OBJECTIVES: We estimated increases in medical expenditures due to medication nonadherence and potential savings of increasing adherence for members of a prescription-drug benefit plan taking medications in four drug therapy classes (TCs).
METHODS: We used data from the Medical Expenditure Panel Survey (MEPS) to estimate functional relationships between adherence and resource utilization for patients taking medications in four TCs. Resource use included all-cause and disease-specific annual hospitalizations and emergency room (ER) visits. TCs included depression, diabetes, high blood cholesterol (cholesterol), and high blood pressure or heart disease (heart). Adherence was estimated using the medication possession ratio (MPR). MPR less than 80% was considered nonadherence. Average medication expenditures, by TC, was obtained from a large prescription-drug database. Expenditures per hospitalization and ER visit were estimated from MEPS. Unit costs and functional relationships between adherence and resource use were applied to estimate annual resource use and medication expenditure. Increased expenditures due to nonadherence were estimated for nonadherent patients ver- sus those with 80% adherence. Total expenditures considered expenditures from inpatient admissions, ER visits, and medications. Potential savings was defined as reduction in total expenditures due to increasing adherence.
RESULTS: Nonadherence resulted in increased all-cause total expenditures in diabetes, cholesterol, and heart by $240 million (M), $150M, and $47M, respectively. Increasing adherence by 2% reduced increases in all-cause expenditure by 11% to 21%. Nonadherence resulted in increased disease-specific hospitalization and ER visit expenditure for depression ($6M), diabetes ($44M), and cholesterol ($5M). However, increases in the disease-specific hospitalization and ER expenditures were offset by lower medication expenditure, thus resulting in overall lower disease-specific expenditure among the nonadherent patients. Overall, increases in medication adherence resulted in savings in all-cause expenditure but not in disease-specific expenditure.
CONCLUSIONS: Medication nonadherence can be costly to payers. Increasing adherence even by small amounts may result in significant savings.
