Improving the Identification of Diabetic Retinopathy and Related Conditions in the Electronic Health Record Using Natural Language Processing Methods.

Purpose: To compare the performance of 3 phenotyping methods in identifying diabetic retinopathy (DR) and related clinical conditions. Design: Three phenotyping methods were used to identify clinical conditions including unspecified DR, nonproliferative DR (NPDR) (mild, moderate, severe), consolidated NPDR (unspecified DR or any NPDR), proliferative DR, diabetic macular edema (DME), vitreous hemorrhage, retinal detachment (RD) (tractional RD or combined tractional and rhegmatogenous RD), and neovascular glaucoma (NVG). The first method used only International Classification of Diseases, 10th Revision (ICD-10) diagnosis codes (ICD-10 Lookup System). The next 2 methods used a Bidirectional Encoder Representations from Transformers with a dense Multilayer Perceptron output layer natural language processing (NLP) framework. The NLP framework was applied either to free-text of provider notes (Text-Only NLP System) or both free-text and ICD-10 diagnosis codes (Text-and-International Classification of Diseases [ICD] NLP System). Subjects: Adults ≥18 years with diabetes mellitus seen at the Wilmer Eye Institute. Methods: We compared the performance of the 3 phenotyping methods in identifying the DR related conditions with gold standard chart review. We also compared the estimated disease prevalence using each method. Main Outcome Measures: Performance of each method was reported as the macro F1 score. The agreement between the methods was calculated using the kappa statistic. Prevalence estimates were also calculated for each method. Results: A total of 91 097 patients and 692 486 office visits were included in the study. Compared with the gold standard, the Text-and-ICD NLP System had the highest F1 score for most clinical conditions (range 0.39-0.64). The agreement between the ICD-10 Lookup System and Text-Only NLP System varied (kappa of 0.21-0.81). The prevalence of DR and related conditions ranged from 1.1% for NVG to 17.9% for DME (using the Text-and-ICD NLP System). Conclusions: The prevalence of DR and related conditions varied significantly depending on the methodology of identifying cases. The best performing phenotyping method was the Text-and-ICD NLP System that used information in both diagnosis codes as well as free-text notes. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

SURGICAL OUTCOMES OF RETINAL DETACHMENT ASSOCIATED WITH PROLIFERATIVE SICKLE CELL RETINOPATHY.

PURPOSE: To evaluate the long-term anatomic and visual outcomes in eyes with sickle cell retinopathy-related retinal detachments (RDs). METHODS: Patients who underwent surgery for sickle cell retinopathy-related RDs at the Wilmer Eye Institute or Wills Eye Hospital between 2008 and 2020 and followed for at least 6 months postoperatively were retrospectively reviewed. The primary outcome was the rate of single-surgery anatomic success and final reattachment. RESULTS: This study included 30 eyes from 28 patients (16 women and 12 men) with tractional RD (n = 13), rhegmatogenous RD (n = 1), and combined tractional RD/rhegmatogenous RD (n = 16). Mean age was 42.1 ± 15.1 years. The mean follow-up duration was 47.8 ± 34.1 months. Twenty-five (83.3%) eyes underwent pars plana vitrectomy and five (16.7%) eyes underwent pars plana vitrectomy with scleral buckling. Single-surgery anatomic success was achieved in 21 (70.0%) eyes at 6 months. Final reattachment was achieved in 28 (93.3%) eyes (22 eyes [73.3%] without tamponade). Recurrence of RDs was significantly associated with male gender (P = 0.041), absence of previous laser (P = 0.032), iatrogenic breaks (P = 0.035), retinectomy (P = 0.034), and silicone oil tamponade (P = 0.024). Overall, the logarithm of the minimum angle of resolution visual acuity improved from 1.53 ± 0.57 (Snellen equivalent, 20/678) to 1.15 ± 1.01 (20/283) at the final visit (P = 0.03); however, eyes with recurrent RD did not achieve significant visual improvement. CONCLUSION: Pars plana vitrectomy to repair sickle cell retinopathy-related RDs was effective in achieving anatomic success and improving vision in most eyes. Single-surgery anatomic success is critical for optimizing visual outcomes.

Neighborhood-Level Social Determinants of Health and Presenting Characteristics for Rhegmatogenous Retinal Detachments.

Importance: Functional outcomes after repair of rhegmatogenous retinal detachments (RRDs) are highly dependent on baseline visual acuity and foveal status. Adverse social determinants of health (SDOH) can present barriers to timely presentation for repair and limit vision outcomes. Objective: To evaluate the association between neighborhood-level SDOH with baseline severity (visual acuity and fovea status) of RRD. Design, Setting, and Participants: This was a retrospective cohort study that included adult patients 18 years and older who underwent primary repair of uncomplicated RRD at the Wilmer Eye Institute from January 2008 to December 2018. Study data were analyzed from December 2023 to April 2024. Exposures: The census block group of patient home addresses were matched to multiple neighborhood-level SDOH including the Area Deprivation Index (ADI), per capita income, percentage of renters, percentage of rent burden, percentage of people using a food assistance program, percentage of uninsured individuals, mode of transportation to work, distance to the nearest transit stop, total road density, National Walkability Index, Index of Medical Underservice score, and aggregate cost of medical care. Main Outcomes and Measures: Odds of presenting with vision worse than 20/40 or fovea-involving RRD using multivariable logistic regression adjusting for age, sex, race and ethnicity, and insurance. Results: A total of 700 patients (mean [SD] age, 57.9 [12.4] years; 432 male [61.7%]) were included. Every decile increase in ADI, indicating more socioeconomic disadvantage, was associated with an increased odds of presenting with worse visual acuity and fovea-involving RRD (odds ratio [OR], 1.14; 95% CI, 1.04-1.24; P = .004 and OR, 1.13; 95% CI, 1.04-1.22; P = .005, respectively). Each $1000 increase in per capita income was associated with lower odds of presenting with worse vision (OR, 0.99; 95% CI, 0.98-0.99; P = .001). Every 1% increase in percentage of workers who drove to work was associated with an increased odds of presenting with worse vision and fovea-involving RRD (OR, 1.02; 95% CI, 1.01-1.03; P = .005 and OR, 1.01; 95% CI, 1.00-1.03; P = .04, respectively). Conclusions and Relevance: Results of this cohort study suggest that patients with a residence in neighborhoods with more socioeconomic deprivation or a higher percentage of workers who drove to work were more likely to present with more severe RRD even after accounting for multiple individual-level characteristics. These findings support consideration of public policy changes to address the barriers faced by patients residing in certain neighborhoods who seek prompt surgical intervention for RRD to reduce health disparities in RRD outcomes.

A Case of a Bumpy Retina.

A 77-year-old male had 3 months of progressive vision decline in the left eye. Dilated fundus examination revealed diffuse hyperpigmented lesions involving the macula and superior fundus that corresponded to areas of nodular retinal pigment epithelium thickening on optical coherence tomography. What would you do next?

Management of uncomplicated rhegmatogenous retinal detachments: a comparison of practice patterns and clinical outcomes in a real-world setting.

OBJECTIVE: To investigate practice patterns and clinical outcomes in the repair of uncomplicated rhegmatogenous retinal detachments (RRD) in a real-world setting over a 10-year period. METHODS: We compared preferences for scleral buckling (SB), pars plana vitrectomy (PPV), PPV/SB, or pneumatic retinopexy (PR) over time, and examined the 1-year single surgery anatomic success (SSAS) and best-corrected visual acuity (BCVA) at a tertiary academic institution from 2008-2018. RESULTS: Eight hundred eight eyes had RRD repair between 2008-2011 (n = 240), 2012-2014 (n = 271), and 2015-2017 (n = 297). Compared to 2008-2011, PPV was preferred over SB in 2012-2014 (OR: 2.93; 95% CI: 1.86-4.63) and 2015-2017 (OR: 5.94; 95% CI: 3.76-9.38), and over PPV/SB in 2012-2014 (OR: 2.74; 95% CI: 1.65-4.56) and 2015-2017 (OR: 3.16; 95% CI: 31.96-5.12). PR was uncommonly utilized (<10%). Younger surgeons (graduating 2010-2017) favored PPV over SB when compared to older surgeons [graduating 1984-2000 (OR: 1.77; 95% CI: 1.18-2.65) and 2001-2009 (OR 1.73; 95% CI: 1.14-2.65)], but similarly selected PPV vs. PPV/SB as their older counterparts (p > 0.05). Compared to PPV, SSAS was higher with SB (OR: 1.53; 95% CI: 1.03-2.26) and PPV/SB (OR: 2.55; 95% CI: 1.56-4.17). One-year BCVA was markedly improved compared to baseline only for eyes that achieved SSAS (p < 0.001). CONCLUSIONS: Over the past 10 years, PPV has become the favored approach to repair uncomplicated RRD and this appears to be driven by younger surgeons' preferences. Given the superior long-term SSAS in SB and PPV/SB as compared to PPV, SB and PPV/SB should be more frequently considered when determining the appropriate repair strategy for uncomplicated RRD.

Vitrectomy versus Vitrectomy with Scleral Buckling in the Treatment of Giant Retinal Tear Related Retinal Detachments: An International Multicenter Study.

PURPOSE: To determine the practice pattern for treating giant retinal tear (GRT) related detachments, and their anatomic and visual outcomes with pars plana vitrectomy (PPV) with or without scleral buckling (SB). DESIGN: Retrospective cohort study. SUBJECTS: Eyes with GRT detachments repaired from 2008 to 2020 with at least 6 months of follow-up from 7 institutions in North and South America, Europe, and Asia. METHODS: Eyes repaired using PPV versus PPV/SB were compared. MAIN OUTCOME MEASURES: Anatomic and functional outcomes. RESULTS: A comparable number of eyes underwent PPV (n = 101) and PPV/SB (n = 99). Except for history of developmental abnormalities, prior intraocular surgery, and lens status, no differences in baseline demographics, ocular characteristics, or intraoperative surgical adjuncts were observed. The overall single surgery anatomic success (SSAS) at 6 months and 1 year were similar between the groups (82.2% and 77.2% of PPV, and 87.9% and 85.7% of PPV/SB). When stratified by age, the 1-year SSAS rate was higher for PPV/SB (88.5%) than PPV (56.3%) (P = 0.03) for children <18 years. For both children and adults, the mean best-corrected visual acuity (BCVA) at baseline did not differ between the PPV and PPV/SB groups. However, for children, mean BCVA at 1 year was better in the PPV/SB than PPV groups (P = 0.001) while for adults, no difference was found between the 2 groups. The mean time to the first redetachment was 7.9 months in the PPV group and 5.5 months in the PPV/SB group (P = 0.8). Proliferative vitreoretinopathy was the most common cause for redetachment (70.4% of PPV and 93.8% of PPV/SB in redetached eyes; P = 0.1). Postoperative complications were also similar between the 2 groups, including ocular hypertension, epiretinal membrane, and cataract formation. CONCLUSIONS: PPV and PPV/SB are equally popular among surgeons globally for managing GRT detachments and have comparable anatomic and visual outcomes in adults. In children, PPV/SB is superior to PPV for anatomic and functional success at 1 year. In adults, the relief of traction by the GRT may reduce peripheral traction and obviate the need for an SB. However, in children, a supplemental SB can be beneficial as complete vitreous shaving and posterior hyaloid detachment, and postoperative positioning, are difficult in this group.

Prevalence of SARS-CoV-2 in Conjunctival Swab Samples Among Patients Presenting with Conjunctivitis During the COVID-19 Pandemic.

PURPOSE: To investigate the prevalence of SARS-CoV-2-associated conjunctivitis among patients presenting with conjunctivitis to ophthalmology clinics in the absence of respiratory symptoms suggesting COVID-19 during the first year of the pandemic. PATIENTS AND METHODS: This prospective, observational, cross-sectional study enrolled patients aged 18 years and older who presented with acute conjunctivitis between May 2020 and May 2021. After reviewing demographics, ocular and systemic symptoms, a slit lamp examination was performed. Five samples were collected (conjunctival swab from each eye, nasal swab from each nostril, and a nasopharyngeal swab) in separate tubes for a reverse transcription-polymerase chain reaction test for SARS-CoV-2. Patients were called at 1 and 2 weeks to follow up on their ocular and systemic symptoms and collect information about conjunctival adenovirus test results and COVID-19 test results if performed outside of the study. RESULTS: A total of 36 patients were enrolled. The most common ocular symptom was redness (35/36, 97%), and 14 patients (39%) had symptoms in both eyes at the time of the presentation. SARS-CoV-2 RNA was not detected in any samples collected during the study visit (95% confidence interval [CI] from 0 to 0.08). None of the study participants were diagnosed with COVID-19 following conjunctivitis during the 2-week follow-up period. We found that 25 patients had a conjunctival adenovirus test done on the day of their study visit, 9 of which reported positive results. CONCLUSION: Our data suggest that it may not be valuable to obtain routine testing of SARS-CoV-2 in patients presenting with conjunctivitis in the absence of COVID-19.