Impact of electronic chemotherapy order forms on prescribing errors at an urban medical center: results from an interrupted time-series analysis

OBJECTIVE: To evaluate the impact of electronic standardized chemotherapy templates on incidence and types of prescribing errors. DESIGN: A quasi-experimental interrupted time series with segmented regression. SETTING: A 700-bed multidisciplinary tertiary care hospital with an ambulatory cancer center. PARTICIPANTS: A multidisciplinary team including oncology physicians, nurses, pharmacists and information technologists. INTERVENTION(S): Standardized, regimen-specific, chemotherapy prescribing forms were developed and implemented over a 32-month period. MAIN OUTCOME MEASURE(S): Trend of monthly prevented prescribing errors per 1000 chemotherapy doses during the pre-implementation phase (30 months), immediate change in the error rate from pre-implementation to implementation and trend of errors during the implementation phase. Errors were analyzed according to their types: errors in communication or transcription, errors in dosing calculation and errors in regimen frequency or treatment duration. Relative risk (RR) of errors in the post-implementation phase (28 months) compared with the pre-implementation phase was computed with 95% confidence interval (CI). RESULTS: Baseline monthly error rate was stable with 16.7 prevented errors per 1000 chemotherapy doses. A 30% reduction in prescribing errors was observed with initiating the intervention. With implementation, a negative change in the slope of prescribing errors was observed (coefficient = -0.338; 95% CI: -0.612 to -0.064). The estimated RR of transcription errors was 0.74; 95% CI (0.59-0.92). The estimated RR of dosing calculation errors was 0.06; 95% CI (0.03-0.10). The estimated RR of chemotherapy frequency/duration errors was 0.51; 95% CI (0.42-0.62). CONCLUSIONS: Implementing standardized chemotherapy-prescribing templates significantly reduced all types of prescribing errors and improved chemotherapy safety.