LAUSR

Learning Objectives: Little is known about the medication profiles of critically ill children. Our aim was to describe the medications prescribed to these children at a national level. Methods: Using the Health Facts® (Cerner Corporation, Kansas City, MO) database, initial pediatric intensive care unit (PICU) encounters from 1/2009 to 6/2016 for patients >0 hours to 22 years were analyzed for demographic, clinical, medication, and outcome data. Medication data included generic and brand names, national drug code, and pharmacy-based dispensed and stop times. 96% of the Health Facts® drug codes were classified by Multum, a pharmaceutical industry standard for medication classification. The remainder were classified by two physicians. IV carrier solutions, including saline, were excluded. Data are presented as median (IQR). Results: 66,443 PICU encounters from 160 hospitals were evaluated. Median age was 1.3 years (0–14.5). Median ICU length of stay was 1.5 days (0.5–3.7). Hospital mortality was 2.7%. The most common individual medications (% encounters, median duration [IQR]) were acetaminophen (33.3%, 35 hrs [12–86]), parenteral nutrition solutions (including D10W) (24.4%, 29 hrs [16–62]), ampicillin (20.3%, 52 hrs [37–109]), and morphine (19.8%, 32 hrs [7–78]). The percent of encounters prescribed medication categories commonly associated with PICUs included antibiotics (27.8%); anticonvulsants (22.4%); glucocorticoids (17.2%); and vasopressors and/or inotropes (17.2%, of which 42.5% were epinephrine, 26.0% dopamine, and 5.5% norepinephrine). Children receiving invasive or non-invasive mechanical ventilation (n=10,851) were prescribed sedation (89.4%, median duration 29 hrs [11–97] of which 83.9% were benzodiazepines, 29.1% propofol, 11.4% barbiturates, 9.5% dexmedetomidine), narcotic analgesia (100%, 41 hrs [9–111] of which 63% were fentanyl, 60.7% morphine, 12.1% hydromorphone), and neuromuscular blockade (50.1%, 2 hrs [1–15] of which 50.4% were rocuronium, 16.6% succinylcholine, 7.6% pancuronium). Conclusions: These data present a unique opportunity to examine medication profiles in PICU patients. Provider preferences often dictate these patients’ pharmacological management.