The cornerstone of modern obstetrics is to employ an evidence-based estimated balance of risks and benefits to identify and manage the at-risk fetus. The small-for-gestational-age (SGA) newborn is the result of a wide spectrum of compromising events in utero and fetal adaptations (intrauterine growth restriction [IUGR]). IUGR carries with it subsequent increased risk for short- and long-term adverse outcomes postnatally. The correct decision as to the timing of the delivery, based on awareness of potential hazards of delivery either too early or too late, is crucial for optimizing the outcome of these newborns. There is evidence that at term, mortality and morbidity are increased among infants whose birthweight (BW) is ≤3rd percentile for gestational age (GA). Very-low-BW (VLBW) infants at any GA are at risk for increased morbidity and mortality. A traditional clinical approach assumes that SGA infants have accelerated maturation due to intrauterine stress and thus are likely to experience fewer short-term complications of prematurity, especially less respiratory morbidity, when compared with age-matched appropriate for GA (AGA) premature infants. However, conflicting evidence has accumulated regarding mortality as well as the entire spectrum of major neonatal morbidities in preterm SGA infants: bronchopulmonary dysplasia (BPD), retinopathy of prematurity (ROP), intraventricular-periventricular hemorrhage (IVH), periventricular leukomalacia (PVL), and necrotizing enterocolitis (NEC).

As such, the objective of this study was to determine whether the severity of growth restriction was associated with increased neonatal mortality and morbidity among VLBW premature SGA infants.

Materials and Methods

This is a population-based observational study that analyzed data collected by the Israel Neonatal Network on VLBW infants (≤1500 g) born in Israel during January 1995 through December 2007. All 28 neonatal departments in Israel are included in data collection, which comprises the Israel National VLBW Infant Database ( Appendix ). The data collected included parental demographic information, maternal pregnancy history and antenatal care, details of delivery, infants’ status at delivery, diagnoses, procedures, complications during hospitalization, and status at discharge. A prestructured form was completed for each infant. All departments use an operation manual and standard definitions based on the Vermont Oxford Trials Network (Vermont Oxford Network Database Manual of Operation, release 2.0; Vermont Oxford Network, Burlington, VT).

All liveborn infants receive a unique identification number at birth. Patient information received by the database coordinator is cross-checked with the national birth registry and any missing infant data are requested from the birth hospital. Birth hospital and patient identification subsequently remain confidential by consensus agreement of all participating centers. Data are collected on all infants until death or discharge home.

This study was approved by the Institutional Review Board (Helsinki Committee) of the Sheba Medical Center, Tel Hashomer, Israel (SMC 7674-10).

Study population

During the study period, the database included records of 19,399 infants, representing >99% of all liveborn VLBW births in Israel. For purpose of this analysis we excluded infants of BW >50th percentile (n = 4134), birth at <24 weeks’ (n = 459) or ≥32 weeks’ (n = 4331) gestation, and infants born with major congenital malformations (n = 719), resulting in a study population of 9756 neonates ( Figure ). Since chromosomal disorders and major congenital malformations are responsible for many of the worst outcomes in SGA neonates, babies with malformations were not included in our study. Congenital malformations were recorded according to the list of reportable malformations required in Israel, which includes all major malformations.

Schematic description of study population

BW , birthweight; GA , gestational age; VLBW , very low birthweight.

Grisaru-Granovsky. VLBW small-for-gestational-age infants’ mortality and morbidity. Am J Obstet Gynecol 2012.

Definition of SGA and study groups

The GA in completed weeks was based on the attending neonatologist’s best estimate of the GA based on the last menstrual period, obstetric history and examination, and prenatal ultrasound and postnatal examination performed in the first hours of life. BW percentiles were determined from the sex-specific charts of Kramer et al. SGA was defined as a BW <10th percentile for GA and was stratified into 2 subgroups: <3rd percentile (n = 610) and 3 ^rd to <10th percentile (n = 1302). Infants of BW 10 ^th to <25th percentile (n = 2857) were considered a separate study group while BW 25th-50th percentile (n = 4977) served as the reference group.

Definition of outcome variables

Mortality was defined as death at any time before discharge home. Respiratory distress syndrome (RDS) was diagnosed by a chest radiograph consistent with RDS together with the need for supplementary oxygen or mechanical ventilation therapy. BPD was diagnosed according to the criteria of Bancalari et al, including clinical and radiologic features, together with the requirement of oxygen therapy at 28 days of age. ROP was graded according to the international classification, and was recorded as the most severe stage of ROP in either eye. IVH was diagnosed by ultrasound and graded according to Papile et al. PVL was diagnosed by the presence of multiple periventricular cysts identified by ultrasound after 28 days of life. NEC was determined by the clinical and radiological criteria of Bell et al and only definite NEC (Bell stages II-III) was included.

Covariates

The covariates included in the study were: GA (completed weeks of gestation); sex; multiple birth; existence of preterm premature rupture of membranes (PPROM) for >12 hours prior to birth; antenatal clinical chorioamnionitis; and use of antenatal steroid therapy defined as “partial” (if delivery occurred within 24 hours after the first dose or >1 week after the last dose) or “complete” (if delivery occurred from 24 hours to 7 days after a completion of the course) therapy.

Statistical analyses

Differences in clinical characteristics and outcomes among the 4 study groups were tested by χ ² test and Mantel-Haenszel test for trends. Univariate analysis of mortality and morbidity by BW percentile group was performed. Multivariable logistic regression analyses were used to assess the independent effect of SGA severity on mortality, RDS, IVH grades 3-4, ROP grades 3-4, PVL, BPD, and NEC, after adjusting for confounding variables. Results of the logistic regression are presented as odds ratios with appropriate 95% confidence intervals. Goodness-of-fit for the models was estimated by Hosmer-Lemeshow statistics. The predictive discrimination of the models was estimated using C-statistics, corresponding to the area of a receiver operating characteristic curve. The statistical analyses were performed using SAS software, version 9.1 (SAS Institute, Cary, NC).