Temporal clustering of Kawasaki disease cases around the world.

In a single-site study (San Diego, CA, USA), we previously showed that Kawasaki Disease (KD) cases cluster temporally in bursts of approximately 7 days. These clusters occurred more often than would be expected at random even after accounting for long-term trends and seasonality. This finding raised the question of whether other locations around the world experience similar temporal clusters of KD that might offer clues to disease etiology. Here we combine data from San Diego and nine additional sites around the world with hospitals that care for large numbers of KD patients, as well as two multi-hospital catchment regions. We found that across these sites, KD cases clustered at short time scales and there were anomalously long quiet periods with no cases. Both of these phenomena occurred more often than would be expected given local trends and seasonality. Additionally, we found unusually frequent temporal overlaps of KD clusters and quiet periods between pairs of sites. These findings suggest that regional and planetary range environmental influences create periods of higher or lower exposure to KD triggers that may offer clues to the etiology of KD.

The Electrocardiogram Is a Poor Diagnostic Tool to Detect Left Ventricular Hypertrophy in Children: A Comparison with Echocardiographic Assessment of Left Ventricular Mass.

OBJECTIVE: The electrocardiogram (ECG) is used to detect left ventricular hypertrophy (LVH) in children and young adults, and it is considered an important screening tool for the most common causes of sudden cardiac arrest in the United States. However, the reliability and accuracy of the ECG in the detection of LVH are poorly understood. This study's objective was to assess the reliability and accuracy of ECG parameters predicting LVH in comparison with echocardiographic two-dimensional left ventricular mass (2D-LVM) assessment. DESIGN: Two hundred fifty patients met study criteria with complete 2D-LVM measurements, temporally matched ECGs and no confounding structural heart defects. The echocardiographic diagnosis of LVH was made by 2D-LVM z-score > 2. The electrocardiographic diagnosis of LVH was fulfilled if specific Q, R, or S wave voltages exceeded the normal limits for the respective age groups established by the Davignon, Macfarlane and Rijnbeek datasets, or fulfilled the LVH criteria set forth by adult indices (Sokolow-Lyon, Cornell, Gubner, Romhilt-Estes). RESULTS: None of the ECG parameters showed significant correlation with 2D-LVM z-scores. In the detection of LVH, the most commonly used R in V6 voltages showed the following results: sensitivity 13-29%, specificity 77-96%, positive predictive value (PPV) 29-50%, and negative predictive value (NPV) 77%. The overall accuracy ranged 65-75%. The accuracy of adult indices for LVH ranged from 49% to 78%. CONCLUSIONS: The ECG's prediction of LVH is unreliable with dismal sensitivity, variable specificity, poor PPV, and barely acceptable NPV. With such overall poor reliability, the use of current ECG standards in screening for LVH warrants reassessment.