Early Endophthalmitis Incidence and Risk Factors after Glaucoma Surgery in the Medicare Population from 2016 to 2019.

PURPOSE: To determine early endophthalmitis incidence and risk factors after glaucoma surgeries in the Medicare population. DESIGN: Retrospective, longitudinal study. PARTICIPANTS: Medicare Fee-for-Service (FFS) and Medicare Advantage beneficiaries in the United States aged 65 years or older undergoing glaucoma surgery. METHODS: Medicare claims were used to identify all patients who underwent glaucoma, cataract, or combined cataract/glaucoma surgery from 2016 to 2019. Endophthalmitis cases within 42 days of the index surgery were identified using the diagnostic codes. Multivariable logistic regression models were used to evaluate factors associated with postoperative endophthalmitis. MAIN OUTCOME MEASURES: The 42-day postoperative endophthalmitis incidence and risk factors associated with endophthalmitis after glaucoma surgery. RESULTS: There were 466 928 glaucoma surgeries, of which 310 823 (66.6%) were combined with cataract surgery. Cataract surgeries alone (n = 8 460 360) served as a reference group. Microinvasive glaucoma surgeries constituted most glaucoma procedures performed (67.8%), followed by trabeculectomy (14.0%), tube shunt (10.9%), and other procedures (7.3%). There were 572 cases of endophthalmitis identified after all glaucoma surgeries. Endophthalmitis incidence after glaucoma, combined cataract/glaucoma, and cataract surgeries alone was 1.5 (95% confidence interval [CI], 1.3-1.7), 1.1 (95% CI, 1.0-1.2), and 0.8 (95% CI, 0.8-0.8) per 1000 procedures, respectively. The median day of diagnosis of endophthalmitis was later for glaucoma surgeries (16.5 days) compared with combined cataract/glaucoma or cataract surgeries alone (8 and 6 days, respectively). Compared with microinvasive glaucoma surgery (MIGS), tube shunts were the only surgery type to be a significant risk factor for endophthalmitis for both stand-alone (adjusted odds ratio [aOR], 1.8, P = 0.002) and combined surgery (aOR 1.8, P = 0.047). The other risk factor for both stand-alone (aOR 1.1, P = 0.001) and combined (aOR 1.06, P = 0.049) surgeries was the Charlson Comorbidity Index (CCI). Age (aOR 1.03, P = 0.004) and male gender (1.46, P = 0.001) were significant risk factors for combined cataract and glaucoma surgeries. CONCLUSIONS: Compared with cataract surgery, early endophthalmitis incidence was higher for both glaucoma and combined cataract/glaucoma surgeries, with the highest incidence among tube shunts. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Association Between Open Payments-Reported Industry Transfers of Value and Prostaglandin Analog Prescribing in the US.

Importance: Reported transfers of value (TOV) from pharmaceutical companies have been associated with greater use of branded anti-vascular endothelial growth factor agents by ophthalmologists, but payment under the Medicare Part B buy-and-bill model includes a financial incentive to choose costlier agents, potentially confounding analyses of pharmaceutical TOV and prescribing patterns. How these reported TOV are associated with prescribing patterns for prescription eye drops, not subject to the incentives created by Part B payments, should be considered. Objective: To assess the association between prostaglandin analog (PGA) eye drop prescribing and reported nonresearch TOV by makers of branded PGAs to US vision care professionals. Design, Setting, and Participants: This retrospective cohort analysis used a 20% nationally representative sample of 2018 Medicare Part D claims and industry TOV reported to the Open Payments program. Optometrists and ophthalmologists who had more than 10 claims for PGA drops in the 20% sample were analyzed. Analysis took place from June 2021 to February 2022. Main Outcomes and Measures: Multivariable logistic regression assessing the association between membership in strata of reported TOV and branded PGA prescribing rate, controlling for prescriber demographic factors, local area practices, total PGA prescribing volume, and plan formularies involved. Results: A total of 20 612 ophthalmologists and 5426 optometrists (7449 [29%] female and 18 589 [71%] male) prescribed PGA eye drops. Of these, 9685 (37%) were reported to have received TOV from manufacturers of branded PGAs in 2018, totaling $5 060 346. The median (IQR) reported TOV was $65 ($24-$147). Multivariable logistic regression showed that the predicted probability of primarily prescribing branded PGAs among prescribers who reported receiving no TOV was 12.9% (95% CI, 12.4%-13.4%). This figure increased to 19.6% (95% CI, 18.8%-20.4%) among prescribers receiving TOV, a 50% increase. There was a dose-response association, such that the top 10% of TOV recipients had a 29.2% probability (95% CI, 26.4%-31.9%) of preferential branded use. Conclusions and Relevance: While the median reported TOV to a PGA prescriber was relatively low in this study, there was a positive association between amount of reported nonresearch TOV received from PGA makers and the frequency of branded PGA use. This shows that small reported TOV were associated with differences in prescribing. High rates of branded PGA prescribing may pose a cost burden to patients that affects adherence. Clinicians and policy makers should be aware of these associations.

Outcomes and Revenue Generation of a Community-based Screening at a Center in the United States: The SToP Glaucoma Program.

PRCIS: Of 611 individuals seen at referral clinic visits following community screenings, 76% were diagnosed with ≤1 eye condition needing treatment, generating a total of $213,110 in collections for the institution over 2.5 years. PURPOSE: The purpose of this study was to examine the outcomes and revenue generation of community-based eye screenings. MATERIALS AND METHODS: Individuals aged 50 years and above screened at community sites in Baltimore, MD, with abnormal ophthalmic findings were referred for one free-of-charge definitive eye examination at the Wilmer Eye Institute. Diagnoses, treatment, and billing information were abstracted from electronic medical records of patients subsequently seen at Wilmer from January 1, 2016, to July 31, 2018. RESULTS: A total of 611 individuals attended 3696 encounters at Wilmer during this time period. Most patients were female (60.3%) and African American (83.7%). At the screening event, 82.9% reported difficulty seeing when not wearing corrective eyewear, although only 49.8% reported having visited an eye doctor within the last 2 years. The majority (60.2%) reported having Medicare/Medicaid coverage, and 8.1% reported being uninsured. At the definitive eye examination after the screening, 75.5% of patients were diagnosed with ≥1 eye condition, most commonly cataract (30.3%), suspicion of glaucoma (24.9%), manifest glaucoma (11.9%), diabetic retinopathy (5.4%), and ocular hypertension (2.6%). Overall, 430 (70.4%) individuals required treatment including surgery (n=106), intravitreal injections (n=14), laser procedures (n=9), and medications (n=48). A total of $213,110 was collected for visits and procedures after the initial referral visit during the study period. CONCLUSIONS: A large community-based vision screening program in Baltimore was able to identify ocular conditions requiring treatment in underserved older adults and connect them to eyecare. Our findings also highlight that this model simultaneously generates new revenue streams for the institution organizing the community screenings.

Demographics, clinical interests, and ophthalmology skills confidence of medical student volunteers and non-volunteers in an extracurricular community vision screening service-learning program.

BACKGROUND: Medical school curricular hours dedicated to ophthalmology are low and declining. Extracurricular ophthalmology activities, such as participation in community vision screenings, may serve an important adjunctive role in medical school curricula. The Johns Hopkins University (JHU) Vision Screening In Our Neighborhoods (ViSION) Program is an example of a voluntary medical student-directed community service-learning program. METHODS: We used a mixed-methods cross-sectional approach, including an online survey and semi-structured interviews. JHU School of Medicine students enrolled in MD or MD/PhD programs during the 2019-2020 academic year were surveyed regarding demographics, career and service interests, involvement in ophthalmology-related activities, and confidence in their ophthalmology-related skills. Survey responses were compared between ViSION volunteers and non-volunteers using Fisher's exact chi-square tests. Semi-structured interviews were conducted via webconference with 8 prior or current ViSION volunteers and responses analyzed using inductive thematic analysis. Data were collected when ViSION volunteers were in variable stages of their medical education and involvement with the ViSION program. RESULTS: A total of 118 medical students were included, representing an overall response rate of 24.6% of JHU medical students. ViSION volunteers reported greater involvement in ophthalmology-related research (42% vs. 4%, p < 0.001), intent to apply to ophthalmology residency programs (35% vs. 1%, p = 0.001), and confidence with multiple ophthalmology knowledge and clinical skill domains. In particular, ViSION volunteers were more likely to feel confident estimating cup-to-disc ratio using direct ophthalmoscopy (20% vs. 0%, p < 0.001). In open-ended survey and interview questions, most volunteers attributed at least some degree of their ophthalmology skill development and desire to pursue ophthalmology and public health careers to their ViSION experience. CONCLUSIONS: Medical students who volunteered with a student-led community vision screening program were more likely to have a prior interest in ophthalmology than those who did not volunteer, but only 1/3 of volunteers planned to pursue a career in ophthalmology. Overall, volunteers reported higher confidence performing ophthalmology-related clinical skills, suggesting that student-led community vision screening programs may provide an important avenue for medical students to explore public health aspects of ophthalmology, while practicing ophthalmology exam skills and learning about common ophthalmic pathologies, regardless of their career intentions.

Applicant Characteristics Associated with Glaucoma Fellowship Match from 2010 to 2017.

PURPOSE: To describe trends in the glaucoma fellowship match results, glaucoma applicant characteristics, and applicant characteristics associated with matching. DESIGN: Retrospective study. PARTICIPANTS: Ophthalmology fellowship applicants participating in the San Francisco (SF) Match between 2010 and 2017. METHODS: De-identified, individual applicant data provided by the San Francisco (SF) Match were used to describe glaucoma fellowship match trends and applicant characteristics between 2010 and 2017. Publicly available, conglomerate SF Match data were used to determine trends in the number of programs participating in the glaucoma fellowship match, as well as the number of positions offered and filled from 2014 to 2019. All trends analyses were performed using linear regression models on log-transformed response variables. Summary statistics for applicants who matched in glaucoma were compared with those who did not match. A multivariable logistic regression model was used to evaluate factors associated with matching in glaucoma fellowship. MAIN OUTCOME MEASURES: Fellowship match status (matched in glaucoma vs. did not match in any ophthalmology subspecialty). RESULTS: From 2010 to 2019, the number of matched glaucoma fellowship applicants increased from 54 to 77 (mean 3.3% per year, P = 0.001), and the proportion of applicants to any ophthalmology fellowship who matched in glaucoma increased from 13% to 15.5% (mean, 3% per year; P = 0.041). Compared with applicants who did not match in any specialty between 2010 and 2017, matched glaucoma applicants were more likely to have graduated from a top 10 residency program (10.7% vs. 4.5%, P < 0.001), U.S. medical school (93.1% vs. 39.1%, P < 0.001), or U.S. residency program (95.0% vs. 42.2%, P < 0.001), and complete more interviews (9 vs. 1, P < 0.001). After controlling for potential covariates, factors associated with increased odds of matching in glaucoma included graduating from a U.S. residency program (odds ratio [OR], 9.91; 95% confidence interval [CI], 5.45-18.03), applying to fewer programs (OR, 0.91; 95% CI, 0.88-0.93), and completing a greater number of interviews (OR, 1.49; CI, 1.39-1.60). United States Medical Licensing Examination (USMLE) Step scores were not associated with matching in glaucoma. CONCLUSIONS: The number of applicants who matched in glaucoma fellowship increased from 2010 to 2019. Factors associated with matching in glaucoma were graduating from a U.S. residency program and completing more interviews.

Improvement in the assessment of diastolic function in a clinical echocardiography laboratory following implementation of a quality improvement initiative.

BACKGROUND: The aim of this study was to demonstrate improvement in the characterization of diastolic function in the routine practice of a clinical echocardiography laboratory after the implementation of a quality improvement initiative. The echocardiographic analysis of left ventricular (LV) diastolic dysfunction is an inherently complex process involving the integration of multiple indices for accurate assessment. METHODS: A baseline survey of 50 randomly chosen echocardiographic studies was reviewed for the accuracy of diastolic function assessment. A four-step quality improvement protocol was then initiated: (1) sonographer and physician education; (2) the implementation of data acquisition protocol changes using LV inflow, tissue Doppler velocity of the mitral annulus in early diastole (e'), flow propagation velocity of LV inflow (Vp), and left atrial volume index (LAVI), along with the establishment of uniform criteria for diagnostic interpretation; (3) peer review of performance; and (4) focused interactive case review sessions. RESULTS: At baseline, measurements of LV inflow were most often correct (100% accurate), while measurements of e' (82% accurate), Vp (12% accurate), and LAVI (12% accurate) and the proper classification of diastolic function (44% accurate) were significantly limited. After the quality improvement initiative, there were significant increases in the accuracy of all recorded measurements, with e' 92% accurate (a 10% improvement; P < .10), Vp 67% accurate (a 55% improvement; P < .001), LAVI 80% accurate (a 68% improvement, P < .001), and proper characterization of diastolic function 76% accurate (a 32% improvement, P < .001). CONCLUSIONS: A multifaceted quality improvement protocol including staff education, systematic support with enhanced infrastructure, and peer review with feedback can be effective for improving the clinical performance of a nonacademic echocardiography laboratory in the characterization of diastolic function.