The pitfalls of using birthweight centile charts to audit care

Association Between Iatrogenic Delivery for Suspected Fetal Growth Restriction and Childhood School Outcomes

IMPORTANCE

Timely delivery of infants suspected of having fetal growth restriction (FGR) is a balance between preventing stillbirth and minimizing prematurity, particularly because many infants with suspected FGR have normal growth.

OBJECTIVE

To explore the association between iatrogenic delivery for suspected FGR and childhood school outcomes.

DESIGN, SETTING, AND PARTICIPANTS

A retrospective whole-population cohort study linking perinatal data from births 32 weeks' or more gestation between January 1, 2003, to December 31, 2013, to developmental and educational test scores at preparatory school, and at school grades 3, 5, and 7 in Victoria, Australia. Follow-up was concluded in 2019.

EXPOSURES

Suspicion or nonsuspicion of FGR, presence or absence of iatrogenic delivery (defined as early induction of labor or cesarean delivery prior to labor) for suspected FGR, and presence or absence of small for gestational age (SGA).

MAIN OUTCOMES AND MEASURES

The coprimary outcomes were being in the bottom 10th percentile on 2 or more of 5 developmental domains at school entry and being below the national minimum standard on 2 or more of 5 educational domains in grades 3, 5, or 7.

RESULTS

In the birth population of 705 937 infants, the mean gestation at birth was 39.1 (SD, 1.5) weeks and the mean birth weight was 3426 (SD, 517) grams. The birth population linked to 181 902 children with developmental results and 425 717 children with educational results. Compared with infants with severe SGA (birth weight <3rd percentile) not suspected of having FGR, infants with severe SGA delivered for suspected FGR were born earlier (mean gestation, 37.9 weeks vs 39.4 weeks). They also had a significantly increased risk of poor developmental outcome at school entry (16.2% vs 12.7%; absolute difference, 3.5% [95% CI, 0.5%-6.5%]); adjusted odds ratio [aOR], 1.36 [95% CI, 1.07-1.74]) and poor educational outcomes in grades 3, 5, and 7 (for example, in grade 7: 13.4% vs 10.5%; absolute difference, 2.9% [95% CI, 0.4%-5.5%]); aOR, 1.33 [95% CI, 1.04-1.70]). There was no significant difference between infants with normal growth (birth weight ≥10th percentile) delivered for suspected FGR and those not suspected of having FGR in developmental outcome (8.6% vs 8.1%; absolute difference, 0.5% [95% CI, -1.1% to 2.0%]); aOR, 1.17 [95% CI, 0.95-1.45]) or educational outcome in grade 3, 5 or 7, despite being born earlier (mean gestation, 38.0 weeks vs 39.1 weeks).

CONCLUSIONS AND RELEVANCE

In this exploratory study conducted in Victoria, Australia, iatrogenic delivery of infants with severe SGA due to suspected FGR was associated with poorer school outcomes compared with infants with severe SGA not suspected of having FGR. Iatrogenic delivery of infants with normal growth due to suspected FGR was not associated with poorer school outcomes compared with infants with normal growth not suspected of having FGR.

Maternal region of origin and Small for gestational age: a cross-sectional analysis of Victorian perinatal data

Background

Being born small for gestational age is a strong predictor of the short- and long-term health of the neonate, child, and adult. Variation in the rates of small for gestational age have been identified across population groups in high income countries, including Australia. Understanding the factors contributing to this variation may assist clinicians to reduce the morbidity and mortality associated with being born small. Victoria, in addition to New South Wales, accounts for the largest proportion of net overseas migration and births in Australia. The aim of this research was to analyse how migration was associated with small for gestational age in Victoria.

Methods

This was a cross sectional population health study of singleton births in Victoria from 2009 to 2018 (n = 708,475). The prevalence of being born small for gestational age (SGA; <10th centile) was determined for maternal region of origin groups. Multivariate logistic regression analysis was used to analyse the association between maternal region of origin and SGA.

Results

Maternal region of origin was an independent risk factor for SGA in Victoria (p < .001), with a prevalence of SGA for migrant women of 11.3% (n = 27,815) and 7.3% for Australian born women (n = 33,749). Women from the Americas (aOR1.24, 95%CI:1.14 to 1.36), North Africa, North East Africa, and the Middle East (aOR1.57, 95%CI:1.52 to 1.63); Southern Central Asia (aOR2.58, 95%CI:2.50 to 2.66); South East Asia (aOR2.02, 95%CI: 1.95 to 2.01); and sub-Saharan Africa (aOR1.80, 95%CI:1.69 to 1.92) were more likely to birth an SGA child in comparison to women born in Australia.

Conclusions

Victorian woman’s region of origin was an independent risk factor for SGA. Variation in the rates of SGA between maternal regions of origin suggests additional factors such as a woman’s pre-migration exposures, the context of the migration journey, settlement conditions and social environment post migration might impact the potential for SGA. These findings highlight the importance of intergenerational improvements to the wellbeing of migrant women and their children. Further research to identify modifiable elements that contribute to birthweight differences across population groups would help enable appropriate healthcare responses aimed at reducing the rate of being SGA.

Does detection of fetal growth restriction improve neonatal outcomes?

AIM

Timely delivery of fetal growth restriction (FGR) is a balance between avoiding stillbirth and minimising prematurity. We sought to assess the neonatal outcomes for babies suspected of FGR, both true and false positives.

METHODS

This population cohort study examined all singleton births in Victoria, Australia from 2000 to 2017 (n = 1 231 415). Neonatal morbidities associated with neonatal intensive care unit (NICU) admission were assessed for babies born ≥32 weeks' with severe FGR (<3rd centile) and babies with birthweight ≥10th centile who were iatrogenically delivered for suspected FGR.

RESULTS

Babies with severe FGR iatrogenically delivered for suspected FGR were more likely to require NICU admission than babies with severe FGR who were not detected (3.0% vs. 1.1%, P < 0.001). After adjusting for potential confounders, the odds of NICU admission were increased (adjusted odds ratio (aOR) = 3.00, 95% confidence interval = 2.45-3.67; P < 0.001). Rates of NICU admission were also higher in ≥10th centile babies iatrogenically delivered for suspected FGR than for ≥10th centile babies who entered labour spontaneously (1.8% vs. 0.5%, P < 0.001). After adjustments, the odds of NICU admission were increased (aOR = 3.91, 95% confidence interval = 3.40-4.49; P < 0.001). NICU admissions were associated with morbidities related to iatrogenic prematurity.

CONCLUSIONS

Detection and planned delivery of FGR reduces stillbirth but may be associated with increased neonatal morbidity related to iatrogenic prematurity.

Preventing harm: A balance measure for improving the detection of fetal growth restriction

BACKGROUND

Increasing the detection of fetal growth restriction (FGR), while reducing stillbirth, also leads to unnecessary early intervention, and associated morbidity, for normally grown babies who are incorrectly suspected of FGR.

AIMS

We sought to design a balance measure that addresses the specificity of FGR detection.

METHODS

A retrospective cohort study on all singleton births ≥32 weeks gestation in 2016 and 2017 in Victoria. We compared two balance measures for the detection of FGR, defined as the proportion of all babies iatrogenically delivered before 39 weeks gestation for suspected FGR that had a birthweight ≥10th centile (balance measure 1) or ≥25th centile (balance measure 2). Hospital level performance on each balance measure was derived and compared to an existing performance measure for severe FGR detection in Victoria.

RESULTS

Of the 38 hospitals analysed, 12 (32%) had a favourable performance on an existing indicator of FGR detection, seven (18%) hospitals had a favourable performance on balance measure 1, and 15 (39%) had a favourable performance on balance measure 2. There was a moderate correlation between hospital performance on the existing indicator and on balance measure 1 (r = 0.447, P = 0.005) but not balance measure 2 (r = -0.063, P = 0.71). There was no difference in perinatal mortality between high performing hospitals and low performing hospitals.

CONCLUSION

Introducing a balance measure into routine reporting may bring greater awareness to the unintended harm associated with increased detection of FGR.