May 2026· Ciência & Saúde Coletiva· Vol 31 5, pp.
e01132026
· 0 citations· 7 references
Medicine
TL;DR
It is concluded that current LLMs do not offer safety for autonomous operation in PHC, requiring rigorous professional curation, continuous validation, and escalation mechanisms to human care to ensure user safety.
Abstract
Obesity represents a growing challenge for Primary Health Care (PHC), demanding the use of new technologies and innovative solutions. This study evaluated the adherence of Large Language Models (LLMs) to Brazilian guidelines for obesity management in PHC. An in silico experimental study was conducted in two phases. Initially, 62 LLMs underwent latency screening, with 24 meeting the usability criterion (<10s) and proceeding to content evaluation. Clinical adherence was measured by 16 questions extracted verbatim from the Clinical Protocol (PCDT), covering diagnosis, monitoring, and treatment, and was peer-reviewed (Kappa 0.68-1.00). The exclusion of 38 models (61%) due to latency highlighted infrastructure barriers. Among the 24 viable models, adherence was heterogeneous and limited, with the best performance reaching only 61.1% compliance. The correlation between model size and accuracy was moderate, but multimodality proved to be a predictor of better performance. It is concluded that current LLMs do not offer safety for autonomous operation in PHC, requiring rigorous professional curation, continuous validation, and escalation mechanisms to human care to ensure user safety.
The increasing prevalence of non-communicable diseases, such as diabetes, hypertension, and metabolic disorders, highlights the need for accessible health screening services. However, existing health screening systems generally provide examination results without offering automated educational recommendations to support preventive healthcare. This study proposes an Artificial Intelligence (AI)-based health screening information system that automatically generates fitness and nutrition recommendations based on categorized health screening results. The research employed a Research and Development (R&D) approach using the Extreme Programming (XP) software development methodology. The proposed system was developed using the Laravel framework and the Filament administration panel and integrates a Large Language Model (LLM) through the OpenAI API. The proposed architecture combines structured prompt engineering, predefined AI guardrails, SHA-256 prompt hashing, and recommendation caching to improve recommendation consistency and computational efficiency. The system was evaluated using User Acceptance Testing (UAT) and an AI Recommendation Consistency Evaluation. The UAT results showed that all functional requirements were successfully fulfilled. The consistency evaluation demonstrated that repeated processing of identical health screening data produced stable recommendations, achieving an average qualitative consistency score of 90.4% and an average TF–IDF cosine similarity of 0.784. These findings indicate that the proposed architecture is capable of generating reliable and consistent educational recommendations while reducing redundant AI requests through recommendation caching. This study contributes to the development of intelligent health information systems by introducing an efficient AI recommendation architecture that supports digital health screening and preventive healthcare services.
Nasrul, Henry Saptono, Rusmanto· Komputasi· 0 citations
Aim: To evaluate the effectiveness of artificial intelligence (AI)-based chatbots in virtual nursing care at home, comparing them with traditional care, in terms of therapeutic adherence and improvement of clinical outcomes. Methods: A systematic review was conducted on March 7, 2025, according to the the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. The databases consulted were PubMed (Medline) and the Cochrane Library. Studies published since 2020 in Italian or English were included. After the screening process, six studies dealing directly or indirectly with the use of AI in home nursing consultation were included. Results: The analysis shows that chatbots and AI systems improve monitoring, therapeutic adherence, and access to healthcare services. Specifically, AI-based nutritional models showed improved anthropometric and biochemical indicators and a reduction in nutritional risk in patients with chronic kidney disease. Smart home systems have improved the timeliness of clinical event detection and the nurse’s decision-making role; medication support robots have reduced nursing workload and improved therapeutic safety. Critical issues related to training, professional acceptance, and technological infrastructure have emerged, limiting the full integration of AI into home care. Conclusion: This review highlights a growing interest in integrating AI into home nursing care, with promising preliminary results. The study fills a gap in the literature by identifying current trends, benefits, and limitations and highlights the need for further clinical research to validate the effectiveness of AI-based nursing chatbots.
This first scoping review of AI applications in hypertension health education identified a mismatch between rapid advances in generative AI and the limited availability of rigorous clinical evidence.
Haoran Chen, Shenglan Xiao, Tong Wan et al.· Journal of Medical Internet...· 0 citations
Recent studies suggest that health literacy (HL) may play an important role in management of individuals living with severe obesity (BMI ≥ 40 kg/m2), but to date, there has been no attempt to synthesize characteristics and outcomes of these studies. To explore the prevalence of low HL in people with severe obesity, and how HL may impact on health outcomes and management of severe obesity, six databases and two independent journals were searched. Seventeen studies were included in this review, mostly undertaken in high income countries, among females and candidates of bariatric surgery. Low HL ranged from 3.7% to 57.9% in cohorts studied, possibly explained by the heterogeneity of HL tools (i.e., seven different validated tools). Eight studies found higher HL to lead to better treatment outcomes (e.g., weight loss, attending appointments). Thematic synthesis of qualitative studies identified a lack of access to health information and support from healthcare providers (domains of HL) as barriers in improving health in patients with severe obesity. There is notable gap in interventional studies on improving HL in people with severe obesity. Overall, HL seems to be an important predictor of health outcomes in management of severe obesity. Further research is needed to assess the impact of HL on a broader range of health outcomes in nonsurgical patients with severe obesity.
Andrea Xia, Supreet Saluja, D. Muscat et al.· Obesity Reviews· 0 citations
BACKGROUND
Obesity is a major public health challenge worldwide, driving chronic disease and health care costs. Primary health care (PHC) is a key setting for prevention, early identification, and management of obesity, yet implementation of evidence-based interventions remains limited. Identifying the barriers and enablers that influence adoption of obesity care in PHC is critical to strengthen delivery.
OBJECTIVE
This scoping review will systematically map and synthesize barriers and enablers to implementing obesity interventions in PHC, considering behavioral, nutritional, pharmacological, digital, and organizational approaches, across both adult and pediatric populations.
METHODS
The review will follow the Joanna Briggs Institute (JBI) methodology and be reported according to PRISMA-ScR guidelines. Searches will be conducted in PubMed, Cochrane Library, Scopus, Web of Science, and ScienceDirect, complemented by gray literature from OpenGrey, ResearchGate, Google Scholar, and organizational websites. Eligible sources will include peer-reviewed and gray literature published between 2019 and 2025 in English, French, German, Italian, Portuguese, Spanish, or Turkish. Two reviewers will independently screen, extract, and analyze data, with discrepancies resolved by consensus. Data will be synthesized descriptively and thematically, supported by NVivo, and results presented in tables, thematic maps, and conceptual diagrams. Expected Results Anticipated barriers include time constraints, limited provider training, stigma, scarce resources, and inadequate reimbursement models. Enablers are expected to include supportive health policies, training, patient engagement strategies, and team-based care models.
CONCLUSIONS
This review will synthesize implementation factors affecting obesity care in PHC, informing scalable strategies and policy to improve obesity management worldwide.
Sandra León-Herrera, Ileana Gefaell, M. Guisado-Clavero et al.· Obesity Reviews· 0 citations
A scoping review of 24 PubMed-indexed studies published between 2023 and 2026 was conducted to assess current applications, benefits, limitations, and future directions of LLMs in healthcare.
Antoni Klamka, Paulina Kawalec, Kamil Bronikowski et al.· Quality in Sport· 0 citations