ABSTRACT
OBJECTIVE: This study is directed to determine whether the concentrations of serum amyloid A (SAA) in maternal serum could be used to predict a tocolytic failure in preterm delivery, by comparing with other factors associated with inflammation. METHODS: A total of 100 pregnant women from September, 2000 to August, 2001 received continuous prenatal care and underwent delivery in our hospital was enrolled in the study. Gestational age was ranged between 20 and 37 weeks. Subjects were divided into four groups (group I, no preterm labor and no premature rupture of membranes [n=38]; group II, premature rupture of membranes and no preterm labor [n=12]; group III, preterm labor and no premature rupture of membranes [n=34]; Group IV, preterm labor and premature rupture of membranes [n=16]). The levels of SAA, CRP, ESR, and WBC count were measured in maternal serum. RESULTS: SAA levles, CRP levels, and WBC count in patients with tocolytic failure were significantly higher than those in patients without tocolytic failure. SAA and CRP appeared to be significant factors by logistic regression analysis. From the ROC curve analysis of maternal SAA for the prediction of tocolytic failure, we set 6 mg/L as a cut-off value in this study. Sensitivity, specificity, positive predictive value, and negative predictive value were 76%, 72%, 47.5%, and 90%, respectively. As for CRP, 0.59 mg/dL was set as a cut-off value, and sensitivity, specificity, positive predictive value, and negative predictive value were 72%, 81.3%, 56.3%, and 89.7%, respectively. When cut-off values for both SAA and CRP were applied at the same time, sensitivity, specificity, positive predictive value, and negative predictive value were 60%, 92%, 71.4%, and 87.3%, respectively. CONCLUSION: This study showed that the measurement of maternal serum amyloid A may be a fast, non-invasive diagnostic method in the prediction of tocolytic failure in preterm delivery.