Large language models approach expert-level clinical knowledge and reasoning in ophthalmology: A head-to-head cross-sectional study.

Large language models (LLMs) underlie remarkable recent advanced in natural language processing, and they are beginning to be applied in clinical contexts. We aimed to evaluate the clinical potential of state-of-the-art LLMs in ophthalmology using a more robust benchmark than raw examination scores. We trialled GPT-3.5 and GPT-4 on 347 ophthalmology questions before GPT-3.5, GPT-4, PaLM 2, LLaMA, expert ophthalmologists, and doctors in training were trialled on a mock examination of 87 questions. Performance was analysed with respect to question subject and type (first order recall and higher order reasoning). Masked ophthalmologists graded the accuracy, relevance, and overall preference of GPT-3.5 and GPT-4 responses to the same questions. The performance of GPT-4 (69%) was superior to GPT-3.5 (48%), LLaMA (32%), and PaLM 2 (56%). GPT-4 compared favourably with expert ophthalmologists (median 76%, range 64-90%), ophthalmology trainees (median 59%, range 57-63%), and unspecialised junior doctors (median 43%, range 41-44%). Low agreement between LLMs and doctors reflected idiosyncratic differences in knowledge and reasoning with overall consistency across subjects and types (p>0.05). All ophthalmologists preferred GPT-4 responses over GPT-3.5 and rated the accuracy and relevance of GPT-4 as higher (p<0.05). LLMs are approaching expert-level knowledge and reasoning skills in ophthalmology. In view of the comparable or superior performance to trainee-grade ophthalmologists and unspecialised junior doctors, state-of-the-art LLMs such as GPT-4 may provide useful medical advice and assistance where access to expert ophthalmologists is limited. Clinical benchmarks provide useful assays of LLM capabilities in healthcare before clinical trials can be designed and conducted.

Factors influencing adherence to non-communicable disease medication in India: secondary analysis of cross-sectional data from WHO - SAGE2.

Background: Non-communicable diseases (NCDs) are a leading cause of death globally and disproportionately affect those in low- and middle-income countries lower-middle-income countries. Poor medication adherence among patients with NCDs is prevalent in India due to lack of initiation, missed dosing or cessation of treatment, and represents a growing healthcare and financial burden. Objective: This study aimed to identify factors influencing medication adherence in adults with NCDs in India. Methods: We performed a cross-sectional study, conducting secondary data analysis on the second wave of the World Health Organisation's 'Study on global AGEing and adult health (SAGE)', a survey that collected data from predominantly older adults across India. Bivariate analysis and multivariate logistic regression modelling were conducted to specifically interrogate the reasons for lack of initiation and cessation of treatment. Reporting of this study was informed by the STROBE guidelines. Results: The average medication adherence rate was 51% across 2,840 patients with one or more NCDs, reflecting non-initiation and lack of persistence of treatment. The strongest factor significantly predicting non-adherence to medication across these components was multimorbidity (odds ratio 0.47, 95% CI 0.40-0.56). Tobacco use (OR = 0.76, CI 0.59-0.98) and never having attended school (OR = 0.75, CI 0.62-0.92) were significantly associated with poor medication adherence (p < 0.05) while rural living (OR = 0.70, CI 0.48-1.02), feelings of anxiety (OR = 0.84, CI 0.66-1.08) and feelings of depression (OR = 0.90, CI 0.70-1.16) were factors lacking statistically significant association with medication adherence on multivariate analysis. Older age (OR = 2.02, CI 1.51-2.71) was significantly associated with improved medication adherence whilst there was a weak association between increased wealth and improved medication use. Limitations: The SAGE2 survey did not capture whether patients were taking their medication doses according to prescribed instructions-as a result our findings may under-estimate the true prevalence of medication non-adherence. Conclusion: Our analysis provides evidence that poor medication adherence in India is multifactorial, with distinct socioeconomic and health-system factors interacting to influence patient decision making. Future large-scale surveys interrogating adherence should assess all components of adherence specifically, whilst public health interventions to improve medication adherence should focus on barriers that may exist due to multimorbidity, comorbid depression and anxiety, and low educational status.

Interventions to promote medication adherence for chronic diseases in India: a systematic review.

Introduction: Cost-effective interventions that improve medication adherence are urgently needed to address the epidemic of non-communicable diseases (NCDs) in India. However, in low- and middle-income countries like India, there is a lack of analysis evaluating the effectiveness of adherence improving strategies. We conducted the first systematic review evaluating interventions aimed at improving medication adherence for chronic diseases in India. Methods: A systematic search on MEDLINE, Web of Science, Scopus, and Google Scholar was conducted. Based on a PRISMA-compliant, pre-defined methodology, randomized control trials were included which: involved subjects with NCDs; were located in India; used any intervention with the aim of improving medication adherence; and measured adherence as a primary or secondary outcome. Results: The search strategy yielded 1,552 unique articles of which 22 met inclusion criteria. Interventions assessed by these studies included education-based interventions (n = 12), combinations of education-based interventions with regular follow up (n = 4), and technology-based interventions (n = 2). Non-communicable diseases evaluated commonly were respiratory disease (n = 3), type 2 diabetes (n = 6), cardiovascular disease (n = 8) and depression (n = 2). Conclusions: Although the vast majority of primary studies supporting the conclusions were of mixed methodological quality, patient education by CHWs and pharmacists represent promising interventions to improve medication adherence, with further benefits from regular follow-up. There is need for systematic evaluation of these interventions with high quality RCTs and their implementation as part of wider health policy. Systematic review registration: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022345636, identifier: CRD42022345636.

Accuracy and reliability of self-administered visual acuity tests: Systematic review of pragmatic trials.

BACKGROUND: Remote self-administered visual acuity (VA) tests have the potential to allow patients and non-specialists to assess vision without eye health professional input. Validation in pragmatic trials is necessary to demonstrate the accuracy and reliability of tests in relevant settings to justify deployment. Here, published pragmatic trials of these tests were synthesised to summarise the effectiveness of available options and appraise the quality of their supporting evidence. METHODS: A systematic review was undertaken in accordance with a preregistered protocol (CRD42022385045). The Cochrane Library, Embase, MEDLINE, and Scopus were searched. Screening was conducted according to the following criteria: (1) English language; (2) primary research article; (3) visual acuity test conducted out of eye clinic; (4) no clinical administration of remote test; (5) accuracy or reliability of remote test analysed. There were no restrictions on trial participants. Quality assessment was conducted with QUADAS-2. RESULTS: Of 1227 identified reports, 10 studies were ultimately included. One study was at high risk of bias and two studies exhibited concerning features of bias; all studies were applicable. Three trials-of DigiVis, iSight Professional, and Peek Acuity-from two studies suggested that accuracy of the remote tests is comparable to clinical assessment. All other trials exhibited inferior accuracy, including conflicting results from a pooled study of iSight Professional and Peek Acuity. Two studies evaluated test-retest agreement-one trial provided evidence that DigiVis is as reliable as clinical assessment. The three most accurate tests required access to digital devices. Reporting was inconsistent and often incomplete, particularly with regards to describing methods and conducting statistical analysis. CONCLUSIONS: Remote self-administered VA tests appear promising, but further pragmatic trials are indicated to justify deployment in carefully defined contexts to facilitate patient or non-specialist led assessment. Deployment could augment teleophthalmology, non-specialist eye assessment, pre-consultation triage, and autonomous long-term monitoring of vision.

Trialling a Large Language Model (ChatGPT) in General Practice With the Applied Knowledge Test: Observational Study Demonstrating Opportunities and Limitations in Primary Care.

BACKGROUND: Large language models exhibiting human-level performance in specialized tasks are emerging; examples include Generative Pretrained Transformer 3.5, which underlies the processing of ChatGPT. Rigorous trials are required to understand the capabilities of emerging technology, so that innovation can be directed to benefit patients and practitioners. OBJECTIVE: Here, we evaluated the strengths and weaknesses of ChatGPT in primary care using the Membership of the Royal College of General Practitioners Applied Knowledge Test (AKT) as a medium. METHODS: AKT questions were sourced from a web-based question bank and 2 AKT practice papers. In total, 674 unique AKT questions were inputted to ChatGPT, with the model's answers recorded and compared to correct answers provided by the Royal College of General Practitioners. Each question was inputted twice in separate ChatGPT sessions, with answers on repeated trials compared to gauge consistency. Subject difficulty was gauged by referring to examiners' reports from 2018 to 2022. Novel explanations from ChatGPT-defined as information provided that was not inputted within the question or multiple answer choices-were recorded. Performance was analyzed with respect to subject, difficulty, question source, and novel model outputs to explore ChatGPT's strengths and weaknesses. RESULTS: Average overall performance of ChatGPT was 60.17%, which is below the mean passing mark in the last 2 years (70.42%). Accuracy differed between sources (P=.04 and .06). ChatGPT's performance varied with subject category (P=.02 and .02), but variation did not correlate with difficulty (Spearman ρ=-0.241 and -0.238; P=.19 and .20). The proclivity of ChatGPT to provide novel explanations did not affect accuracy (P>.99 and .23). CONCLUSIONS: Large language models are approaching human expert-level performance, although further development is required to match the performance of qualified primary care physicians in the AKT. Validated high-performance models may serve as assistants or autonomous clinical tools to ameliorate the general practice workforce crisis.

Greater Occipital Nerve Block as an Effective Intervention for Medically Refractory Pediatric Migraine: A Retrospective Study.

Background Migraine is a common complaint worldwide, spanning all ages, but is not so well investigated in children and adolescents. Pediatric migraines incur significant health and social consequences with the most incisive effects seen in school performance, physical health, and mental well-being, making early effective management of pediatric migraine desirable. However, unlike adult migraine, the treatment for pediatric migraine has not been well established, which is especially true for the treatment of medically refractory migraine in children. Methodology We conducted a retrospective study to assess the feasibility of greater occipital nerve (GON) blocks as a therapeutic option for medically refractory migraine in children. We recruited subjects under 18 years old with a severe medically refractory headache that was affecting day-to-day life and who consented to receive a GON block. GON block effectivity was assessed through follow-up consultations and a post-procedural questionnaire. Results Six subjects received a first-time GON block for medically refractory migraine (mean age = 12 years and three months old, age range = 10 to 15 years old, three out of six subjects were female). We found that GON block was effective in all six patients for treatment of medically refractory migraine as assessed through follow-up consultations and a questionnaire sent out six weeks post-intervention. GON block was generally tolerable with only one patient reporting side effects (migraine flare-up for three days) post-intervention. Repeat injection was performed in four out of six patients, all of whom reported a repeat benefit. Conclusions We conclude that GON block is a feasible therapeutic option for the management of medically refractory migraine in children.

Analgesia for retinopathy of prematurity screening: A systematic review.

BACKGROUND AND AIMS: Premature neonates require regular ophthalmological examination, generally indirect ophthalmoscopy, to screen for retinopathy of prematurity (ROP). Conventional analgesia is provided with topical anesthetic eyedrops and oral sugar solution, but neonates still experience significant pain. Here, the literature base was examined to evaluate the usefulness of other pharmacological analgesics. MATERIALS AND METHODS: A systematic review was undertaken, adhering to a PROSPERO preregistered protocol in accordance with PRISMA guidelines (identifier CRD42022302459). Electronic databases were searched for primary research articles on pharmacological pain interventions used for ROP screening in neonates. The primary outcome measure was pain scores recorded using validated pain scoring tools, with and without pharmacological interventions in neonates during eye examination. For analysis, studies were separated into two categories: topical anesthesia and alternative pharmacological treatments. RESULTS: Eleven studies met the inclusion criteria. Topical analgesia, oral paracetamol, and intranasal fentanyl were found to be effective in reducing the pain of eye examination. Oral morphine and inhaled nitrous oxide had no significant effect on premature infant pain profile (PIPP) scores during indirect ophthalmoscopy. DISCUSSION: In addition to topical anesthesia, premedication with oral paracetamol is recommended during screening examination for ROP. The routine use of fentanyl is not recommended due to the risk of potential side effects. Non-pharmacological measures, such as sweet oral solutions and comfort techniques should also be employed. Further research is required to determine whether the use of nitrous oxide has a role, and to develop a safe and effective analgesic strategy to fully ameliorate the pain of ROP screening.