{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T17:48:28Z","timestamp":1782841708177,"version":"3.54.5"},"reference-count":25,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JMIR Form Res"],"abstract":"Abstract<\/jats:title>\n \n Background<\/jats:title>\n Multimodal large language models (MLLMs) capable of integrating visual and textual information represent a promising advancement for clinical applications requiring image interpretation. Wound care assessment, which demands simultaneous analysis of wound photographs and clinical data, provides an ideal domain to evaluate multimodal vs unimodal artificial intelligence capabilities against human expertise.<\/jats:p>\n <\/jats:sec>\n \n Objective<\/jats:title>\n This study aims to compare the performance of MLLMs, unimodal ChatGPT 5.0, and human clinical experts on a standardized wound care certification examination.<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n This cross-sectional comparative study evaluated 3 participant groups on a 25-question wound care certification examination spanning 4 clinical domains (Diagnosis, Treatment, Complication Management, and Wound Subtype Knowledge). Participants included 3 MLLMs (Med-PaLM 2, LLaVA-Med, and BioGPT), 1 unimodal large language model (ChatGPT 5.0), and 4 human clinical experts (general surgeon, wound care nurse, and 2 internal medicine physicians). Statistical analyses included one-way ANOVA with Tukey post hoc tests and domain-specific Kruskal-Wallis comparisons.<\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n \n Human experts achieved the highest accuracy (mean 86%, SD 9.1%), followed by MLLMs (mean 78.7%, SD 12.2%), while ChatGPT 5.0 achieved 64% accuracy, failing the 70% certification threshold. Significant overall group differences were observed (\n F<\/jats:italic>\n 2,5<\/jats:sub>\n =8.42,\n P<\/jats:italic>\n =.02, \u03b7\u00b2=0.74). MLLMs significantly outperformed ChatGPT 5.0 (difference=14.7 percentage points,\n P<\/jats:italic>\n =.03, Cohen\n d<\/jats:italic>\n =1.38), with the multimodal advantage most pronounced in visually dependent domains: Diagnosis (81% vs 43%,\n P<\/jats:italic>\n =.008) and Complication Management (72% vs 50%,\n P<\/jats:italic>\n =.03). No multimodal advantage was observed for text-based Wound Subtype Knowledge (both 67%). Med-PaLM 2 achieved 92% accuracy, matching that of the wound care nurse, while the general surgeon achieved the highest overall performance (96%).\n <\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n MLLMs demonstrate significant performance advantages over unimodal artificial intelligence in wound care assessment, particularly for visually dependent clinical tasks. While human experts with specialized wound care experience maintain overall superiority, the point estimate of the top-performing MLLM (Med-PaLM 2, 92%) fell within the observed range of human scores; however, the underpowered comparison (power=0.52) and wide CIs preclude definitive conclusions regarding noninferiority or equivalence to human experts. These findings support the potential role of MLLMs as clinical decision-support tools, warranting further adequately powered validation studies.<\/jats:p>\n <\/jats:sec>","DOI":"10.2196\/88618","type":"journal-article","created":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T19:30:07Z","timestamp":1771875007000},"page":"e88618-e88618","source":"Crossref","is-referenced-by-count":0,"title":["The Power of Multimodality in Multimodal Large Language Models, Unimodal ChatGPT 5.0, and Human Clinical Experts on a Wound Care Certification Examination: Cross-Sectional Comparative Study"],"prefix":"10.2196","volume":"10","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-1763-5857","authenticated-orcid":false,"given":"Mete","family":"Ucdal","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5405-0084","authenticated-orcid":false,"given":"Melike Elif","family":"Celik","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2125-3536","authenticated-orcid":false,"given":"Guliz","family":"Evik","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0008-7614-2730","authenticated-orcid":false,"given":"Saniye Beyza","family":"Kuru","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-9284-1023","authenticated-orcid":false,"given":"Saadet","family":"Ozer","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-7360-8533","authenticated-orcid":false,"given":"Sultan","family":"Gungor","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2026,4,27]]},"reference":[{"issue":"3","key":"R1","doi-asserted-by":"publisher","first-page":"57","DOI":"10.3390\/informatics11030057","article-title":"Large language models in healthcare and medical domain: a review","volume":"11","author":"Nazi","journal-title":"Informatics (MDPI)"},{"issue":"3","key":"R2","doi-asserted-by":"publisher","DOI":"10.1016\/j.omtn.2024.102255","article-title":"Large language model to multimodal large language model: a journey to shape the biological macromolecules to biological sciences and medicine","volume":"35","author":"Bhattacharya","journal-title":"Mol Ther Nucleic Acids"},{"issue":"11","key":"R3","doi-asserted-by":"publisher","DOI":"10.1111\/iwj.70109","article-title":"Non-invasive techniques for wound assessment: a comprehensive review","volume":"21","author":"Zhao","journal-title":"Int Wound J"},{"issue":"12","key":"R4","doi-asserted-by":"publisher","DOI":"10.3390\/diagnostics15121529","article-title":"The role of ChatGPT in dermatology diagnostics","volume":"15","author":"Khamaysi","journal-title":"Diagnostics (Basel)"},{"issue":"3","key":"R5","doi-asserted-by":"publisher","first-page":"943","DOI":"10.1038\/s41591-024-03423-7","article-title":"Toward expert-level medical question answering with large language models","volume":"31","author":"Singhal","journal-title":"Nat Med"},{"key":"R6","doi-asserted-by":"crossref","unstructured":"Gao J Li C Liu H et al. LLaVA-med: training a large language-and-vision assistant for biomedicine in one day. Presented at: NIPS \u201923: Proceedings of the 37th International Conference on Neural Information Processing Systems; Dec 10-16, 2023:28541-28564; New Orleans, Louisiana, USA. [doi: 10.52202\/075280-1240]","DOI":"10.52202\/075280-1240"},{"issue":"1","key":"R7","doi-asserted-by":"publisher","DOI":"10.1038\/s43856-025-01116-x","article-title":"Introducing mCODEGPT as a zero-shot information extraction from clinical free text data tool for cancer research","volume":"5","author":"Zhang","journal-title":"Commun Med (Lond)"},{"issue":"11","key":"R8","doi-asserted-by":"publisher","DOI":"10.7759\/cureus.97155","article-title":"Can large language models draft safe, reliable patient leaflets on driving after stroke in comparison to the Stroke Association leaflet?","volume":"17","author":"Suleman","journal-title":"Cureus"},{"issue":"6","key":"R9","doi-asserted-by":"publisher","first-page":"807","DOI":"10.1038\/s41591-020-0941-1","article-title":"Developing specific reporting guidelines for diagnostic accuracy studies assessing AI interventions: the STARD-AI steering group","volume":"26","author":"Sounderajah","journal-title":"Nat Med"},{"issue":"7","key":"R10","doi-asserted-by":"publisher","first-page":"737","DOI":"10.1016\/j.jclinepi.2010.02.006","article-title":"The COSMIN study reached international consensus on taxonomy, terminology, and definitions of measurement properties for health-related patient-reported outcomes","volume":"63","author":"Mokkink","journal-title":"J Clin Epidemiol"},{"key":"R11","unstructured":"Wound care certification practice questions. AppleTree CEU. URL: https:\/\/appletreeceu.com\/wound-care-certification-practice-questions [Accessed 07-04-2026]"},{"key":"R12","unstructured":"ABWM-CWS practice exam. Kill Exams. URL: https:\/\/killexams.com\/demo-download\/ABWM-CWS.pdf#page=2.44 [Accessed 21-04-2026]"},{"key":"R13","unstructured":"WOCNCB\u00ae examination handbook. Wound, Ostomy and Continence Nursing Certification Board (WOCNCB); 2026. URL: https:\/\/www.wocncb.org\/UserFiles\/file\/exam_handbook.pdf#page=43.04 [Accessed 13-04-2026]"},{"issue":"2","key":"R14","doi-asserted-by":"publisher","first-page":"51","DOI":"10.1016\/j.jtv.2019.01.001","article-title":"Prevention and treatment of pressure ulcers\/injuries: the protocol for the second update of the International Clinical Practice Guideline 2019","volume":"28","author":"Kottner","journal-title":"J Tissue Viability"},{"issue":"1","key":"R15","doi-asserted-by":"publisher","first-page":"34","DOI":"10.1111\/wrr.13133","article-title":"WHS (Wound Healing Society) guidelines update: diabetic foot ulcer treatment guidelines","volume":"32","author":"Lavery","journal-title":"Wound Repair Regen"},{"issue":"Sup12","key":"R16","doi-asserted-by":"publisher","first-page":"S10","DOI":"10.12968\/jowc.2022.31.Sup12.S10","article-title":"IWII Wound Infection in Clinical Practice consensus document: 2022 update","volume":"31","author":"Swanson","journal-title":"J Wound Care"},{"issue":"17","key":"R17","doi-asserted-by":"publisher","first-page":"9398","DOI":"10.3390\/app15179398","article-title":"An intelligent docent system with a small large language model (sLLM) based on retrieval-augmented generation (RAG)","volume":"15","author":"Jung","journal-title":"Appl Sci (Basel)"},{"issue":"1","key":"R18","doi-asserted-by":"publisher","DOI":"10.1038\/s41746-024-01185-7","article-title":"Hidden flaws behind expert-level accuracy of multimodal GPT-4 vision in medicine","volume":"7","author":"Jin","journal-title":"NPJ Digit Med"},{"issue":"9","key":"R19","doi-asserted-by":"publisher","first-page":"1425","DOI":"10.1093\/jamia\/ocaf116","article-title":"Diagnostic accuracy differences in detecting wound maceration between humans and artificial intelligence: the role of human expertise revisited","volume":"32","author":"K\u00fccking","journal-title":"J Am Med Inform Assoc"},{"issue":"1","key":"R20","doi-asserted-by":"publisher","DOI":"10.1007\/s10916-025-02153-8","article-title":"Artificial intelligence methods for diagnostic and decision-making assistance in chronic wounds: a systematic review","volume":"49","author":"Reifs Jim\u00e9nez","journal-title":"J Med Syst"},{"issue":"3","key":"R21","doi-asserted-by":"publisher","first-page":"929","DOI":"10.1177\/00031348251385104","article-title":"Artificial Intelligence for Postoperative Wound Monitoring: An Integrative Review of Digital Innovation and Clinical Feasibility","volume":"92","author":"Grunhut","journal-title":"Am Surg"},{"issue":"6","key":"R22","doi-asserted-by":"publisher","first-page":"1561","DOI":"10.1111\/iwj.13755","article-title":"Reshaping wound care: evaluation of an artificial intelligence app to improve wound assessment and management amid the COVID-19 pandemic","volume":"19","author":"Barakat-Johnson","journal-title":"Int Wound J"},{"issue":"1","key":"R23","doi-asserted-by":"publisher","DOI":"10.1186\/s12911-025-03144-y","article-title":"Developing an AI-powered wound assessment tool: a methodological approach to data collection and model optimization","volume":"25","author":"Stefanelli","journal-title":"BMC Med Inform Decis Mak"},{"key":"R24","doi-asserted-by":"publisher","DOI":"10.3389\/fdgth.2025.1657583","article-title":"Feasibility of a multimodal AI-based clinical assessment platform in emergency care: an exploratory pilot study","volume":"7","author":"Russ","journal-title":"Front Digit Health"},{"issue":"7","key":"R25","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0271742","article-title":"A time motion study of manual versus artificial intelligence methods for wound assessment","volume":"17","author":"Mohammed","journal-title":"PLoS ONE"}],"container-title":["JMIR Formative Research"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/formative.jmir.org\/2026\/1\/e88618","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T17:15:07Z","timestamp":1782839707000},"score":1,"resource":{"primary":{"URL":"https:\/\/formative.jmir.org\/2026\/1\/e88618"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,4,27]]},"references-count":25,"URL":"https:\/\/doi.org\/10.2196\/88618","relation":{},"ISSN":["2561-326X"],"issn-type":[{"value":"2561-326X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4,27]]},"article-number":"v10i3e88618"}}