About

Brian VanderBeek, MD, MPH, MSCE, is an Associate Professor of Ophthalmology at the Hospital of the University of Pennsylvania within the Department of Ophthalmology. He is based at the Scheie Eye Institute in Philadelphia, PA. Dr. VanderBeek completed his undergraduate degree in Biology at the University of Michigan in 1999, followed by a Master of Public Health in Toxicology from the same institution in 2001. He earned his MD from the University of Cincinnati College of Medicine in 2005 and subsequently obtained a Master of Science in Clinical Epidemiology (MSCE) from the University of Pennsylvania School of Medicine in 2015. His clinical expertise includes diabetic retinopathy, age-related macular degeneration, and retinal vein occlusion. His research focuses on epidemiology, observational study design, and big data analysis related to ophthalmology. Dr. VanderBeek has contributed to numerous studies and presentations on diabetic retinal disease, racial and ethnic disparities in ophthalmic conditions, and trends in prevalence and incidence of retinal diseases. His work emphasizes understanding the epidemiology and disparities in eye health, with a particular focus on diabetic retinal conditions.

Research topics

• Ophthalmology
• Surgery
• Medicine

Selected publications

• Racial and Ethnic Differences in Prevalence and Incidence of Diabetic Retinal Disease

  Retina · 2026-03-31

  articleSenior author

  PURPOSE: To determine prevalence and incidence trends of diabetic retinal disease (DRD) and its vision-threatening forms over the last 20 years among patients with diabetes mellitus (DM) among US racial and ethnic groups. METHODS: A retrospective cohort study of members of commercial and Medicare Advantage health plans between 2000 and 2022 was conducted, with cohorts of White(W), Black/African American(B/AA), Hispanic(H), and Asian(A) DM patients identified using ICD-9/10 codes. Outcomes included annual prevalence and incidence of DRD, diabetic macular edema (DME), and proliferative diabetic retinopathy (PDR). Multivariable logistic and Poisson regression models analyzed trends in prevalence odds ratios and incidence rate ratios, respectively. RESULTS: B/AA patients had higher prevalence rates of DRD every year analyzed compared with White patients (2021 B/AA:23.1%; W:19.0%; p<0.001). Both Hispanic (2001 H:12.3%) and Asian (2001:11.9%) patients initially had lower DRD prevalence than White patients (2001:13.1%; p<0.001 for both); both are now higher with Hispanic patients having the highest rates (2021 H:26.0%; A:21.2%;W:19.0%, p<0.001). DME and PDR prevalence increased across all groups through 2015/2016, then decreased through 2021 (2021 DME:W:4.5%, B/AA:5.9%; H:5.9%, A:4.7%; 2021 PDR:W:2.9%, B/AA:4.3%, H:5.0%, A:2.9%).Since 2009, incidence rates for DRD, DME, and PDR in Hispanic and B/AA patients have been higher than for White patients (IRR=1.08-1.85; p<0.001 for all comparisons). Asian patients initially had higher DRD incidence rates than White patients, but that difference disappeared in 2021 before increasing again in 2022 (2022 IRR=1.07, 95%CI=1.01-1.14). CONCLUSION: Disparities in prevalence and incidence of DRD, DME, and PDR persist for B/AA and have worsened for Hispanic patients.

  PublisherDOI

• Epidemiology of retinal disease

  Handbook of clinical neurology · 2026-01-01

  book-chapterSenior author
  PublisherDOI

• Improving Fairness and Mitigating Bias in Multicenter Electronic Health Records Models to Predict Glaucoma Outcomes

  Ophthalmology Science · 2026-02-17

  articleOpen access

  Purpose: To evaluate the effectiveness and generalizability of bias mitigation methods in glaucoma progression prediction models across a multicenter electronic health records (EHRs) repository and to propose a novel evaluation metric that balances performance and fairness in clinical artificial intelligence (AI). Design: A cohort study. Participants: A total of 50 656 glaucoma patients drawn from seven participating institutions in the SOURCE consortium, a harmonized EHR repository spanning ophthalmology departments in the United States. Methods: We trained five model architectures (e.g., XGBoost, neural networks, and transformers) to predict progression to surgery. Each model was evaluated with and without five bias-mitigation methods across preprocessing, inprocessing, and postprocessing. Performance and fairness were assessed on 1 internal and 2 external test sets. We introduced FairOdds-AUC, a composite metric that adjusts area under the receiver operating curve (AUROC) by equalized odds gaps across sex and race/ethnicity. The FairOdds-AUC metric was implemented in Python and is available as an open-source package for reproducibility and future use. Main Outcome Measures: Area under the receiver operating curve, equalized odds for sex and race/ethnicity, and FairOdds-AUC. Results: Inprocessing methods, particularly inverse propensity weighting (IPW) and the adversarial fairness classifier, achieved more favorable fairness-performance tradeoffs than baseline and other mitigation approaches across all evaluation sets. For example, on the internal test set, IPW improved FairOdds-AUC from 0.562 (95% confidence interval 0.540, 0.581) to 0.600 (0.575, 0.629) for the transformer model and from 0.556 (0.534, 0.577) to 0.5922 (0.53, 0.61919) for a fully connected network, while maintaining essentially the same discrimination. Adversarial fairness classifier achieved the highest FairOdds-AUC in several settings (up to 0.613 [0.595, 0.629] for the deep learning fully connected network) with substantial reductions in equalized odds difference for sex. Postprocessing and preprocessing bias mitigation strategies yielded more variable FairOdds-AUC changes (-0.009 to +0.021) and showed weaker generalizability across external sites. FairOdds-AUC consistently reflected the balance between AUROC and equalized odds, with the optimal mitigation strategy depending on fairness-utility priorities. Conclusions: Across a large, diverse glaucoma cohort, inprocessing bias methods provided the most consistent performance across evaluation sites in promoting fairness. FairOdds-AUC offers a flexible, interpretable way to evaluate clinical AI where fairness matters. Our findings support the recommendation to incorporate fairness evaluations and fairness-aware model training for future ophthalmic AI applications. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

  PublisherDOI

• Data Volume and the Need for Clinical Decision Support in Glaucoma Care

  Ophthalmology Science · 2026-05-01

  articleOpen access
  PublisherDOI

• Evaluating Metrics Assessing Surgical Success in Patients Undergoing Cataract Surgery

  JAMA Ophthalmology · 2026-04-16

  articleOpen access

  Importance: As health insurers, payers, and policymakers look to Merit-Based Incentive Payment System (MIPS) measures to judge eye care quality, compare surgeon performance, and make decisions about reimbursement, it is essential to understand the validity and generalizability of these metrics. Objective: To assess the validity and generalizability of 2024 MIPS measure 191, Cataracts: 20/40 or Better Visual Acuity Within 90 Days Following Cataract Surgery. Design, Setting, Participants: This was a retrospective cohort study of patients who underwent at least 1 cataract surgery from 2010 through 2023 at any of 16 participating health systems in the Sight Outcomes Research Collaborative (SOURCE). The study assessed whether patient demographic characteristics and nonclinical and clinical factors were associated with achieving success, which MIPS measure 191 defines as a best recorded visual acuity of at least 20/40 within 90 days following cataract surgery. Data analysis was conducted from June 2024 to December 2025. Exposures: Cataract surgery. Main Outcomes and Measures: The percentage of patients undergoing cataract surgery achieving success and the odds of surgical success were determined; for patients with 2 operative eyes, only the first eye surgery was assessed. Success rates were evaluated among all surgery recipients and separately among patients with no preexisting chronic ocular diseases as specified by the metric. Success rates were quantified using more stringent cutoffs and adjusting the follow-up duration. Logistic regression models assessed how nonbiological determinants of health and clinical factors influenced the odds of success. Results: A total of 55 132 patients underwent cataract surgery (mean [SD] age, 70.3 [9.3] years; 32 240 [58.5%] female; 1973 [3.6%] Asian American; 7053 [12.8%] Black; 1993 [3.6%] Hispanic; 42 178 [76.5%] White [race and ethnicity were self-reported]). Among all patients undergoing surgery, 49 979 (90.7%) achieved surgical success. Excluding patients with ocular comorbidities (25 563 patients [46.4%]), 28 242 of 29 569 (95.5%) achieved surgical success. Living in the least (vs most) affluent community (odds ratio [OR], 0.81; 95% CI, 0.72-0.91), undergoing complex surgery (OR, 0.82; 95% CI, 0.75-0.89) or a combination of cataract with another intraocular surgery (OR, 0.32; 95% CI, 0.29-0.35), and having diabetes (OR, 0.90; 95% CI, 0.84-0.98) were associated with lower odds of surgical success. Conclusions and Relevance: In this cohort study, most patients undergoing cataract surgery achieved success as defined by MIPS measure 191; however, the existing measure excluded nearly half of patients undergoing surgery, and older patients and Black patients were more likely to be excluded. Case-mix adjustment of patients' sociodemographic characteristics and clinical factors may be necessary to ensure fairness when comparing surgeons' performance.

  PublisherOA PDFDOI

• Visual Deficits Due to Diabetic Retinal Disease Across Race and Ethnicity

  Journal of VitreoRetinal Diseases · 2026-01-23 · 1 citations

  articleOpen accessSenior author

  Purpose: To understand the current disparities in visual deficits across race and ethnicity in patients with diabetic retinal disease (DRD). Methods: The Sight Outcomes Research Collaborative was used to identify all patients with DRD, diabetic macular edema (DME), and proliferative diabetic retinopathy (PDR) from 2015 to 2023. We created descriptive cohorts for each of the 3 diseases based on patients self-describing as White (non-Hispanic), Black/African American (non-Hispanic), Hispanic, or none of these categories (grouped as Other). We implemented a linear regression model with generalized estimating equations to estimate differences in characteristics between groups. The primary outcome was percentage of eyes with visual deficits to a visual acuity (VA) level of 20/200 or worse. Results: Percentages of eyes with VA of 20/200 or worse at inclusion were as follows: for DRD, 6.7% (1208 of 17 909 eyes) in the White (non-Hispanic) group, compared with 10.2% (779 of 7644 eyes) in the Black/African American group, 11.6% (499 of 4295 eyes) in the Hispanic group, and 7.8% (314 of 4040 eyes) in the Other group (each P &lt; .001); for DME, 5.1% (214 of 4218 eyes) in the White (non-Hispanic) group, compared with 4.9% (81 of 1670 eyes) in the Black/African American group, 3.4% (29 of 844 eyes) in the Hispanic group, and 4.9% (43 of 877 eyes) in the Other group (each P = .21); for PDR, 16.4% (958 of 5838 eyes) in the White (non-Hispanic) group, compared with 23.3% (684 of 2935 eyes) in the Black/African American group, 24.4% (459 of 1879 eyes) in the Hispanic group, and 19.0% (242 of 1271 eyes) in the Other group (each P &lt; .001). Conclusions: Disparities in visual deficits between races and ethnicities continue to exist.

  PublisherDOI

• Trends in Prevalence and Incidence of Diabetic Retinal Disease across Age Cohorts

  Ophthalmology Retina · 2026-03-01

  articleOpen accessSenior authorCorresponding

  OBJECTIVE: Two recent studies have suggested that younger patients with diabetes mellitus (DM) are progressing to diabetic retinal disease (DRD) at higher rates than in previous years. This study aims to determine how the rates of DRD and vision-threatening diabetic retinopathy (VTDR), comprised of diabetic macular edema (DME) and proliferative diabetic retinopathy (PDR), among patients with DM have changed across age groups over time. DESIGN: Retrospective cohort study comprised of members of commercial and Medicare Advantage health plans between 2000 and 2022. SUBJECTS: Cohorts of patients aged ≤25, 26 to 45, 46 to 64, and ≥65 years were created from all patients with DM identified using International Classification of Diseases 9/10 or ICD-9/10 codes. METHODS: Logistic and Poisson regression models were used to estimate prevalence and incidence, respectively. MAIN OUTCOME MEASURES: The main outcomes were the unadjusted prevalence and incidence of DRD, VTDR, DME, and PDR. RESULTS: Age cohorts remained proportional with respect to DRD prevalence (P) and incidence (I) over the 20 years analyzed. The ≤25 age cohort was always the lowest, followed by the 26 to 45, 46 to 64, and ≤65 cohorts, respectively (2001-2021 ≤25 P = 3.0%-4.5%; I = 5.1-11.1 [cases/1000-person-years]; 26-45 P = 6.3%-10.9%; I = 9.7-21.0; 46-64 P = 10.6%-15.7%; I = 14.7-30.9; ≥65 P = 13.2%-23.1%; I = 20.0-39.2). Similar proportional relationships were seen for VTDR and DME prevalence. However, VTDR and DME incidence saw differences with ≥65 age cohort (2022 VTDR I = 6.0 cases/1000 person-years; DME I = 5.0), ending lower than the 44 to 64 age cohort (2022 VTDR I = 7.1, DME = 5.5), whereas the 26 to 45 (2022 VTDR I = 4.7; DME I = 3.2) and ≤25 (2022 VTDR I = 2.0; DME I = 1.7) were the next lowest, respectively (P < 0.001 for all comparisons across age in 2022). Proliferative diabetic retinopathy prevalence was highest in the ≥65 cohort (3.7%) until 2021, when it became similar to the 46 to 64 age cohort (3.5%) (P = 0.40). Proliferative diabetic retinopathy incidence also declined significantly in ≥65 cohort (2022 I = 2.3) ending lower than 46 to 64 cohort (I = 3.8) and similar to the 26 to 44 cohort (I = 2.6), but still higher than the ≤25 cohort (I = 0.6) (P = 0.007 for ≥65 vs. ≤25 and P < 0.001 for ≥65 vs. 46-64; P = 0.07 for ≥65 vs. 26-45). CONCLUSIONS: In contrast to recent studies, no evidence was found to suggest that DRD is progressing faster in those aged ≤25 years. Those aged ≥65 years saw significant declines in DRD prevalence and incidence over the observation period. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

  PublisherDOI

• Risk of non-infectious uveitis associated with disease-modifying therapies for multiple sclerosis

  American Journal of Ophthalmology · 2026-05-01

  articleSenior author
  PublisherDOI

• Influence of Intraocular Pressure on Clinical Decision-Making in Glaucoma Management

  JAMA Ophthalmology · 2026-01-08 · 3 citations

  articleOpen access

  Importance: Understanding of intraocular pressure (IOP) as a continuous risk factor for glaucoma has evolved over time, and IOP reduction is widely acknowledged as the mainstay of treatment. However, the impact of specific IOP levels on clinical decision-making remains an underexplored topic. Objective: To assess how IOP levels influence the decision to initiate or escalate glaucoma therapy in clinical practice. Design, Setting, and Participants: In this retrospective, multicenter cohort study, the Sight Outcomes Research Collaborative (SOURCE) ophthalmology data repository was used to identify clinic encounters between October 2009 and January 2022 for patients with glaucoma with IOPs ranging from 12 mm Hg to 25 mm Hg. Data analysis was performed from July 2024 to September 2025. Main Outcomes and Measures: The primary outcome was whether IOP-lowering therapy was initiated or escalated after each clinic encounter, defined as a new prescription for IOP-lowering medication within 1 week of the encounter, laser treatment within 4 weeks, or glaucoma surgery within 8 weeks. The rate of treatment initiation at different IOP levels was measured, and then mixed-effects logistic regression was used to model the odds of treatment initiation at specific indicator IOP levels. Results: This analysis included 1 866 801 clinic encounters from 184 504 eyes of 94 232 unique patients across 7 sites in SOURCE. Mean (SD) patient age was 69.5 (10.8) years, and of the total clinic encounters, 1 084 827 (58.1%) included female patients. The rate of IOP-lowering treatment increased with higher IOP levels, with the largest acceleration in treatment rate at IOPs of 22 mm Hg or higher. With mixed-effects logistic regression modeling, an indicator IOP of 22 mm Hg had a greater effect on treatment initiation (odds ratio, 1.11; 95% CI, 1.08-1.14) compared with lower indicator IOPs. Conclusions and Relevance: In this cohort study, while clinicians seem to generally use IOP as a continuous risk factor in their treatment patterns, with higher rates of glaucoma therapy at increasing IOP levels, these findings suggest that the historical IOP cutoff of 22 mm Hg may still influence clinician decision-making in glaucoma management. Improved clinical decision support may be useful to assist clinicians with using IOP as a continuous risk factor in their decision-making.

  PublisherOA PDFDOI

• Trends in prevalence and incidence of diabetic retinal disease in patients with type 1 and type 2 diabetes mellitus

  Journal of Diabetes and its Complications · 2026-04-07

  articleOpen accessSenior author

  PURPOSE: To assess the prevalence and incidence of diabetic retinal disease (DRD) among patients with Type 1 (T1DM) and Type 2 diabetes mellitus (T2DM). METHODS: Data were abstracted from patients enrolled in commercial insurance and Medicare Advantage plans. Using ICD-10 coding, the yearly prevalence and incidence of DRD in T1DM and T2DM from January 1, 2016, to June 30, 2022, were analyzed. Assessments were also made into the prevalence and incidence of vision threatening disease (VTDR), defined as having a diagnosis of diabetic macular edema (DME) or proliferative diabetic retinopathy (PDR). Odds (OR) and incidence rate ratios (IRR) were also calculated. RESULTS: Prevalence of DRD in both cohorts increased from 2016 (T1DM:25.1%; T2DM:11.1%) through 2021 (T1DM:34.4%; T2DM:20.7%). T1DM DRD incidence decreased from 55.1/1000py in 2016 to 39.2 in 2022, while T2DM varied from 31.6 to 38.2, ending at 35.5 in 2022. VTDR (yearly OR = 4.61-4.92), DME (OR = 2.79-2.92), and PDR (OR = 6.43-6.72) each had higher prevalences in patients with T1DM vs. T2DM across all years studied (p < 0.001 for all comparisons). A similar pattern was seen for incidence as well (T1DM VTDR IRR = 2.36-2.94; DME IRR = 2.08-2.62; PDR IRR = 2.66-3.67; p < 0.001 for all T1DM vs. T2DM comparisons). CONCLUSION: All forms of DRD had uniformly higher prevalence and incidence in patients with T1DM compared with those with T2DM across all years observed. The incidence of T1DM DRD decreased, approaching the rate of T2DM in 2022, suggesting a reduction in the disparity between the two groups. Prevalence of VTDR, DME, and PDR was also higher in T1DM vs. T2DM. The incidence of these disease states decreased proportionally across both cohorts, maintaining relatively stable ratios.

  PublisherDOI

Recent grants

• Patient Centered Care for Diabetic Macular Edema

  NIH · $221k · 2015–2020

• Patient Centered Care for Diabetic Macular Edema

  NIH · $910k · 2015–2020

Frequent coauthors

• Yinxi Yu

  University of Pennsylvania

  85 shared

• Alexander J. Brucker

  Duke University

  40 shared

• John H. Kempen

  Harvard University

  33 shared

• Lucia Sobrin

  Massachusetts Eye and Ear Infirmary

  30 shared

• Drew Scoles

  Penn Presbyterian Medical Center

  25 shared

• Jonathan C. Tsui

  23 shared

• Brendan McGeehan

  University of Pennsylvania

  23 shared

• Wei Pan

  Central South University

  22 shared

Labs

• VanderBeek LabPI