Abstract: Objective: To construct and validate the risk prediction model and evaluation tool for delayed onset of lactogenesis, to provide an effective tool for early identification of group at risk of delayed onset of lactation. Methods: A Meta-analysis of risk factors for delayed onset of maternal lactation was performed, and the natural logarithm of the combined risk of each risk factor was used as the β coefficient of the model, and the natural logarithm of the ratio of the incidence of delayed onset of maternal lactation to the non-incidence rate at 72 hours after delivery was used as the a coefficient of the model to build the prediction model. Data of 420 maternal cases were collected, and the predictive performance of the model was analyzed. Results: The prediction model for delayed lactation onset was logit (P)= -0.859+0.174×advanced age+0.148×primiparity+0.113×caesarean+0.239×pre-pregnancy overweight or obese+0.182×excessive weight gain during pregnancy+0.166×gestational diabetes mellitus+0.336×hypertensive disorder complicating pregnancy+0.223×breastfeeding started later+0.315×anxiety+0.285×depression. The area under ROC for the model was 0.765, 95%CI (0.717, 0.812), the sensitivity was 0.736, the specificity was 0.712. The risk assessment tool score was 0-20, the score ≥6 was high-risk group, and the area under ROC for the tool was 0.751, 95%CI (0.702, 0.799), the sensitivity was 0.688, and the specificity was 0.719. Conclusion: The risk prediction model and evaluation tool for delayed onset of lactogenesis based on Meta-analysis have good predictive power, and can be used as a tool for early prediction of delayed onset of lactogenesis.

Key words: onset of lactogenesis; risk factor; prediction model; assessment tool; Meta-analysis