A new simple method to estimate the gestational age from chest x-ray measurements in newborns and preterm infants

This poster is published under an open license. Please read the disclaimer for further details.

Congress:

ECR 2017

Poster Number:

C-2712

Type:

Scientific Exhibit

Keywords:

Paediatric, Thorax, Lung, Conventional radiography, Diagnostic procedure, eLearning, Statistics, Congenital, Education and training, Epidemiology

Authors:

J. A. Prat-Matifoll, J. Piqueras, A. Prats-Uribe, C. Parra-Fariñas, E. Maciag, C. D. Hernández Giraldo, C. A. Ortiz Andrade, J. C. Carreño; Barcelona/ES

DOI:

10.1594/ecr2017/C-2712

DOI-Link:

https://dx.doi.org/10.1594/ecr2017/C-2712

Fig. 3: ROC curve for a Dorsal Spine Length of 90 mm to differentiate preterm newborns...

Fig. 4: ROC curve for a triple combination of the 3 measurements (dorsal spine length,...

Results

Multiple threshold measures were obtained in order to establish an optimal value (high sensitivity and specificity) to classify newborns into pre-term (extremely preterm, very preterm and moderate/late preterm), term and post-term.

Threshold values obtained by measurements taken in A-P chest x-ray exams:

1.One key measure to differentiate preterm from term and postterm newborns (Fig.3):

- Dorsal Spine length: 90mm

S: 90.82% (IC 86- 94.5%) PPV: 92.2%

E: 81.48% (IC 71.3-89.2%) NPV: 78.6%

2.Three key measures to differentiate between extremely and very preterm from late preterm and term newborns:

- Dorsal Spine length: 80mm

S: 92.16% (IC 85.1-96.6%)

E: 81.71% (IC 75.2-87.1%)

- Right pulmonary length: 47mm

S: 78.4% (IC 69.2-86%)

E: 89% (83.6-94%)

- Transversal chest width: 93mm

S: 79.4% (IC 70.3-86.8%)

E: 91.4% (IC 86.3-95.1%)

3.The triple combination of the 3 measures mentioned above, helps to improve the PPV for differentiating between extremely/very preterm from late/term newborns (Fig.4):

-1 or more values: PPV: 68.3% / NPV: 96.3%

-2 or more values: PPV: 85.5% / NPV: 93.5%

-3 values: PPV: 93.1% / NPV: 82.1%

4. Other variables

Other variables were recorded without observing significant differences.

Other studied variables were the gestational age, weight, twin pregnancy, persistent ductus arteriosus, congenital malformations, metabolopathies, premature newborn complications (hyaline membrane disease, necrotizing enterocolitis, respiratory distress syndrome or haemolytic disease of the newborn requiring phototherapy), previous pregnancies and maternal complications (pre-eclampsia and eclampsia, chorioamnionitis or threat of premature birth).