肺部超声评分联合血清标志物对新生儿肺炎肺泡灌洗治疗后预后的预测价值

Predictive value of lung ultrasound score combined with serum markers for prognosis after alveolar lavage treatment in neonatal pneumonia

    摘要
  • 摘要:
    目的 探讨肺部超声评分（LUS）与血清白三烯B4清除率（LTB4c）、降钙素原清除率（PCTc）、C-反应蛋白清除率（CRPc）的动态变化对新生儿肺炎支气管肺泡灌洗治疗后预后不良的预测价值。
    方法 选择2022年1月－2025年1月于海南医科大学第二附属医院收治的300例肺炎新生儿作为研究对象，根据治疗30 d预后情况，分为预后良好组（n=205）和预后不良组（n=95）。采用多元线性回归（MLR）模型分析治疗7～14 d LTB4c、PCTc、CRPc与LUS评分的相关性。采用logistic回归模型分析治疗14 d LUS评分与LTB4c、PCTc、CRPc对预后的相关性。采用限制性立方样条（RCS）模型评估治疗14 d时LUS评分及LTB4c、PCTc、CRPc对治疗30 d预后不良的剂量反应关系。采用受试者工作特征（ROC）曲线分析治疗14 d时LUS评分及LTB4c、PCTc、CRPc对治疗30 d预后不良的预测价值。
    结果 与预后良好组相比，预后不良组患儿治疗7～14 d的LTB4c时、PCTc、CRPc均更低（P＜0.05），LUS均更高（P＜0.05）。调整混杂因素，MLR分析结果显示，治疗14 d时LUS评分与LTB4c、PCTc、CRPc均呈负相关（β=−0.423，P=0.013；β=−0.461，P=0.009；β=−0.511，P=0.006）。logistic回归分析结果显示，排除混杂因素后（模型5），LUS评分、LTB4c、PCTc、CRPc与治疗30 d预后不良具有相关性（P＜0.05）。RCS分析结果显示，LUS评分、LTB4c、PCTc、CRPc连续变化与治疗30 d预后不良风险的关联强度呈非线性剂量反应关系。LUS评分及LTB4c、PCTc、CRPc联合判断治疗30 d预后不良的准确性较高（曲线下面积为0.823）。
    结论 LUS评分及LTB4C、PCTC、CRPc均与新生儿肺炎预后不良相关，且LTB4C、PCTc、CRPc降低与LUS评分升高相关。LUS评分与LTB4C、PCTC、CRPc的联合模型可准确预测患儿预后不良风险。

     

    Abstract:
    OBJECTIVE  To explore the predictive value of dynamic changes in lung ultrasound score （LUS）, serum leukotriene B4 clearance rate （LTB4c）, procalcitonin clearance rate （PCTc） and C-reactive protein clearance rate （CRPc） for poor prognosis after bronchoalveolar lavage treatment in neonatal pneumonia.
    METHODS  A total of 300 neonates with pneumonia admitted to the Second Affiliated Hospital of Hainan Medical University from Jan. 2022 to Jan. 2025 were selected as the study subjects. Based on the recovery status at 30 days after treatment, they were divided into a good prognosis group （n=205） and a poor prognosis group （n=95）. A multiple linear regression （MLR） model was used to analyze the correlation between LTB4c, PCTc, CRPc and LUS scores at 7 to 14 days of treatment. A logistic regression model was employed to analyze the correlation between LUS score at 14 days of treatment and LTB4c, PCTc, CRPc with prognosis. A restricted cubic spline （RCS） model was used to evaluate the dose-response relationship between LUS score at 14 days of treatment and LTB4c, PCTc, CRPc with poor prognosis at 30 days of treatment. The receiver operating characteristic （ROC） curve was used to analyze the predictive value of LUS score at 14 days of treatment and LTB4c, PCTc, CRPc for poor prognosis at 30 days of treatment.
    RESULTS  The poor prognosis group had lower levels of LTB4c, PCTc and CRPc （P＜0.05） and higher LUS scores （P＜0.05） at 7 to 14 days of treatment than the good prognosis group. After adjusting for confounding factors, the MLR analysis revealed that LUS scores at 14 days of treatment were negatively correlated with LTB4c, PCTc and CRPc （β=−0.423, P=0.013. β=−0.461, P=0.009. β=−0.511, P=0.006）. Logistic regression analysis indicated that, after excluding confounding factors （model 5）, LUS scores, LTB4c, PCTc and CRPc were correlated with poor prognosis at 30 days of treatment （P＜0.05）. RCS analysis demonstrated a nonlinear dose-response relationship between the continuous changes in LUS scores, LTB4c, PCTc, CRPc and the risk of poor prognosis at 30 days of treatment. The combined assessment of LUS scores, LTB4c, PCTc and CRPc demonstrated high accuracy in predicting poor prognosis at 30 days of treatment, the area under the curve was 0.823.
    CONCLUSIONS  LUS scores, LTB4c, PCTc and CRPc are all correlated with poor prognosis in neonatal pneumonia, and reductions in LTB4c, PCTc and CRPc are correlated to increases in LUS scores. The combined model of LUS scores with LTB4c, PCTc and CRPc can accurately predict the risk of poor prognosis in children.

     

    • HTML全文
    • 参考文献(24)
    • 相关文章
    • 施引文献
  • 资源附件(0)

/

返回文章
返回
下载中心