Am J Perinatol 2019; 36(08): 818-827
DOI: 10.1055/s-0038-1675161
Original Article
Thieme Medical Publishers 333 Seventh Avenue, New York, NY 10001, USA.

Population versus Customized Growth Curves: Prediction of Composite Neonatal Morbidity

Hector Mendez-Figueroa
1   Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Baylor College of Medicine, Houston, Texas
,
Suneet P. Chauhan
2   Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, Texas
,
Tyisha Barrett
1   Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Baylor College of Medicine, Houston, Texas
,
Van Thi Thanh Truong
3   Center for Clinical Research and Evidence-Based Medicine, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, Texas
,
Claudia Pedroza
3   Center for Clinical Research and Evidence-Based Medicine, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, Texas
,
Sean C. Blackwell
2   Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, Texas
› Institutsangaben
Weitere Informationen

Publikationsverlauf

07. Februar 2018

10. September 2018

Publikationsdatum:
02. November 2018 (online)

Preview

Abstract

Objective To assess the ability of customized and population growth nomograms in identifying newborns with composite neonatal morbidity (CNM).

Study Design This study included women who participated in the 10 Maternal-Fetal Medicine Units (MFMU) trials and delivered a nonanomalous singleton with a known gestational age (GA) of 24 weeks or more and documented birthweight. Population nomograms were based on Alexander's nomogram, whereas customized nomograms used publicly available softwares. Random-effect logistic regression was used to estimate the adjusted odds ratio (aOR). Positive and negative likelihood ratios (LRs) were calculated to assess nomogram performance.

Results Of 92,225 women, 85% met the inclusion criteria. Using the population nomogram, 12% were small for gestational age (SGA) and 10% were large for gestational age (LGA), and using customized nomograms, 15% were SGA and 16% LGA. SGA newborns had a higher likelihood of CNM (aOR: 2.62; 95% confidence interval [CI]: 2.48–2.76) for population nomograms and 3.22 (95% CI: 3.07–3.39) for customized nomograms. LGA newborns had a similar CNM with population nomogram but significantly higher with customized nomogram (aOR: 1.42; 95% CI: 1.34–1.50). For the adverse outcomes among SGA and LGA, the positive LRs for the two nomograms were similar with overlapping 95% CI.

Conclusion Though both SGA and LGA are associated with adverse perinatal outcomes, the detection using both nomograms was similar.

Note

This study was presented at the Society of Maternal-Fetal Medicine Annual Meeting in Las Vegas, NV, January 2017.


Condensation

Compared with population nomogram, customized growth nomogram has a slightly improved detection of the risk of stillbirth and neonatal morbidity/mortality.