RESUMO
OBJECTIVE: To determine the prevalence of adverse drug events (ADEs) leading to hospital admission, and to assess those that were potentially preventable, identifying the drug classes involved, types of medication errors and the factors associated with the preventable ADEs. METHOD: An observational study, over a six-month period on ADEs that lead or contributed to hospital admissions, carried out in 6 medical units of a university hospital. RESULTS: A total of 259 ADEs were detected of which 159 (61.4%) were assessed to be potentially preventable. The overall prevalence of admissions directly due to ADEs was of 6.7% (177) and to preventable ADEs of 4.7% (125). In addition, 82 ADEs that contributed to hospital admission were detected. Risk factors for preventable ADEs were patient age of 65-74 (OR = 1.40) or = 75 years (OR = 2.70), self-medication (OR = 15.55), prescription in primary care (OR = 2,88) and the use of narrow therapeutic index drugs (OR = 2.40). The drug classes most frequently involved in preventable ADEs were NSAID and aspirin (32.5%), diuretics (15.3%), antihypertensives (9.1%) and digoxin (7.7%). Inadequate therapy monitoring (20.7%), prescription of an inappropriate drug (15.7%) or of an excessive dosage (12.0%), lack of preventive treatment (15.7%), non-adherence (10.6%) and inappropriate self-medication (10.1%) were the most commonly identified types of error. CONCLUSIONS: A high proportion (4.7%) of hospital admissions are caused by potentially preventable ADEs. Results obtained justified the need to adopt measures directed at improving surveillance and prescription quality, and educating patients in safe drug use, focusing especially on older patients and narrow therapeutic index drugs.