[Preventable adverse drug effects at an emergency department]

To determine the incidence and evaluate the preventability of adverse drug events (ADE) associated with visits to the Emergency Department at our hospital and subsequent hospital admissions.

A six-month observational study was conducted at an Emergency Department in a University Teaching Hospital (October 15th, 1995, to April 15th, 1996). The parameters influencing the preventability were identified by means of a multivariate logistic regression analysis.

A total of 776 ADEs (2.25%) were detected out of a total of 33,975 patients attended at the Emergency Department; 178 patients were admitted. A total of 322 cases (43.3%) were classified as preventable and were graded as mild (37.1%), moderate (32.5%), severe (27.4%), and fatal (3%). The logistic regression analysis showed that preventability was related to drugs with a narrow therapeutic index (NTI) (OR: 10.12; 95%CI: 5.36-19.07), type A ADE (OR: 4.65; 95%CI: 2.79-7.78), age > or = 65 years (OR: 3.04; 95%CI: 2.13-4.34) and self-administered medication (OR: 2.2; 95%CI: 1.32-3.65). Among admitted patients, oral anticoagulants, NSAIDs, digoxin, diuretics, and insulin caused adverse events which were considered as preventable in more than 50% of cases. The errors most frequently associated with preventable ADEs included inappropriate therapy monitoring (22.5%), increased doses with NTI drugs (22.3%), absence of preventive therapy (14.3%), excessive dose according to patient's characteristics (13.4%), and inappropriate self-administered medication (10%).

The incidence of preventable ADEs (medication errors) is high and its severity is higher than that of non-preventable ADEs. A prompt development and implementation of measures leading to avoiding prescription errors and inappropriate treatment monitoring, the factors identified as responsible for preventable ADEs, is clearly warranted.