Use of laboratory data to identify risk factors of giant coronary aneurysms due to Kawasaki disease


Yosikazu Nakamura, MD, MPH, FFPH, Department of Public Health, Jichi Medical School, 3311−1 Yakushiji, Minamikawachi, Tochigi 329−0498 Japan. Email:


AbstractBackground: Although some laboratory findings are known to be indicators of the risk of giant coronary aneurysm formation among Kawasaki disease patients, an appropriate cut-off point to predict aneurysm formation is not clear.

Methods: One hundred and five patients with giant coronary aneurysms were selected from the 15th and 16th nationwide surveys of Kawasaki disease in Japan. A total of 2936 patients without Kawasaki disease were recruited from a single hospital as a control group. Odds ratios were calculated for six laboratory data with specific values as cut-off points. Receiver operating characteristic (ROC) curves were observed to determine the most appropriate laboratory tests and cut-off points.

Results: Hematocrit, leukocyte count, neutrophil proportion, and hemoglobin had one or more peaks of odds ratio for specific cut-off points, but they did not have a clear cut-off point for the predictor according to the receiver operating characteristic curves. Alanine aminotransferase (ALT) increased the risk of giant coronary aneurysms continuously so no clearly appropriate cut-off point was identified. Serum sodium concentration of 135 mEq/L had a peak of odds ratio, and those with <135 mEq/L had the highest odds ratio (4.78). This value seemed appropriate with a sensitivity of 78% and specificity of 57%, although the predictive positive value was as small as 5%.

Conclusion: The author's propose that a serum sodium concentration of <135 mEq/L at the patient's first visit to hospital may be a predictor of giant coronary aneurysms due to Kawasaki disease.