The accuracy of emergency department medication history as determined by mass spectrometry analysis of urine: a pilot study.

Background: Emergency Department (ED) medication lists (ML) are considered inaccurate based on previous comparisons of ED ML with patients' self-reporting of medications and reviews of patients' pharmacy and medical records.

Objective: To determine the accuracy of ED ML using mass spectrometry analysis of urine samples.

Methods: This was a prospective observational study conducted at an urban tertiary care university hospital. Convenience sampling of patients who underwent ED triage was done. Included were patients 18 years or older who were capable of providing informed consent and who reported use of at least one medication. Excluded were patients unable to consent, prisoners, non-English-speaking patients, and patients unwilling or unable to provide a urine sample. Mass spectrometry analysis was performed on enrolled patients' urine, and their ED triage ML were recorded. Urinalysis results were compared to ED triage ML. Concordance between respective ED triage ML and urinalysis results was determined. Medications were grouped by medication class. The top five discrepant medication classes were identified.

Results: There were 100 patients enrolled; 21 patients, although eligible, did not provide a urine sample and were excluded, and one patient withdrew. Mean age was 51 years, and 54 patients were male. Twenty-two medication classes were identified. No patient had 100% concordance of ED triage ML and urinalysis results. Opioid analgesic, sedative hypnotics, cardiac, psychiatric, and nonopioid analgesic medications were the top discrepant medication classes.

Conclusions: ED triage ML obtained by patient recall are inaccurate when compared to medications detected in urine using mass spectrometry analysis.