|| Checking for direct PDF access through Ovid
First, the aim of the study was to assess the prevalence, characteristics, and severity of unintended medication discrepancies (UMDs) and medication errors (MEs) at admission and discharge of hospitalization. Second, the aim of the study was to identify clinical and hospitalization factors associated with risk of UMDs as well as characteristics of the medication reconciliation process associated with UMDs detection.This prospective observational study included all adult patients admitted from 2013 to 2015 in the Endocrinology-Diabetology-Nutrition Department of Montpellier Hospital, France. Clinical pharmacists conducted medication reconciliation by collecting the best possible medication history from different sources and comparing it with admission and discharge prescriptions to identify discrepancies. Unintended medication discrepancies corrected by the physician were considered as MEs. Risk factors of UMDs were identified with logistic regression.Of 904 patients included, 266 (29.4%) had at least one UMD, at admission or at discharge. In total, 378 (98.2%) of 385 UMDs were considered to be MEs. Most MEs were omissions (59.3%). Medication errors were serious or very serious in 36% of patients and had potentially moderate severity in almost 40% of patients. The risk of UMDs increased constantly with the number of treatments (P < 0.001). Thyroid (adjusted odds ratio [OR] = 1.79, 95% CI = 1.12–2.86) and infectious diseases (adjusted OR = 1.80, 95% CI = 1.17–2.78) were associated with UMDs risk at admission. The best type of source for the detection of UMDs was the general practitioner or nurse (OR = 2.64, 95% CI = 1.51–4.63).Unintended medication discrepancies are frequent at hospital and depend on intrinsic clinical parameters but also on practice of medication reconciliation process, such as number and type of sources used.