Background

We sought to determine if fever in the early postintravenous immunoglobulin (IVIG) time period (first 36 hours after IVIG completion) for Kawasaki disease, with or without additional infliximab, can predict IVIG resistance and coronary artery abnormalities (CAA).

Methods

Acute Kawasaki disease subjects enrolled in a clinical trial of infliximab plus IVIG (n = 96) versus placebo/IVIG (n = 94) had temperatures recorded every 6 hours after completion of IVIG infusion. Fever was defined as temperature >38.0°C; patients with persistent or recrudescent fever >36 hours after completion of IVIG were classified as IVIG resistant. Multivariable logistic regression by fever pattern was performed to predict outcomes (IVIG resistance and CAA).

Results

There was no difference in the time to defervescence between the infliximab/IVIG group (n = 96) versus placebo/IVIG group (n = 94). There was no fever after completion of IVIG in the majority of subjects [66% of those with no CAA (n = 139) and 76.5% of those with CAA, (n = 51)]. Although subjects with at least 1 fever 24-36 hours post-IVIG had a higher probability of IVIG resistance [odds ratio = 30.6 (95% confidence interval: 6.7-139.8); P < 0.0001], fever at 24-36 hours was not associated with higher likelihood of CAA. There were also 11% (n = 19) of IVIG responders who had fever at 24-36 hours post-IVIG. The majority of subjects with CAA (43 of 51, 84.3%) were identified by the initial echocardiogram, so the effect of fever on development of CAA could not be assessed.

Conclusions

Fever in the first 36 hours after IVIG completion is not predictive of CAA. Our data support refraining from retreatment until 36 hours after completion of IVIG.