Accurate gestational age (GA) determination is vital for effective management of complicated pregnancies and mitigating the risks associated with prematurity. The widely accepted gold standard involves combining the Last Menstrual Period (LMP) and first-trimester ultrasound measurements. Prenatal ultrasound is known for its precision of ± 5-7 days in early pregnancy, but this accuracy diminishes in later trimesters, reaching ± 21-30 days in the third trimester. However, in low- and middle-income countries (LMICs), prenatal ultrasonography suffers from limited availability, high costs, and insufficient access to healthcare services. Moreover, LMP calculations and clinical tools like the Ballard Score also demonstrate poor accuracy for determining GA, making it crucial to explore alternative methods.
This study aims to assess the effectiveness of a previously published equation for estimating GA based on fetal transverse cerebellar diameter (TCD) measurements, particularly in LMIC settings where the standard methods are not reliably applicable.
The study employs a simulation approach, leveraging data from existing studies focused on fetal TCD measurements in LMICs. The equation under evaluation is derived from a previous study: GA in weeks = 0.470 × TCD in mm + 13.162.
To frame the analysis, only studies meeting specific criteria were considered, including those conducted in LMICs that reported both mean and standard deviation of fetal TCD measurements for each week of GA. These studies had to base GA assessment on a reliable standard, such as validated LMP or first-trimester ultrasound.
Data was collected through PubMed searches utilizing terms related to fetal biometry and TCD in connection with GA estimation in developing countries. Relevant study features such as country, sample size, and TCD measurements were systematically retrieved, culminating in the formation of four distinct datasets for simulation.
Simulations generated TCD values for each dataset based on clinical means and standard deviations. These values were processed to estimate GA using the aforementioned equation. Various statistical tools, including calibration plots and Bland-Altman plots, were applied to assess the accuracy of GA estimations. Pooled biases and 95% limits of agreement (LoA) were also computed to provide a thorough evaluation of the equation's performance across the data subsets.
From the analysis of the four datasets (sample sizes: 400, 257, 450, and 500), findings indicated varied performance of the TCD-based equation:
The results underscore that while the TCD-based equation can assist in estimating GA, particularly for younger gestations (23-32 weeks) in low-resource contexts, its utility diminishes beyond this range. The simulation nature of the study indicates that these findings need further validation in actual newborn cohorts. Future research is encouraged to explore the equation's application in real-world settings where precise GA estimation is crucial for improving neonatal outcomes.
This study proposes that utilizing TCD measurements via ultrasound could serve as a valuable tool for estimating gestational age in resource-limited settings, with potential benefits observed particularly in the lower ranges of GA. However, the variability in precision for later gestational periods highlights the necessity for ongoing evaluation and validation of this method.