Abstract
PURPOSE: The study aimed to test the hypothesis that hospital electronic prescribing and medicine administration system (HEPMA) implementation impacted patient discharge letter quality, nature and frequency of prescribing errors. METHOD: A quasi experimental before and after retrospective case note review was conducted in one United Kingdom district general hospital. The total sample size was 318 (random samples of 159 before and after implementation), calculated to achieve a 10% error reduction with a power of 80% and p < 0.05. Adult patients discharged after ≥24-h inpatient stay were assessed for discharge information documentation quality using a modified validated discharge document template. Prescribing errors were classified as medicine omissions, commissions, incorrect dose/frequency/duration, drug interactions, therapeutic duplications or missing/inaccurate allergy information. Post-implementation assessments were undertaken 4 months following HEPMA implementation. Error severity was determined by a multidisciplinary panel consensus using the Medications at Transitions and Clinical Handoffs (MATCH) study validated scoring system. RESULTS: There were no statistically significant differences in patient demographics between the pre- and post-implementation groups. Discharge information documentation quality improved; allergy documentation increased from 11 to 159/159 (p < 0.0001). The number of patients with prescribing errors reduced significantly from 158 to 37/159 (p < 0.001). Prescribing error category incidence identified in pre-implementation patients was reduced (e.g. omission incidence from 66 to 18/159 (p < 0.001)), although a new error type (sociotechnical [errors caused by the system]) was identified post-implementation (n = 8 patients). Post-implementation prescribing errors severity rating identified 8/37 as likely to cause potential patient harm. CONCLUSION: HEPMA implementation was associated with improved discharge documentation quality, statistically significant prescribing error reduction and prescribing error type alteration. There remains a need to be alert for potential prescribing errors.