Sonographic growth curves versus neonatal birthweight growth curves for the identification of fetal growth restriction.

OBJECTIVE

Fetal growth restriction is suspected when the estimated fetal weight is <10th percentile for gestational age. Using a regional sonographic estimated fetal weight growth curve to diagnose fetal growth restriction has no known benefits; however, the traditional approach of using birthweight curves is misleading, since a large proportion of preterm births arise from pathological pregnancies. Our aim was to compare the diagnostic accuracies of sonographic versus birthweight curves in diagnosing fetal growth restriction. Our secondary aim was to compare maternal, fetal and neonatal outcome based on these two approaches.

METHODS

Retrospective study based on computerized medical records. Included were women with a singleton pregnancy, that underwent fetal biometry between 24 and 36.6 weeks' gestation (January 2010-February 2016) and delivered in our center. Each pregnancy was assigned to one of three groups based on the earliest sonographic estimated fetal weight performed: -Appropriate for gestational age, -fetal growth restriction based on sonographic but not birthweight curves; or -fetal growth restriction based on birthweight growth curves. Demographics, obstetric characteristics, ultrasound data, and neonatal data were retrieved and compared between groups. Primary outcome: rate of small for gestational age neonates in each group. Secondary outcomes were various adverse maternal and neonatal outcomes.

RESULTS

Six thousand and five pregnancies met inclusion criteria. Of these 5386 (89.6%) were categorized as G1, 300 (5%) as G2 and 319 (5.3%) as G3. The rate of small for gestational age neonates differed significantly between groups: G1 9.2%, G2 39.7% and G3 70%. Multivariable logistic regression modeling reiterated these rates: the odds ratios for small for gestational age were 6.47 [95% CI 4.99-8.40] and 23.99 [95% CI 18.26-31.51] for G2 and G3 respectively. Prediction of small for gestational age based on sonographic EFW curves increased the sensitivity for detection of SGA from 26% to 41% with a slight decrease in specificity from 98% to 95%, and a decrease of the positive likelihood ratio from 18.4 to 7.7, however there was no significant change in the overall test accurcy; 88.5% to 87.1%.Secondary outcomes also differed between groups: G2 and G3 had similar rates of maternal and neonatal morbidities and most parameters were higher than G1. G2 and G3 showed lower mean gestational age at delivery (36.2 weeks and 35.9 weeks vs.37.8;  < .0001), and higher rates of preterm delivery (40% and 51.7% vs. 21.5%;  < .001), as well as higher rates of intrauterine fetal demise 3% in G2, 6.9% in G3 and 0.9% in G1,  < .0001.

CONCLUSION

Pregnancies that are currently managed as appropriate for gestational age based on birthweight curves, but classified as growth restricted when prenatal sonographic curves are used, are associated with higher rates of small for gestational age and poor perinatal outcomes, at rates comparable to pregnancies that are classified as growth restricted based on birthweight curves. Furthermore, applying sonographic curves increases the sensitivity for detection of small for gestational age neonates. Consequently, consideration should be given to the use of sonographic biometry curves for defining fetal growth restriction.