Predictors of Medication-Refill Adherence in an Indigent Rural Population

Med Care. 2002 Dec;40(12):1294-300. doi: 10.1097/00005650-200212000-00016.


Objective: We evaluated the association of medication refill adherence with demographic and prescription characteristics to determine whether such factors could guide intervention strategies in an indigent rural population.

Methods: The study was conducted at a University-based internal medicine practice serving an indigent rural population. Refill data for diabetes, hypertension, and hypercholesterolemia drugs from a closed pharmacy system were used to calculate mean adherence (for all drugs taken by each patient) and minimum adherence (that of the least adhered to drug) for 1984 patients during a 9-month period.

Results: Mean refill adherence was <80% for 33% of the population and minimum refill adherence was <80% for 55% of the patients. Increasing age, race (white), and prescription length were associated with higher mean and minimum adherence, independent of income, prescription copay, and insurance status. Number of drugs taken had a positive mean but negative minimum adherence association. Gender, number of primary care visits, and dosage schedule were not independently associated with adherence. The model explained 6.8% of the variance in mean adherence.

Conclusions: In a rural indigent population, medication refill adherence was associated with race, age, and prescription length, though these factors explained only a small amount of adherence variability. Although ingestion adherence is the goal, refill adherence is a necessary condition for ingestion adherence. To enhance adherence, physicians need better predictors to target their efforts to patients most in need of attention. Prescription claims data could serve this purpose.

MeSH terms

  • Diabetes Mellitus / drug therapy
  • Drug Prescriptions*
  • Female
  • Humans
  • Hypercholesterolemia / drug therapy
  • Hypertension / drug therapy
  • Male
  • Medical Indigency
  • Middle Aged
  • Patient Compliance*
  • Regression Analysis
  • Rural Population
  • Social Class
  • Treatment Refusal
  • Virginia