Abstract
BACKGROUND
We have previously demonstrated that a competing risks model for prediction of small for gestational age (SGA) neonates, has a superior performance to that of the traditional risk-scoring methods. The Fetal Medicine Foundation (FMF) fetal and neonatal population weight charts are derived from the sonographic estimated fetal weight (EFW), rather than birthweight, because a large proportion of babies born preterm arise from pathological pregnancy. The individualized risk assessment for SGA at mid-gestation could be the basis of an antenatal plan that aims to improve the management of preterm SGA pregnancies with minimum resources.
OBJECTIVE
To stratify subsequent assessment after 24 weeks' gestation based on the estimated risk for SGA neonates delivering <28, <32 and <36 weeks' gestation by the combination of maternal risk factors with EFW and uterine artery pulsatility index (triple test), assessed at mid-gestation. The rationale of the study is that those at high risk for SGA <28, <32 and <36 weeks would require ultrasound examinations at 26, 30 and 33 weeks' gestation, respectively.
METHODS
Our cohort was derived from a prospective, non-intervention study in women with singleton pregnancies attending for a routine ultrasound scan at 19+0 to 23+6 weeks' gestation in two UK maternity hospitals. The competing risks model was used to estimate the individual patient-specific risks of delivery with SGA at <36 weeks' gestation from the triple test. Different risk cut-offs were used with the intention of detecting about 80%, 85% and 90% of cases of delivery with SGA at <28, <32 and <36 weeks' gestation. Discrimination measures by means of sensitivity, specificity, positive and negative predictive values were computed for different risk cut-offs. Calibration for risks for delivery with SGA at <36 weeks' gestation was assessed by plotting the observed incidence of SGA against the predicted incidence.
RESULTS
The study population of 134,443 singleton pregnancies contained 16,813 (12.51%) pregnancies that subsequently delivered SGA neonates <10th percentile, as defined by the FMF chart, including 196 (0.15%), 566 (0.42%), and 1787 (1.33%) that delivered at <28, <32 and <36 weeks, respectively. If the objective of screening was to identify about 80% of cases of delivery of SGA neonates <10th percentile at <28, <32 and <36 weeks' gestation, using the FMF chart to define SGA, the respective screen positive rates would be 9.5%, 19.6% and 29.6%. If the objective of screening was to identify about 80% of cases of delivery of SGA neonates <3rd percentile at <28, <32 and <36 weeks' gestation, using the FMF chart to define SGA, the respective screen positive rates would be 6.5%, 13.0% and 21.6%. The calibration plots demonstrated good agreement between the predicted risk and the observed incidence of SGA.
CONCLUSIONS
Assessment of risk for birth of SGA neonates at mid-gestation is useful to identify the subgroups that require monitoring at 26, 30 and 33 weeks' gestation, in addition to a routine scan at 36 weeks. The FMF competing risk model for SGA can be customised to the desired detection rate and to the clinical resources availability.