ABSTRACT
INTRODUCTION: Overutilization of point-of-care (POC) testing may reduce the overall value of care due to high-cost cartridges, need for staff training, and quality assurance requirements. METHODS: The Diagnostic Stewardship group at Cincinnati Children's Hospital Medical Center assembled a multidisciplinary team to reduce the use of POC blood gas testing by 20% in the pediatric intensive care unit (PICU). Key drivers of test overutilization included poor knowledge of cost, concern with testing turnaround time, and a lack of a standard definition of when a POC test was appropriate. We calculated weekly the outcome measure of POC blood gas tests per PICU patient-day and a balancing measure of blood gas result turnaround time using data extracted from the electronic medical record. Interventions focused on staff education, the establishment of a standard practice guideline for the use of POC testing, and improving turnaround time for laboratory blood gas testing. RESULTS: Over the baseline period starting July 2016, a median of 0.94 POC blood gas tests per PICU patient-day was ordered. After initial staff training, the rate was reduced to 0.60 tests per PICU patient-day and further reduced to 0.41 tests per PICU patient-day after a formal policy change was adopted. We have sustained this rate for 15 months through June 2018. Institutional direct cost savings were estimated to be $19,000 per year. CONCLUSIONS: Our improvement initiative was associated with a significant and rapid reduction in the use of POC testing in the PICU. Interventions focused on cost awareness, and a formal guideline helped establish a consensus around appropriate utilization.
ABSTRACT
Multidrug resistance in Gram-negative bacteria has led to a resurgence in colistin use. No pharmacokinetic data exist for burn patients. A 17-year-old boy suffered a 71% TBSA full-thickness burn with deep necrosis and compartment syndrome. He developed multidrug-resistant Acinetobacter baumannii burn wound sepsis/septic shock with acute renal failure requiring dialysis. The organism was resistant to all tested antibiotics except colistin. He received colistin 2.5 mg/kg every 24 hours. Peak and trough serum concentrations, area under the concentration-time curve, and elimination half-lives of colistin were 3.6 ± 1.0 µg/ml, 0.9 ± 0.5 µg/ml, 47.1 ± 14.4 mg · hr/L, and 12.3 ± 9.4 hours (mean ± SD), respectively. Serum levels were at or above the minimum inhibitory concentration for >90% of therapy. Nevertheless, salvage therapy with colistin proved futile as the patient developed acidosis, coagulopathy, and was vasopressor-dependent without any wound healing. He died on hospital day 52. Microbiologically, the serum levels of colistin were seemingly adequate, as repeat cultures were negative. Given the peak and trough levels of colistin relative to the minimum inhibitory concentration for the organism (0.5 µg/ml), it would seem that the dosage of 2.5 mg/kg administered every 24 hours for this patient on dialysis was appropriate. Patients on dialysis infected with an organism possessing a higher inhibitory concentration (≥1 µg/ml) should probably receive the same dosage every 12 hours to avoid subtherapeutic concentrations. Large-scale study of the pharmacokinetics of colistin in patients with burn injury is urgently needed.
