早产儿坏死性小肠结肠炎影响因素分析及发病预测模型的构建

Risk factors for necrotizing enterocolitis and establishment of prediction model of necrotizing enterocolitis in preterm infants

摘要:

目的 探讨早产儿坏死性小肠结肠炎（necrotizing enterocolitis，NEC）发生的影响因素，制定一个可以预测NEC发生并指导预防的评分表。 方法 回顾性收集2011年1月至2020年12月吉林大学白求恩第一医院新生儿科收治的早产儿的临床资料，分为NEC组（Bell Ⅱ期及以上）（n=298）和非NEC组（n=300），对NEC影响因素进行单因素及多因素统计分析，明确NEC的独立影响因素，并根据影响因素构建预测NEC的列线图，用受试者工作特征曲线及一致性指数（C指数）测量列线图的预测性能。 结果 多因素logistic回归分析显示：Ⅱ度及以上颅内出血、经外周静脉穿刺中心静脉置管、使用母乳强化剂、输红细胞悬液、红细胞比容>49.65%、平均红细胞体积>114.35 fL、平均血小板体积>10.95 fL是NEC的独立危险因素（P<0.05）；使用肺表面活性物质、使用益生菌、血小板分布宽度>11.8 fL是NEC的保护因素（P<0.05）。列线图预测NEC风险的准确性良好，bootstrap校正的C指数为0.844。预测有无NEC的列线图总分最佳截断值为171.02分，灵敏度、特异度分别为74.7%、80.5%。 结论 NEC发病风险预估列线图在指导NEC的早期预判及有针对性的预防及早期干预方面有一定的临床价值。

Abstract:

Objective To investigate the risk factors for necrotizing enterocolitis (NEC) in preterm infants, and to establish a scoring model that can predict the development and guide the prevention of NEC. Methods A retrospective analysis was performed on the medical data of preterm infants who were admitted to the Department of Neonatology,Bethune First Hospital of Jilin University, from January 2011 to December 2020. These infants were divided into two groups: NEC (298 infants with Bell II stage or above) and non-NEC (300 infants). Univariate and multivariate analyses were performed to identify the factors influencing the development of NEC. A nomogram for predicting the risk of NEC was established based on the factors. The receiver operator characteristic (ROC) curve and the index of concordance (C-index) were used to evaluate the predictive performance of the nomogram. Results The multivariate logistic regression analysis showed that grade ≥2 intracranial hemorrhage, peripherally inserted central catheterization, breast milk fortifier, transfusion of red cell suspension, hematocrit >49.65%, mean corpuscular volume >114.35 fL, and mean platelet volume >10.95 fL were independent risk factors for NEC (P<0.05), while the use of pulmonary surfactant, the use of probiotics, and the platelet distribution width >11.8 fL were protective factors against NEC (P<0.05). The nomogram showed good accuracy in predicting the risk of NEC, with a bootstrap-corrected C-index of 0.844. The nomogram had an optimal cutoff value of 171.02 in predicting the presence or absence of NEC, with a sensitivity of 74.7% and a specificity of 80.5%. Conclusions The prediction nomogram for the risk of NEC has a certain clinical value in early prediction, targeted prevention, and early intervention of NEC.