The SCRIPT project: a knowledge translation approach to improve prescription practice in a general intensive care unit

BACKGROUND: Medication errors can be dangerous and are an intensive care unit quality indicator. Few studies have assessed interventions aimed at improving prescription practice. Anecdotal reports in our ICU indicated a high incidence of prescription errors, including illegible handwriting, and lack of prescriber identity and antibiotic indications. Knowledge translation (KT) is an emerging tool that uses collaborative stakeholder participation and focuses on education and inclusiveness rather than punitive audits. OBJECTIVE: To evaluate interventions aimed at improving prescription practice. DESIGN: A KT quality assurance project with three phases: (1) measuring baseline performance; (2) education; and (3) measuring post-education performance. SETTING AND PARTICIPANTS: Doctors and nurses working in the 19-bed general ICU of a tertiary referral hospital. INTERVENTIONS: Education protocols were developed by senior doctors, nurses and the pharmacist. "Walking the process" with all doctors and nurses identified local barriers and issues. Ten weeks were allocated for the baseline and post-education audits, and 8 weeks for education. The project was then delayed for a few months to prevent the Hawthorne effect. MAIN OUTCOME MEASURE: Incidence of prescription errors in the post-education phase. RESULTS: There was a highly significant improvement in many error rates - illegible prescriptions reduced from 39.4% to 13.4% (P < 0.001); absent or illegible prescriber name from 38.3% to 27.0% (P < 0.001); not documenting antibiotic indication from 64.8% to 32.5% (P < 0.001). CONCLUSION: A systematic KT process of collaborative education can reduce ICU prescription errors.