RESUMEN
Background Prescribing discharge medications is a potential "next step" for pharmacists in Australian hospitals, however, safety must be demonstrated via a randomised controlled study. Objective To determine if a collaborative, pharmacist led discharge prescribing model results in less patients with medication errors than conventional prescribing for both handwritten and digital prescriptions. Setting Geriatric Medical ward in a quaternary hospital, Australia Methods A prospective, single-blinded randomised controlled study of patients randomised to conventional (control) or a pharmacist-led prescribing (intervention) arms at discharge from hospital. This study had 2 phases; (1) handwritten prescribing and (2) digital prescribing. In addition, the two prescribing methods were compared. Main outcome measures The primary outcome was the percentage of patients with a medication error on their discharge prescription. Results In phase 1, 45 patients were recruited; 21 (control) and 24 (intervention). 95% of control patients and 29% in the intervention arm had at least one medication error, p < 0.0002, relative risk (RR) 0.31, confidence interval (CI) 0.16-0.58. The number of items with at least 1 error reduced from 69 to 4%; p < 0.0001, RR 0.06, CI 0.03-0.11 and fewer items had at least 1 clinically significant error (11% vs 2%, p = 0.0004, RR 0.15, CI 0.04-0.30). In phase 2, 39 patients were recruited; 18 (control) and 21 (intervention). 100% of control patients and 62% in the intervention arm had at least one medication error (p = 0.005, RR 0.62, CI 0.44-0.87). Items with at least 1 error decreased from 21 to 7% (p < 0.0001, RR 0.34, CI 0.44-0.56), there were fewer items with at least 1 clinically significant error (13% vs 5%, p < 0.003, RR 0.4, CI 0.22-0.72). There was no significant change in the primary outcome between handwritten and digital (60% vs 79%, p < 0.055). Conclusion In a geriatric setting, pharmacist-led partnered discharge prescribing results in significantly less patients with medication errors than the conventional method for both handwritten and digital methods.