| Title | To What Extent Do Pediatricians Accept Computer-Based Dosing Suggestions? |
| Author(s) | Brigid K. Killelea, MD, MPH, Rainu Kaushal, MD, MPH, Mary Cooper, MD, JD and Gilad J. Kuperman, MD, PhD |
| Source | Pediatrics, Vol. 119, No. 1, Pages 69-75 |
| Publication Date | Jan-07 |
| Abstract | OBJECTIVE. Pediatric medication errors occur frequently among hospitalized patients and are often related to dosing. Computerized physician order entry systems with decision support can decrease dosing errors, as well as other types of errors; however, their use in pediatrics has not been extensively studied. Our objective was to determine physician acceptance of dosing and frequency decision support elements in an inpatient pediatric computerized physician order entry system at 1 academic medical center. PATIENTS AND METHODS. We performed a retrospective analysis of all electronic medication orders entered for pediatric inpatients at a large, urban teaching hospital between April 15, 2004, and December 31, 2004. Rates of physician acceptance of computerized physician order entry system–generated dosing and frequency suggestions were determined. RESULTS. We analyzed 54413 orders in the computerized physician order entry system, of which 27313 orders had dosing or frequency decision support. Of the orders with decision support, approximately one third (8822) were accepted exactly by prescribers. Of the 18491 remaining orders, 8708 were changed for dose, 2466 for frequency, and 7317 for both. Among the 18491 orders that were changed, the majority 11322 deviated by a substantial amount (>50%) from the total daily dose initially suggested by the decision support feature. Overall, patient weight was missing 31.3% of the time, although patient age alone sometimes was sufficient for the computer to make a dosing suggestion. CONCLUSIONS. Although dosing-decision support systems have the potential to improve care, more work needs to be done to determine and optimize their effectiveness. Commercial vendors of dosing knowledge bases need to deliver effective products, because most health care organizations will not have the resources to customize decision support rules. |