{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T14:53:13Z","timestamp":1791298393682,"version":"4.1.0"},"reference-count":24,"publisher":"Wiley","issue":"7","license":[{"start":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T00:00:00Z","timestamp":1791244800000},"content-version":"vor","delay-in-days":5,"URL":"http:\/\/onlinelibrary.wiley.com\/termsAndConditions#vor"},{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"tdm","delay-in-days":0,"URL":"http:\/\/doi.wiley.com\/10.1002\/tdm_license_1.1"}],"content-domain":{"domain":["onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Evaluation Clinical Practice"],"published-print":{"date-parts":[[2026,10]]},"abstract":"<jats:title>ABSTRACT<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Surgical decision\u2010making in low back pain requires balancing diagnostic uncertainty, asymmetric risk and guideline adherence, particularly when distinguishing non\u2010specific conditions from pathologies warranting timely surgical referral. As large language models (LLMs) are increasingly considered as decision\u2010support tools, their reliability in guideline\u2010concordant surgical triage requires systematic evaluation.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>To assess and compare the performance of two LLMs (GPT and Gemini) in guideline\u2010adherent surgical decision\u2010making across a range of lumbar spine presentations.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>This comparative cross\u2010sectional study was conducted between January and February 2026. Sixty standardised clinical scenarios representing non\u2010specific low back pain, lumbar disc herniation with radiculopathy, lumbar spinal stenosis, and red\u2010flag emergency conditions were developed. Guideline\u2010based management decisions served as the reference standard. Both models were queried using an identical fixed prompt to determine diagnosis, management strategy, and urgency. Sensitivity, specificity, overall accuracy, and area under the receiver operating characteristic (ROC) curve (AUC) were calculated. Discriminative performance was compared using the DeLong test.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      Both models demonstrated high adherence in clearly defined emergency scenarios. GPT showed higher sensitivity for identifying surgical indications than Gemini (95.8% versus 75%), while specificity was 100% for both. Overall accuracy was 98.3% for GPT and 90% for Gemini. ROC analysis confirmed superior discriminative performance for GPT (AUC 0.99 versus 0.88;\n                      <jats:italic>p<\/jats:italic>\n                      \u2009&lt;\u20090.05). Differences were primarily observed in non\u2010emergent surgical cases.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>LLMs demonstrated strong alignment with guideline\u2010based management in unambiguous scenarios, but variability emerged in sensitivity for non\u2010emergent surgical indications. These findings underscore the importance of structured validation when integrating AI\u2010assisted tools into clinical decision\u2010making pathways.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1111\/jep.70630","type":"journal-article","created":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T14:32:04Z","timestamp":1791297124000},"update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Guideline Adherence and Variability in AI\u2010Assisted Surgical Decision\u2010Making for Low Back Pain"],"prefix":"10.1111","volume":"32","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-3893-797X","authenticated-orcid":false,"given":"Yi\u011fit","family":"\u00d6nalo\u011flu","sequence":"first","affiliation":[{"name":"Department of Orthopaedics and Traumatology Ba\u015fak\u015fehir \u00c7am and Sakura City Hospital Istanbul T\u00fcrkiye"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2026,10,6]]},"reference":[{"issue":"6","key":"e_1_2_10_2_1","doi-asserted-by":"crossref","first-page":"299","DOI":"10.21037\/atm.2020.02.175","article-title":"Global Low Back Pain Prevalence and Years Lived With Disability From 1990 to 2017: Estimates From the Global Burden of Disease Study 2017","volume":"8","author":"Wu A.","year":"2020","journal-title":"Annals of Translational Medicine"},{"key":"e_1_2_10_3_1","doi-asserted-by":"crossref","first-page":"21","DOI":"10.1016\/j.jot.2018.07.009","article-title":"Informed Appropriate Imaging for Low Back Pain Management: A Narrative Review","volume":"15","author":"W\u00e1ng Y. X. J.","year":"2018","journal-title":"Journal of Orthopaedic Translation"},{"key":"e_1_2_10_4_1","doi-asserted-by":"publisher","DOI":"10.29058\/mjwbs.1881313"},{"key":"e_1_2_10_5_1","doi-asserted-by":"publisher","DOI":"10.2196\/72062"},{"key":"e_1_2_10_6_1","doi-asserted-by":"crossref","first-page":"195","DOI":"10.1186\/s12916-019-1426-2","article-title":"Key Challenges for Delivering Clinical Impact With Artificial Intelligence","volume":"17","author":"Kelly C. J.","year":"2019","journal-title":"BMC Medicine"},{"issue":"1","key":"e_1_2_10_7_1","doi-asserted-by":"crossref","first-page":"44","DOI":"10.1038\/s41591-018-0300-7","article-title":"High\u2010Performance Medicine: The Convergence of Human and Artificial Intelligence","volume":"25","author":"Topol E. J.","year":"2019","journal-title":"Nature Medicine"},{"issue":"2","key":"e_1_2_10_8_1","doi-asserted-by":"crossref","first-page":"e188","DOI":"10.7861\/fhj.2021-0095","article-title":"Artificial Intelligence in Healthcare: Transforming the Practice of Medicine","volume":"8","author":"Bajwa J.","year":"2021","journal-title":"Future Healthcare Journal"},{"key":"e_1_2_10_9_1","doi-asserted-by":"crossref","first-page":"344","DOI":"10.1186\/s12891-024-07468-0","article-title":"Recent Clinical Practice Guidelines for the Management of Low Back Pain: A Global Comparison","volume":"25","author":"Zhou T.","year":"2024","journal-title":"BMC Musculoskeletal Disorders"},{"issue":"9898","key":"e_1_2_10_10_1","doi-asserted-by":"crossref","first-page":"1121","DOI":"10.1016\/S0140-6736(13)61215-5","article-title":"Understanding of Regional Variation in the Use of Surgery","volume":"382","author":"Birkmeyer J. D.","year":"2013","journal-title":"Lancet"},{"key":"e_1_2_10_11_1","doi-asserted-by":"publisher","DOI":"10.14245\/ns.2448388.194"},{"key":"e_1_2_10_12_1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0350402"},{"key":"e_1_2_10_13_1","doi-asserted-by":"crossref","first-page":"31","DOI":"10.1038\/s41746-021-00385-9","article-title":"Do as AI Say: Susceptibility in Deployment of Clinical Decision\u2010Aids","volume":"4","author":"Gaube S.","year":"2021","journal-title":"npj Digital Medicine"},{"key":"e_1_2_10_14_1","doi-asserted-by":"crossref","first-page":"108","DOI":"10.1038\/s41398-021-01224-x","article-title":"How Machine\u2010Learning Recommendations Influence Clinician Treatment Selections: The Example of Antidepressant Selection","volume":"11","author":"Jacobs M.","year":"2021","journal-title":"Translational Psychiatry"},{"key":"e_1_2_10_15_1","doi-asserted-by":"crossref","first-page":"794","DOI":"10.1056\/NEJMoa0707136","article-title":"Surgical Versus Nonsurgical Therapy for Lumbar Spinal Stenosis","volume":"358","author":"Weinstein J. N.","year":"2008","journal-title":"New England Journal of Medicine"},{"key":"e_1_2_10_16_1","doi-asserted-by":"crossref","first-page":"1424","DOI":"10.1056\/NEJMoa1508788","article-title":"Laminectomy Plus Fusion Versus Laminectomy Alone for Lumbar Spondylolisthesis","volume":"374","author":"Ghogawala Z.","year":"2016","journal-title":"New England Journal of Medicine"},{"key":"e_1_2_10_17_1","doi-asserted-by":"crossref","first-page":"1515","DOI":"10.1097\/00007632-200006150-00010","article-title":"Cauda Equina Syndrome Secondary to Lumbar Disc Herniation: A Meta\u2010Analysis of Surgical Outcomes","volume":"25","author":"Ahn U. M.","year":"2000","journal-title":"Spine"},{"key":"e_1_2_10_18_1","doi-asserted-by":"crossref","first-page":"11","DOI":"10.1186\/s13063-020-04951-6","article-title":"Reporting Guidelines for Clinical Trials of Artificial Intelligence Interventions: The SPIRIT\u2010AI and Consort\u2010AI Guidelines","volume":"22","author":"Ibrahim H.","year":"2021","journal-title":"Trials"},{"issue":"11","key":"e_1_2_10_19_1","doi-asserted-by":"crossref","first-page":"e745","DOI":"10.1016\/S2589-7500(21)00208-9","article-title":"The False Hope of Current Approaches to Explainable Artificial Intelligence in Health Care","volume":"3","author":"Ghassemi M.","year":"2021","journal-title":"Lancet Digital Health"},{"issue":"6","key":"e_1_2_10_20_1","doi-asserted-by":"crossref","first-page":"517","DOI":"10.1001\/jama.2017.7797","article-title":"Unintended Consequences of Machine Learning in Medicine","volume":"318","author":"Cabitza F.","year":"2017","journal-title":"Journal of the American Medical Association"},{"key":"e_1_2_10_21_1","doi-asserted-by":"crossref","first-page":"25","DOI":"10.1038\/s43856-021-00028-w","article-title":"Mitigating Bias in Machine Learning for Medicine","volume":"1","author":"Vokinger K. N.","year":"2021","journal-title":"Communications Medicine"},{"key":"e_1_2_10_22_1","doi-asserted-by":"publisher","DOI":"10.25259\/NMJI_327_2022"},{"key":"e_1_2_10_23_1","doi-asserted-by":"publisher","DOI":"10.12996\/gmj.2023.77"},{"key":"e_1_2_10_24_1","doi-asserted-by":"publisher","DOI":"10.1177\/02683555261460252"},{"key":"e_1_2_10_25_1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0353355"}],"container-title":["Journal of Evaluation in Clinical Practice"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1111\/jep.70630","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/full-xml\/10.1111\/jep.70630","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1111\/jep.70630","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T14:32:07Z","timestamp":1791297127000},"score":1,"resource":{"primary":{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/10.1111\/jep.70630"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10]]},"references-count":24,"journal-issue":{"issue":"7","published-print":{"date-parts":[[2026,10]]}},"alternative-id":["10.1111\/jep.70630"],"URL":"https:\/\/doi.org\/10.1111\/jep.70630","archive":["Portico"],"relation":{},"ISSN":["1356-1294","1365-2753"],"issn-type":[{"value":"1356-1294","type":"print"},{"value":"1365-2753","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10]]},"assertion":[{"value":"2026-05-08","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-09-18","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-10-06","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"e70630"}}