Fatal adverse drug events: the paradox of drug treatment

Abstract

Objective: Study patient characteristics, morbidity patterns and drug regimens associated with fatal adverse drug events (FADEs) amongst medical department inpatients.

Design: An observational, descriptive study using aggregated medical records, autopsies and pre and postmortem drug analyses.

Setting: A department of internal medicine at a Norwegian county hospital.

Subjects: All patients dying in the department over a 2-year period.

Results: The incidence of FADEs were 18.2% (133/732). Compared with non-FADE cases, FADE cases were older, used more drugs both on admission and at death, and had higher comorbidity (P < 0.001). Drugs suspected to cause or contribute to fatal outcome were mainly those used for treating chronic pulmonary diseases (terbutaline, theophylline), antithrombotic drugs (aspirin, warfarin, heparines) and drugs for treating coronary heart disease and heart failure (e.g. diuretics, nitrates, angiotensin converting enzymes (ACE) inhibitors, calcium channel blockers). Bronchodilatory drugs, antithrombotic drugs and cardiovascular drugs account for 26, 31 and 30 FADE cases, respectively. Patients dying from gastrointestinal diseases had the highest relative FADE occurrence (42%), cancer patients the lowest occurrence (4%). Serious drug-drug and drug-disease interactions were frequently suspected. Various degrees of inappropriateness in choice of drug, dosage or administration route were seen in 50% of FADE cases.

Conclusions: This study shows a high incidence of FADEs associated with high age, high comorbidity and polypharmacy, and partly to inappropriate drug prescribing or use. Treatments frequently associated with FADEs were bronchodilatory treatment of patients with both chronic obstructive lung disease and coronary heart disease, vasodilatory treatment in patient with endstage heart failure and the combination of several antithrombotic drugs. A systematic strategy is needed to avoid unnecessary adverse drug events (ADEs).