Institutional members access full text with Ovid®

External Validation of a Risk Score to Predict Intravenous Immunoglobulin Resistance in Patients With Kawasaki Disease

Seki, Mitsuru MD*†; Kobayashi, Tohru MD*; Kobayashi, Tomio MD†; Morikawa, Akihiro MD*; Otani, Tetsuya MD‡; Takeuchi, Kazuo MD, MPH§; Ayusawa, Mamoru MD¶; Tsuchiya, Keiji MD∥; Yasuda, Kenji MD**; Suzuki, Takahiro MD*; Shimoyama, Shinya MD†; Ikeda, Kentaro MD†; Ishii, Yoichiro MD*; Arakawa, Hirokazu MD*

Pediatric Infectious Disease Journal: February 2011 - Volume 30 - Issue 2 - pp 145-147
doi: 10.1097/INF.0b013e3181f386db
Original Studies

Background: We previously developed a new risk score to predict intravenous immunoglobulin (IVIG) resistance in Kawasaki disease. However, the IVIG dosage used in that study (1 g/kg/d for 2 consecutive days) differs from the single infusion of 2 g/kg recommended in the United States and elsewhere. Our aim was to assess the validity and applicability of our risk score in patients treated with a single infusion.

Methods: We used a database of 1626 patients with Kawasaki disease given initial IVIG treatment at a dose of 1 g/kg/d for 2 consecutive days (n = 990; IVIG- 1 g/kg × 2) or 2 g/kg/d for 1 day (n = 636; IVIG- 2 g/kg × 1) across 17 hospitals in Japan. Patients received the total IVIG dose within 36 hours in IVIG- 1 g/kg × 2 and 24 hours in IVIG- 2 g/kg × 1. We stratified the patients according to a risk scoring system developed to predict IVIG unresponsiveness, based on scores of ≥5 points. We compared the accuracy of prediction between the 2 groups using receiver operating characteristic analysis.

Results: Baseline characteristics and clinical outcomes were similar between both groups. The areas under the receiver operating characteristic curve in IVIG- 2 g/kg × 1 were similar to those of IVIG- 1 g/kg × 2. Using a cut-off risk score of ≥5 points, we could identify IVIG resistance in terms of coronary artery abnormalities within 1 month and coronary artery abnormalities at 1 month with equivalent sensitivity and specificity in both groups.

Conclusion: Our risk score can be used to predict IVIG unresponsiveness to a regimen based on a single infusion of 2 g/kg IVIG.

From the *Department of Pediatrics, Gunma University Graduate School of Medicine, Gunma, Japan; †Department of Cardiology, Gunma Children's Medical Center, Gunma, Japan; ‡Department of Health Policy, National Research Institute for Child Health and Development, Tokyo, Japan; §Faculty of Education, Saitama University, Saitama, Japan; ¶Department of Pediatrics and Child Health, Nihon University School of Medicine, Tokyo, Japan; ∥Department of Pediatrics, Japan Red Cross Medical Center, Tokyo, Japan; and **Department of Pediatrics, Shimane University, Shimane, Japan.

Accepted for publication July 23, 2010.

Supported by Grants-in-Aid for Clinical Research for New Medicine from the Ministry of Health, Labour, and Welfare of Japan.

Address for correspondence: Tohru Kobayashi, MD, Department of Pediatrics and Developmental Medicine, Gunma University Graduate School of Medicine, 3–39–15 Showa, Maebashi, Gunma 371–8511, Japan. E-mail:

© 2011 Lippincott Williams & Wilkins, Inc.