Purpose: The intake of medications (drugs) without the knowledge of the treating physician (unknown co-medication) and nonadherence strongly influence drug safety. The aim of our study was to objectively assess unknown co-medication and nonadherence in hospitalized patients by screening urine for a large number of drugs using highly sensitive full scan gas chromatograpy/mass spectrometry (GC/MS). Secondary objectives were to determine the relationship of co-medication and nonadherence to the number of drugs prescribed and to compare history-taking by a pharmacist versus a physician.
Methods: In 152 patients, the drug histories taken by physicians, patients' self-reported adherence, and information compiled during as many as three structured interviews conducted by a trained pharmacist on days 1-2, 3-4, and 7-11 of the hospital stay were compared with the GC/MS results from urine samples collected after each interview.
Results: In the interviews performed by the pharmacist, 235 additional drugs were identified that were not documented in the chart. Of all the drugs indicated in any interview, 16.9% were identified only by the physician, 24.1% only by the pharmacist, and 59% by both. Overall, in 78% of the patients at least one additional drug was identified by urine screening. The findings suggest overall nonadherence to at least one drug in 13.0% of patients on admission and in 23.3% of patients at any time during hospitalization. Nonadherence was less frequent for critical dose drugs and correlated with the number of prescribed drugs.
Conclusions: The drug history among hospitalized patients is often incomplete, and nonadherence and unknown co-medication are alarmingly frequent. This lack of knowledge might impact the overall success of drug therapies in the hospital setting.