In describing the definition for incomplete KD, the authors have presented the diagnostic approach, which is largely adapted from American Heart Association (AHA) scientific statement on Kawasaki disease [2]. While AHA algorithm considers evaluation for incomplete KD in children with fever ³5 days and 2 or 3 compatible clinical criteria, the algorithm by Shenoy, et al. [1] triggers KD evaluation if fever ³5 days is accompanied by less than four compatible clinical features. Although these two statements appear similar, this approach loses specificity by including children who present with fever and just one compatible clinical feature. Individually, the clinical features like rash, lymphadenopathy, conjunctival injection, oral or extremity changes are nonspecific and may occur with various childhood infections in India. This approach risks huge number of children with underlying infections being referred for echocardio-graphic evaluation.

Treatment with intravenous immunoglobulin is recommended if 3 of the 5 laboratory features (anemia for age, platelet ³450×109/L, albumin <3 g/dL, elevated alanine aminotransferase, leucocyte count ³15 ×109/L, urine >10 WBC/hpf) are met in a child lacking echocardiographic abnormalities. Compared to Western cohorts, these criteria should be carefully defined in a low- and middle-income setting like India, with a high prevalence of iron deficiency anemia [3] and associated thrombocytosis (present in up to a quarter of those with iron deficiency) [4]. Iron deficiency when associated with infection accounted for more than half of all cases of reactive thrombocytosis in Indian children [5]. Given these findings, the current definition is likely to overestimate the burden of incomplete KD in Indian children, risking increased cost and potentially delaying the diagnosis of underlying infections. For example, as per the algorithm, a child with undifferentiated fever ³5 days due to measles or a rickettsial infection that has a rash, iron deficiency anemia (and associated thrombo-cytosis) and hypoalbuminemia (negative acute phase reactant) would be treated for Kawasaki disease even if the echocardiogram is normal. In the absence of a ‘gold standard’ for diagnosis, we believe that grading recommendations based on available quality of evidence may be more useful for the readers to make informed decisions [6].