Should nurses or clinical pharmacists perform medication reconciliation? A randomized controlled trial

Purpose To study differences in outcomes of medication reconciliation (MR) when performed by clinical pharmacists compared to nurses. Methods 201 patients (21-92 years) admitted to the Department of Cardiology at the University Hospital of North Norway, autumn 2012, were randomized into a pharmacist group (PG) and a nurse group (NG). The nurses and the pharmacists were trained for performing the MR process by an independent clinical pharmacist. Medication discrepancies (MDs) were discussed with the physicians. Time spent during the MR was recorded. An independent expert group rated clinical relevance of the MDs retrospectively. Results At least one MD was identified in 78 % and 84 % of patients in PG and NG, respectively (P=0.269) with a mean number of MDs per patient 3.1 (SD 2.1) and 2.8 (SD 2.2), respectively (P = 0.528). Mean time spent/patient on the MR process was 22.9 min (SD 11.6) in the PG and 32.2 min (SD 20.3) in the NG (P < 0,001). Physicians agreed significantly more often to act upon the MDs presented by pharmacists compared to nurses (P=0.001). The expert group finally assessed 48 % and 49 % of the MDs to be of the clinical relevance in the PG and the NG, respectively. Conclusions By applying a structured method for MR, a small however not statistically significant difference in identified MDs between nurses and clinical pharmacists was revealed. The pharmacists spent significantly less time than the nurses, and physicians agreed significantly more often with the pharmacist that action should be taken on the MDs. This is important in the discussion of who to perform MR.