Comparison of matching by body volume or gestational age for calculation of observed to expected total lung volume in fetuses with isolated congenital diaphragmatic hernia

OBJECTIVE: To determine the bias induced by matching fetuses according to gestational age (GA) or fetal body volume (FBV) when calculating the observed to expected total fetal lung volume (o/e TFLV) in cases of isolated congenital diaphragmatic hernia (CDH).

METHODS: This was a single-center, retrospective study on archived magnetic resonance (MR) images of fetuses with isolated CDH over a 10-year period. We retrieved the TFLV, GA and o/e TFLVGA , and delineated FBV to obtain TFLVFBV in each case. We evaluated the relationship between o/e TFLVFBV and o/e TFLVGA by Bland-Altman analysis. All outliers were manually identified, and their specific clinical features were retrieved.

RESULTS: Records of a total of 377 MR examinations of 225 fetuses were identified and included in the analysis. The mean ( ± SD) time spent on FBV measurement was 16.12 ± 4.95 min. On reproducibility analysis of FBV measurement (n = 10), the intraobserver intraclass correlation coefficient (ICC) was 0.998 and the interobserver ICC was 0.999. FBV was highly correlated with GA (R(2)  = 0.899; P < 0.0001). There was good agreement between o/e TFLVGA and o/e TFLVFBV , with a mean difference of -1.10% and 95% limits of agreement of -8.58 to 6.39. There were no outliers in fetuses that had an o/e TFLV < 25%. Discrepancies induced by different methods were more likely in women with a body mass index ≥ 25 kg/m(2) (+16.5%), fetuses with an estimated fetal weight (EFW) ≤ 10(th) centile (+21.3%) or an EFW > 90(th) centile (+14.7%).

CONCLUSIONS: Discrepancies in matching by FBV and GA when calculating o/e TFLV are more likely in fetuses with an abnormal EFW or in fetuses carried by overweight women. The clinical relevance of using FBV rather than GA for calculation of the o/e TFLV might be limited, as there was no discrepancy between the two methods in fetuses with small lungs ( < 25%), which is the group of most interest for lung volume assessment.