{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,7]],"date-time":"2026-09-07T23:40:49Z","timestamp":1788824449504,"version":"build-2803163510"},"reference-count":31,"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 The global incidence of blindness has continued to increase, despite the enactment of a Global Eye Health Action Plan by the World Health Assembly. This can be attributed, in part, to an aging population, but also to the limited diagnostic resources within low- and middle-income countries (LMICs). The advent of generative artificial intelligence (AI) within health care could pose a novel solution to combating the prevalence of blindness globally.<\/jats:p>\n <\/jats:sec>\n \n Objective<\/jats:title>\n The objectives of this study are to quantify the effect the addition of a complex prompt has on the diagnostic accuracy of a commercially available LLM, and to assess whether such LLMs are better or worse at diagnosing conditions that are more prevalent in LMICs.<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n Ten clinical vignettes representing globally and LMIC-prevalent ophthalmological conditions were presented to GPT-4\u20100125-preview using simple and complex prompts. Diagnostic performance metrics, including sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), were calculated. Statistical comparison between prompts was conducted using a chi-square test of independence.<\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n \n The complex prompt achieved a higher diagnostic accuracy (90.1%) compared to the simple prompt (60.4%), with a statistically significant difference (\n \u03c7<\/jats:italic>\n 2<\/jats:sup>\n =428.86;\n P<\/jats:italic>\n <.001). Sensitivity, specificity, PPV, and NPV were consistently improved for most conditions with the complex prompt. The simple prompt struggled with LMIC-prevalent conditions, diagnosing only 1 of 5 accurately, while the complex prompt successfully diagnosed 4 of 5.\n <\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n The study established that overall, the inclusion of a complex prompt positively affected the diagnostic accuracy of GPT-4\u20100125-preview, particularly for LMIC-prevalent conditions. This highlights the potential for LLMs, when appropriately tailored, to support clinicians in diverse health care settings. Future research should explore the generalizability of these findings across other models and specialties.<\/jats:p>\n <\/jats:sec>","DOI":"10.2196\/64986","type":"journal-article","created":{"date-parts":[[2025,5,15]],"date-time":"2025-05-15T01:00:08Z","timestamp":1747270808000},"page":"e64986-e64986","source":"Crossref","is-referenced-by-count":5,"title":["Enhancing Diagnostic Accuracy of Ophthalmological Conditions With Complex Prompts in GPT-4: Comparative Analysis of Global and Low- and Middle-Income Country (LMIC)\u2013Specific Pathologies"],"prefix":"10.2196","volume":"9","author":[{"ORCID":"https:\/\/orcid.org\/0009-0001-7804-0788","authenticated-orcid":false,"given":"Shona Alex Tapiwa","family":"M'gadzah","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7725-4082","authenticated-orcid":false,"given":"Andrew","family":"O'Malley","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2025,6,30]]},"reference":[{"issue":"2","key":"R1","doi-asserted-by":"publisher","first-page":"e144","DOI":"10.1016\/S2214-109X(20)30489-7","article-title":"Causes of blindness and VISION impairment in 2020 and trends over 30 years, and prevalence of avoidable blindness in relation to VISION 2020: the Right to Sight: an analysis for the Global Burden of Disease Study","volume":"9","author":"GBD 2019 Blindness and Vision Impairment Collaborators","journal-title":"Lancet Glob Health"},{"key":"R2","doi-asserted-by":"crossref","unstructured":"World Population Prospects 2019: Highlights. United Nations; 2019. [doi: 10.18356\/13bf5476-en]","DOI":"10.18356\/13bf5476-en"},{"key":"R3","unstructured":"Universal eye health: a global action plan 2014\u20132019. World Health Organization; 2013."},{"issue":"62","key":"R4","first-page":"32","volume":"20","author":"Parikshit","journal-title":"Community Eye Health"},{"issue":"12","key":"R5","doi-asserted-by":"publisher","first-page":"1129","DOI":"10.1136\/archdischild-2011-300093","article-title":"Visual impairment in children in middle- and lower-income countries","volume":"96","author":"Courtright","journal-title":"Arch Dis Child"},{"issue":"4","key":"R6","doi-asserted-by":"publisher","first-page":"387","DOI":"10.1136\/bjo.83.4.387","article-title":"Measuring the burden of childhood blindness","volume":"83","author":"Rahi","journal-title":"Br J Ophthalmol"},{"key":"R7","unstructured":"What is a chatbot? IBM. 2024. URL: https:\/\/www.ibm.com\/topics\/chatbots [Accessed 26-06-2025]"},{"key":"R8","unstructured":"Sanjeev S . SightBot: ChatGPT-powered research insights with pubmed citations. Brilliantly. 2023. URL: https:\/\/www.brilliantly.ai\/blog\/sightbot [Accessed 26-06-2025]"},{"key":"R9","unstructured":"Abhinav Venigalla JF Carbin M . Mosaic research. Databricks. 2022. URL: https:\/\/www.mosaicml.com\/blog\/introducing-pubmed-gpt [Accessed 26-06-2025]"},{"issue":"6","key":"R10","doi-asserted-by":"publisher","first-page":"3121","DOI":"10.1007\/s40123-023-00805-x","article-title":"The use of ChatGPT to assist in diagnosing glaucoma based on clinical case reports","volume":"12","author":"Delsoz","journal-title":"Ophthalmol Ther"},{"key":"R11","doi-asserted-by":"publisher","DOI":"10.1016\/j.jfop.2023.100005","article-title":"Conversational AI models for ophthalmic diagnosis: comparison of ChatGPT and the Isabel pro differential diagnosis generator","volume":"1","author":"Balas","journal-title":"JFO Open Ophthalmology"},{"issue":"3","key":"R12","doi-asserted-by":"publisher","first-page":"563","DOI":"10.1177\/02646196211055924","article-title":"Prevalence of blindness and its major causes in sub-Saharan Africa in 2020: a systematic review and meta-analysis","volume":"40","author":"Xulu-Kasaba","journal-title":"Br J Vis Impair"},{"issue":"3","key":"R13","doi-asserted-by":"publisher","first-page":"262","DOI":"10.1136\/bjo.2005.081224","article-title":"The number of people with glaucoma worldwide in 2010 and 2020","volume":"90","author":"Quigley","journal-title":"Br J Ophthalmol"},{"issue":"4","key":"R14","doi-asserted-by":"publisher","first-page":"353","DOI":"10.4314\/ejhs.v24i4.11","article-title":"Prevalence of refractive error and visual impairment among rural school-age children of Goro District, Gurage Zone, Ethiopia","volume":"24","author":"Kedir","journal-title":"Ethiop J Health Sci"},{"key":"R15","unstructured":"Chang RT . Cataracts. BMJ Best Practice. 2023. URL: https:\/\/bestpractice.bmj.com\/topics\/en-gb\/499 [Accessed 26-06-2025]"},{"key":"R16","unstructured":"What are the clinical features of cataracts? National Institute for Health and Care Excellence. 2022. URL: https:\/\/cks.nice.org.uk\/topics\/glaucoma\/diagnosis\/ocular-hypertension-primary-open-angle-glaucoma [Accessed 26-06-2025]"},{"key":"R17","unstructured":"Amerasinghe N Serov-Volach I . Open-angle glaucoma. BMJ Best Practice. 2023. URL: https:\/\/bestpractice.bmj.com\/topics\/en-gb\/373 [Accessed 26-06-2025]"},{"key":"R18","unstructured":"Trachoma. World Health Organization. URL: https:\/\/www.who.int\/news-room\/fact-sheets\/detail\/trachoma [Accessed 26-06-2025]"},{"key":"R19","unstructured":"Lansingh VC Callahan K . Trachoma. BMJ Best Practice. 2024. URL: https:\/\/bestpractice.bmj.com\/topics\/en-gb\/958 [Accessed 26-06-2025]"},{"key":"R20","unstructured":"Diabetic retinopathy. NHS. 2021. URL: https:\/\/www.nhs.uk\/conditions\/diabetic-retinopathy [Accessed 26-06-2025]"},{"key":"R21","unstructured":"Dowler J . Diabetic retinopathy. BMJ Best Practice. 2024. URL: https:\/\/bestpractice.bmj.com\/topics\/en-gb\/532 [Accessed 26-06-2025]"},{"key":"R22","unstructured":"Murray TG Villegas VM . Retinoblastoma. BMJ Best Practice. 2023. URL: https:\/\/bestpractice.bmj.com\/topics\/en-gb\/1055 [Accessed 26-06-2025]"},{"key":"R23","doi-asserted-by":"publisher","DOI":"10.2196\/50638","article-title":"Prompt Engineering as an Important Emerging Skill for Medical Professionals: Tutorial","volume":"25","author":"Mesk\u00f3","journal-title":"J Med Internet Res"},{"key":"R24","unstructured":"Atoms. School of Medicine at the University of St Andrews. 2022. URL: https:\/\/medicine.st-andrews.ac.uk\/arclight\/teaching-materials\/atoms-cards\/ [Accessed 11-06-2026]"},{"key":"R25","unstructured":"Models. OpenAI. 2024. URL: https:\/\/platform.openai.com\/docs\/models [Accessed 11-06-2025]"},{"key":"R26","unstructured":"Openai-python. GitHub. URL: https:\/\/github.com\/openai\/openai-python [Accessed 27-06-2025]"},{"key":"R27","doi-asserted-by":"publisher","DOI":"10.3389\/fpubh.2017.00307","article-title":"Sensitivity, specificity, and predictive values: foundations, pliabilities, and pitfalls in research and practice","volume":"5","author":"Trevethan","journal-title":"Front Public Health"},{"key":"R28","doi-asserted-by":"publisher","DOI":"10.2196\/58275","article-title":"Ensuring appropriate representation in AI-generated medical imagery: a methodological approach to address skin tone bias","volume":"3","author":"O\u2019Malley","journal-title":"JMIR AI"},{"key":"R29","unstructured":"Veenhuizen M O\u2019Malley A . Investigating and combating gender bias in generative large language models. Presented at: The 1st International Conference on Artificial Intelligence in Medical Education with MAEMHS (Malaysia Association of Education in Medical and Health Science); Dec 6-8, 2024; Kuala Lumpur, Malaysia. URL: https:\/\/medicineandhealthukm.com\/sites\/default\/files\/article\/2024\/AI%20in%20Medical%20Education.pdf [Accessed 11-06-2025]"},{"issue":"1","key":"R30","doi-asserted-by":"publisher","first-page":"S61","DOI":"10.12669\/pjms.36.ICON-Suppl.1720","article-title":"Retinoblastoma presentation and survival: a four-year analysis from a tertiary care hospital","volume":"36","author":"Zia","journal-title":"Pak J Med Sci"},{"key":"R31","doi-asserted-by":"crossref","unstructured":"M\u2019Gadzah SAT O\u2019Malley AS . Enhancing diagnostic accuracy of ophthalmological conditions with complex prompts in GPT-4: a comparative analysis of global and LMIC-specific pathologies (dataset). University of St Andrews; 2025.","DOI":"10.2196\/64986"}],"container-title":["JMIR Formative Research"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/formative.jmir.org\/2025\/1\/e64986","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T22:30:00Z","timestamp":1785537000000},"score":1,"resource":{"primary":{"URL":"https:\/\/formative.jmir.org\/2025\/1\/e64986"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,30]]},"references-count":31,"URL":"https:\/\/doi.org\/10.2196\/64986","relation":{},"ISSN":["2561-326X"],"issn-type":[{"value":"2561-326X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,6,30]]},"article-number":"v9i2e64986"}}