Background According to the WHO, 50% of medication errors (ME) occur on admission or discharge from hospital. At his arrival in the emergency department, the patient does not provide much information on his treatment. Medication reconciliation (MR) can provide solutions to this issue.

Purpose To estimate the severity of potential consequences of ME detected by MR in an emergency department and follow the correction rate for ME.

Material and methods A prospective observational study was conducted. Retroactive MR on admission was performed in an 8 bed ward over 30 weeks. Inclusion criteria were, first, age (65 years and older) and second, a planned transfer to a medical or surgical department. MR was performed by a pharmacist, using different sources of information (electronic medical file, patient’s prescriptions, community pharmacist, etc). Each unintentional discrepancy was recorded as ME. Data were analysed with 2 software programmes (Excel (Microsoft) and Epi-Info (CDC, Atlanta)). ME severity was assessed at monthly multidisciplinary meetings (physicians and pharmacists). The rating (minor, significant, major, critical, catastrophic) was based on national recommendations. ME follow-up (corrected or not) was performed.

Results 129 patients were included (41% men, 59% women). Mean age was 81 years. On average, 2.6 sources of information per patient were used. 167 ME were detected. At least 1 ME occurred in 61 patients. The most prevalent ME was omission (59%). 72% of ME were assessed as minor, 21% as significant, 7% as major and 0% as critical or catastrophic. Physicians and pharmacists who evaluated ME severity were also those involved in patient care. Thus the ME severity may have been underestimated. The rates of ME corrected by physicians was, respectively, 62% of minor, 77% of significant and 91% of major. The more severe the ME , the more it was corrected after MR (p<0.05).

Conclusion This study shows that the severity of potential consequences of ME detected by MR is more often minor. It reveals that every ME detected with MR on admission was not systematically corrected by physicians but the correction rate increased with the severity of potential harm.

No conflict of interest