RATIONALE, AIMS AND OBJECTIVES: Medication-prescribing errors with adverse drug events impose substantial harms on patients and health systems. Medication errors resulting in preventable adverse drug events most commonly occur at the ordering stage. Electronic prescribing may prevent such errors but its impact has not been rigorously evaluated. METHODS: We conducted a pragmatic cluster randomized controlled trial in academic hospital ambulatory clinics to evaluate the effects of a commercially available electronic prescribing software system on total prescription error ratio. Secondary outcomes included the number of callbacks for clarification from community pharmacies to physicians' clinics. RESULTS: Twenty-six physicians used the electronic prescribing system, writing 1980 prescriptions during 44 intervention weeks when the electronic prescribing system was available (7.6% of these were electronic, the remainder handwritten) and 973 prescriptions during 22 control weeks while the system was switched off (1.4% electronic, prescribed in the previous intervention week, but issued with delay). The total prescription error rate was 118/1980 (6.0%) in intervention weeks and 57/973 (5.9%) in control weeks (P = 0.91). During the intervention period more callbacks requesting clarification were made to clinic administrators (n = 83, 1.89 per week) than during control weeks (n = 32, 1.45 per week; P < 0.001). CONCLUSION: Implementation of the electronic prescribing system had no impact on total prescription error, and increased the callback rate. In spite of intensive user support, few prescriptions in intervention weeks were made using the electronic system. Given the costs, training requirements, workflow redesigns and regulatory hurdles, additional evaluations of outpatient prescribing on clinically important outcomes are needed.
RCT Entities:
RATIONALE, AIMS AND OBJECTIVES: Medication-prescribing errors with adverse drug events impose substantial harms on patients and health systems. Medication errors resulting in preventable adverse drug events most commonly occur at the ordering stage. Electronic prescribing may prevent such errors but its impact has not been rigorously evaluated. METHODS: We conducted a pragmatic cluster randomized controlled trial in academic hospital ambulatory clinics to evaluate the effects of a commercially available electronic prescribing software system on total prescription error ratio. Secondary outcomes included the number of callbacks for clarification from community pharmacies to physicians' clinics. RESULTS: Twenty-six physicians used the electronic prescribing system, writing 1980 prescriptions during 44 intervention weeks when the electronic prescribing system was available (7.6% of these were electronic, the remainder handwritten) and 973 prescriptions during 22 control weeks while the system was switched off (1.4% electronic, prescribed in the previous intervention week, but issued with delay). The total prescription error rate was 118/1980 (6.0%) in intervention weeks and 57/973 (5.9%) in control weeks (P = 0.91). During the intervention period more callbacks requesting clarification were made to clinic administrators (n = 83, 1.89 per week) than during control weeks (n = 32, 1.45 per week; P < 0.001). CONCLUSION: Implementation of the electronic prescribing system had no impact on total prescription error, and increased the callback rate. In spite of intensive user support, few prescriptions in intervention weeks were made using the electronic system. Given the costs, training requirements, workflow redesigns and regulatory hurdles, additional evaluations of outpatient prescribing on clinically important outcomes are needed.