AJO International最新文献

{"title":"Evaluating patient-facing eye disease information: ChatGPT-5 vs. Pfizer health answers","authors":"Adriana Kaganovski ,&nbsp;Anisha Kasi ,&nbsp;Aliya Grinberg ,&nbsp;Michael Kozlov ,&nbsp;Riya Patel ,&nbsp;Minwoo Kwon ,&nbsp;Carolyn Lai ,&nbsp;Inci Dersu","doi":"10.1016/j.ajoint.2025.100215","DOIUrl":"10.1016/j.ajoint.2025.100215","url":null,"abstract":"<div><h3>Purpose</h3><div>To evaluate and compare the quality, accuracy, understandability, and actionability of responses generated by a health-specific AI chatbot (Health Answers by Pfizer) and a general-purpose AI chatbot (ChatGPT/GPT-5) to ophthalmology-related patient queries.</div></div><div><h3>Methods</h3><div>We input the top five Google Trends search queries on three leading causes of blindness worldwide — glaucoma, cataract, and age-related macular degeneration — into both chatbots. We evaluated each chatbot response with the Flesch Reading Ease test, Flesch-Kincaid Grade Level, Patient Education Materials Assessment Tool, and DISCERN tool to assess the quality, accuracy, understandability, and actionability of each response.</div></div><div><h3>Results</h3><div>ChatGPT-5 produced responses that were easier to read (Flesch Reading Ease of 48.1 vs 39.0, <em>p</em> = 0.02) and written at a lower grade level (Flesch-Kincaid Grade Level of 8.9 vs 12.2, <em>p</em> = 0.003). ChatGPT-5 also scored higher for understandability (PEMAT-P understandability scores of 83.8% vs 80.5%, <em>p</em> = 0.024) and information quality (DISCERN scores of 41.3 vs 36.4, <em>p</em> = 0.047). In contrast, Health Answers by Pfizer produced content that was significantly more actionable (PEMAT-P actionability scores of 41.3% vs 23.3%, <em>p</em> = 0.004).</div></div><div><h3>Conclusions</h3><div>This study highlights the utility of Health Answers by Pfizer in producing content of higher actionability, compared to ChatGPT-5, which produced content of greater understandability, quality, and readability. It is paramount for effective patient education in ophthalmology to improve AI chatbots to balance clarity with actionability, especially given the critical nature of silently progressing diseases like glaucoma.</div></div>","PeriodicalId":100071,"journal":{"name":"AJO International","volume":"3 1","pages":"Article 100215"},"PeriodicalIF":0.0,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145840816","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Anatomical and visual outcomes of pneumatic retinopexy in primary rhegmatogenous retinal detachment","authors":"Hassan Moutei,&nbsp;Meriem Abdellaoui,&nbsp;Fouad Chraibi,&nbsp;Idriss Benatiya","doi":"10.1016/j.ajoint.2025.100218","DOIUrl":"10.1016/j.ajoint.2025.100218","url":null,"abstract":"<div><h3>Purpose</h3><div>To evaluate pneumatic retinopexy (PnR) outcomes for primary rhegmatogenous retinal detachment (RRD) in a real-world North African cohort and identify factors associated with surgical failure.</div></div><div><h3>Design</h3><div>Retrospective observational case series.</div></div><div><h3>Methods</h3><div>A total of 112 eyes from 112 patients treated for primary RRD with PnR (January 2023–March 2025) were reviewed. Patient selection followed PIVOT trial criteria, including single retinal break or clustered breaks ≤1 clock hour, located within superior 8 clock hours (between 8 and 4 o'clock), with documented postoperative compliance. Preoperative demographics, ocular characteristics, and postoperative complications were systematically recorded. The primary anatomical outcome was defined as complete retinal reattachment without additional surgery at 6 months. Anatomical failure was defined as subretinal fluid persistence exceeding 4 weeks requiring secondary intervention. The primary visual outcome was best-corrected visual acuity (BCVA) at 6 months and its change over time. Logistic regression was used to identify factors associated with anatomical and visual failure.</div></div><div><h3>Results</h3><div>Single-operation anatomical success was 67 % (75/112 eyes); secondary procedures achieved 98 % final reattachment (110/112). Mean final BCVA was 0.12 logMAR, with 80 % of visual improvement occurring within the first postoperative month (0.62 logMAR gain; <em>p</em> &lt; 0.001). Pseudophakia (OR=3.49, <em>p</em> &lt; 0.001), high myopia (OR=1.82, <em>p</em> = 0.02), and prolonged symptom duration (OR=1.12 per day, <em>p</em> = 0.002) were independent factors associated with anatomical failure. New or missed retinal breaks (OR=2.16, <em>p</em> = 0.03) and proliferative vitreoretinopathy (PVR) (OR=1.96, <em>p</em> = 0.02) were independent postoperative factors of failure. For visual outcomes, pseudophakia (OR=3.2, <em>p</em> = 0.001), high myopia (OR=2.49, <em>p</em> = 0.004), worse preoperative BCVA (OR = 2.23, <em>p</em> = 0.002), and prolonged symptom duration (OR=1.35 per day, <em>p</em> = 0.002) were independent factors associated with poor outcomes. Postoperative complications (missed breaks OR=2.8; PVR OR=1.8) also independently contributed to poor visual recovery.</div></div><div><h3>Conclusions</h3><div>PnR achieves competitive anatomical and visual outcomes in real-world practice when rigorous patient selection and meticulous surgical technique are applied. Success depends equally on preoperative patient selection and surgical execution. Optimization through timely intervention, comprehensive preoperative examination, enhanced visualization, and rigorous surveillance may increase real-world success rates toward trial-level performance while maintaining procedural simplicity and cost-effectiveness in resource-limited settings.</div></div>","PeriodicalId":100071,"journal":{"name":"AJO International","volume":"3 1","pages":"Article 100218"},"PeriodicalIF":0.0,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841443","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"In memoriam, Nobuhiko Matsuo, 1931-2025","authors":"Toshihiko Matsuo","doi":"10.1016/j.ajoint.2025.100217","DOIUrl":"10.1016/j.ajoint.2025.100217","url":null,"abstract":"","PeriodicalId":100071,"journal":{"name":"AJO International","volume":"3 1","pages":"Article 100217"},"PeriodicalIF":0.0,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145840817","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Assessing demographic variation in large language model outputs for patient education materials in cataract surgery","authors":"Angel Gao ,&nbsp;Abu Bakar Butt ,&nbsp;Fred Min ,&nbsp;Amin Hatamnejad ,&nbsp;Keean Nanji ,&nbsp;Husayn Gulamhusein","doi":"10.1016/j.ajoint.2025.100216","DOIUrl":"10.1016/j.ajoint.2025.100216","url":null,"abstract":"<div><h3>Purpose</h3><div>To evaluate whether large language model (LLM)-generated patient education materials for cataract surgery vary in readability, length, and accuracy based on demographic modifiers including race, gender, geography, and insurance status.</div></div><div><h3>Design</h3><div>Cross-sectional study</div></div><div><h3>Methods</h3><div>This study analyzed 7,000 responses from five LLMs (ChatGPT, Claude, Copilot, DeepSeek, and Gemini) between March-May 2025 using 280 standardized prompts that varied by race, gender, province/territory, and insurance coverage. Each prompt was submitted five times. Readability was assessed using the Flesch-Kincaid Grade Level (FKGL), Flesch Reading Ease (FRE), and SMOG index. Accuracy was assessed by dual blinded reviewers against AAO clinical guidelines. ANOVA was performed (α = 0.05).</div></div><div><h3>Results</h3><div>LLM outputs differed significantly across all metrics (<em>p</em> &lt; 0.001). Gemini generated the longest (876 ± 143 words) and among the most complex text (FKGL 11.9 ± 1.2). Race, insurance status, and geography significantly impacted readability. Prompts referencing Indigenous patients were the most complex (FKGL 11.1 ± 1.8, FRE 36.5 ± 7.9). Insured prompts were longer and more complex (11.0 ± 1.7 vs. 10.8 ± 1.7; 429 vs. 399 words; <em>p</em> &lt; 0.001). Prompts from Nunavut and Manitoba were the least readable (FKGL ≥ 11.1), while Quebec and PEI were most readable. Gender had minimal impact. No outputs contained clinically unsafe information, but most lacked sufficient depth. None of the responses met the AMA’s sixth-grade readability recommendation.</div></div><div><h3>Conclusion</h3><div>LLM-generated patient education for cataract surgery varies by patient demographics. These disparities may hinder equitable access to health information and highlight the need for bias-aware development of AI tools in healthcare.</div></div>","PeriodicalId":100071,"journal":{"name":"AJO International","volume":"3 1","pages":"Article 100216"},"PeriodicalIF":0.0,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841441","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Intravitreal injection history and cataract surgery complications: A multicenter retrospective cohort","authors":"Justus Zemberi ,&nbsp;Julian Peregoff ,&nbsp;Matthew Santos","doi":"10.1016/j.ajoint.2025.100214","DOIUrl":"10.1016/j.ajoint.2025.100214","url":null,"abstract":"<div><h3>Purpose</h3><div>To evaluate whether a recent history of intravitreal injection is associated with postoperative complications after cataract surgery in a large, real-world cohort.</div></div><div><h3>Design</h3><div>Retrospective, propensity score–matched cohort study.</div></div><div><h3>Methods</h3><div>Adults undergoing cataract surgery with or without at least one intravitreal injection within 90 days before surgery were identified from the U.S. TriNetX research network. Cohorts were matched for demographic characteristics and ocular and systemic comorbidities. Postoperative complications at 14, 30, and 90 days were compared using risk ratios with 95% confidence intervals.</div></div><div><h3>Results</h3><div>After matching, 10,107 patients with recent intravitreal injection and 10,107 controls were included. A recent intravitreal injection history was associated with higher rates of postoperative macular edema at 14, 30, and 90 days (RR range 2.07–2.69) and vitreous hemorrhage at 90 days (RR 1.74). Absolute risk differences were modest. No significant associations were observed for infection, retinal detachment, anterior vitrectomy, lens dislocation, anterior uveitis, or secondary lens procedures.</div></div><div><h3>Conclusion</h3><div>In this large retrospective analysis, patients undergoing cataract surgery with a recent history of intravitreal injection had higher observed rates of postoperative macular edema and vitreous hemorrhage, while other major complications were not increased. These associations likely reflect the presence of underlying retinal disease rather than a direct effect of the injection procedure itself and should be interpreted as non-causal.</div></div>","PeriodicalId":100071,"journal":{"name":"AJO International","volume":"3 1","pages":"Article 100214"},"PeriodicalIF":0.0,"publicationDate":"2025-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841444","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"“Indocyanine green enhanced TTT Vs TTT for treatment of thicker tumors in Retinoblastoma- a randomised control trial”","authors":"Ipsita Barman ,&nbsp;Rachna Meel ,&nbsp;Neelam Pushker ,&nbsp;Vinod Kumar ,&nbsp;Neiwete Lomi ,&nbsp;Anjolie Chhabra ,&nbsp;Mandeep S. Bajaj","doi":"10.1016/j.ajoint.2025.100213","DOIUrl":"10.1016/j.ajoint.2025.100213","url":null,"abstract":"<div><h3>Background</h3><div>Transpupillary thermotherapy (TTT) is used as a focal treatment modality for smaller retinoblastoma tumors, however its efficacy in thicker tumors is limited. This study evaluated whether indocyanine green (ICG) augmentation of TTT (ICGeTTT) improves tumor regression compared to TTT alone for residual chemoreduced retinoblastomas &gt;2 mm in height.</div></div><div><h3>Methods</h3><div>In this randomized controlled trial, simple randomisation table was used to allocate 28 chemoreduced tumors to treatment with either TTT (group 1) or ICGeTTT (group 2). Treatment was administered every 3–4 weeks until tumor height was &lt;2 mm or a maximum of four sessions were completed.</div></div><div><h3>Results</h3><div>The baseline tumor height was higher for the ICGeTTT group. Adjusted analysis for this baseline difference showed a statistically significant greater reduction in tumor height for group 2 (44 %) as compared to group 1 (21 %) (<em>p</em> = 0.018). The percentage of tumors achieving complete regression was higher in group 2 but this difference did not reach statistical significance. The cumulative energy used and side effects from treatment were similar between the groups.</div></div><div><h3>Conclusion</h3><div>Despite the limitations of the study including the small sample size and the baseline difference, the study still demonstrated a greater tumor height reduction with the use of ICGeTTT as compared to TTT in chemoreduced thick residual tumors and hence ICGeTTT may be preferred in such cases.</div></div>","PeriodicalId":100071,"journal":{"name":"AJO International","volume":"3 1","pages":"Article 100213"},"PeriodicalIF":0.0,"publicationDate":"2025-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841388","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multimodal large language models for IOL power calculation in cataract surgery: A feasibility study","authors":"Hye Won Jun ,&nbsp;Sun Young Ryu ,&nbsp;Tae Keun Yoo","doi":"10.1016/j.ajoint.2025.100205","DOIUrl":"10.1016/j.ajoint.2025.100205","url":null,"abstract":"<div><h3>Purpose</h3><div>To assess whether multimodal large language models (MLLMs) can reproduce classical intraocular lens (IOL) power calculations and provide usable backup guidance when standard calculators are unavailable or for educational purposes.</div></div><div><h3>Design</h3><div>Methods-comparison feasibility study.</div></div><div><h3>Methods</h3><div>From a public IOLMaster 700 dataset, 180 eyes were sampled and stratified by axial length and anterior chamber depth. IOL powers were computed by device Sanders–Retzlaff–Kraff/Theoretical (SRK/T) (reference), Barrett Universal II (APACRS), and three MLLMs (ChatGPT-5, Gemini-2.5-Pro, ChatGPT-4o) using identical biometry. Agreement metrics included mean absolute error (MAE) and the proportion within ±0.25, ±0.50, and ±1.00 D versus device SRK/T.</div></div><div><h3>Results</h3><div>ChatGPT-5 and Gemini-2.5-Pro implemented SRK/T using effective lens position estimation and a vergence approach, whereas ChatGPT-4o defaulted to SRK I unless tightly constrained. ChatGPT-5 showed near-reference agreement with SRK/T (MAE 0.30 D; 78.8% within ±0.50 D vs SRK/T), while Gemini-2.5-Pro and ChatGPT-4o had larger errors. Subgroup analyses across axial length–anterior chamber depth strata showed that ChatGPT-5 did not differ significantly from SRK/T after Bonferroni correction, whereas Gemini-2.5-Pro and ChatGPT-4o exhibited significant positive biases in long and short eyes. Agreement patterns were similar when Barrett Universal II was used as the secondary comparator.</div></div><div><h3>Conclusion</h3><div>MLLMs, particularly ChatGPT-5, may provide a portable, stepwise backup for classical IOL power calculation (SRK/T). MLLMs may offer educational value and serve as a secondary check on conventional calculations but should not replace validated biometry platforms, especially in eyes with extreme biometry. Prospective, outcome-based validation and usability testing are warranted before any real-world clinical deployment.</div></div>","PeriodicalId":100071,"journal":{"name":"AJO International","volume":"2 4","pages":"Article 100205"},"PeriodicalIF":0.0,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145736585","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Special issue on evidence-based global health in ophthalmology – AJO international","authors":"Mapa Prabhath Piyasena","doi":"10.1016/j.ajoint.2025.100169","DOIUrl":"10.1016/j.ajoint.2025.100169","url":null,"abstract":"","PeriodicalId":100071,"journal":{"name":"AJO International","volume":"2 4","pages":"Article 100169"},"PeriodicalIF":0.0,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145736584","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Association between sociodemographic factors and visual impairment at initial presentation: A SOURCE data repository analysis","authors":"Xi Dai ,&nbsp;Muhammad Jehanzeb Khan ,&nbsp;Ahmed Sabit ,&nbsp;Mustafa Iftikhar ,&nbsp;Jiangxia Wang ,&nbsp;Jessica Chow ,&nbsp;Joshua D. Stein ,&nbsp;Fasika A. Woreta","doi":"10.1016/j.ajoint.2025.100210","DOIUrl":"10.1016/j.ajoint.2025.100210","url":null,"abstract":"<div><h3>Purpose</h3><div>To assess sociodemographic factors associated with the extent of VI at initial presentation from four of the most common causes of vision loss in the US.</div></div><div><h3>Design</h3><div>Retrospective cohort analysis</div></div><div><h3>Participants</h3><div>Patients ≥ age 40 in the Sight Outcomes Research Collaborative (SOURCE) Ophthalmology Data Repository presenting with cataract, glaucoma, diabetic retinopathy, or age-related macular degeneration ≥ 1 eye between Jan 1, 2012 and Dec 31, 2020.</div></div><div><h3>Methods</h3><div>VI was categorized into none/mild (20/40 or better), moderate (&lt; 20/40 to &gt; 20/200), and severe (20/200 or worse) based on best recorded better acuity in the better seeing eye at initial presentation. A multinomial logistic regression model was used to evaluate risk factors associated with VI at initial presentation.</div></div><div><h3>Main Outcome Measures</h3><div>Odds of moderate and severe VI at initial visit.</div></div><div><h3>Results</h3><div>We included 331,473 patients across seven participating sites. The mean (SD) age was 67.5 (10.5) years and 58.8% were female. Across the four disease categories, compared with White patients, Black, Asian, Hispanic, and Native American patients had higher odds of moderate VI (aOR 1.32, 1.35, 1.62, 1.51) and severe VI (aOR 1.59, 1.07, 1.50, 1.56), respectively. Non English speaking patients had higher odds of presenting with moderate and severe VI (aOR 1.17 and 2.25). Patients living in less affluent communities were also more likely to present with moderate (aOR 1.12, 1.22, 1.29, 1.12) and severe VI (aOR 1.30, 1.37, 1.47, 1.42) across the four disease categories. For Native American patients with severe VI, the adjusted odds ratio was 1.56 (95% CI 1.48 to 1.65).</div></div><div><h3>Conclusions</h3><div>These findings highlight the importance of targeted public health strategies, including language-concordant care to address disparities in vision health and promote earlier diagnosis and treatment for vulnerable populations in the US.</div></div>","PeriodicalId":100071,"journal":{"name":"AJO International","volume":"3 1","pages":"Article 100210"},"PeriodicalIF":0.0,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145798362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Corrigendum to “The Brazilian experience on screening, visual assessment, and prescription of glasses in childhood: A perspective of global interest” [AJO International (2025) Volume 2, Issue 3, 100158]","authors":"ACSB Carneiro ,&nbsp;JM Furtado ,&nbsp;A Leite ,&nbsp;ALF Darcie ,&nbsp;AC Ribeiro ,&nbsp;LM Hopker ,&nbsp;C Nakanami ,&nbsp;SP Donahue ,&nbsp;JD Rossetto","doi":"10.1016/j.ajoint.2025.100164","DOIUrl":"10.1016/j.ajoint.2025.100164","url":null,"abstract":"","PeriodicalId":100071,"journal":{"name":"AJO International","volume":"2 4","pages":"Article 100164"},"PeriodicalIF":0.0,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145736586","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}