Management guidelines for refractory Kawasaki disease (KD) are vague. We sought to assess practice variation and identify factors associated with large/complex coronary artery aneurysms (LCAA) and resource utilization in refractory KD. This retrospective cohort study identified patients aged ≤18 years with KD (2004 to 2014) using the Pediatric Health Information System. Refractory KD was defined as receiving >1 dose of intravenous immunoglobulin. Demographics, medications, concomitant infections, length of stay (LOS), and charges were collected. Antithrombotic therapy was a surrogate for LCAA. LOS and hospital charges assessed resource utilization. Multivariate regression identified factors associated with LOS, charges, and LCAA. Of 14,194 patients with KD, 2,974 (21%) had refractory KD and 203 of those 2,974 (7%) had LCAA. Additional intravenous immunoglobulin was the sole medication in 77%. Other medications added were steroids (18%), infliximab (2%), and both (3%). Warfarin, low-molecular-weight heparin, tissue plasminogen activator, and clopidogrel were prescribed with equal frequency (2%). Male gender (adjusted relative risk 1.52, 95% confidence interval [CI] 1.08 to 2.16, p <0.01), admission to an intensive care unit (4.79, 95% CI 3.40 to 6.74, p <0.001), arrhythmia (3.00, 95% CI 1.94 to 4.65, p <0.001), and concomitant viral infection (2.29, 95% CI 1.49 to 3.52, p <0.001) were associated with LCAA. Severe illness, race, region, and payer were independently associated with increased charges (p <0.05 for all). In conclusion, treatment for refractory KD varies widely. Concomitant viral infection was associated with a greater risk of LCAA in refractory KD. Better understanding of optimal management may improve outcomes and decrease both variability in management and resource utilization for refractory KD.
Kawasaki disease (KD) is an acute inflammatory vasculitis with a predilection for coronary arteries. Treatment with high-dose aspirin and intravenous immunoglobulin (IVIG) decreases the incidence of coronary artery aneurysms to <5%. Despite this, about 20% of children are refractory to a single dose of IVIG and have evidence of ongoing inflammation with persistence of fever beyond 36 hours after completion of the initial IVIG infusion. Because continued inflammation increases the risk of large/complex coronary artery aneurysms (LCAA), persistent fever is aggressively treated with a multitude of anti-inflammatory therapies of unproven efficacy. Factors reported to be associated with the formation of LCAA in KD are varying and inconsistent. Resource utilization is greater in hospitalized patients with KD compared to children hospitalized for other common febrile childhood illnesses, but no studies to date have compared utilization of hospital resources in patients with responsive and refractory KD. The aims of this study were to assess practice variation, to compare resource use between patients with responsive and refractory KD, and to evaluate factors associated with LCAA and resource utilization in refractory KD in a large multicenter cohort. We hypothesized that practice would vary widely and that compared to patients with responsive KD, patients with refractory KD would be resource intensive. In addition, we hypothesized that there would be independent, modifiable risk factors associated with LCAA and resource utilization.
Methods
For this retrospective, observational cohort study, data were obtained from the Pediatric Health Information System (PHIS). The PHIS database was queried for patients aged ≤18 years discharged with KD ( International Classification of Diseases, Ninth Revision , code: 446.1) and a medication charge for IVIG from January 1, 2004, to March 31, 2014. PHIS is a large, inpatient, administrative database which obtains information from 45 tertiary care, free-standing, not-for-profit pediatric hospitals in the United States. These hospitals are affiliated with the Child Health Corporation of America (Shawnee Mission, KS), a business alliance of children’s hospitals. Bimonthly coding consensus meetings, coding consistency reviews, and quarterly data quality reports are performed for evaluation of data quality and reliability. This study was exempted from review by the University of Utah Institutional Review Board as all data were deidentified at the time of collection.
Data collected included patient demographics (age, gender, ethnicity, and region), length of stay (LOS), diagnoses, procedures, services billed to the patient (pharmacy, imaging/radiology, laboratories, clinical services, and supplies), and hospital charges. Primary outcomes included antithrombotic use, LOS, and hospital charges.
The study population was divided into 2 groups. Patients were in the refractory KD group if they received >1 dose of IVIG (with or without other immunomodulators) within 30 days of initial diagnosis ( Figure 1 ). Those who received only 1 dose of IVIG were in the responsive KD group. Race is categorized in the PHIS database as Asian, Hispanic, non-Hispanic white, non-Hispanic black, and other ( Table 1 ). Several discharge diagnoses were flagged from the PHIS database based on International Classification of Diseases, Ninth Revision, codes: seizure activity (flagged as “neurologic sequelae”), resultant heart failure (flagged as “heart failure medications” which included a pharmacy charge for milrinone, nesiritide, carvedilol, or angiotensin-converting enzyme inhibitors), arrhythmia, renal failure, and shock or cardiac arrest. Patients with known baseline genetic disorders or congenital defects were also identified (flagged as “congenital defects”).
