Construction and validation of risk prediction model for delayed onset of lactogenesis stage Ⅱ following cesarean section / 中华围产医学杂志

ABSTRACT

Objective:To explore the risk factors of delayed onset of lactogenesis stage Ⅱ(DOL Ⅱ) after cesarean section, and to establish a risk prediction model.Methods:This study involved 330 women who underwent cesarean section in the First Affiliated Hospital of Zhengzhou University from September 2021 to January 2022 and were further divided into DOL Ⅱ group ( n=104) or non-DOL Ⅱ group ( n=226). All clinical data were compared. Univariate analysis and multivariate logistic regression were used to analyze the independent risk factors of DOL Ⅱ after cesarean section to establish the risk prediction model and draw nomogram. The predictive validity of the model was evaluated by the area under the receiver operating curve (AUC) and the goodness of fit was verified by Hosmer-Lemeshow test. Another 129 women who underwent cesarean section in our hospital from February to March 2022 were recruited for external validation. Results:The incidence of DOL Ⅱ in the model development and validation cohort were 31.5%(104/330) and 31.0%(40/129), respectively . In the model developing cohort, significant differences were shown in the constituent ratio of different groups of pre-pregnant body mass index {lean[9.7%(22/226) vs 3.8%(4/104)]; fit [66.8%(151/226) vs 62.5%(65/104)]; overweight or obsess[23.5%(53/226) vs 33.7%(35/104)]}, the proportion of primiparas [50.4%(114/226) vs 61.5%(64/104)], breastfeeding education during pregnancy [64.2%(145/226) vs 40.4%(42/104)] and maternal separation [36.3%(82/226) vs 50.0%(52/104)], and the gestational age at delivery[38.0(36.0-39.0) vs 37.0(35.0-38.5) weeks] and frequency of breastfeeding within 48 h post-delivery [6.0(3.0-9.0) vs 2.0(0.5-5.0)] between the DOL Ⅱ and non-DOI Ⅱ group (all P<0.05). The predictors included in the model were pre-pregnancy overweight or obese ( OR=4.040, 95% CI1.196-13.651), primipara ( OR=1.866, 95% CI1.079-3.227), breastfeeding education during pregnancy ( OR=0.582, 95% CI0.339-1.000), and frequency of breastfeeding within 48 h after delivery ( OR=0.791, 95% CI0.720-0.857). The Hosmer-Lemeshow test showed that P=0.814 and the AUC was 0.784(95% CI0.731-0.837). The Youden index was 1.504 with a sensitivity of 0.769 and a specificity of 0.735. While for the external validation, Hosmer-Lemeshow test showed that P=0.260 and the AUC was 0.751(95% CI0.661-0.840). The Youden index was 1.460 with a sensitivity of 0.775 and a specificity of 0.685. Conclusion:The risk prediction model for DOL Ⅱ developed in this study has a good consistency and predictive performance, which can provide a reference for clinical screening of mothers at high risk of DOL Ⅱ following cesarean section.