Supplementary Material for: Unisex and Sex-Specific Prescriptive Fetal Growth Charts for Improved Detection of Small-for-Gestational-Age Babies in a Low-Risk Population: a post hoc analyses of a cluster randomized study

Introduction
Our aim was to develop and evaluate the performance of population-based sex-specific and unisex prescriptive fetal abdominal circumference growth charts in predicting small-for-gestational-age (SGA) birthweight, severe SGA (sSGA) birthweight and severe adverse perinatal outcomes (SAPO) in a low-risk population.

Methods
This is a post hoc analysis of the Dutch nationwide cluster-randomised IRIS study, encompassing ultrasound data of 7,704 low-risk women. IRIS prescriptive unisex and IRIS sex-specific abdominal circumference (AC) fetal growth charts were derived using quantile regression. As comparison we used the descriptive unisex Verburg chart, which is commonly applied in the Netherlands. Diagnostic parameters were calculated based on the 34-36 weeks ultrasound.

Results
Sensitivity rates for predicting SGA and sSGA birthweights were more than twofold higher based on the IRIS prescriptive sex-specific (respectively SGA 43%; sSGA 59%) and unisex (SGA 39%; sSGA55%) charts, compared to the Verburg chart (SGA16%; sSGA23% both p < 0.01). Specificity rates were highest for Verburg (SGA 99%; sSGA98%) and lowest for IRIS sex-specific (SGA 94%; sSGA 92%). Results for predicting SGA with SAPO were similar for the prescriptive charts (44%), and again higher than the Verburg chart (20%). The IRIS sex-specific chart identified significantly more males as SGA and sSGA (resp. 42%; 60%, p<0.001) than the IRIS unisex chart (resp. 35%; 53% p < 0.01).

Conclusion
Our study demonstrates improved performance of both the IRIS sex-specific and unisex prescriptive fetal growth compared to the Verburg descriptive chart, doubling detection rates of SGA, sSGA and SGA with SAPO. Additionally, the sex-specific chart outperformed the unisex chart in detecting SGA and sSGA. Our findings suggest potential benefits of using prescriptive AC fetal growth charts in low-risk populations and emphasize the importance of considering customizing fetal growth charts for sex. Nevertheless, the increased sensitivity of these charts should be weighed against the decrease in specificity.

引言
本研究旨在开发并评估基于人群的性别特异性与通用预设胎儿腹围生长曲线，在低风险人群中预测小于胎龄儿（small-for-gestational-age, SGA）出生体重、重度小于胎龄儿（severe SGA, sSGA）出生体重以及重度不良围产结局（severe adverse perinatal outcomes, SAPO）的效能。

方法
本研究为荷兰全国整群随机IRIS研究的事后分析，共纳入7704名低风险孕妇的超声检查数据。本研究采用分位数回归法构建IRIS通用预设胎儿腹围（abdominal circumference, AC）生长曲线与IRIS性别特异性胎儿腹围生长曲线。以荷兰临床常用的描述性通用Verburg生长曲线作为对照。诊断效能指标基于孕34~36周的超声检查结果计算得出。

结果
相较于Verburg生长曲线（预测小于胎龄儿灵敏度为16%，重度小于胎龄儿灵敏度为23%，二者P均<0.01），IRIS性别特异性生长曲线（分别为43%、59%）与IRIS通用预设生长曲线（分别为39%、55%）预测小于胎龄儿与重度小于胎龄儿出生体重的灵敏度均提升一倍以上。Verburg生长曲线的特异度最高（小于胎龄儿99%，重度小于胎龄儿98%），IRIS性别特异性生长曲线的特异度最低（小于胎龄儿94%，重度小于胎龄儿92%）。预设生长曲线预测合并重度不良围产结局的小于胎龄儿的效能相似（灵敏度为44%），同样高于Verburg生长曲线（20%）。相较于IRIS通用预设生长曲线（分别为35%、53%，P<0.01），IRIS性别特异性生长曲线可识别出更多男性小于胎龄儿与重度小于胎龄儿（分别为42%、60%，P<0.001）。

结论
本研究表明，相较于Verburg描述性生长曲线，IRIS性别特异性与通用预设胎儿生长曲线的效能均得到提升，小于胎龄儿、重度小于胎龄儿以及合并重度不良围产结局的小于胎龄儿的检出率均提升一倍。此外，在检测小于胎龄儿与重度小于胎龄儿时，性别特异性生长曲线的表现优于通用生长曲线。本研究结果提示，在低风险人群中使用预设胎儿腹围生长曲线可能存在临床获益，同时强调了针对胎儿生长曲线进行性别定制的重要性。但需注意，此类生长曲线灵敏度的提升需与特异度的下降进行权衡。