Accuracy of medication information sources compared to the best possible medication history for patients presenting to the emergency department

OBJECTIVE: To assess the accuracy of medication information sources available for adult patients presenting to the ED, compared to a best possible medication history (BPMH). METHODS: This prospective observational study was undertaken in the ED of a major tertiary-referral teaching hospital. A convenience sample of consecutive adult patients taking one or more regular medications was included. A BPMH was ascertained using patient/carer interviews, where available, and confirmed with one or more other sources. For residential care facility (RCF) patients, the RCF medication chart and at least one other source were used. Information sources compared with the BPMH were community pharmacy dispensing history, patient's own medications, patient's medication list, general practitioner letter, medications stored in and labelled on dose administration aids (DAAs) and the RCF chart. Number of discrepancies per patient for each source was determined by comparing medications and dose regimens to those documented in the BPMH. RESULTS: A total of 455 patients (median age 71 years) took a median of six 'regular' and two 'as required' medications. The median number (range) of discrepancies per patient for regular medication names and dosages were RCF chart 0 (0-3), DAA contents 2.0 (0-9), patient's medication list 2.5 (0-16), DAA medications label 3.0 (0-7), community pharmacy history 3.0 (0-19), general practitioner letter 3.0 (0-18) and patient's own medications 4.0 (0-16). Overall, 40.4% of discrepancies were deemed 'moderate' or 'high' clinical significance. Omission errors accounted for 55.6% of discrepancies. CONCLUSIONS: A combination of sources is essential to determine the BPMH. RCF charts provided the most accurate information. Other sources had two to four regular medication-related discrepancies per patient.