Development of a Cloud-Based Clinical Decision Support System for Ophthalmology Triage Using Decision Tree Artificial Intelligence.

Purpose: Clinical decision support systems (CDSS) are an emerging frontier in teleophthalmology, drawing on heuristic decision making to augment processes such as triage and referral. We describe the development and implementation of a novel cloud-based decision tree CDSS for on-call ophthalmology consults. The objective was to standardize the triage and referral process while providing a more accurate provisional diagnosis and urgency. Design: Prospective comparative cohort study. Subjects: On-call referrals to a Canadian community ophthalmology clinic. Methods: A web-based decision tree algorithm was developed using current guidelines and expert opinion. The algorithm collected tailored information on the patient's ophthalmic concern, and outputted a provisional diagnosis and urgency before sending an electronic referral to the on-call ophthalmology clinic. Data were described using descriptive statistics. Spearman-rho correlations and Cohen's kappa coefficient were used to characterize the observed relationships. Post hoc analysis was conducted using analysis of contingency tables and adjusted residuals. Main Outcome Measures: Diagnostic category, provisional diagnosis, and urgency for the referring provider, CDSS, and ophthalmologist. Results: Ninety-six referrals were processed. Referring providers included medical doctors (76.0%, n = 73), optometrists (20.8%, n = 20), and nurse practitioners (3.1%, n = 3). The CDSS (κ = 0.5898; 95% confidence interval [CI], 0.4868-0.6928; P < 0.0001) performed equally well with 66.7% agreement in determining category when compared with referring providers (κ = 0.5880; 95% CI, 0.4798-0.6961; P < 0.0001). The CDSS (agreement = 53.1%; κ = 0.4999; 95% CI, 0.4021-0.5978; P < 0.0001) performed better than referring providers (agreement = 43.8%; κ = 0.4191; 95% CI, 0.3194-0.5188; P < 0.0001) in determining a diagnosis. The CDSS (ρ = 0.5014; 95% CI, 0.3092-0.6935; P < 0.0001) also performed better than referring providers (ρ = 0.4035; 95% CI, 0.2406-0.5665; P < 0.0001) in determining urgency. The CDSS assigned a lower level of urgency in 22 cases (22.9%) compared with referring providers in 6 cases (6.3%). Conclusions: To our knowledge, this is the first cloud-based CDSS in ophthalmology designed to augment the triage and referral process. The CDSS achieves a more accurate diagnosis and urgency, standardizes information collection, and overcomes antiquated paper-based consults. Future directions include developing a random forest model or integrating convolutional neural network-based machine learning to refine the speed and accuracy of triage and referral processes, with emphasis on increasing sensitivity of the CDSS.

Intravitreal injection anesthesia--comparison of different topical agents: a prospective randomized controlled trial.

PURPOSE: To compare the anesthetic effectiveness of 3 topical agents used for intravitreal injections. DESIGN: Randomized, triple-armed, double-blinded, prospective, single-centered trial in patients receiving intravitreal ranibizumab for neovascular age-related macular degeneration. METHODS: Patients were randomized 1:1:1 to receive 0.5% tetracaine hydrochloride drops and a 4% lidocaine pledget (n = 31), 0.5% tetracaine hydrochloride drops alone (n = 31), or 4% cocaine (+ epinephrine 1/100,000) drops alone (n = 31). Patients were asked to score their pain experience using a visual analogue scale (VAS) immediately following and 15 minutes after their injection. The average of these scores was used as the primary outcome. The physician performing the procedure separately scored his perception of the patients' pain using the Wong-Baker FACES scale. RESULTS: Means of the averaged VAS pain score for Groups 1, 2, and 3 were: 19 (95% confidence interval [CI] 12-26), 21 (95% CI 13-29), and 21 (95% CI 16-27) respectively. Mean Wong-Baker pain scores for Groups 1, 2, and 3 were 1.9 (95% CI 1.3-2.6), 2.1 (95% CI 1.4-2.7), and 2.3 (95% CI 1.6-3.1) respectively. There was no significant difference (P = .549) between groups for average VAS pain score. Similarly, there was no significant difference (P = .790) for the physician-perceived pain score between groups. CONCLUSIONS: There was no clinical difference in patient pain experience between the 3 anesthetic options tested. The addition of a 4% lidocaine pledget offered no clinical advantage in pain relief compared to 0.5% tetracaine or 4% cocaine (+ epinephrine 1/100,000) drops alone.

Newfoundland rod-cone dystrophy, an early-onset retinal dystrophy, is caused by splice-junction mutations in RLBP1.

Some isolated populations exhibit an increased prevalence of rare recessive diseases. The island of Newfoundland is a characteristic geographic isolate, settled by a small number of families primarily during the late 1700s and early 1800s. During our studies of this population, we identified a group of families exhibiting a retinal dystrophy reminiscent of retinitis punctata albescens but with a substantially lower age at onset and more-rapid and distinctive progression, a disorder that we termed "Newfoundland rod-cone dystrophy" (NFRCD). The size of one of these families was sufficient to allow us to perform a genomewide screen to map the NFRCD locus. We detected significant linkage to markers on the long arm of chromosome 15, in a region encompassing RLBP1, the gene encoding the cellular retinaldehyde-binding protein. Previously, mutations in RLBP1 have been associated with other retinal dystrophies, leading us to hypothesize that RLBP1 mutations might also cause NFRCD. To test this hypothesis, we sequenced all coding exons and splice junctions of RLBP1. We detected two sequence alterations, each of which is likely to be pathogenic, since each segregates with the disease and is predicted to interfere with mRNA splicing. In contrast to some previously reported RLBP1 mutations, which yield a protein that may retain some residual activity, each NFRCD mutation is likely to give rise to a null allele. This difference may account for the severe phenotype in these families and exemplifies the molecular continuum that underlies clinically distinct but genetically related entities.