Taiwan Baby data collection update on December 17th, 2021

Ksenia Olisova

Data collection for the Taiwan Baby study started on June 10, 2020; the results presented in the current update are up until December 17, 2021. In total, 225 women returned for 764 biometry scans between 14 and 38 weeks of gestation. The majority were recruited (Table 1) in the Taiji clinic (n=501) and 大千(n=220). Data collection for our study coincided with COVID-19 pandemic, as a result most of the research centers were unable to make a  noticeable input. To check intrarater variability we used the concordance correlation coefficient (CCC above 0.95 suggests high concordance between first and second scan) for all measurements in our study CCC was 0.99.

Table 1 Number of scans per study site between June 10, 2020 and December 17, 2021 （Property of Taiji Clinic） 

Along with providing local data about fetal growth patterns in healthy singleton pregnancies, one of the major objectives for Taiwan Baby study is to compare the data from Taiwanese population with international prescriptive fetal growth standards provided by the INTERGROWTH 21-st. It is important to add our findings to the growing body of research regarding fetal growth patterns, as the debate about universal international growth standards continues up to date. The INTERGROWTH team is hypothesized that if provided optimal conditions, regardless ethnicity fetuses grow similarly. Their main argument is that genetics plays minimal role in defining the fetal growth, nutrition, socio-economic situation, environment, and access to quality health care have more significant impact. Thus, if put in optimal conditions, fetuses can reach their growth potential, which is similar between different locations. On the other hand, there are multiple studies pointing out at the importance of local standards and noticeable differences between ethnicities. Following the methodology provided by the INTERGROWTH team, head circumference (HC) was the primary measurement for comparing distribution between the studies. As we can see in Figure 1, centiles built using INTERGROWTH's formula and TWBB data for HC - the 2.5th, 97.5th, and mean are very similar. 

Figure 1 Comparison between centiles produced by Taiwan Baby and INTERGROWTH（Property of Taiji Clinic）

To further support our findings, we calculated Z-scores using the INTERGROWTH reference range and checked the normality of the Z-score distribution by the Kolmogorov-Smirnov test, also checked if 95% of our data lay within 2.5th and 97.5th centiles. In  Figure 2 we can see, Z-scores followed a normal distribution with Kolmogorov-Smirnov test’s p-value=0.56, mean=-0.16, SD=0.93, and 96.7% of our HC measurements are within the 95% confidence interval. As a result, we can suppose that our data follows similar distribution as INTERGROWTH.

Figure 2 Distribution of head circumference Z-scores,  95% confidence interval（Property of Taiji Clinic）

We can see analogous tendencies for other measurements as well, biparietal diameter (outer-to-outer, BPDoo), occipitofrontal diameter, abdominal circumference, femur length measurements in our sample all have 2.5th, 50th, and 97.5th centiles similar with such of respective INTERGROWTH measures. In addition to the abovementioned measurements, Taiwan Baby data collection includes humerus length and biparietal diameter measured outer-to-inner (BPDoi). 

In conclusion, our results suggest an acceptable intrarater reproducibility and head circumference measured in TWBB study follows distribution like INTERGROWTH 21-st. Further data collection and continuous quality control are crucial for the success of our study. We hope to increase the number of pregnancies recruited in other study sites to improve the representativeness of our sample.