100 Effect of a Brief Educational Intervention on Prescription Writing by Pediatric Residents and Hospitalists

To measure the effect of a brief educational intervention on medication ordering by pediatric residents and hospitalists. Currently, at the Alberta Children's Hospital (ACH) medication orders are handwritten in each patient's chart. Such orders are commonly incomplete, unclear, and contain abbreviations with multiple possible meanings. Incorrect prescription writing is a significant source of medication error and a threat to patient safety (Bates, 1999; Kaushal et al., 2001).

Medication ordering by pediatric residents and hospitalists was investigated over several weeks before and after a seminar on prescribing errors (Powerpoint^® presentation, 30 minutes in length). A sequential chart review (n=21) measured frequency of ordering errors: defined as an order which was illegible, incomplete, contained unclear abbreviations, or whose dosage was calculated incorrectly (from mg/kg or mg/m²). Prescribing drugs to which the patient had a known allergy was also counted as an ordering error. As drug-drug interactions were not readily apparent on the order sheet, prescribing combinations of drugs with known interactions was not considered. Orders were considered individually according to guidelines published by the Institute for Safe Medication Practices (ISMP).

Error frequency (total errors/total orders) decreased 27.5% (64.8% to 37.3%) following the presentation. The two phases had slight differences in: the number of orders reviewed (444 versus 314), orders per chart (22.2 versus 17.6), and standard deviation in error frequency (19.5% versus 22.3%). Though the number of orders per chart varied, influenced by length of hospitalization as well as number of medications prescribed, it should not have influenced error frequency. Overall, the two samples were similar enough to allow comparison and the difference in error frequency was found to be statistically significant using the t-test (p<0.01).

Error frequency decreased from the first review to the second (64.9% to 37.3%) and was statistically significant (p<0.01). Currently, evidence suggests a brief educational intervention (t=30min) can alter prescribing behaviour and lower error frequency for orders written by hospitalists and residents. Whether this will reduce medication errors is unknown, though it is reasonable to infer.