Prediction of ILD by Multivariate regression analysis and ROC curve analysis:
The most sensitive variable to detect ILD in CTD child was KL-6 and FVC%. According to the Receiver Operating Characteristic Curve (ROC) analysis, the area under the curve was of 0.977 and the cut-off value of serum KL-6 to distinguish ILD was 63.4 U/ml, with sensitivity of 95.2 %, specificity of 89.7%, positive predictive value of 83.26% and negative predictive value of 97.2% (figure 2). Moreover, FVC% Receiver Operating Characteristic Curve (ROC) analysis shows that the area under the curve was of 0.979 and the cut-off value of serum FVC% to distinguish ILD was 79, with sensitivity of 95.2%, specificity of 100%, positive predictive value of 100% and negative predictive value of 95.4% (figure 3). Comparison between KL6 and FVC% Roc curves shows that there is no statistical difference (P = 0.7385) between KL6 and FVC% ROC curves.