IMPORTANCE: ChatGPT is an artificial intelligence (AI) chatbot that has significant societal implications. Training curricula using AI are being developed in medicine, and the performance of chatbots ...in ophthalmology has not been characterized. OBJECTIVE: To assess the performance of ChatGPT in answering practice questions for board certification in ophthalmology. DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used a consecutive sample of text-based multiple-choice questions provided by the OphthoQuestions practice question bank for board certification examination preparation. Of 166 available multiple-choice questions, 125 (75%) were text-based. EXPOSURES: ChatGPT answered questions from January 9 to 16, 2023, and on February 17, 2023. MAIN OUTCOMES AND MEASURES: Our primary outcome was the number of board certification examination practice questions that ChatGPT answered correctly. Our secondary outcomes were the proportion of questions for which ChatGPT provided additional explanations, the mean length of questions and responses provided by ChatGPT, the performance of ChatGPT in answering questions without multiple-choice options, and changes in performance over time. RESULTS: In January 2023, ChatGPT correctly answered 58 of 125 questions (46%). ChatGPT’s performance was the best in the category general medicine (11/14; 79%) and poorest in retina and vitreous (0%). The proportion of questions for which ChatGPT provided additional explanations was similar between questions answered correctly and incorrectly (difference, 5.82%; 95% CI, −11.0% to 22.0%; χ21 = 0.45; P = .51). The mean length of questions was similar between questions answered correctly and incorrectly (difference, 21.4 characters; SE, 36.8; 95% CI, −51.4 to 94.3; t = 0.58; df = 123; P = .22). The mean length of responses was similar between questions answered correctly and incorrectly (difference, −80.0 characters; SE, 65.4; 95% CI, −209.5 to 49.5; t = −1.22; df = 123; P = .22). ChatGPT selected the same multiple-choice response as the most common answer provided by ophthalmology trainees on OphthoQuestions 44% of the time. In February 2023, ChatGPT provided a correct response to 73 of 125 multiple-choice questions (58%) and 42 of 78 stand-alone questions (54%) without multiple-choice options. CONCLUSIONS AND RELEVANCE: ChatGPT answered approximately half of questions correctly in the OphthoQuestions free trial for ophthalmic board certification preparation. Medical professionals and trainees should appreciate the advances of AI in medicine while acknowledging that ChatGPT as used in this investigation did not answer sufficient multiple-choice questions correctly for it to provide substantial assistance in preparing for board certification at this time.
A 68-year-old male presented with a macula-off rhegmatogenous retinal detachment (RRD) in the right eye that was treated with pars plana vitrectomy (PPV). Shiragami hypothesized that RVPs commonly ...occur after PPV and represent retinal displacement.1 RVP's are thought to occur because of differences in the composition and characteristics of the fluorophores in the RPE cells or differences in the metabolic activity of RPE cells that are naïve to light exposure.1,2 Retinal displacement is thought to occur because of a stretch of the retina that occurs because of the movement of residual subretinal fluid at the completion of surgery (often inferiorly), induced by the buoyant force of the tamponade in a direction related to head position and gravity.3,4 This case, for the first time proves that RVPs are in fact aligned with the prior location of retinal vessels before the retinal detachment occurred as hypothesized by Shiragami. There is increasing evidence to support that the integrity of anatomic outcomes following retinal detachment can vary with surgical technique and this may impact functional outcomes.4 This case highlights how extensive retinal displacement can be, even in cases with subtle post-operative RVPs and suggests that patients may benefit from surgical techniques that are refined to minimize the risk of LIRA.
The optimal surgery to repair rhegmatogenous retinal detachment (RRD) is unknown. The purpose of this trial was to compare outcomes of pneumatic retinopexy (PnR) versus pars plana vitrectomy (PPV) ...for the management of primary RRD.
Prospective, randomized controlled trial.
Patients with RRD demonstrating a single retinal break or a group of breaks in detached retina within 1 clock hour above the 8- and 4-o'clock meridians, with any number, location and size of retinal breaks or lattice degeneration in attached retina.
Patients were randomized to undergo either PnR or PPV. Macula-on and macula-off patients were assigned to intervention group by stratified randomization and were treated within 24 and 72 hours, respectively.
The primary outcome was 1-year Early Treatment Diabetic Retinopathy Study (ETDRS) visual acuity (VA). Important secondary outcomes were subjective visual function (25-item National Eye Institute Visual Function Questionnaire), metamorphopsia score (M-CHARTS), and primary anatomic success.
One hundred seventy-six patients were recruited between August 2012 and May 2016. ETDRS VA after PnR exceeded that after PPV by 4.9 letters at 12 months (79.9±10.4 letters vs. 75.0±15.2 letters; P = 0.024). Mean ETDRS VA also was superior for the PnR group compared with the PPV group at 3 months (78.4±12.3 letters vs. 68.5±17.8 letters) and 6 months (79.2±11.1 letters vs. 68.6±17.2 letters). Composite 25-item National Eye Institute Visual Function Questionnaire scores were superior for PnR at 3 and 6 months. Vertical metamorphopsia scores were superior for the PnR group compared with the PPV group at 12 months (0.14±0.29 vs. 0.28±0.42; P = 0.026). Primary anatomic success at 12 months was achieved by 80.8% of patients undergoing PnR versus 93.2% undergoing PPV (P = 0.045), with 98.7% and 98.6%, respectively, achieving secondary anatomic success. Sixty-five percent of phakic patients in the PPV arm underwent cataract surgery in the study eye before 12 months versus 16% in the PnR group (P < 0.001).
Pneumatic retinopexy should be considered the first line treatment for RRD in patients fulfilling Pneumatic Retinopexy versus Vitrectomy for the Management of Primary Rhegmatogenous Retinal Detachment Outcomes Randomized Trial (PIVOT) recruitment criteria. Pneumatic retinopexy offers superior VA, less vertical metamorphopsia, and reduced morbidity when compared with PPV.
This cross-sectional study assesses the accuracy of answers generated by an updated version of a popular chatbot to board certification examination preparation questions.
We reviewed the literature on the efficacy and safety of pars plana vitrectomy (PPV), scleral buckle (SB), and pneumatic retinopexy (PR) for the management of rhegmatogenous retinal detachments ...(RRDs). A systematic search was performed on three databases from inception to September 2020. Randomized controlled trials (RCTs) comparing RRD management options were included. Meta-analysis was performed using a random effects model. Eighteen RCTs and 2,751 eyes were included. For PPV versus SB, early postoperative corrected distance visual acuity (CDVA) favored SB (weighted mean <1 month postoperatively: ~counting fingers for PPV versus ~20/260 for SB, P = 0.02), but differences were nonsignificant at other time points. There was no difference for primary reattachment (P = 0.08). PPV had a lower incidence of choroidal detachment (P = 0.004), hypotony (P = 0.01), and strabismus/diplopia (P = 0.04) but a higher incidence of iatrogenic breaks (P = 0.003) and cataract development/progression (P = 0.05) relative to SB. Combination management was nonsignificantly different relative to PPV alone for CDVA, complications and reattachment rate.
In closing, PPV is associated with a slower visual recovery, but similar final visual acuity and primary reattachment rate relative to SB. Combination procedures did not improve primary reattachment rates or vision relative to standalone PPV. Heterogeneity was seen across the included trials, and further randomized trials are needed to reduce the uncertainty of these estimates.
To describe the characteristics of rhegmatogenous retinal detachments (RD) that developed after endophthalmitis and to provide insight regarding incidence, management, and outcomes.
Retrospective ...chart review of consecutive patients diagnosed with endophthalmitis at two Canadian institutions and from one surgeon at a third center over a 10-year period.
316 eyes presented with endophthalmitis of which 32 (10%) were complicated by subsequent RD. The mean age (SD) of patients was 72.0 (13.77). Intravitreal injections (33%) and cataract surgery (31%) were the most common inciting events for the endophthalmitis. Cases were managed with tap and inject (16%), vitrectomy (30%) or both (47%). There were no differences in the rate of RD when comparing the inciting events, microbial growth, or management approaches. Baseline visual acuity was the only predictor of RD (p=0.009). There was an improvement in VA following management of the RD (logMAR 2.33 vs 1.69 or Snellen HM vs 20/980, p=0.009) but final VA in these patients was worse than in those who did not develop RD (logMAR 1.69 vs 1.27 or Snellen 20/980 vs 20/372, p=0.029).
All patients with RD post endophthalmitis required additional surgery and functional outcomes remained poor. Nevertheless, repair of these RDs can preserve some vision and be globe saving.
ChatGPT-4 is an upgraded version of an artificial intelligence chatbot. The performance of ChatGPT-4 on the United States Medical Licensing Examination (USMLE) has not been independently ...characterized. We aimed to assess the performance of ChatGPT-4 at responding to USMLE Step 1, Step 2CK, and Step 3 practice questions.
Practice multiple-choice questions for the USMLE Step 1, Step 2CK, and Step 3 were compiled. Of 376 available questions, 319 (85%) were analyzed by ChatGPT-4 on March 21
, 2023. Our primary outcome was the performance of ChatGPT-4 for the practice USMLE Step 1, Step 2CK, and Step 3 examinations, measured as the proportion of multiple-choice questions answered correctly. Our secondary outcomes were the mean length of questions and responses provided by ChatGPT-4.
ChatGPT-4 responded to 319 text-based multiple-choice questions from USMLE practice test material. ChatGPT-4 answered 82 of 93 (88%) questions correctly on USMLE Step 1, 91 of 106 (86%) on Step 2CK, and 108 of 120 (90%) on Step 3. ChatGPT-4 provided explanations for all questions. ChatGPT-4 spent 30.8 ± 11.8 s on average responding to practice questions for USMLE Step 1, 23.0 ± 9.4 s per question for Step 2CK, and 23.1 ± 8.3 s per question for Step 3. The mean length of practice USMLE multiple-choice questions that were answered correctly and incorrectly by ChatGPT-4 was similar (difference = 17.48 characters, SE = 59.75, 95%CI = -100.09,135.04,
= 0.29,
= 0.77). The mean length of ChatGPT-4's correct responses to practice questions was significantly shorter than the mean length of incorrect responses (difference = 79.58 characters, SE = 35.42, 95%CI = 9.89,149.28,
= 2.25,
= 0.03).
ChatGPT-4 answered a remarkably high proportion of practice questions correctly for USMLE examinations. ChatGPT-4 performed substantially better at USMLE practice questions than previous models of the same AI chatbot.