INTERGROWTH-21st compared with GROW customized centiles in the detection of adverse perinatal outcomes at term

Background: INTERGROWTH-21st charts provide standards for infants born under optimal pregnancy conditions. However, their validity in a general obstetric population is unclear. We aimed to identify whether INTERGROWTH-21st charts, compared with gestation related optimal weight (GROW) charts customized on maternal height, weight, and parity, better identified the at-risk infant. Methods: We performed a retrospective cohort analysis of all term women who gave birth at a single tertiary obstetric center during the period 1994-2016. Routinely collected maternity data was used for analysis. The primary outcome was an Apgar score <7 at 5 min. Secondary outcomes included Apgar score <5 at 5 min, stillbirth or admission to the neonatal intensive care unit (NICU). Populations of newborns were identified as SGA by: (a) INTERGROWTH-21st <10th centile (SGA(IG10th)); (b) INTERGROWTH-21st z-score < -1 (SGA(IGzscore)); and (c) GROW customized charts <10th (SGA(cust)). The subgroups identified by only one chart were also specifically examined. Each SGA group was compared to infants appropriate for gestational age (AGA) on all charts (non-SGA). Results: Data for 71,487 births were available for analysis after exclusion of women with missing height or weight data. Only 3280 (4.6%) newborns were considered SGA(IG10th), with 5878 (8.2%) SGA(IGzscore) and 7599 (10.6%) SGA(cust). INTERGROWTH-21st identified only 110 additional infants (0.15%) that were not identified by customized charts; none of these experienced any adverse outcomes. Customized centiles identified a further 4429 (6.2%) SGA infants (SGA(cust-only)) that were not identified as SGA(IG10th), and who did demonstrate an increased risk of Apgar score <7 (OR 1.33, 95%CI 1.08-3.28) and stillbirth (OR 2.47, 95%CI 1.41-4.44) compared to the non-SGA infant. Significantly more obese women had infants considered SGA(cust) (19.3%) than SGA(IG10th) (10.0%) or SGA(IGzscore) (9.9%). Conclusions: Amongst our general obstetric study population, the 10th centile of INTERGROWTH-21st identified only 4.6% of infants as SGA and was less likely to identify infants of obese women as SGA. Customized centiles identified almost all SGA-IG infants, including an additional group (SGA(cust-only)) at higher risk of stillbirth and adverse outcomes compared with non-SGA infants.