高级检索

    基于Nomgram模型对老年阵发性房颤病人合并高尿酸血症临床预测模型的初步探索

    Preliminary exploration of clinical prediction model for elderly patients with paroxysm atrial fibrillation complicated with hyperuricemia based on Nomgram model

      摘要
    • 摘要:
      目的 分析老年阵发性房颤(PAF)病人合并高尿酸血症(HUA)的危险因素，并构建Nomgram预警模型。
      方法 回顾性选取老年PAF病人471例作为研究对象，根据是否合并HUA将其分为HUA组45例和非HUA组426例，收集病人的临床资料，分析老年PAF病人合并HUA的危险因素，根据独立危险因素构建Nomgram预警模型并进行预测效能验证。
      结果 2组病人性别、体质量指数(BMI)、血肌酐(Scr)水平和糖尿病、高血压、高血脂、非酒精性脂肪肝情况差异均有统计学意义(P < 0.05~P < 0.01)；logistic回归分析显示，BMI>28 kg/m2、糖尿病、高血压、高血脂、非酒精性脂肪肝及Scr>96 μmol/L均为PAF病人合并HUA的独立危险因素(P < 0.05~P < 0.01)。基于6项独立危险因素建立PAF合并HUA的风险列线图预警模型，验证结果显示，模型一致性指数(C-index)为0.797(95%CI：0.763~0.832)，校正曲线预测值与实测值基本一致，内部验证PAF合并HUA的风险列线图模型AUC为0.803(95%CI：0.779~0.827)，决策曲线显示阈值概率在5%~93%范围内时具有较高的净获益值。
      结论 BMI>28 kg/m2、糖尿病、高血压、高血脂、非酒精性脂肪肝及Scr>96 μmol/L是PAF病人合并HUA的独立危险因素，基于上述危险因素建立的列线图模型可用于评估和量化PAF合并HUA风险。

       

      Abstract:
      Objective To analyze the risk factors of elderly patients with paroxysm atrial fibrillation (PAF) complicated with hyperuricemia (HUA) and construct a Nomgram early warning model.
      Methods A total of 471 elderly PAF patients were retrospectively selected as the study subjects, and divided into HUA group (n=45) and non-HUA (n=426) according to whether they were complicated with HUA.Clinical data were collected to analyze the risk factors of elderly patients with PAF complicated with HUA.A Nomgram warning model was constructed based on independent risk factors, and predictive efficacy was validated.
      Results There were significant differences in sex, body mass index (BMI), serum creatinine (Scr) level and diabetes mellitus, hypertension, hyperlipidemia, nonalcoholic fatty liver between the two groups (P < 0.05 to P < 0.01).Logistic regression analysis showed that BMI>28 kg/m2, diabetes mellitus, hypertension, hyperlipidemia, nonalcoholic fatty liver and Scr>96 μmol/L were independent risk factors for PAF patients complicated with HUA (P < 0.05 to P < 0.01).Based on six independent risk factors, a risk Nomogram warning model for PAF complicated with HUA was established.The validation results showed that the C-index was 0.797 (95%CI: 0.763-0.832), the predicted value of the calibration curve was basically consistent with the measured value, the internal validation of the risk Nomogram model for PAF complicated with HUA had an AUC of 0.803 (95%CI: 0.779-0.827), and the decision curve showed a high net benefit value when the threshold probability was within the range of 5% to 93%.
      Conclusions BMI>28 kg/m2, diabetes mellitus, hypertension, hyperlipidemia, nonalcoholic fatty liver and Scr>96 μmol/L are independent risk factors for PAF patients complicated with HUA.The Nomgram model based on the above risk factors can be used to evaluate and quantify the risk of PAF complicated with HUA.

       

      • HTML全文
      • 参考文献(27)
      • 相关文章
      • 施引文献
    • 资源附件(0)

    /

    返回文章
    返回