Unveiling inequities in retinoblastoma treatment: A SEER-based analysis of race, ethnicity, geography, and income (2000–2021).

Abstract

e13794 Background: Retinoblastoma is a rare childhood intra-ocular malignancy. The impact of geographic and demographic variables on treatment time and options for retinoblastoma patients remains poorly understood. Using the Surveillance Epidemiology and End Results (SEER) database, this study investigates associations between the time from diagnosis to treatment and in the treatment options provided to retinoblastoma patients by race, ethnicity, geography, and income. Methods: 1394 retinoblastoma patients from 2000 - 2021 were identified with the ICD-O-3 morphologic code 9510 on SEER. Patients older than 18 and incomplete cases were excluded. SEER provided data on race/ethnicity (Non-Hispanic White, Non-White), sex, income, age, and county classification (metropolitan, nonmetropolitan). These variables were studied in relation to retinoblastoma treatment options provided using chi-square tests, odds ratios (ORs), and regression. ANOVA compared treatment times across groups within each variable, and multiple linear regression evaluated the variables’ combined effect on treatment time. Results: Chi-square tests revealed surgical treatments differed by race/ethnicity (p = 0.014) and geography (p = 0.006), but not income (p = 0.566). Non-Hispanic White patients were less likely to undergo enucleation (OR: 0.679, 95% CI: 0.548 - 0.842), as were metropolitan patients (OR: 0.652, 95% CI: 0.459 - 0.926). Metropolitan patients were more likely to receive chemotherapy (OR: 1.523, 95% CI: 1.082 - 2.144). Multinomial regression identified geography (p = 0.002), race/ethnicity (p = 0.010), and age (p < 0.001) as predictors of surgical treatment. Binary regression identified age (p < 0.001) and geography (p = 0.015) as predictors of chemotherapy access. ANOVA found no differences in treatment time across income, race/ethnicity, and geography. Multiple linear regression confirmed these factors do not predict treatment time. Conclusions: This study identifies a previously unknown geographical disparity in retinoblastoma treatment options, as patients in metropolitan counties were less likely to receive enucleation and more likely to receive chemotherapy. No differences in time between diagnosis and treatment were observed across race/ethnicity, income, and geography. This finding may imply that delays in initiating treatment are unlikely to explain disparities in enucleation rates. ANOVA of treatment time for retinoblastoma patients across race/ethnicity, income, and geography (2000-2021). ​ Posterior Mean Treatment Time (Days)​ Variance​ 95% Confidence Interval ​ Non-Hispanic White​ 10.27​ 0.63​ 8.71 - 11.82​ Non-White​ 11.12​ 0.46​ 9.80 - 12.45​ >$60,000 Income​ 11.00​ 0.30​ 8.65 - 13.35​ <$60,000 Income​ 10.71​ 2.40​ 9.60 - 11.83​ Metropolitan​Counties​ 10.68​ 1.44​ 9.61 - 11.75​ Non-Metropolitan​Counties​ 11.45​ 0.32​ 8.41 - 14.49​