COMPUTERIZED PN ORDERING OPTIMIZES TIMELY NUTRITION THERAPY IN A NEONATAL INTENSIVE-CARE UNIT

Objective To determine whether automation could accelerate the parente ral nutrition (PN) ordering and delivery process with concurrent impro vements in the quality of nutrition therapy. Design The time required to order, process, and deliver PN orders and specific nutrient composi tion of the PN solution were collected prospectively for 2 weeks on al l neonatal intensive care unit (NICU) patients receiving PN during bot h the manual phase (before automation) and computer phase of the study . Subjects/Setting A total of 81 newborn infants in the NICU receiving PN for more than 5 days completed the study. Statistical analyses Stu dent's unpaired t test was used to evaluate differences between comput er and manual methods for all outcome variables of interest. Results T he time required to write and deliver PN orders was significantly lowe r using computer rather than manual methods (1.4+/-0.2 vs 4.5+/-0.5 mi nutes; P=.0001). Significant improvements in the nutrient composition of the PN solution resulted from use of computer ordering for energy ( 93.4+/-1.48 vs 79.2+/-1.8 kcal/kg per day; P=.0001), protein (2.92+/-0 .02, vs 2.7+/-0.03 g protein per kilogram per day; P=.0001), calcium ( 2.3+/-0.1 vs 1.8+/-0.1 mEq/kg per day; P=.0005), and phosphate (1.3+/- 0.06 vs 0.9+/-0.06 mM/kg per day; P=.0001). Ln addition, alkaline phos phatase levels improved (272+/-11 vs 404+/-25 U/L: P=.0001) and calori c and protein goals were achieved sooner (5.9+/-0.4 vs 8.7+/-0.8 days; P=.0045) when computer ordering rather than the manual method of orde ring PN was used. Implications Our findings indicate that automating t he process of writing and delivering PN orders saved time because it e liminated repetitive tasks and tedious calculations previously require d of neonatologists, dietitians, and pharmacists. Patient care in our population of neonates nas enhanced by improving the nutrient content of the PN solution.