Changes in Medication Management After a Diagnosis of Cancer Among Medicare Beneficiaries With Diabetes

PURPOSE: A new cancer diagnosis commonly initiates a cascade of health care decisions that have potentially important consequences for management of other chronic conditions such as diabetes. We sought to determine whether a new cancer diagnosis is associated with changes in medication adherence among Medicare beneficiaries with diabetes, and whether the relationship is affected by life expectancy and generosity of drug coverage. METHODS: The study population was drawn from a 5% random sample of Medicare beneficiaries with diabetes enrolled in Medicare Part D in 2007 and 2008. Patients had cancer newly diagnosed between January and December 2007 (n = 4,348) and were compared with a cancer-free control group (N = 28,507) assigned a pseudo-diagnosis date. Adherence (proportion of days covered [PDC]) with oral hypoglycemic agents, renin-angiotensin-aldosterone system inhibitors, and statins was tracked for 6 months before and after the diagnosis date. Multivariable regression models assessed the independent impact of a cancer diagnosis, life expectancy (proxy measure: died 7 to 12 months after index date), and coverage generosity (proxy measure: low-income subsidy recipient) on PDC, controlling for individual characteristics. RESULTS: Relatively larger declines in medication adherence (3 to 5 percentage points; P < .001) were observed overall for patients with cancer versus controls. Short life expectancy was associated with between 8% and 11% lower PDC (P < .001) in the cancer subgroup relative to controls. Low-income subsidy status had no differential effect on changes in drug adherence. CONCLUSION: A cancer diagnosis among patients with diabetes reduced adherence with evidence-based medications, particularly if patients' life expectancy was short.