In this study, we showed that the incidence of medication errors related to pre-admission medication which occurred during hospitalization was reduced after the initiation of pharmacist-led MR for pre-admission patients. In the process of pre-admission MR, healthcare professionals collect BPMH with the involvement of patients, carers, primary care physician and/or pharmacies, and adapt it to prescriptions during hospitalization. Medication errors often occur at transitions of care due to discrepancies in medication information [5, 12,13,14]. In this regard, several studies have evaluated pharmacists’ clinical interventions in hospitalized patients [15,16,17]. We previously reported the benefits of pharmacist intervention on adverse events in hospitalized patients in the otolaryngology ward and in the respiratory medicine ward [18, 19]. Here, we evaluated the impact of pharmacist-led MR for pre-admission patients on the incidence of medication errors in the early post-admission period.

In this study, 94.2% of inpatients in the post-initiation group received pharmacist-led MR prior to admission. The remaining patients did not have the opportunity to visit an outpatient clinic before admission due to emergency admission or transfer from another hospital, and received MR after admission. After initiation of the intervention, physicians were able to increase the number of prescriptions which were alternatives to the pre-admission medication and which were required during hospitalization the day before admission, based on the medication history prepared by the pharmacist. As a result, the efficiency of post-hospitalization procedures, such as nurse receipt of physicians’ orders and preparation of medicines in the pharmacy department, was considered to have improved. The incidence of medication errors decreased significantly after the initiation of pharmacist-led MR for pre-admission patients, despite an increase in the number of inpatients. In particular, the number of errors related to administration by healthcare professionals decreased. Pharmacist-led MR prior to admission was a significant protective factor against incidents related to pre-admission medication. In the present study, it was not necessarily the case that pharmacist-led MR directly influenced prevention of medication errors or change of prescriptions. However, medication errors analyzed in this study were related to medications prescribed pre-admission only, and occurred in the early phase of hospitalization. It was considered that pharmacist-led MR for pre-admission patients indirectly contributed to a reduction in medication errors via the acceleration of tasks related to medicines during hospitalization.

The primary outcome of this study was the number of medication errors related to pre-admission medications that occurred within five days of admission. In an exploratory case study of 30 patients, Frydenberg et al. reported that the majority of medication errors occurred on admission, and that half of these were due to an incomplete medication list on the referral letter for hospitalization [20]. In a retrospective observational study, Dei Tos et al. reported the identification of unintentional medication discrepancies on admission in 53 of 144 patients [21]. Accordingly, transitions in care – such as admission – are associated with a risk of medication error and adverse events. Several reports have investigated the effect of MR performed early in hospitalization. In a study of older patients, Mazhar et al. reported that the most significant predictors of unintentional medication error were the number of medications prescribed on admission (OR 1.32, 95% CI 1.09–1.54, p < 0.034), number of sources consulted to obtain a better medication history (OR 1.53, 95% CI 1.38–1.76, p < 0.01), and completion of a medication history within 24 h of admission (OR 0.89, 95% CI 0.86–0.94, p < 0.03) [21].

Ouweini et al. reported the impact of pharmacist-led MR within 48 h of orthopedic admission for surgical treatment [22], and found that 84.5% of interventions based on pharmacist-led MR were accepted by surgical residents and fellows. In contrast, in a prospective cohort study of adult patients admitted to the emergency department, pharmacist-led MR was conducted prior to the preparation of a physician’s admission order. The findings indicated that serious errors occurred at similar proportions in the intervention and control groups [23]. Trends in clinical benefit were inconsistent across these reports. Possible reasons for this include differences in departments, age groups and study design. In a meta-analysis study evaluating the impact of MR at transitions of care, Redmond et al. reported that the implemented interventions reduced the number of medication discrepancies at transitions of care. However, they also noted that the quality of the evidence was very low [24]. Studies about MR have reported the effects of different interventions, practice processes and practice systems. Our present study is the first to evaluate the impact of pharmacist-led MR on pre-admission patients. Pharmacist-led MR in pre-admission patients contributed to a high proportion of acceptance by physicians of recommendations regarding antithrombotic medicine interruption and patient safety after admission.

This study has several limitations. First, it was conducted under a retrospective design. Factors that may affect work efficiency and patient safety after admission, such as differences in staffing structure before and after the initiation of the intervention, were not fully considered. There were no changes in the fixed number of physicians, nurses, or pharmacists on the orthopedic ward during the study period. In particular, the number of pharmacists assigned to the ward remained constant at two throughout the period. However, it was difficult to account for detailed transfers of physicians and nurses during the study period. Second, changes in the time taken to obtain medication histories were not reflected in the assessment of post-admission work efficiency or patient safety. Third, potential medication errors were not detected because the medication errors analysis was based on spontaneous reports from healthcare professionals. In general, the proportion of medication errors reported in previous reports varies depending on how the errors are detected [13].