Oct 15 2012
By Piriya Mahendra, medwireNews Reporter
A clinical prediction rule composed of clinical birth variables could be used to predict histologic chorioamnionitis (HC) and histologic chorioamnionitis with fetal involvement (HCF) in preterm babies, a study shows.
Jasper Been (Maastricht University Medical Centre, the Netherlands) and co-authors say that the models tested in their study are comprise clinical variables readily available in clinical practice and therefore are "broadly applicable and simple to use."
They developed HC and HCF prediction rules with preference for high sensitivity using clinical variables available at birth, including ethnicity, maternal age, and parity among others.
The study showed that HC and HCF were present in 29% and 24% of the derivation cohort and in 44% and 22% of the validation cohort, respectively. HC was predicted with 87% accuracy, yielding an area under the receiver-operating characteristic (ROC) curve of 0.95, a positive predictive value of 80%, and a negative predictive value of 93%.
The corresponding figures for HCF were 83% for accuracy, 0.92 for area under the ROC curve, 59% for positive predictive value, and 97% for negative predictive value.
The authors note that, as expected, external validation resulted in some loss of test performance, preferentially affecting positive predictive rather than negative predictive values.
The study involved the clinical and placental pathology data of 216 singleton preterm newborn babies (gestational age ≤32 weeks), born between 2001 and 2003, in the derivation cohort, and 206 preterm babies, born between 2009 and 2010, in the validation cohort.
"The prediction rules presented here carry future potential in facilitation of subgroup-targeted early intervention strategies," Been et al remark in PLoS ONE.
"Moreover, the prediction rules are easy to use, cheap, and their results are readily available," the researchers add.
They conclude that the models require further evaluation to assess their value in supporting clinical decision-making.
Licensed from medwireNews with permission from Springer Healthcare Ltd. ©Springer Healthcare Ltd. All rights reserved. Neither of these parties endorse or recommend any commercial products, services, or equipment.