RESUMO
OBJECTIVES: To assess if observed higher observed risks of cardiac complications for patients with Kawasaki disease (KD) treated earlier may reflect bias due to confounding from initial disease severity, as opposed to any negative effect of earlier treatment. STUDY DESIGN: We used data from Japanese nationwide KD surveys from 1997 to 2004. Receipt of additional intravenous immunoglobulin (IVIG) (data available all years) or any additional treatment (available for 2003-2004) were assessed as proxies for initial disease severity. We determined associations between earlier or later IVIG treatment (defined as receipt of IVIG on days 1-4 vs days 5-10 of illness) and cardiac complications by stratifying by receipt of additional treatment or by using logistic modeling to control for the effect of receiving additional treatment. RESULTS: A total of 48 310 patients with KD were included in the analysis. In unadjusted analysis, earlier IVIG treatment was associated with a higher risk for 4 categories of cardiac complications, including all major cardiac complications (risk ratio, 1.10; 95% CI, 1.06-1.15). Stratifying by receipt of additional treatment removed this association, and earlier IVIG treatment became protective against all major cardiac complications when controlling for any additional treatment in logistic regressions (OR, 0.90; 95% CI, 0.80-1.00). CONCLUSIONS: Observed higher risks of cardiac complications among patients with KD receiving IVIG treatment on days 1-4 of the illness are most likely due to underlying higher initial disease severity, and patients with KD should continue to be treated with IVIG as early as possible.