JSJosé Ferreira Santos

José Ferreira Santos

*This is an algorithmically generated author record based on publicly available information.

José Ferreira Santos has a total of 19 co-authors. Here are the top co-authors by number of shared publications:

  • Hélder Dores

    Hospital da Luz Lisboa , Luz Saúde, Lisboa

    4 shared publications
  • Hélder Dores

    Department of Cardiology, Hospital da Luz Lisboa, Luz Saúde, 1600-209 Lisbon, Portugal

    3 shared publications
  • Francisca Leite

    Católica Medical School , Sintra Campus, Estrada OctávioPato, 2635-631 Rio de Mouro, Lisboa

    2 shared publications
  • Ricardo Ladeiras-Lopes

    Hospital da Luz Guimarães, Luz Saúde, Portugal; Cardiovascular Research and Development Centre-UnIC@RISE, Department of Surgery and Physiology, Faculty of Medicine of the University of Porto, Porto, Portugal

    2 shared publications
  • Bernardo Neves

    Católica Medical School , Sintra Campus, Estrada OctávioPato, 2635-631 Rio de Mouro, Lisboa

    1 shared publication
  • Diana Teixeira

    CHRC, NOVA Medical School, Lisboa, Portugal

    1 shared publication

Latest Publications

  • Open Access

    Large language models approach clinician performance in ESC cardiovascular risk stratification: a vignette-based benchmark study

    by José Ferreira Santos, Regina de Brito Duarte, Inês Mota, Rita Carvalheira Santos, José Maria Moreira, Joana Campos, Nuno André Silva, Bernardo Neves, Ricardo Ladeiras-Lopes, Francisca Leite, et al.

    Aims: Guideline-based cardiovascular risk stratification requires three distinct competencies: extracting risk factor data from clinical text, computing a validated risk score, and applying guideline-defined thresholds to assign a final risk category. We evaluated contemporary large language models (LLMs) on each of these tasks within the European Society of Cardiology (ESC) SCORE2 framework and compared LLM performance against a pooled individual clinician benchmark to contextualize findings against real-world human reproducibility. Methods and results: Eleven LLMs were evaluated using 30 simulated outpatient clinical vignettes presented in both Portuguese and English. For each vignette, models extracted cardiovascular risk factors, determined SCORE2 applicability, generated 10-year risk estimates where appropriate, and assigned a final three-class ESC risk category. A committee of three cardiologists established the reference standard; eight independent clinicians provided an individual-level human benchmark. Traditional risk-factor extraction was near-perfect across all models (micro-F1 0.97–0.99). Agreement with expert-assigned final risk categories was moderate and variable (best: GPT-4o, quadratic-weighted κw 0.69, 95% CI 0.44–0.84), with 10 of 11 models more often underestimating than overestimating risk. To isolate the source of classification error, post hoc deterministic recalculation of SCORE2 was performed using model-extracted variables in eligible vignettes; this markedly improved agreement across all models (κw 0.85–0.90), demonstrating that extraction was largely intact and computational execution was the primary failure mode. The pooled individual clinician benchmark showed moderate agreement with the reference standard (κw 0.52, 95% CI 0.28–0.67), indicating that the best-performing LLMs matched or exceeded the average individual clinician on this guideline-based task. Performance was broadly consistent across Portuguese and English. Conclusion: Contemporary LLMs reliably extract cardiovascular risk information from clinical text, and the best-performing systems achieved agreement within the range of average individual clinicians on this structured task. Their principal limitation lies in downstream computation and rule application.

  • Open Access

    Inconsistent risk stratification thresholds across European guidelines: A call for harmonization

    by José Ferreira Santos, Ricardo Ladeiras-Lopes, Hélder Dores

  • Open Access

    Large Language Models in Cardiovascular Prevention: A Narrative Review and Governance Framework

    by José Ferreira Santos, Hélder Dores

    Background: Large language models (LLMs) are becoming progressively integrated into clinical practice; however, their role in cardiovascular (CV) prevention remains unclear. This review synthesizes current evidence on LLM applications in preventive cardiology and proposes a governance framework for their safe translation into practice. Methods: We conducted a comprehensive narrative review of literature published between January 2015 and November 2025. Evidence was synthesized across three functional domains: (1) patient applications for health literacy and behavior change; (2) clinician applications for decision support and workflow efficiency; and (3) system applications for automated data extraction, registry construction, and quality surveillance. Results: Evidence suggests that while LLMs generate empathetic, guideline-concordant patient education, they lack the nuance required for unsupervised, personalized advice. For clinicians, LLMs effectively summarize clinical notes and draft documentation but remain unreliable for deterministic risk calculations and autonomous decision-making. System-facing applications demonstrate potential for automated phenotyping and multimodal risk prediction. However, safe deployment is constrained by hallucinations, temporal obsolescence, automation bias, and data privacy concerns. Conclusions: LLMs could help mitigate structural barriers in CV prevention but should presently be deployed only as supervised “reasoning engines” that augment, rather than replace, clinician judgment. To guide the transition from in silico performance to bedside practice, we propose the C.A.R.D.I.O. framework (Clinical validation, Auditability, Risk stratification, Data privacy, Integration, and Ongoing vigilance) as a roadmap for responsible integration.


  • Open Access

    Cardiovascular Prevention: Current Gaps and Future Directions

    by Hélder Dores, José Ferreira Santos, Victor Gil, Pedro de Araújo Gonçalves

    Cardiovascular Disease (CVD) remains the leading cause of morbidity and mortality worldwide. Despite significant advances in diagnosis and treatment, the global burden of CVD remains high, underscoring the crucial need for more effective and comprehensive prevention strategies. This narrative overview aims to critically evaluate the current pillars of cardiovascular prevention, identify the gaps in approaches and outline promising future directions. Challenges and barriers in lifestyle adherence and pharmacological management are addressed, while the increasing role of non-traditional and emerging risk factors is discussed. Future directions include maximizing the value of digital health to improve patient engagement and adherence, adopting precision medicine to refine risk stratification and implementing public health policies for population-level interventions. The optimization of cardiovascular prevention requires a multi-level approach that integrates clinical strategies with personalized solutions and environmental policies to ultimately reduce the global impact of CVD.


  • Open Access

    Large Language Models in Cardiovascular Prevention: From Potential to Practice

    by José Ferreira Santos, Hélder Dores

    Background: Large language models (LLMs) are becoming progressively integrated into clinical practice; however, their role in cardiovascular (CV) prevention remains unclear. This review synthesises current evidence on LLM applications in preventive cardiology and proposes a governance framework for their safe translation into practice. Methods: We conducted a comprehensive narrative review of literature published be-tween January 2015 and November 2025. Evidence was synthesised across three func-tional domains: (1) patient applications for health literacy and behaviour change; (2) clinician applications for decision support and workflow efficiency; and (3) system ap-plications for automated data extraction, registry construction, and quality surveillance. Results: Evidence suggests that while LLMs generate empathetic, guideline-concordant patient education, they lack the nuance required for unsupervised, personalized advice. For clinicians, LLMs effectively summarise clinical notes and draft documentation but remain unreliable for deterministic risk calculations and autonomous decision-making. System-facing applications demonstrate potential for automated phenotyping and mul-timodal risk prediction. However, safe deployment is constrained by hallucinations, temporal obsolescence, automation bias, and data privacy concerns. Conclusions: LLMs could help mitigate structural barriers in CV prevention but should presently be deployed only as supervised “reasoning engines” that augment, rather than replace, clinician judgment. To guide the transition from in silico performance to bedside practice, we propose the C.A.R.D.I.O. framework (Clinical validation, Auditability, Risk stratification, Data privacy, Integration, and Ongoing vigilance) as a roadmap for re-sponsible integration.

  • Top Cited Papers
    Open Access

    Applications of large language models in cardiovascular disease: a systematic review

    by José Ferreira Santos, Ricardo Ladeiras-Lopes, Francisca Leite, Hélder Dores

    Cardiovascular disease (CVD) remains the leading cause of morbidity and mortality worldwide. Large language models (LLMs) offer potential solutions for enhancing patient education and supporting clinical decision-making. This study aimed to evaluate LLMs’ applications in CVD and explore their current implementation, from prevention to treatment. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, this systematic review assessed LLM applications in CVD. A comprehensive PubMed search identified relevant studies. The review prioritized pragmatic and practical applications of LLMs. Key applications, benefits, and limitations of LLMs in CVD prevention were summarized. Thirty-five observational studies met the eligibility criteria. Of these, 54% addressed primary prevention and risk factor management, while 46% focused on established CVD. Commercial LLMs were evaluated in all but one study, with 91% (32 studies) assessing ChatGPT. The LLM applications were categorized as follows: 72% addressed patient education, 17% clinical decision support, and 11% both. In 68% of studies, the primary objective was to evaluate LLMs’ performance in answering frequently asked patient questions, with results indicating accurate, comprehensive, and generally safe responses. However, occasional misinformation and hallucinated references were noted. Additional applications included patient guidance on CVD, first aid, and lifestyle recommendations. Large language models were assessed for medical questions, diagnostic support, and treatment recommendations in clinical decision support. Large language models hold significant potential in CVD prevention and treatment. Evidence supports their potential as an alternative source of information for addressing patients’ questions about common CVD. However, further validation is needed for their application in individualized care, from diagnosis to treatment.


  • Open Access

    A digital tool for self-reporting cardiovascular risk factors: The RADICAL study

    by José Ferreira Santos, Inês Castela, Sara Gamboa Madeira, Sofia Furtado, Hugo Vieira Pereira, Diana Teixeira, Hélder Dores