RESUMEN
Objective: The aim of this study was to evaluate the accuracy, comprehensiveness, and safety of a publicly available large language model (LLM)-ChatGPT in the sub-domain of glaucoma. Design: Evaluation of diagnostic test or technology. Subjects participants and/or controls: We seek to evaluate the responses of an artificial intelligence chatbot ChatGPT (version GPT-3.5, OpenAI). Methods intervention or testing: We curated 24 clinically relevant questions in the domain of glaucoma. The questions spanned four categories: pertaining to diagnosis, treatment, surgeries, and ocular emergencies. Each question was posed to the LLM and the responses obtained were graded by an expert grader panel of three glaucoma specialists with combined experience of more than 30 years in the field. For responses which performed poorly, the LLM was further prompted to self-correct. The subsequent responses were then re-evaluated by the expert panel. Main outcome measures: Accuracy, comprehensiveness, and safety of the responses of a public domain LLM. Results: There were a total of 24 questions and three expert graders with a total number of responses of n = 72. The scores were ranked from 1 to 4, where 4 represents the best score with a complete and accurate response. The mean score of the expert panel was 3.29 with a standard deviation of 0.484. Out of the 24 question-response pairs, seven (29.2%) of them had a mean inter-grader score of 3 or less. The mean score of the original seven question-response pairs was 2.96 which rose to 3.58 after an opportunity to self-correct (z-score - 3.27, p = 0.001, Mann-Whitney U). The seven out of 24 question-response pairs which performed poorly were given a chance to self-correct. After self-correction, the proportion of responses obtaining a full score increased from 22/72 (30.6%) to 12/21 (57.1%), (p = 0.026, χ2 test). Conclusion: LLMs show great promise in the realm of glaucoma with additional capabilities of self-correction. The application of LLMs in glaucoma is still in its infancy, and still requires further research and validation.
RESUMEN
PRCIS: In our case series, the 3-year failure for Paul Glaucoma Implant (PGI) implantation was 14.6%. At 3 years postoperatively, there was a significant reduction in mean intraocular pressure (IOP) and the number of glaucoma medications used. OBJECTIVE: To determine the 3-year efficacy and safety of the PGI, a novel glaucoma tube shunt in patients with glaucoma. METHODS: Retrospective review of all patients who had undergone PGI implantation in a single tertiary institution in Singapore between May 1, 2017 and January 1, 2022. Data were extracted from electronic health records (Computerized Patient Support System 2 and Epic). The primary outcome measure was failure, defined as IOP >18 mm Hg or <6 mm Hg on 2 consecutive visits after 3 months, reoperation for IOP-related indication, explantation of implant, or loss of light perception vision. Complete success was defined as the absence of failure without medications at 36 months, and qualified success similarly, but with medications. Postoperative mean IOP, mean number of IOP-lowering medications used, and visual acuity were also assessed. RESULTS: Forty-eight eyes in 48 patients were identified. Thirty-one patients (64.6%) had primary open angle and angle closure glaucoma, and 18 (37.5%) had previous existing tube implants or trabeculectomy. At 3 years postoperatively, 7 cases (14.6%) fulfilled the criteria for failure and 36 (75%) met the criteria for complete success. The mean IOP at 36 months was 14.9 ± 4.11 mm Hg, from the mean preoperative IOP of 20.6 ± 6.13 mm Hg ( P < 0.001). The mean number of IOP-lowering medications used was reduced from 3.13 ± 0.959 preoperatively to 0.167 ± 0.476 at 36 months ( P < 0.001). The most common postoperative complication was hypotony (n = 17, 35.4%), of which the majority were self-limiting, followed by hyphema (n = 5, 10.4%) and tube exposure (n = 4, 8.3%). CONCLUSION: The PGI demonstrated sustained IOP reduction and a reduction of medication burden at 3 years postoperatively.