【摘要】 目的 基于血清学标志物构建早产儿视网膜病变(ROP)患儿预后的Nomogram预测模型，并验证模型的预测价值。方法 选取2022年1月至2024年1月河北医科大学第二医院收治的195例(390眼)ROP患儿，治疗后随访3个月，根据预后情况分为预后不良组(n=41)、预后良好组(n=154)。比较两组患儿一般资料、治疗前血清学标志物[血管内皮生长因子(VEGF)、胰岛素样生长因子-1(IGF-1)、谷氨酸(Glu)、信号转导和转录激活因子3(STAT3)、低氧诱导因子3α(HIF-3α)]水平，通过LASSO-Logistic回归分析ROP患儿预后不良的影响因素，根据影响因素构建ROP患儿预后不良的Nomogram预测模型，通过受试者工作特征曲线、校准曲线及决策曲线验证模型的预测价值。结果 预后不良组患儿1 min Apgar、5 min Apgar、病情程度重度占比、支气管肺发育不良占比、败血症占比及治疗前血清VEGF、Glu、STAT3、HIF-3α水平均高于预后良好组，胎龄、出生体重、血清IGF-1水平均低于预后良好组(均为P<0.05);LASSO-Logistic回归分析显示，胎龄、病情程度、支气管肺发育不良、败血症及治疗前血清VEGF、IGF-1、Glu、STAT3、HIF-3α水平均为ROP患儿预后不良的影响因素(均为P<0.05);根据影响因素构建ROP患儿预后不良的Nomogram预测模型，该模型预测ROP患儿预后不良的曲线下面积为0.943(95%CI:0.907～0.978),具有较高预测效能，该模型的校准度良好，预测结果与实际观测结果有较好的一致性，且在预测ROP患儿预后不良方面拥有良好的临床效用。结论 血清VEGF、IGF-1、Glu、STAT3、HIF-3α水平均为ROP患儿预后的影响因素，基于以上血清学标志物构建的ROP患儿预后的Nomogram预测模型具有较高应用价值。更多还原

【Abstract】 Objective To construct a Nomogram prediction model for prognosis of retinopathy of prematurity(ROP) based on serological markers and to validate the predictive value of the model. Methods A total of 195 ROP patients(390 eyes) admitted to the Second Hospital of Hebei Medical University from January 2022 to January 2024 were selected. After 3-month follow-up post-treatment, the patients were divided into poor-prognosis(n=41) and good-prognosis(n=154) groups. General data and pre-treatment serological markers [vascular endothelial growth factor(VEGF), insulin-like growth factor-1(IGF-1), glutamate(Glu), signal transduction and transcriptional activator 3(STAT3), hypoxic inducible factor 3α(HIF-3α)] were compared between the two groups. LASSO-Logistic regression was used to identify risk factors for poor prognosis in ROP patients. A Nomogram prediction model for poor prognosis in ROP patients was constructed based on these factors. The predictive value of the model was validated using receiver operating characteristic(ROC) curves, calibration curves, and decision-curve analysis. Results The 1-min Apgar score, 5-min Apgar score, proportion of severe disease, proportion of bronchopulmonary dysplasia(BPD), proportion of sepsis, and pre-treatment serum levels of VEGF, Glu, STAT3, and HIF-3α were higher in the poor-prognosis group than in the good-prognosis group, while gestational age, birth weight, and serum IGF-1 levels were lower(all P<0.05). LASSO-Logistic regression analysis showed that gestational age, disease severity, BPD, sepsis, and pre-treatment serum levels of VEGF, IGF-1, Glu, STAT3, and HIF-3α were risk factors for poor prognosis in ROP patients(all P<0.05). A Nomogram prediction model for poor prognosis in ROP patients was constructed based on these factors. The model had an area under the curve(AUC) of 0.943(95%CI: 0.907-0.978) for predicting poor prognosis, indicating high predictive efficacy. The model had good calibration, with good consistency between predicted and actual results, and demonstrated good clinical utility in predicting poor prognosis in ROP patients. Conclusion Serum levels of VEGF, IGF-1, Glu, STAT3, and HIF-3α are all prognostic factors for ROP patients. The Nomogram prediction model for ROP prognosis based on these serological markers has high application value.更多还原