Background Medicines reconciliation is known to minimise medicines errors and reduce morbidity in hospitalised patients. Its role in the Intensive Care Unit (ICU) has not been widely studied.

Purpose To analyse the number, type, and importance of pharmacist contributions in the ICU of a tertiary level hospital.

Material and methods Prospective study conducted in July 2014 to assess the accuracy of in-patient prescription charts. Drug history and medicines reconciliation were undertaken within the first 24 h by a specialist pharmacist. The patient’s own medicines, information from relatives or carers and community pharmacy records, were used as sources of information. Discrepancies between the patient’s regular medicines and prescribed medicines were conveyed to the medical team. Discrepancies were considered medicines errors when they required further intervention by the responsible doctor.

Results 48 patients were included, percentage of females 39.6%, mean age 62 ± 14.04. Average number of regular medicines per patient: 5.41 ± 3.5. Discrepancies were found in 62.5% of patients, of which 79.9% required pharmacy intervention (the rest of them were obviated due to the patient’s clinical condition – mostly inability to swallow). These involved 98 out of 260 drugs prescribed. Of them, the contribution was accepted in 51.3% of cases (40 prescriptions), whereas in the remainder 48.7% of cases, the changes were intentional. According to their pharmacological class, the highest number of contributions was found in anti-hypertensives (9 contributions), followed by statins (6) and diuretics (6). In relation with the type of medicine involved, a total of 33 medicines errors were accountable as omissions, whilst wrong directions were found in 4 cases, and wrong dose occurred in 3 cases.

Conclusion Medicines errors in the ICU have a similar incidence to those in other non-acute clinical settings. This study is in line with previous publications, suggesting that the ICU might benefit from the regular input from a pharmacist, which in turn would result in a reduction of medicines errors.

References and/or acknowledgements No conflict of interest.