CoChat
Sign Up
Cited by 14Artificial Intelligence in Healthcare and EducationArtificial Intelligence in HealthcareElectronic Health Records Systems

Barriers and Facilitators to Artificial Intelligence Implementation in Diabetes Management from Healthcare Workers’ Perspective: A Scoping Review

Open access
Cangelosi, Giovanni, Conti, Andrea, Caggianelli, Gabriele, Panella, Massimiliano, Petrelli, Fabio, Mancin, Stefano, Ratti, Matteo, Masini, Alice·2025·Medicina
DOI 10.3390/medicina61081403PMC PMC12387811
Open this paper in CoChat — free accountOpen PDF

Abstract

Background and Objectives: Diabetes is a global public health challenge, with increasing prevalence worldwide. The implementation of artificial intelligence (AI) in the management of this condition offers potential benefits in improving healthcare outcomes. This study primarily investigates the barriers and facilitators perceived by healthcare professionals in the adoption of AI. Secondarily, by analyzing both quantitative and qualitative data collected, it aims to support the potential development of AI-based programs for diabetes management, with particular focus on a possible bottom-up approach. Materials and Methods: A scoping review was conducted following PRISMA-ScR guidelines for reporting and registered in the Open Science Framework (OSF) database. The study selection process was conducted in two phases—title/abstract screening and full-text review—independently by three researchers, with a fourth resolving conflicts. Data were extracted and assessed using Joanna Briggs Institute (JBI) tools. The included studies were synthesized narratively, combining both quantitative and qualitative analyses to ensure methodological rigor and contextual depth. Results: The adoption of AI tools in diabetes management is influenced by several barriers, including perceived unsatisfactory clinical performance, high costs, issues related to data security and decision-making transparency, as well as limited training among healthcare workers. Key facilitators include improved clinical efficiency, ease of use, time-saving, and organizational support, which contribute to broader acceptance of the technology. Conclusions: The active and continuous involvement of healthcare workers represents a valuable opportunity to develop more effective, reliable, and well-integrated AI solutions in clinical practice. Our findings emphasize the importance of a bottom-up approach and highlight how adequate training and organizational support can help overcome existing barriers, promoting sustainable and equitable innovation aligned with public health priorities.

AI snapshot

Written from the paper’s full text.

TL;DR

This scoping review synthesized 7 studies (2019–2024) on healthcare workers’ perceived barriers and facilitators to AI implementation in diabetes management, mapped through the CFIR framework. It identifies trust, performance, cost, data security, training, and integration concerns as key barriers, while efficiency, usability, time savings, and organizational support emerge as facilitators.

Key findings

  • From 3451 records retrieved, 593 duplicates removed, 3143 titles/abstracts screened, 32 full texts reviewed, and 7 studies included; sample sizes ranged from 10 to 207 participants in the text, with ages 40–60 years.
  • Barriers included unsatisfactory clinical performance, high acquisition costs, data security/liability concerns, black-box decision-making, limited training, poor integration with hospital information systems, lack of collaborative networks, fear of job displacement or lost autonomy, insufficient reimbursement, and increased workload.
  • Facilitators included improved clinical efficiency, ease of use, time savings for patients and physicians, organizational support, engagement of administrators/department heads, positive attitudes toward AI, perceived care coordination, ease of integration, national guidelines, and a stronger primary doctor filter function.
  • Most included studies focused on AI for diabetic retinopathy screening (5 of 7); only two addressed glycemic control or general diagnostic intervention. Designs were 4 qualitative, 2 cross-sectional, and 1 mixed-methods.

Methods

  • Scoping review following PRISMA-ScR, registered in OSF, using the PCC framework and searching PubMed Medline, Scopus, CINHAL, and Embase without temporal restrictions, updated to 31 January 2025.
  • Two-phase selection (title/abstract then full text) conducted independently by three researchers, with a fourth resolving conflicts; data extraction and quality appraisal used JBI tools and MMAT, with narrative synthesis guided by CFIR and SWiM because meta-analysis was not feasible.

Limitations

  • Stated by the paper: the exploratory scoping review design prevented quantitative synthesis due to methodological heterogeneity; many findings rely on subjective perceptions of healthcare workers; future longitudinal and randomized studies are needed to assess real effectiveness and efficiency.
  • Evident from the text: only 7 studies were included, most focused on diabetic retinopathy screening rather than broader diabetes management; sample sizes varied widely; quality appraisal rated several studies as low or medium quality, and no studies were excluded for insufficient quality.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/medicina61081403

Cites

112 of 112 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

51 of 101 cited sources checked · 32 supported · 0 flagged · accuracy not yet measured

  • Study design and conductreasoning

    A narrative synthesis was chosen because a combined quantitative synthesis (meta-analysis) was not feasible due to heterogeneity of included studies.

    “In this review, while the benefits of meta-analysis are acknowledged, a combined quantitative synthesis was deemed not feasible due to the heterogeneity of the included studies.”
    • Also stated in: § 2.7. Synthesis of the Results · plain
      “As a result, a detailed narrative synthesis was chosen, following established guidelines for synthesis without meta-analysis (SWiM) [50].”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [50] Campbell, 2020, Synthesis without Meta-Analysis (SWiM) in Systematic Reviews: Reporting Guideline, BMJ source

      The Abstract describes alternatives to meta-analysis, often called narrative synthesis, and Item 5 discusses methods to use when meta-analysis is not possible. The guideline does not assess the citing review’s included studies or establish that their heterogeneity made meta-analysis infeasible; that review-specific reason is not supported by this source. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
  • Findingcitation

    Ease of use is a facilitator to AI adoption in diabetes management.

    “easiness of use [52,54]”
    • Also stated in: § Abstract · plain
      “ease of use”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [52] Liao, 2024, Barriers and Facilitators to Implementing Imaging-Based Diagnostic Artificial Intelligence-Assisted Decision-Making Software in Hospitals in China: A Qualitative Study Using the Updated Consolidated Framework for Implementation Research, BMJ Open source

      The Results section, under “Innovation complexity (+): easiness of use,” identifies ease of use as a facilitator of implementing imaging-based diagnostic AI software in hospitals. The paper does not establish this finding specifically for AI adoption in diabetes management; its mention of diabetic retinopathy does not make that the setting of the ease-of-use finding. (LLM, conf=0.85)

      Experimental — accuracy not yet measured
    • cites [54] Roy, M., Jamwal, M., Vasudeva, S., and Singh, M. (2024). Physicians Behavioural Intentions towards AI-Based Diabetes Diagnostic Interventions in India. J. Public Health. source
  • Findingcitation

    Impact of a grading workload is a barrier in the implementation process domain.

    “and the impact of a grading workload [57].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [57] Wahlich, 2025, Patient and Practitioner Perceptions around Use of Artificial Intelligence within the English NHS Diabetic Eye Screening Programme, Diabetes Res. Clin. Pract. source

      Section 3.1.3 (Impact on the workforce; page unavailable) reports that practitioners expected AI to reduce graders’ workload but worried this could lead to job losses or cuts rather than reinvestment. This supports concern about the impact of changing grading workload, but the paper does not identify grading workload itself as a barrier in an ‘implementation process’ domain. (LLM, conf=0.78)

      Experimental — accuracy not yet measured
  • Findingcitation

    Successful active engagement of the clinicians is a facilitator in the implementation process domain.

    “The only facilitator identified in the implementation process is the successful active engagement of the clinicians [52].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [52] Liao, 2024, Barriers and Facilitators to Implementing Imaging-Based Diagnostic Artificial Intelligence-Assisted Decision-Making Software in Hospitals in China: A Qualitative Study Using the Updated Consolidated Framework for Implementation Research, BMJ Open source
  • Findingcitation

    Improved clinical efficiency is a facilitator to AI adoption in diabetes management.

    “The identified facilitators are an improved clinical efficiency [52,54,55]”
    • Also stated in: § Abstract · plain
      “Key facilitators include improved clinical efficiency”
    Quote in the paperSubstantial claim1 of 3 sources checked
    • cites [52] Liao, 2024, Barriers and Facilitators to Implementing Imaging-Based Diagnostic Artificial Intelligence-Assisted Decision-Making Software in Hospitals in China: A Qualitative Study Using the Updated Consolidated Framework for Implementation Research, BMJ Open source

      The Results table identifies improved clinical efficiency as a facilitator for implementing imaging-based diagnostic AI software, and the Results > Innovation section says participants found the software quicker than human interpretation. The Methods > Innovation selection section shows that diabetic-retinopathy screening was one application studied, but the efficiency finding is not reported specifically for diabetes management. (LLM, conf=0.87)

      Experimental — accuracy not yet measured
    • cites [54] Roy, M., Jamwal, M., Vasudeva, S., and Singh, M. (2024). Physicians Behavioural Intentions towards AI-Based Diabetes Diagnostic Interventions in India. J. Public Health. source
    • cites [55] Wyatt, 2020, A Lesson in Implementation: A Pre-Post Study of Providers’ Experience with Artificial Intelligence-Based Clinical Decision Support, Int. J. Med. Inform. source
  • Classified as probably not a substantial claim. 20
  • May combine two claims. 12
  • The quote only partly states this claim. 3

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

    Study design and conductcitation

    Data extraction and quality appraisal used Joanna Briggs Institute (JBI) tools.

    “The risk of bias and methodological quality of the included studies were assessed using established guidelines of the JBI framework [43].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [43] Tricco, 2018, PRISMA Extension for Scoping Reviews (PRISMA-ScR): Checklist and Explanation, Ann. Intern. Med. source

      The Introduction says JBI published guidance for conducting scoping reviews, and the PRISMA-ScR checklist (Table 1) addresses how to report critical appraisal if it is done. Neither passage establishes that JBI tools were used to assess the included studies; the checklist is a reporting guideline, not a record of that assessment. (LLM, conf=0.82)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    Risk of bias and methodological quality were assessed using JBI checklists for qualitative and cross-sectional studies and the MMAT tool for mixed-methods studies.

    “The risk of bias and methodological quality of the included studies were evaluated using JBI checklists for qualitative [45] and cross-sectional studies [46] and the MMAT tool for mixed-methods studies [47].”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites [45] Lockwood, 2015, Qualitative Research Synthesis: Methodological Guidance for Systematic Reviewers Utilizing Meta-Aggregation, Int. J. Evid.-Based Healthc. source
    • cites [46] Aromataris, E., and Munn, Z. (2017). Chapter 7: Systematic Reviews of Etiology and Risk. Joanna Briggs Institute Reviewer’s Manual, The Joanna Briggs Institute.

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [47] Hong, 2018, The Mixed Methods Appraisal Tool (MMAT) Version 2018 for Information Professionals and Researchers, EFI source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    The scoping review was conducted following PRISMA-ScR guidelines for reporting.

    “This review followed the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews guidelines (PRISMA-ScR) (PRISMA-ScR checklist available in Supplementary Table S1) [43].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [43] Tricco, 2018, PRISMA Extension for Scoping Reviews (PRISMA-ScR): Checklist and Explanation, Ann. Intern. Med. source

  • The quote only partly states this claim.

    Findingcitation

    Poorly tailored reporting of AI systems is a barrier to AI adoption.

    “Romero et al. [55] and Liao et al. [52] also reported the poorly tailored reporting of AI systems.”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites [55] Wyatt, 2020, A Lesson in Implementation: A Pre-Post Study of Providers’ Experience with Artificial Intelligence-Based Clinical Decision Support, Int. J. Med. Inform. source
    • cites [52] Liao, 2024, Barriers and Facilitators to Implementing Imaging-Based Diagnostic Artificial Intelligence-Assisted Decision-Making Software in Hospitals in China: A Qualitative Study Using the Updated Consolidated Framework for Implementation Research, BMJ Open source

  • Classified as probably not a substantial claim.

    Study design and conductcitation

    The CFIR framework was used to structure the description and synthesis of barriers and facilitators.

    “The description of the identified barriers and facilitators is structured according to the Consolidated Framework for Implementation Research (CFIR) [49]”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [49] Damschroder, 2009, Fostering Implementation of Health Services Research Findings into Practice: A Consolidated Framework for Advancing Implementation Science, Implement. Sci. source

  • May combine two claims.

    Interpretationcitation

    Fragmented care coordination and lack of shared governance structures limit the effective deployment of AI across healthcare levels.

    “Additionally, fragmented care coordination and lack of shared governance structures were seen as limiting factors in the effective deployment of AI across healthcare levels [101,102].”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [101] Nair, 2023, Impact of Holistic, Patient-Centric Yoga & Naturopathy-Based Lifestyle Modification Program in Patients with Musculoskeletal Disorders: A Quasi-Experimental Study, Adv. Integr. Med. source
    • cites [102] Baines, 2022, Meaningful Patient and Public Involvement in Digital Health Innovation, Implementation and Evaluation: A Systematic Review, Health Expect. source

      The Results and Discussion passages address barriers to patient and public involvement in digital health, including power imbalances and a disconnect between developers and patients. They do not address fragmented care coordination, shared governance across healthcare levels, or AI deployment. Searches for those terms and close paraphrases found no passage supporting the attributed claim; page numbers were unavailable in our copy. (LLM, conf=0.76)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Two included studies were performed inside community healthcare centers.

    “Notably, two studies were performed inside community healthcare centers [52,57].”
    Quote in the paperNot a substantial claim2 of 2 sources checked
    • cites [52] Liao, 2024, Barriers and Facilitators to Implementing Imaging-Based Diagnostic Artificial Intelligence-Assisted Decision-Making Software in Hospitals in China: A Qualitative Study Using the Updated Consolidated Framework for Implementation Research, BMJ Open source
    • cites [57] Wahlich, 2025, Patient and Practitioner Perceptions around Use of Artificial Intelligence within the English NHS Diabetic Eye Screening Programme, Diabetes Res. Clin. Pract. source

      The Subjects, materials and methods section describes an online survey and says validation interviews took place at three diabetic eye screening programme centres. It does not identify those centres as community healthcare centres or say the study itself was performed inside them; searches for community health centres yielded no further passage establishing that setting. (LLM, conf=0.70)

      Experimental — accuracy not yet measured

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductcitation

    The use of the CFIR framework allowed for an in-depth analysis of factors influencing AI adoption by healthcare workers and a coherent classification from a public health perspective.

    “The use of the CFIR framework [49] allowed for an in-depth analysis of the factors influencing AI adoption by healthcare workers, facilitating a coherent and useful classification from a public health perspective.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [49] Damschroder, 2009, Fostering Implementation of Health Services Research Findings into Practice: A Consolidated Framework for Advancing Implementation Science, Implement. Sci. source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Two quantitative cross-sectional studies were based on surveys.

    “Two quantitative cross-sectional studies were based on surveys [54,56]”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites [54] Roy, M., Jamwal, M., Vasudeva, S., and Singh, M. (2024). Physicians Behavioural Intentions towards AI-Based Diabetes Diagnostic Interventions in India. J. Public Health. source
    • cites [56] Wewetzer, 2023, Determinants of the Implementation of Artificial Intelligence-Based Screening for Diabetic Retinopathy-a Cross-Sectional Study with General Practitioners in Germany, Digit. Health source

  • Classified as probably not a substantial claim.

    Interpretationcitation

    The lack of adequate training on AI tools is a systemic barrier that must be urgently addressed.

    “From an internal perspective, the lack of adequate training on AI tools is a systemic barrier that must be urgently addressed—especially if the goal is to adopt AI-based care systems across multiple levels of healthcare delivery [77].”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [77] Ahmed, 2023, A Systematic Review of the Barriers to the Implementation of Artificial Intelligence in Healthcare, Cureus source

  • May combine two claims.

    Study design and conductcitation

    The study adopted JBI methodology, PRISMA-ScR guidelines, and OSF protocol registration, which ensure transparency and international rigor.

    “The study’s strengths include its solid methodological framework, facilitated by the adoption of the JBI methodology, the PRISMA-ScR guidelines for reporting [43,44], and the registration of the protocol on the OSF database, elements that ensure transparency and international rigor.”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites [43] Tricco, 2018, PRISMA Extension for Scoping Reviews (PRISMA-ScR): Checklist and Explanation, Ann. Intern. Med. source
    • cites [44] Pollock, 2023, Recommendations for the Extraction, Analysis, and Presentation of Results in Scoping Reviews, JBI Evid. Synth. source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingreasoning

    Wewetzer et al. was reported as medium-quality owing to the lack of managing confounding factors.

    “while the study by Wewetzer et al. [56] was reported as medium-quality owing to the lack of managing confounding factors.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [56] Wewetzer, 2023, Determinants of the Implementation of Artificial Intelligence-Based Screening for Diabetic Retinopathy-a Cross-Sectional Study with General Practitioners in Germany, Digit. Health source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingnone given

    Liao et al. and Peterson et al. were well designed.

    “The studies by Liao et al. [52] and Peterson et al. [53] were well designed.”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites [52] Liao, 2024, Barriers and Facilitators to Implementing Imaging-Based Diagnostic Artificial Intelligence-Assisted Decision-Making Software in Hospitals in China: A Qualitative Study Using the Updated Consolidated Framework for Implementation Research, BMJ Open source
    • cites [53] Petersen, 2024, How to Improve Attendance for Diabetic Retinopathy Screening: Ideas and Perspectives From People with Type 2 Diabetes and Health-Care Professionals, Can. J. Diabetes source

  • May combine two claims.

    Premisecitation

    In chronic disease management, AI systems can support transitions between care settings, enhance remote monitoring, and offer predictive insights that guide therapeutic decisions.

    “For chronic disease management, where continuity and coordination of care are essential, AI systems can support transitions between care settings, enhance remote monitoring, and offer predictive insights that guide therapeutic decisions [106,107,108].”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites [106] Awad, 2021, Connected Healthcare: Improving Patient Care Using Digital Health Technologies, Adv. Drug. Deliv. Rev. source

      The Introduction describes remote monitoring and chronic-condition management outside traditional care settings, and §2 (Sensors) says AI can improve interpretations of real-time sensor data. The paper does not establish that AI supports transitions between care settings. Section 5.1 describes predictive analysis for its diabetes-monitoring example as a future possibility, not an existing source of insights guiding treatment decisions. Page numbers were not provided in our copy. (LLM, conf=0.85)

      Experimental — accuracy not yet measured
    • cites [107] Neborachko, 2019, Current Trends of Digital Solutions for Diabetes Management, Diabetes Metab. Syndr. Clin. Res. Rev. source
    • cites [108] Yang, 2024, Harnessing the Potential of Artificial Intelligence: Affordances, Constraints, and Strategic Implications for Professional Services, J. Strateg. Inf. Syst. source
    • Rests on an assumption: Continuity and coordination of care are essential for chronic disease management.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    One included study adopted a mixed-methods design.

    “and finally, one study adopted a mixed-methods design [55].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [55] Wyatt, 2020, A Lesson in Implementation: A Pre-Post Study of Providers’ Experience with Artificial Intelligence-Based Clinical Decision Support, Int. J. Med. Inform. source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    The involved healthcare professionals were mainly nurses, healthcare assistants, and doctors.

    “The involved healthcare professionals were mainly nurses, healthcare assistants, and doctors (i.e., ophthalmologists, diabetologists, endocrinologists, physicians, and general practitioners).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    All included studies described research conducted in primary care settings, except the study conducted by Roy et al., which was conducted in a clinic.

    “All the included studies described research conducted in primary care settings, with the exception of the study conducted by Roy et al. [54], which was conducted in a clinic.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [54] Roy, M., Jamwal, M., Vasudeva, S., and Singh, M. (2024). Physicians Behavioural Intentions towards AI-Based Diabetes Diagnostic Interventions in India. J. Public Health. source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    The population of included studies varied from 10 participants to 207, with a range of ages from 40 to 60 years.

    “The population varied from 10 participants to 207, with a range of ages from 40 to 60 years.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingnone given

    Romero et al. mixed-methods study was well designed and conducted, with a high-quality rating.

    “Finally, the mixed-methods study performed by Romero et al. [55] all was well designed and conducted, with a high-quality rating.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [55] Wyatt, 2020, A Lesson in Implementation: A Pre-Post Study of Providers’ Experience with Artificial Intelligence-Based Clinical Decision Support, Int. J. Med. Inform. source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    The review protocol was registered in the Open Science Framework database.

    “The protocol for this review was registered in the Open Science Framework database (https://osf.io/xgy2z; accessed on 12 June 2025).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

    Study design and conductown data

    Study selection was conducted in two phases-title/abstract screening and full-text review-independently by three researchers, with a fourth resolving conflicts.

    “The study selection process for this review followed a two-phase procedure: an initial screening of titles and abstracts, followed by a detailed evaluation of full-text articles.”
    • Also stated in: § 2.4. Study Selection Process · plain
      “The initial screening was conducted independently and blind by three authors (G.C., A.M., and M.R.), who evaluated titles and abstracts based on their relevance to the study and in accordance with the predefined inclusion criteria. A fourth independent researcher (A.C.) resolved any disagreements at this stage.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    No studies were excluded for insufficient quality due to the exploratory nature of the work.

    “However, due to the exploratory nature of the present work, no studies were excluded for insufficient quality.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationreasoning

    Healthcare professionals' proximity to patients and deep understanding of local care pathways position them as crucial actors in fostering responsible innovation.

    “Their proximity to patients and deep understanding of local care pathways position them as crucial actors in fostering responsible innovation.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Healthcare professionals have proximity to patients and deep understanding of local care pathways.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    These findings point toward the value of a bottom-up approach that prioritizes the perspectives and practical needs of frontline professionals as a promising pathway to support successful implementation of AI in clinical practice.

    “Crucially, these findings point toward the value of a bottom-up approach—one that prioritizes the perspectives and practical needs of frontline professionals—as a promising pathway to support the successful implementation of AI in clinical practice.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The perspectives and practical needs of frontline professionals are important for successful implementation of AI in clinical practice.

  • May combine two claims.

    Interpretationreasoning

    Lack of coordination between levels of care and communication difficulties among professionals represent critical barriers to AI implementation.

    “In terms of governance and shared decision-making, it becomes clear that the lack of coordination between levels of care (e.g., primary care, specialists, hospitals) and potential communication difficulties among professionals represent critical barriers to AI implementation.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductreasoning

    Many findings are based on subjective perceptions of the involved workers in the included studies, requiring caution in interpreting the data.

    “Additionally, many findings are based on subjective perceptions of the involved workers in the studies included, which requires caution in interpreting the collected data and highlights the need for future integration with objective data and evidence-driven research.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Active and continuous involvement of healthcare professionals represents a key opportunity to develop more effective, reliable, and context-aware AI solutions better integrated into everyday clinical workflows.

    “Their active and continuous involvement represents a key opportunity to develop more effective, reliable, and context-aware AI solutions that are better integrated into everyday clinical workflows.”
    Quote in the paperNot a substantial claim
27 of 27

  • The paper relies on this without support.

    The issue of potential barriers and facilitators perceived by healthcare professionals in the effective implementation of AI-based technologies in routine clinical practice remains largely unexplored.

    “Nonetheless, the issue of potential barriers and facilitators perceived by healthcare professionals in the effective implementation of these AI-based technologies in routine clinical practice remains largely unexplored—a gap this study aims to address in a bottom-up view”

  • The paper relies on this without support.

    The barriers and facilitators identified from studies focused mainly on diabetic retinopathy screening generalize to AI implementation in diabetes management more broadly.

    “The findings of this study, interpreted through the CFIR framework [49], clearly show that the barriers and facilitators to AI adoption operate on multiple levels, reflecting organizational, technological, individual, and systemic dynamics that influence full implementation and value realization.”

  • The paper relies on this without support.

    Continuity and coordination of care are essential for chronic disease management.

    “For chronic disease management, where continuity and coordination of care are essential”

  • The paper relies on this without support.

    Healthcare professionals have proximity to patients and deep understanding of local care pathways.

    “Their proximity to patients and deep understanding of local care pathways”

  • The paper relies on this without support.

    The perspectives and practical needs of frontline professionals are important for successful implementation of AI in clinical practice.

    “one that prioritizes the perspectives and practical needs of frontline professionals—as a promising pathway to support the successful implementation of AI in clinical practice.”
5 of 5

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Record count after duplicate removal does not add up

    Major

    Removing 593 duplicates from 3451 records leaves 2858, not the reported 3143 records screened.

    Reported 3143 · recomputed 2858

  • Hospital study contradicts primary-care setting claim

    Minor

    The text says Roy et al. was the only study outside primary care, but Table 2 lists Petersen et al. as conducted in a hospital.

  • Female percentage column contains a value above 100%

    Minor

    Table 2 labels the parenthetical figure as percent female, but reports 107 for a sample of 209. If 107 is the number of female participants, the percentage is approximately 51.2%, not 107%.

    Reported 107 · recomputed 51.196172

  • Cross-sectional studies receive conflicting quality ratings

    Minor

    The text calls both cross-sectional studies low-quality, then rates one of them, Wewetzer et al., medium-quality.

  • Reported sample-size range excludes a listed study

    Minor

    The stated maximum sample size is 207, but Table 2 lists 209 participants for Wewetzer et al.

5 of 5

Cites

Flagged first, then by citations

7 of the 119 references couldn’t be matched to a known paper.

Research works printed but not linked: 1

  1. Perais, 2023, Prognostic Factors for the Development and Progression of Proliferative Diabetic Retinopathy in People with Diabetic Retinopathy, Cochrane Database Syst. Rev., 2023, CD013775

Cited by

Sorted by citations

Related

Authors
Cangelosi, G., Conti, A., Caggianelli, G., Panella, M., Petrelli, F., Mancin, S., Ratti, M., & Masini, A.
Affiliations
  • Università di Camerino (IT)
  • Università degli Studi del Piemonte Orientale “Amedeo Avogadro” (IT)
  • Azienda Ospedaliera San Giovanni Addolorata (IT)
  • IRCCS Humanitas Research Hospital (IT)
Corresponding author
  • Panella, Massimiliano
  • Petrelli, Fabio
ORCID
Venue
Medicina
Publisher
MDPI AG
Published
2025-08-01
Volume
61
Issue
8
Pages
1403
ISSN
  • 1010-660X (ISSN-L)
  • 1648-9144
DOI
10.3390/medicina61081403
PMC
PMC12387811
OpenAlex
W4412943687
Semantic Scholar
432cdb9e064f20cea4e8aed6c8ef1e5fea14d3aa
Cited by
14
References
119
Citations by year
2026: 11 · 2025: 3
FWCI
2.89
Type
journal article
Language
en
Open access
Licence
Locations
  • Medicina (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Europe PMC (PubMed Central) (submittedVersion, cc-by)
  • PubMed Central (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Unicam Scientific Publications (University of Camerino) (submittedVersion, cc-by)
Topics
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
  • Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
  • Health Sciences › Health Professions › Health Information Management › Electronic Health Records Systems
Keywords
artificial intelligence, diabetes management, technology adoption, barriers and facilitators, healthcare workers, bottom-up approach, organizational support, clinical teaching, data security, transparency in decision-making, clinical efficiency, AI acceptance
MeSH terms
Artificial Intelligence, Diabetes Mellitus, Health Personnel, Humans, Qualitative Research

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-13.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-13.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review