Grant: Travel Awards - Institute Community Support

Electronic pharmacy data are often used to examine how drugs are used in the real-world. In order to identify how long patients take their medications researchers use a field called the days supply that is entered at the pharmacy where the medication is dispensed. While pharmacy data is considered valid, the days supply field has not received great attention. To identify if the days supply values reported for osteoporosis medications we described the days supply values and identified if they matched an expected value. We used data submitted to the Ontario Drug Benefits program to examine days supply values reported for osteoporosis medications from 1997-2011. The number and proportion of days supply values submitted were summarized by dosing routine (daily-, weekly-, and monthly-oral; nasal spray; cyclical -- 14 days active drug + 76 days calcium; and yearly infusion), and residence status (community vs. long-term care [LTC] resident). We defined the expected days supply by the dosing routine: daily in 7-, 30-, or 100-day intervals, weekly in 7- or 30-day intervals, monthly and nasal spray in 28- or 30-day intervals, and cyclical as 90-day supply. We identified 17,615,364 osteoporosis prescriptions dispensed to community (78%) or LTC (22%). Most daily oral prescriptions (97%) were dispensed for an expected days supply value. However, distinct differences were observed for other dosing routines with the expected days supply more common in community when compared to LTC. In both settings, annual zoledronic acid infusion was most commonly dispensed as 1-day supply (62%). Results suggest that there may be errors in how the days supply values are reported in Ontario pharmacy data, particularly among prescriptions dispensed in LTC. The errors in the days supply field may impact how researchers identify exposure and may lead to incorrect assessment of risk and benefit in studies of drug effects.