A large observational study from the US found that patients were more likely to discontinue prescribed medication in the first month of treatment with a new drug.

The authors of this study comment that there is limited research on health services outcomes in the US, and that while there are studies on discontinuation rates in community-dwelling patients these have focussed on single medication classes. They used data from four large US retail pharmacy chains to identify discontinuation rates in patients taking medications in nine groups used for chronic therapy of potentially serious conditions: inhalers for asthma, oral medications for asthma, oral drugs for breast cancer treatment, cardiovascular medications, insulin, oral anti-diabetic drugs, eyedrops for glaucoma, drugs for osteoporosis, and statins. Patients were identified who filled a prescription for one of the study medications during January 2007 (the index fill). They divided patients into two groups - those who had not recently (within 180 days) filled a prescription for the medication ('medication-naïve') and those refilling a recent previous prescription ('medication-experienced'). Primary outcome measure was median time to discontinuation (defined as being at least 30 days late in refilling a prescription) over a one-year period.

There were data available for 2.17 million patients of 3,821 pharmacies; 57.4% were women, 85% were aged >50, and 10.2% were medication naïve. Absolute discontinuation rates for those who were medication-naïve were high, between 29.6% and 71.8%; rates were lower for those who were medication-experienced, between 7.8% and 29.2%. Highest rate for both groups was for asthma inhalers; lowest rates were for statins for medication-naïve and cardiovascular drugs. Differential rates between the groups ranged from 42.6% for asthma inhalers to 22.0% for osteoporosis drugs.

Median times to discontinuation were significantly shorter for medication-naïve patients in all drug groups ranging from 30 days to 90 days, compared to 122 to >360 for medication-experienced patients (14.2% to 28.9% as long where available). If a 60-day period to next refill was used, median discontinuation times for medication-experienced patients increased further to >330 days for most drug groups.

The authors conclude that patients newly started on medication have a much shorter time to discontinuation, and a much higher risk of discontinuation than patients who are medication-experienced. This difference persists across a wide range of medications for potentially serious chronic conditions. They suggest from their data a need for healthcare professionals to monitor particularly those patients newly started on medication, and discuss some of the factors that may lead to greater adherence among experienced patients.