Critically ill children often require continuous intravenous infusions of life-supporting medications. The complexity of ordering such infusions makes this an error prone process, and such errors can result in serious adverse events. A CPOE system was developed and evaluated to assess its impact on the safety and efficiency of prescribing continuous medication infusions.

Pediatric and neonatal ICU patients are particularly vulnerable to medication errors.¹ Delivering continuous medication infusions introduces a risk for errors because of the use of hand calculation and the need for frequent dose titrations. To improve safety of continuous drug infusions, the Joint Commission on Accreditation for Hospital Organizations (JCAHO) has mandated the use of standardized concentrations in pediatric patients. We achieved this mandate by developing a CPOE for use in a pediatric intensive care unit (PICU). The effect of a CPOE system on decreasing errors of continuous drug infusions in a PICU has not been previously studied.

To compare the safety and efficiency of a custom-designed CPOE system with a handwritten, hand-calculated method for prescribing continuous drug infusions for pediatric ICU patients.

A CPOE system was designed specifically for ordering continuous drug infusions for pediatric ICU patients. The system generates 2 to 4 standardized concentrations and selects the ideal concentration based on a given patient weight and daily fluid intake. It eliminates the need for complex hand calculations and has additional clinical decision support to enhance safety. Using a simulated test environment, resident physicians were asked to generate orders for 9 commonly used continuous drug infusions. In a crossover study design, 26 volunteer physicians ordered drips using both the handwritten method and the CPOE method. Time to completion was recorded. Order sheets were analyzed for errors according to a predetermined classification system. User satisfaction was measured using a web-based questionnaire with a 5-point Likert type scale at the end of the testing session.

A total of 234 orders were generated using each method by 26 physicians ordering 9 drips each. The CPOE method required significantly less time (5.5 minutes ± 2 minutes) as compared to the handwritten method (26 minutes ± 8 minutes), p= .0001. In addition, the CPOE method had significantly less errors, 10 of 243 drip orders (4.3%), as compared to the handwritten method where 170 of 243 drip orders (73%) contained one or more errors, p=.0001. Amongst the handwritten errors, 25% were judged to be ‘high-risk’ with the potential for serious adverse effects, while none of the errors in the CPOE group were high-risk, all 10 being missing signatures. The 170 handwritten drip orders with errors contained a total of 333 distinct erroneous elements, of which 54% were missing data, 16% were calculation errors, 15% were exceeding the maximum drug concentration, and 15% were classified as ‘other’ errors. All 26 physicians committed at least one or more errors when using the handwritten method. User-satisfaction survey revealed that physicians were significantly more satisfied with the CPOE method than the handwritten method, p= .0001.

A custom-built CPOE system for ordering continuous drug infusion in pediatric patients can improve the safety and efficiency of this high-risk process. In addition, a well designed system will enhance user-acceptance and rapid adoption by the physicians.