Accuracy of information on medicines in hospital discharge summaries.

BACKGROUND: At Auckland Hospital, patients have been given handwritten summaries on discharge from hospital with a copy posted to the general practitioner. A typed summary is often not completed so the list of medicines on the handwritten summary needs to be accurate and complete.

METHODS: We selected 100 patient charts from the general medical service and 100 charts from the general surgical service and recorded the medicines on admission, on the inpatient medication chart and on the discharge summary. We noted errors in the recording of medicines on the discharge summary and rated the severity of errors.

RESULTS: Surgical patients were discharged on 7.88 (95% confidence interval (CI) 7.40-8.64) medicines and medical patients on 8.58 (95% CI 7.87-9.29) medicines. During the admission there were 0.59 (95% CI 0.38-0.80) changes to the medicines for the surgical patients and 1.70 (95% CI 1.39-2.01) for the medical patients (P < 0.0001). There were 0.81 (95% CI 0.65-1.02) errors per surgical discharge summary and 1.42 (95% CI 1.20-1.67) errors per medical summary (P = 0.006). Four errors were graded as having the potential to cause readmission to hospital, 24 as potentially serious, 83 as potentially troublesome and 111 as minor.

DISCUSSION: Error rates were high and although the majority were minor, a number of them had the potential to cause serious consequences. There were more medication changes in the medical patients and this may contribute to higher error rates in this group. There is a need to improve the accuracy of recording medicines on discharge summaries. Strategies to improve this problem are discussed.