ABSTRACT
OBJECTIVE: To compare the medical costs associated with risk stratification criteria used to evaluate febrile infants 29-90 days of age. STUDY DESIGN: A cost analysis study was conducted evaluating the Boston, Rochester, Philadelphia, Step-by-Step, and PECARN criteria. The percentage of infants considered low risk and rates of missed infections were obtained from published literature. Emergency department costs were estimated from the Centers for Medicare and Medicaid Services. The Health Care Cost and Utilization Project databases were used to estimate the number of infants ages 29-90 days presenting with fever annually and costs for admissions related to missed infections. A probabilistic Markov model with a Dirichlet prior was used to estimate the transition probability distributions for each outcome, and a gamma distribution was used to model costs. A Markov simulation estimated the distribution of expected annual costs per infant and total annual costs. RESULTS: For low-risk infants, the mean cost per infant for the criteria were Rochester: $1050 (IQR $1004-$1092), Philadelphia: $1416 (IQR, $1365-$1465), Boston: $1460 (IQR, $1411-$1506), Step-by-Step $942 (IQR, $899-$981), and PECARN $1004 (IQR, $956-$1050). An estimated 18 522 febrile 1- to 3-month-old infants present annually and estimated total mean costs for their care by criteria were: Rochester, $127.3 million (IQR, $126.1-$128.5); Philadelphia, $129.9 million (IQR, $128.7-$131.1); Boston, $128.7 million (IQR, $127.5-$129.9); Step-by-Step, $ 126.6 million (IQR, $125.4-$127.8); and PECARN, $125.8 million (IQR, $124.6-$127). CONCLUSIONS: The Rochester, Step-by-step, and PECARN criteria are the least costly when evaluating infants 29-90 days of age with a fever.
Subject(s)
Bacterial Infections/diagnosis , Clinical Decision Rules , Emergency Service, Hospital/economics , Fever/etiology , Health Care Costs/statistics & numerical data , Bacterial Infections/complications , Bacterial Infections/economics , Bacterial Infections/therapy , Databases, Factual , Decision Trees , Female , Fever/diagnosis , Fever/economics , Humans , Infant , Infant, Newborn , Male , Markov Chains , Risk Assessment , United StatesABSTRACT
The use of herpes simplex virus (HSV) polymerase chain reaction for diagnosis of HSV disease involving the central nervous system has not translated into widespread use for the detection of DNAemia. We report our 6-year experience using blood polymerase chain reaction testing for HSV infection in neonates and older children with HSV disease.