings in implementation research (O'Donovan et al., 2023 Conclusion section as well.The research question is focused on HCWs' perspectives on barriers and facilitators to DPP referrals and recruitment, but in the Conclusion/Discussion section the authors refer to HCWs' behavior: "Using theory to understand the HCWs behaviour can give insights into why referral to DPPs is not routine practice …" Hill-Briggs F, Adler NE, Berkowitz SA, Chin MH, et al.: Social Determinants of Health and Diabetes: A Scientific Review.Diabetes Care.2020; 44 (1): 258-79 PubMed Abstract | Publisher Full Text 6. Islam N, Shapiro E, Wyatt L, Riley L, et al.: Evaluating community health workers' attributes, roles, and pathways of action in immigrant communities.Prev Med.2017; 103: 1-7 PubMed Abstract | Publisher Full Text 7. Malcarney MB, Pittman P, Quigley L, Horton K, et al.: The Changing Roles of Community Health Workers.Health Serv Res.2017; 52 Suppl 1 : O'Donovan, B., Kirke, C., Pate, M., McHugh, S., Bennett, K., & Cahir, C. (2023).Mapping the Theoretical Domain Framework to the Consolidated Framework for Implementation Research: do multiple frameworks add value?.Implementation Science Communications, 4(1), 100.Birken, S. A., Powell, B. J., Presseau, J., Kirk, M. A., Lorencatto, F., Gould, N. J., ... & Damschroder, L. J. (2017).Combined use of the Consolidated Framework for Implementation Research (CFIR) and the Theoretical Domains Framework (TDF): a systematic review.Implementation science, 12, 1-14.Reviewer 1 5.Another reason to account for social and environmental factors is to be alert to gaps in the data.A relevant and notable finding from the 2017 systematic review examining the delivery and uptake of diabetes prevention interventions was that socio-demographic datasuch as on patients' income, ethnicity, and education --were poorly reported in its reviewed literature(Messina et al., 2017).This made it difficult to fully understand the context behind study findings(Messina et al., 2017).