Listen to your Doctor, or else!: Medication Under-use and Overuse and Long-term Health Outcomes of Danish Diabetes Patients

Abstract

We use Danish diabetes registry and health insurance data to analyze the extent, consequences, and determinants of under-use and overuse of oral anti-diabetic drugs.Less than half of patients consume the appropriate amount of medication--between 90% and 110% of the amount prescribed by their doctors.The life expectancy of patients consuming the appropriate amount is 2.5 years greater than that of patients consuming less than 70% of the prescribed amount, and 3.2 years greater than that of patients consuming more than 130% of the prescribed amount, controlling for time since diagnosis, insulin dependence, comorbidities, age, gender and education. Patients consuming the appropriate amount are also less likely to be hospitalized than under-or over-users.Pharmaceutical innovation appears to have reduced medication under-use and overuse: patients using newer drugs are significantly more likely to consume the appropriate amount, controlling for socioeconomic factors, average number of pills and average daily out-of-pocket costs.Defined Daily Doses published by the World Health Organization substantially overstate the appropriate level of consumption of these medications.Patients who don't adhere to recommended medication regimens may also disregard other physician instructions. Medication under-use and overuse could easily be monitored to identify patients at risk and enact interventions. Gisela IntroductionMedication non-adherence is regarded as one of the main factors that reduce the effectiveness of drug therapies in clinical practice as compared to the expected effect from clinical trials. For many diseases including diabetes, medication non-adherence has been shown to be associated with long-term complications, hospitalizations, premature mortality, and billions of dollars per year in avoidable direct health care costs (1-5).It is therefore an area of concern to health professionals, insurers, policy makers, and researchers alike.Adherence is the extent to which the patient's behavior (medication-taking and lifestyle practices) coincides with medical or health advice (6). Even though a great deal of prior research has been done in the area, there is no gold standard for identifying identify non-adherence. Many studies have used a cut-off point of ≤80% of the targeted adherence level to classify patients as non-adherent patients across various diseases (4; 7; 8). However, ideally the adherence measure should take the specific epidemiology of the disease adequately into account and may therefore be disease-and even context-specific, depending on the quality of the available data.In health policy making there is an increasing desire to use administratively collected data, such as claims or registry data, as a controlling tool in order to monitor quality of care for example to design disease management programs (DMPs) in a more effective way. In order to be a useful tool for quality management, a meaningful indicator of patient adherence should be a good predictor of health outcomes. It should be easy to col...