Abstract

Objective To explore the predictive values of serum procalcitonin (PCT) and C reactive protein (CRP) levels on coronary artery lesion (CAL) in acute phase of Kawasaki disease. Methods A total of 217 children with Kawasaki disease who hospitalized in Children′s Hospital of Chongqing Medical University from November 1, 2015 to December 31, 2016 were selected as research subjects. All the subjects have received standard Kawasaki disease treatments and all cases were in accord with the inclusion and exclusion criteria in this study. The gender, age, serum PCT and CRP levels, and echocardiographic findings within 30 d after the onset of Kawasake disease of all subjects were collected by retrospective method. According to children with acute Kawasaki disease complicated with CAL or not, they were divided into CAL group (n=22) and non-CAL group (n=29). Mann-Whitney U test was used to analyzed the differences of serum PCT and CRP levels between two groups. Chi-square test was used to compare the positive rates of serum PCT and CRP between two groups. Binary logistic regression analysis model was used to construct a new combined predictor L by serum PCT and CRP levels for predicting CAL in children with Kawasaki disease. And receiver operating characteristics (ROC) curves for new combined predictor L, serum PCR and CRP levels were built respectively, to evaluate the predictive values of combined predictor L and serum PCT, CRP levels on CAL in children with Kawasaki disease. The area under ROC curve (ROC-AUC) for combined predictor L and serum PCT, CRP levels in predicting CAL in children with Kawasaki disease all were calculated. The optimal cut-off point of combined predictor L and serum PCT, CRP levels predicting CAL in children with Kawasaki disease was obtained where the Youden index reached the maximum value. And diagnostic parameters such as sensitivity and specificity were also calculated for comparison.There were no statistically significant differences between two groups in the aspects of gender ratio, age, and so on (P>0.05). Results ①The medians of serum PCT level in CAL group and non-CAL group were 1.16 μg/L and 0.04 μg/L, respectively. The medians of serum CRP level in CAL group and non-CAL group were 25.0 mg/L and 16.4 mg/L, respectively. The levels of serum PCT and CRP in CAL group both were statistically higher than those in non-CAL group, and both the differences were statistically significant (Z=-2.454, P=0.046; Z=-6.412, P<0.001). There were no statistically significant differences in the positive rates of serum PCT and CRP between two groups (χ2=1.222, P=0.269; χ2=0.665, P=0.415). ②Binary logistic regression analysis model with complicated with CAL or not as a dichotomous outcome variable, and levels of serum CRP and PCT as independent variables showed that elevated levels of serum CRP and PCT were independent risk factors for children with Kawasaki disease complicated with CAL (OR=1.077, 95%CI: 1.027-1.190, P<0.001; OR=1.046, 95%CI: 1.015-1.078, P=0.003). The expression of combined predictor L was L=x1+ 0.608x2, in which x1 and x2 referred to the levels of serum CRP and PCT, respectively. ③ROC curve analysis indicated the values of ROC-AUC for combined predictor L, levels of serum PCT and CRP in predicting children with Kawasaki disease complicated with CAL were 0.773 (95%CI: 0.700-0.845, P<0.001), 0.560 (95%CI: 0.475-0.645, P=0.148), and 0.767 (95%CI: 0.695-0.839, P<0.001), respectively. The optimal cutoff values of combined predictor L, levels of serum PCT and CRP were 23.068, 2.32 μg/L, and 22.0 mg/L, respectively, and the sensitivities were 68.1%, 30.6% and 68.1%, respectively, the specialties were 83.4%, 86.9% and 80.7%, respectively. Conclusions Monitoring levels of serum PCT, CRP and combined predictor L (L=x1+ 0.608x2, x1 and x2 referred to levels of serum CRP and PCT, respectively) have predictive values to evaluate the occurrence of children with Kawasaki disease complicating CAL in acute phase. Key words: Procalcitonin; C reactive protein; Mucocutaneous lymph node syndrome; Coronary artery lesion; Predictive value; Combined predictive indicator; Logistic regression model; Child

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call