Accuracy of fetal growth estimation among women with type 1 and type 2 diabetes in pregnancy from a multi-ethnic population — ASN Events
  1. Schedule
  2. Lunch & Poster Session
  3. Accuracy of fetal growth estimation among women with type 1 and type 2 diabetes in pregnancy from a multi-ethnic population

Accuracy of fetal growth estimation among women with type 1 and type 2 diabetes in pregnancy from a multi-ethnic population (#110)

Akhil Gupta 1 , Minha Lee 2 , Raiyomand Dalal 2 , David Simmons 2
  1. Western Sydney University, Macarthur Clinical School, Campbelltown, NSW, Australia
  2. Western Sydney University, Campbelltown, NSW, Australia

Background and Introduction:

Decisions regarding mode and timing of delivery are often influenced by estimated fetal weight (EFW) using ultrasonography. Genetic, ethnic, lifestyle, environmental and other factors determine “normal ranges” for EFW, warranting region-specific centiles. This study is a pragmatic, retrospective audit of ultrasound reports at an outer-metropolitan hospital in Sydney, Australia between 2012 and 2025 among women with type 1 and type 2 diabetes.

Methods:

Estimated Date of Confinement (EDC) from early dating ultrasound reports were recorded. Raw fetal biometric measurements from third trimester singleton pregnancy ultrasounds were extracted: head circumference (HC), abdominal circumference (AC), femur length (FL), biparietal diameter (BPD), EFW by Hadlock.  Antenatal demographic data and final birthweight at delivery were collected from electronic medical records. EFW centiles at scanning were standardised using two methodologies: (1) The Fetal Medicine Foundation (FMF) Calculator AND (2) The GROW Centile Calculator. Birthweight centiles at delivery were standardised using the same Calculators. High precision scanning was defined as no more than 10% difference between EFW centile at the time of scanning and birthweight centile at time of delivery. FMF and GROW calculators were compared using Cohen’s kappa. Scan precision between subgroups (public hospital vs others, obese vs non-obese) was compared using chi-square.

Results:

In a diverse ethnic population (31.9% European, 13.5% South Asian, 13.5% Pacific Islands), 255 ultrasound reports from two public hospitals and >20 external providers were included. Mean age was 31±6 years, mean BMI 31.7±7.6 Kg/m2 and 28.5% had type 1 diabetes.

The FMF vs GROW calculators showed weak correlation among both public hospital sonographers (Cohen’s k=0.564) and external providers (Cohen’s k=0.596).  Correlation remained weak regardless of BMI with Cohen’s k = 0.577 in those with BMI <30 kg/m2 and 0.576 in those with BMI >30kg/m2. The highest precision scans were achieved in the public hospital with 50% accuracy (51/101 scans) compared to external providers with 37% accuracy (58/154 scans) when using GROW calculators (p=0.043).

Conclusions:

In women with pre-existing diabetes mellitus, an experienced maternal fetal medicine unit achieved more precise ultrasound scanning than other providers. The choice of fetal growth calculator affects scanning performance, particularly in multi-ethnic populations.