Background In the past few years, many hospitals have implemented medication reconciliation procedures (MRP) at hospital admission as a strategy to enhance the safety of medication management in this care transition.

Purpose To analyse the results of a MRP led by a pharmacist in patients at risk at admission by emergency department (ED) in order to develop improvement strategies.

Material and methods The study was performed in a referral area hospital of 330 beds. The MRP at admission to the ED selects patients with a higher risk of reconciliation error (RE) based on criteria established in the literature (pluripathology, polypharmacy, high risk medications, chronic kidney disease, etc), with the direct participation of the emergency pharmacist in every step of the process. The results of its implementation were analysed from its implementation to date (2012, 2013, 2014). To analyse the MRP, the following process indicators were established: coverage ratio (number of patients included in the MRP regarding the total patients in the ED), patients with RE regarding the total of reconciled patients, number of drugs with RE regarding the total of reconciled drugs, average REs per patient, error types (omission/medication error, omission/wrong dose or frequency, and other (duplication, interaction, commission).

Results The overall numbers of patients admitted through the ED were 10 900 in 2012,11 300 in 2013 and 11 500 in 2014, with a coverage ratio of the MRP of 15%,10.2% and 13%, respectively. The number of patients with REs for each year were 63.4%, 80% and 65.1% and the number of drugs with REs were 23%,18.7% and 26.6%, with an average number of REs per patient of 2.45, 2.6 and 2.5. Regarding the evolution of the different types of error over the 3 years, the majority were omission/medication errors, increasing in proportion over time (66.7%, 72.3%, 83%), followed by omission/wrong dose or frequency, that remained similar over time (20%, 20.5%, 22.9%). Other types of error tended to decrease (13.3%, 7.2%, 2.9%).

Conclusion Although pharmaceutical intervention manages to avoid a large number of REs, the prevalence of patients with errors and of REs has not diminished over time but remains very high, even tending to increase, suggesting that for improvements in these indicators we should target the improvement plan towards the training of prescribers in medication reconciliation, a strategy that would also allow an increase in the number of patients in whom such errors are avoided.

No conflict of interest.