{
  "abstract": "Background\n\nLarge language models (LLMs) achieve high scores on health-professions examinations. Whether these scores support the same interpretations as human scores remains uncertain. Spain uses parallel multiple-choice examinations for entry to specialty training in medicine (Médico Interno Residente [MIR]), nursing (Enfermero Interno Residente [EIR]), psychology (Psicólogo Interno Residente [PIR]), and pharmacy (Farmacéutico Interno Residente [FIR]).Objective\n\nUsing the January 2026 MIR, EIR, PIR, and FIR examinations, this study aimed to quantify the correspondence between human and LLM item facility and to describe model accuracy, distractor selection, and performance associated with measured surface features.Methods\n\nWe evaluated 839 valid items using 12 configurations: GPT-5.2 at 3 reasoning efforts and Claude Opus, Sonnet, and Haiku 4.5 (Anthropic), each under neutral and expert prompts, with one run per configuration. The primary measure was the Spearman correlation (ρ) between human and machine facility on 828 complete paired items. We examined sensitivity to panel composition, scoring, pooling, and assumed reliability; compared distractor choices with a post hoc plausibility-weighted baseline; and evaluated a surface-feature classifier. Image analyses distinguished supplied from withheld images. An exploratory LLM-judge audit assessed generated explanations. Bootstrap CIs resample items or, for distractor concordance, error events; they exclude run-to-run uncertainty.Results\n\nThe highest pooled accuracy was 94.8% (795/839); all 12 configurations answered 668 of 828 paired items correctly (80.7%). Scores clustered near the ceiling among the strongest configurations. Human and machine facility correlated at 0.40 (95% CI 0.34-0.45), with estimates of 0.31 to 0.40 across panel definitions and 0.39 (95% CI 0.33-0.45) after exam-stratified pooling. The observed correlation was lower than predicted by a specified one-axis simulation with a uniform model advantage (0.63; predictive interval 0.59 to 0.66; P<.001). Among 672 model errors, dominant human-distractor selection differed from the plausibility-weighted baseline by -1.25 percentage points (95% CI -4.53 to 2.19), compatible with no difference. Surface-feature accuracy was 34.4% (289/839), similar to selecting the longest option (34.6%, 290/839). The adjusted association with model correctness was uncertain (odds ratio [OR] 0.87, 95% CI 0.72-1.04). Image-associated accuracy was lower both when images were supplied (OR 0.33, 95% CI 0.21-0.53) and when withheld (OR 0.11, 95% CI 0.08-0.16).Conclusions\n\nIn this single 2026 cohort, high LLM scores coexisted with limited separation among the strongest configurations and moderate correspondence with human item difficulty. These findings caution against interpreting model scores as equivalent evidence of human competence. They do not establish clinical performance or invalidate the examinations for human candidates.",
  "abstract_shown": true,
  "affiliations": [
    "Radiology Department, Hospital General Universitario Morales Meseguer, Av Marqués de los Vélez S/N, Murciaº, Murcia, 30008, Spain, 34 666817007.",
    "Endocrinology Department, Santa Lucía University General Hospital, Cartagena, Murcia, Spain.",
    "Department of Preventive Medicine and Epidemiology, Clinical Institute of Medicine and Dermatology (ICMiD), Hospital Clínic de Barcelona, Barcelona, Catalonia, Spain.",
    "Barcelona Institute for Global Health (ISGlobal), Universitat de Barcelona, Hospital Clínic de Barcelona, Barcelona, Catalonia, Spain.",
    "Department of Otolaryngology, Hospital Universitario Ramón y Cajal, Madrid, Madrid, Spain.",
    "Innovation and Digital Projects Academic Department, Healthcademia, Madrid, Madrid, Spain.",
    "Academia AMIR, Madrid, Madrid, Spain.",
    "Department of Ophthalmology, Hospital Universitario, Madrid, Spain."
  ],
  "article_type": "research article",
  "arxiv": null,
  "author_total": 6,
  "authors": [
    {
      "affs": [
        0
      ],
      "name": "Clemente García-Hidalgo",
      "orcid": "0009-0001-6672-2714"
    },
    {
      "affs": [
        1
      ],
      "name": "Manuel Carpio-Salmerón",
      "orcid": "0009-0009-8772-4565"
    },
    {
      "affs": [
        2,
        3
      ],
      "name": "Pau Benito",
      "orcid": "0000-0002-2480-9133"
    },
    {
      "affs": [
        4
      ],
      "name": "Carlos Carazo-Casas",
      "orcid": "0000-0001-7568-7140"
    },
    {
      "affs": [
        5
      ],
      "name": "Beatriz Carratalá",
      "orcid": "0009-0000-1751-0048"
    },
    {
      "affs": [
        6,
        7
      ],
      "name": "Jaime Campos Pavón",
      "orcid": "0000-0001-7852-2675"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/6954q8nkgrkn3.bib",
  "cited_by": 0,
  "doi": "10.2196/106199",
  "fulltext_url": "https://europepmc.org/articles/PMC13718280",
  "id": "6954q8nkgrkn3",
  "issns": [
    "2369-3762"
  ],
  "landing_url": "https://europepmc.org/article/MED/42954916",
  "language": "eng",
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "heading",
      "Artificial Intelligence"
    ],
    [
      "heading",
      "Artificial Intelligence (AI) in Medical Education"
    ],
    [
      "heading",
      "Generative Language Models Including ChatGPT"
    ],
    [
      "heading",
      "Original Paper"
    ],
    [
      "heading",
      "Testing and Assessment in Medical Education"
    ],
    [
      "heading",
      "Validity, Reliability, and Bias"
    ],
    [
      "openalex",
      "2718|Artificial Intelligence in Healthcare and Education"
    ]
  ],
  "openalex": "W7224467135",
  "pdf_url": "https://europepmc.org/articles/PMC13718280?pdf=render",
  "pmcid": "PMC13718280",
  "pmid": "42954916",
  "published": "2026-10-29",
  "publisher": "JMIR Publications Inc.",
  "ris": "https://research.draconitic.com/paper/6954q8nkgrkn3.ris",
  "sources": [
    {
      "first_seen": "2026-11-01T00:20:36Z",
      "native_id": "MED:42954916",
      "source": "europepmc",
      "url": "https://europepmc.org/article/MED/42954916"
    },
    {
      "first_seen": "2026-11-02T12:22:20Z",
      "native_id": "PMC13718280",
      "source": "pmc",
      "url": "https://doi.org/10.2196/106199"
    }
  ],
  "subjects": [
    {
      "field": "medicine",
      "origin": "openalex:2718|Artificial Intelligence in Healthcare and Education",
      "subfield": "epidemiology-public-health"
    }
  ],
  "title": "Construct Validity of Frontier Large Language Models on 4 Spanish Health Professions Specialty-Access Examinations (MIR, EIR, PIR, and FIR): Cross-Sectional Benchmarking and Psychometric Study.",
  "title_text": "Construct Validity of Frontier Large Language Models on 4 Spanish Health Professions Specialty-Access Examinations (MIR, EIR, PIR, and FIR): Cross-Sectional Benchmarking and Psychometric Study.",
  "type": "peer-reviewed",
  "updated_at": "2026-11-02T12:22:20Z",
  "url": "https://research.draconitic.com/paper/6954q8nkgrkn3",
  "venue": "JMIR medical education",
  "versions": []
}