Detection of adverse drug reactions in a neurological department: comparison between intensified surveillance and a computer-assisted approach.

Adverse drug reactions (ADRs) leading to hospitalisation or occurring during hospital stay contribute significantly to patient morbidity and mortality as well as representing an additional cost for healthcare systems. The aim of this study was to provide data about the type and incidence of ADRs in a neurological department and to compare two different methodological approaches to collecting information on ADRs. The two methods used were intensified surveillance of neurological wards by daily ward rounds and computer-assisted screening for ADRs by means of pathological laboratory parameters. Of admissions to the neurological department, 2.7% were caused by an ADR and 18.7% of patients experienced at least one ADR during hospitalisation. The positive predictive values of pathological laboratory parameters ranged between 0% (eosinophil count) and 100% for increased drug serum concentrations and international normalised ratio, i.e. the latter were always accompanied by a clinically relevant ADR. However, only half of all ADR could be detected by pathological laboratory parameters, the sensitivity of this method came to 45.1% with a specificity of 78.9%. Similar to departments of internal medicine, a high number of ADRs occur on neurological wards. The predominant ADRs were those typical of neurotropic medications such as dyskinesia and increased sedation. Due to the age of the patients involved, cardiovascular co-medication is often prescribed and represents an additional risk factor for ADRs. By measuring pathological laboratory parameters the majority of ADRs could not be detected in neurological patients.