# Longitudinal progression of cost-related medication non-adherence among Medicare patients with diabetes at high risk of hospitalization: The role of dual eligibility

## TL;DR

This study tracks how diabetes patients in Medicare skip medications due to cost over time, finding that dual eligibility reduces this risk while depression increases it.

## Contribution

The study provides longitudinal insights into cost-related medication non-adherence in high-risk diabetes patients and identifies dual eligibility and depression as key factors.

## Key findings

- Dual-eligible patients had a lower hazard ratio for cost-related medication non-adherence (HR = 0.67).
- Patients with depression had a higher likelihood of reporting cost-related medication non-adherence (coefficient = 0.32).
- Dual eligibility was associated with lower income and lower likelihood of cardiovascular disease.

## Abstract

Little is known about the longitudinal progression of cost-related medication non-adherence (CRN) among the high-need, high-cost diabetes population. We aim to document the longitudinal aspect of CRN among Medicare diabetes patients at high risk of hospitalization and the role of Medicare-Medicaid dual eligibility in CRN.

617 Medicare diabetes patients at high risk of hospitalization were followed up at 3-month intervals for a total of 16 surveys. Patients’ socio-demographic and health characteristics by dual eligibility were compared using Chi-square tests. The progression of CRN was documented using a Kaplan-Meier Survival Curve. A Cox Survival Regression analysis and a Generalized Estimating Equation (GEE) analysis were conducted to evaluate the adjusted hazard ratio (HR) and population-averaged effect of dual eligibility on CRN, controlling for socio-demographic and health characteristics.

303 patients (49.1%) reported dual eligibility, among whom 151 (49.8%) reported CRN; they were more likely to be under 65 (p < 0.01), had lower income (p < 0.01), were less likely to report cardiovascular disease (p = 0.05), and were less likely to report CRN (p < 0.01) compared to those who did not report dual eligibility. Those with dual eligibility had a lower hazard ratio (HR = 0.67, p < 0.01) and lower likelihood of reporting CRN (coefficient = −0.40, p < 0.01), and those with depression had higher hazard ratio (HR = 1.31, p = 0.03) and higher likelihood of reporting CRN (coefficient = 0.32, p < 0.01) in the Cox model and GEE, respectively.

While insurance coverage enables patients to overcome their major deficiency in income, many patients fall through the cracks as their disease progresses. Depression is a major risk factor for CRN. Health policy addressing CRN needs to be implemented in tandem with clinical intervention, targeting those at the increasing risk of CRN.