Analysis of allergy alerts within a computerized prescriber-order-entry system

Abstract

Allergy alerts within a computerized prescriber-order-entry (CPOE) system were analyzed. A retrospective analysis was performed to evaluate inpatient CPOE orders that triggered an allergy alert for one month in November 2007. Data were extracted from a computer-generated report to identify the total number of prescription orders and the number of orders triggering an allergy alert during the study time frame. When overriding an allergy alert from the CPOE system, prescribers had to choose one of three rationales: the benefit outweighs the risk, the patient previously tolerated the medication, and the medication is therapeutically appropriate. Data collected included the number of allergy alerts, allergy alert overrides, override rationale, and patients' demographic information. Descriptive statistics were used to summarize the results. Allergy alerts were triggered on 643 (1.3%) of the 49,887 total orders entered during the month (95% confidence interval [CI], 1.2-1.4%). Alerts were triggered on a total of 289 patients, with a mean of 2 orders triggering alerts per patient. The median age was 65 years old. The median hospital stay for patients with an order triggering an alert was three days. Overall, 625 of 643 alerts (97%) were overridden for rationales as following: the patient previously tolerated the medication (49%), the benefit outweighed the risk (29%), and the medication was therapeutically appropriate (24%), and a free text explanation (8%). While a small portion of prescription orders generated through a CPOE system triggered an allergy alert, most of the allergy alerts were overridden by prescribers.