Background To improve drug treatment in an emergency unit, we tested Medicines Reconciliation (MR) based on the WHO High5 Medication Reconciliation program method.

Purpose To assess the contribution of MR and to measure the pharmaceutical time required for this activity.

Material and methods This study lasted from June to October 2014. Patients had to meet the following eligibility criteria: admission to Emergency unit followed by admission to hospital, aged between 65 and 90, poly-medicated, with their medicines presenting a risk. Discussions with the patient (or his family if he could not be questioned), his pharmacist, his GP are all information sources (IS) enabled the pharmacist to write an Optimised Medicines Appraisal (OMA). The MR compared the OMA to the first hospital prescription. The prescriber categorised differences found as intentional (ID) or non-intentional (NID) differences, which equate to medicines errors (ME). NID were categorised according to potential danger (minor, significant, major, critical) and type (oversight, duplication, dose error, dosage error, computer error). Time spent at each stage was measured in minutes.

Results Amongst eligible patients, 1 to 2 (n = 44; 8.76%) were randomly selected daily. Average population age was 79 years (M/F = 1.38) with 9.6 ± 2.8 (median = 9) medicines prior to admission. At least one difference (ID + NID) was found in all patients (44/44). Amongst them, 29 (66%) presented at least one NID: 19/29 with 1 NID, 8/29 with 2 NID and 2/29 with 3 NID. Total NID detected was 39: major 23/39 (59%), significant 8/29 (20.5%) and minor 8/39 (20%). Moreover, 74.4% (29/39) of errors consisted of a medicines oversight, 20.5% (8/39) dosage errors, 5.1% (2/39) dose errors. MR took 36.6 min ± 12 min (median = 36) per patient. Time necessary to avoid 1 major ME was 65 min.

Conclusion Finding more than half of patients with major NID demonstrated the benefits of MR. We have to find the resources to establish MRs in the Emergency unit.

References and/or acknowledgements No conflict of interest.