Comparison of six Z-score formulas based on echocardiography for coronary artery lesions in Kawasaki disease.

GOAL: Kawasaki disease (KD) patients are at risk of developing the serious complication of coronary artery dilation (CAD). To diagnose CAD caused by KD, various Z-Score formulas are used worldwide. This paper aims to evaluate the differences and inclusiveness among the six most commonly used Z-Score formulas in diagnosing CAD in Suzhou, China. Additionally, the study seeks to compare the differences in CAD diagnosis among different high-risk factor groups. By doing so, this research provides a valuable reference for accurately diagnosing CAD in KD patients.

METHOD: This paper presents a retrospective analysis of 1509 patients diagnosed with KD at the Children's Hospital of Soochow University between January 2018 and December 2020. We collected the patients' clinical and echocardiographic data and used six Z-Score formulas (Kobayashi et al., de Zorzi et al., Kurotobi et al., McCrindle et al., Olivieri et al., and Dallaire et al.) to diagnose the degree of CAD in different segments. We then compared the diagnostic differences and inclusiveness of these formulas, especially the diagnostic differences in medium to giant CAA. To achieve this, we divided the patients into groups based on their age (≤12 months, 13-30 months, and > 30 months) and fever duration (≤5 days, 6-7 days, 8-9 days, and ≥ 10 days). Using the McNemar test and the Kappa test, we compared the differences and the consistencies of CDA diagnosis among the six Z-Score formulas. Moreover, we used the Friedman test and Chi-square segmentation formula to compare the differences in age and number of fever duration between groups and to compare each Z-Score formula pair within the group.

RESULTS: Except for the LMCA segment, where there were no statistically significant differences between de Zorzi formula and McCrindle formula, the Z-score formulas showed statistically significant differences in the degree of CAD diagnosis across all other segments. Inclusiveness assessment revealed that Kobayashi formula and Dallaire formula showed significantly higher rates of dilatation (6.58% and 5.32%), or of small aneurysms (6.52% and 4.52%) compared to other formulas (1.0%-1.73%). Medium aneurysms were also more likely to be identified with Kobayashi and Dallaire formulas (0.8% and 0.8%) compared to the remaining formulas (0.13-0.40%). There are significant differences in the diagnoses of medium to giant CAA made by these six formulas in LAD and RCA. The longer the duration of fever and the younger the age, the higher the diagnosis rates of CAD and CAA. There were no statistically significant differences between de Zorzi formula and McCrindle formula, de Zorzi formula and Oliveri formula, and Kurotobi formula and Dallaire formula within the four groups based on the duration of fever. Similarly, there were no statistically significant differences between Kobayashi formula and Dallaire formula, and between de Zorzi formula and Oliveri formula in the age groups of ≤12 months and 13-30 months.

CONCLUSION: There are diagnostic differences among these six Z-score formulas, considering the aforementioned statistics. Kobayashi formula and Dallaire formula are more inclusive, and less likely to under-diagnose significant CAD. They perform evenly for dilatation only, for small aneurysms and the median size aneurysms, and that is for segments of LMCA, LAD and RCA. In addition, McCrindle formula joins the "inclusive" pack for LAD and RCA in the matter of CAD. The younger the age of the patients and the longer the duration of fever, the higher the diagnosis rates of CAD and CAA. Furthermore, the younger the age of the patients and the shorter the duration of fever, the greater the differences between the various formulas.