Pharmacies Sometimes Dispense Discontinued Medications

When physicians do not inform pharmacies that a patient’s medication has been discontinued, potentially harmful medication errors can result.

Pharmacies at times dispense medication to patients even after a physician has discontinued the prescription, leading to potentially harmful medication errors, according to the results of a study published online on November 20, 2012, in Annals of Internal Medicine.

The study aimed to determine the frequency of and potential harm due to pharmacy dispensing of discontinued medications to ambulatory patients. The researchers analyzed electronic health record (EHR) data on patients at Harvard Vanguard Medical Associates (HVMA), a multispecialty group practice including 15 ambulatory health centers in eastern Massachusetts. EHRs allowed the researchers to easily measure the prevalence of medication dispensing after discontinuation.

The study included 30,406 adult patients who had an electronic discontinuation order that was not reordered within 48 hours of discontinuation for antihypertensive, antiplatelet, anticoagulant, oral hypoglycemic, and statin medications between November 2008 and October 2009. Participants were limited to those who had their prescriptions filled at an HVMA pharmacy and who had an outpatient visit at HVMA between November 2006 and November 2008. The targeted medications were selected because they are used for long periods and pose a potential for significant harm when misused.

The researchers found that of 83,902 targeted medications that were electronically discontinued, 1218 (1.5%) were later dispensed to 1128 patients by a pharmacy on average 1.0 times within 12 months of being discontinued. Among the 10 most frequently discontinued medications, the rate of subsequent dispensing ranged from 0.9% for metformin to 2.1% for amlodipine and 2.5% for metoprolol. Compared with patients who were not dispensed a discontinued medication, those who were dispensed one were more likely to be black (odds ratio of 1.5) or insured by Medicare (odds ratio of 1.7) or Medicaid (odds ratio of 1.4).

A manual chart review of 416 medication-dispensing events defined as high-risk identified potential harm in 50 (12%) cases. These included 18 documented clinical reactions, 17 laboratory anomalies, 8 instances of duplicated medication dispensing, and 7 instances in which medication was dispensed in the presence of a documented allergy.

The researchers note that just under half of the potential participants in the study could not be included because they filled their prescriptions at non-HVMA pharmacies and thus lacked pharmacy dispensing information. In addition, chart review was conducted for just a subset of medication-dispensing events at higher risk for adverse events.

The researchers conclude that dispensing of discontinued medications is an important safety concern and that EHRs should be used to facilitate better communication between health care providers and pharmacies. They note that the EHRs used by the providers included in the study automatically transmit prescriptions to pharmacies, but do not transmit discontinuations of prescriptions. More broadly, they argue that EHRs should be used to open up 2-way communication between physician offices and retail pharmacies to help ensure that new prescription orders are transmitted, current medications are verified, and discontinued medications are removed from the pharmacy list.