Background Medicines Reconciliation on Admission (MRA) aims to identify and solve Unintended Medicines Discrepancies (UMDs) defined as differences between the home treatment prescription and the first hospital prescription.

Purpose To assess the impact of MRA on UMDs and to identify risk factors for the development of UMDs.

Material and methods This retrospective study was conducted in six services: vascular surgery, geriatrics, haematology, infectious diseases, nephrology and urology. Pharmacy students, who were supervised by clinical pharmacists, performed MRA within 48 h of the patient’s admission. Students asked the patients about the treatment prescribed at home, self-medication, allergies, adverse events and therapeutic adherence. Students called community pharmacists to get information on the drugs dispensed in the last 3 months and collected further information from other sources: the patient’s medical record, an interview with the family, letters from community doctors. Two pharmacists retrospectively assessed the potential harm of each UMD (high, moderate or minor risk). Correlations between the parameters were tested statistically.

Results In 2013, 645 patients were enrolled (5,673 medicines analysed). 23.4% of patients had at least one UMD and 10.1% of patients had at least three UMDs. Of the UMDs detected, 47.3% were associated with a moderate to major risk for the patient. The number of lines on the home treatment prescription was a risk factor for UMD, as was the type of service in which the activity took place (R = 0.1918; p < 0.0001). Medicines adherence did not correlate with the incidence of UMDs.

Conclusion MRA is a useful process for detecting UMDs. Pharmacy students sometimes have trouble identifying UMDs, so it is necessary to supervise them. MRA will develop in parallel with the use of new information technologies for communications. It plays a role in structuring the “community-hospital” link.

References and/or Acknowledgements I would like to thank all those who worked on this project.

No conflict of interest.