Abstract
Purpose :
To estimate the proportion of the racial and ethnic disparity in glaucoma surgical outcomes eliminated by theoretically intervening on socioeconomic status (SES).
Methods :
A retrospective cohort was constructed using the entire population of 2016-2018 CA fee-for-service Medicare beneficiaries with a claim for incisional glaucoma surgery (trabeculectomy, tube shunt, or EX-PRESS shunt). Exclusion criteria were: non-CA residence, age ≤64 years, or missing eye laterality modifier code. The primary exposure was race and ethnicity: Non-Latino White, Black, Latino, Asian, and Other. The mediator was SES dichotomized to low/non-low based on dual-eligibility for Medicaid coverage. Time to failure event was defined as having a claim for revision of index surgery or reoperation. Follow-up time extended through 2019. Time-to-event was modeled using Cox proportional hazards with age and sex as covariates. Causal mediation analysis was used to estimate the total effect (TE) of the entire racial and ethnic disparity on surgical failure and the controlled direct effect (CDE) for the remaining disparity after fixing SES to non-low for all.
Results :
The final sample consisted of 6,524 beneficiaries, of whom 52.3% (n=3,409) were non-Latino White, 9.7% (n=631) Black, 14.7% (n=959) Asian, 19.7% (n=1,286) Latino, and 3.7% (n=239) Other. About 24.4% (n=1,593) of all beneficiaries had a failure event. The TE estimates showed greater risk of failure in Black (hazard ratio [HR]: 1.12, p=.20), Asian (HR: 1.21, p<.01), Latino (HR: 1.21, p<.01), and Other race and ethnicity beneficiaries (HR: 1.29, p=.04) compared to non-Latino White counterparts. The CDE estimates showed no significant disparities after fixing SES to non-low for Black (HR: 0.96, p=0.75), Asian (HR: 1.19, p=0.08), Latino (HR: 1.13, p=0.22), and Other (HR: 1.22, p=0.20) groups. The proportion of the racial and ethnic disparity that could be eliminated by intervening on SES was 129.7% for Black, 8.6% for Asian, 37.4% for Latino, and 22.5% for Other race and ethnicity beneficiaries.
Conclusions :
There was increased risk of glaucoma surgical failure in racially and ethnically minoritized CA Medicare beneficiaries. This disparity may be mitigated by intervening on SES, though the proportion eliminated was different for each minoritized group. Future studies are needed to identify additional targets to reduce disparities.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.