Medicaid versus commercial insurance: Association with quality of end-of-life care among patients with blood cancers Free

Introduction: Patients with blood cancers often receive suboptimal end-of-life (EOL) care, including high rates of healthcare utilization near death and low rates of hospice enrollment. Modifiable factors such as health insurance type may contribute to these care patterns; however, prior studies have largely held insurance constant, precluding the ability to evaluate its potential impact. We sought to examine quality of EOL care among patients with blood cancers who had Medicaid versus commercial coverage.

Methods: We conducted a retrospective cohort study using the MarketScan Commercial and Multi-State Medicaid claims databases. We included patients <65 years of age who were diagnosed with a blood cancer, survived ≥30 days from their diagnosis date, died between 2015 and 2021, and were insured by either Medicaid (excluding those who were dually eligible for Medicare) or commercial insurance. We examined established EOL care quality measures acceptable for patients with blood cancers (Odejide, JCO 2016) including (1) high-intensity healthcare utilization (≥2 emergency department (ED) visits, ≥2 hospital admissions, or any intensive care unit (ICU) admission in the last month of life and receipt of chemotherapy in the last 14 days of life), (2) in-hospital death, and (3) hospice enrollment. We conducted univariable and multivariable analyses examining the impact of insurance type on EOL care. Multivariable logistic regression models were adjusted for age, sex, geographic region, area-level education status, comorbidity, blood cancer type, hematopoietic stem cell transplantation (HCT) status, and transfusion dependence (≥2 blood transfusions in the last month of life).

Results: Of 2,669 eligible patients, 31.6% were covered by Medicaid and 68.4% by commercial insurance. The median age at death was 57 years and 60.7% were male. The most common blood cancer diagnosis was lymphoma (43.5%), followed by leukemia (32.5%), myeloma (15.9%), myelodysplastic syndromes (12.2%), and myeloproliferative neoplasms (3.5%). The median Charlson Comorbidity Index (Klabunde, J Clin Epidemiol 2000) was 2. Approximately 13.6% of patients had undergone HCT in the 12 months prior to death and 17% were transfusion dependent.

Of the overall cohort, 58.2% of patients had at least one indicator of high-intensity healthcare utilization at the EOL. Specifically, 4.2% and 16.4% had at least two emergency department or hospital admissions in the last 30 days of life respectively; 50.5% had at least one ICU admission in the last 30 days of life; and 12.7% received chemotherapy in the last 14 days of life. Almost half (48.7%) of the cohort died in the hospital. The rate of hospice enrollment was 26.0%; 23.5% of the total population had a hospice length of stay of >3 days.

In multivariable analysis, patients with commercial insurance had greater odds of high-intensity healthcare utilization near death (Odds ratio [OR]: 1.69, 95% Confidence Interval [CI]: 1.44 to 1.99) and in-hospital death (OR: 2.71, 95% CI: 2.29 to 3.21). Conversely, patients with commercial insurance had lower odds of enrolling in hospice compared to those with Medicaid (OR: 0.70, 95% CI: 0.58 to 0.83), and lower odds of hospice length of stay >3 days (OR: 0.66, 95% CI: 0.55 to 0.79).

Conclusions: In this large cohort of blood cancer decedents under 65 years, Medicaid beneficiaries were less likely to experience high-intensity healthcare utilization at the EOL, less likely to die in the hospital, and more likely to have timely hospice use compared to patients with commercial insurance. These data suggest potentially higher quality EOL care with Medicaid compared with commercial insurance. This difference is not trivial; indeed, hospice use and less intensive healthcare utilization at the EOL have been shown to potentially improve overall quality of life near death (Wright, JAMA, 2008), while also reducing healthcare system costs (Egan, Blood Adv, 2020). These findings are provocative, especially in the context of ongoing national efforts that are poised to impact Medicaid coverage.