Currently, the options for treatment of hematologic malignancies continues to expand with resulting positive outcomes in overall survival, but there is still concern that not all populations have the opportunity for cancer treatment due to lack of health insurance. We have taken a cohort of patients with health-insurance related disparities in a majority-minority population whose immigration status often precludes them from standard insurance but can still utilize a payment plan (PP) associated with hospital care. The main objective of our study was to see how insurance status in a majority-minority population affected treatment response.
Doing a retrospective analysis of a cohort of 364 patients with hematologic malignancies at a single center institution, the patients identified received care exclusively with Mays Cancer Center at UT Health San Antonio, between 1998-2017. Variables for each patient measured included age, gender, date of diagnosis, treatment received, date of first treatment, initial response to treatment, stage, > 2 comorbidities, vitality status, HIV status, and insurance status. The patient's insurance status was either identified as funded (to include Medicare, Medicaid, and private insurance), unfunded (self pay) or unfunded-payment plan (PP) associated with the hospital. Statistical significance was assessed with Pearson's Chi-Square, Fisher's Exact test, and a logistic regression model with a generalized logit link with a 3-level response (CR, PR, F) where CR was designated as the referent. For each effect, the Odds Ratio (OR) and its 95% confidence interval (95% CI) are reported. All statistical testing was two-sided with a significance level of 5%. SAS Version 9.4 for Windows (SAS Institute, Cary NC) was used throughout.
Our patient population (n=346) was shown to have a median of 56, female patients (n=174, 50.3%), males patients (n=172, 49.7%), Hispanics (n=180, 52%), uninsured (n=107, 30.9%), and HIV (n=22, 6%). Diagnoses studied included aggressive lymphomas (Burkitt's, 1ry CNS, Hodgkin's, NHL, PTLD; n=252) and indolent lymphomas (Marginal Zone, Follicular; n=94). The odds of treatment failure (F) and of Partial Response (PR) in Hispanic patients was not significantly different from the odds in non-Hispanic patients (F OR=1.43, 95% CI 0.62 to 3.3, p=0.40, PR OR=1.36, 95% CI 0.73 to 2.54, p=0.33). Despite having a slightly larger insured patient population (68.9%), insurance status did not affect outcome for our patients. Less than 30% of patients used Carelink or had no insurance in this cohort [Carelink CR 58 (24.7%) PR 17 (24.3%) F 10 (24.4%), No insurance CR 15 (6.4%) PR 5 (7.1%) F 2 (4.9%)]. Twenty percent of Hispanic patients (36/180) and 29.5% (49/166) of non-Hispanic patients reported using Carelink. After combining Carelink with no insurance, the odds of treatment failure (F) and of Partial Response (PR) in patients with insurance was not significantly different from the odds in those without insurance (F OR=1.12, 95% CI 0.44 to 2.84, p=0.81, PR OR=0.95, 95% CI 0.49 to 1.85, p=0.89). In an examination of variation in the relation between treatment response and insurance status with ethnicity, we found no overall association between treatment response and insurance after adjustment for ethnicity (p=0.97) and no association between treatment response and insurance status among Hispanic or non-Hispanic patients (p=0.75 and p=0.69 respectively).
This study directly challenges the idea that Hispanic populations are faced with being the least likely to have health insurance of any racial or ethnic group; among those 18-64 years of age, 37% of Hispanics are uninsured compared to 13% of non-Hispanic whites. In an uninsured minority-majority Hispanic patient population with PP firmly established in the community that allows for ready access to healthcare, insurance status and race did not adversely affect outcome.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.