Nomogram model predicts coronary artery lesions in recurrent KD patients

A newly constructed nomogram model demonstrates acceptable capacity in assessing the risk of coronary artery lesions (CAL) in patients with recurrent Kawasaki Disease (KD), according to a study.

“A subset of patients with KD will suffer recurrence,” the investigators said. “However, there is still a lack of accurate prediction models for CAL in recurrent KD patients.”

This study was conducted to address such need. Between 2015 and 2021, investigators retrospectively reviewed data from patients with recurrent KD and divided them into training and validation cohorts. They then identified the predictors of CAL using the least absolute shrinkage and selection operator and developed a nomogram based on these predictors using multivariate logistic regression.

The performance of the new model was evaluated using the application of area under the receiver operating characteristic curve (AUC), calibration curves, Hosmer‒Lemeshow test, Brier score, and decision curve analysis.

Of the 159 recurrent KD patients enrolled, 66 (41.5 percent) had CAL. The significant predictors of CAL identified were as follows: haemoglobin levels, intravenous immunoglobulin resistance at recurrence, and CAL at the first episode.

In the training cohort, the model that incorporated the above predictors exhibited good discrimination (AUC, 0.777) and calibration capacities (Hosmer‒Lemeshow: p=0.418; Brier score, 0.190). When applied to the validation cohort, the model consistently showed an acceptable discrimination (AUC, 0.741; Hosmer‒Lemeshow: p=0.623; Brier score, 0.190).

“The decision curve analysis demonstrated that the nomogram model was clinically useful,” the investigators said.