多发性骨髓瘤患者贫血发生风险列线图预测模型的构建及验证

Construction and validation of a nomogram prediction model for anemia risk in multiple myeloma patients

目的分析多发性骨髓瘤患者贫血发生的风险因素,据此构建列线图预测模型,并评估其预测价值。方法收集130例多发性骨髓瘤患者的临床资料,根据是否发生贫血分为贫血组(n=56)与非贫血组(n=74)。多发性骨髓瘤患者贫血发生的影响因素采用多因素Logistic回归分析,依据影响因素构建预测多发性骨髓瘤患者贫血发生风险的列线图预测模型。绘制受试者工作特征(ROC)曲线,计算曲线下面积(AUC),评估列线图预测模型对多发性骨髓瘤患者贫血发生风险的预测价值。采用校准曲线评估列线图预测模型的可靠性和准确性,采用决策曲线分析(DCA)评估列线图预测模型的临床应用价值。结果多因素Logistic回归分析结果显示,血小板计数降低、骨髓浆细胞比例升高、血肌酐水平升高以及白细胞介素-6(IL-6)水平升高均为多发性骨髓瘤患者贫血发生的独立危险因素(P﹤0.05),据此构建列线图预测模型。ROC曲线显示,列线图预测模型预测多发性骨髓瘤患者贫血发生风险的AUC为0.951,预测价值较高。校准曲线显示,该列线图模型预测多发性骨髓瘤患者贫血发生风险的C指数为0.778(95%CI:0.741~0.815),表明该列线图模型预测贫血发生风险的观测值与实际观测结果吻合度较好。DCA曲线显示,该列线图预测模型对多发性骨髓瘤患者贫血发生风险的预测有临床意义,临床净效益较高。结论多发性骨髓瘤患者贫血发生风险与血小板计数降低、骨髓浆细胞比例升高、血肌酐水平和IL-6水平升高密切相关,基于这些风险因素所构建的列线图预测模型对多发性骨髓瘤患者贫血发生风险的预测准确度较高。

Objective To analyze the risk factors for anemia in patients with multiple myeloma,construct a nomogram prediction model based on this,and evaluate its predictive value.Method The clinical data of 130 multiple myeloma patients were collected,and divided into anemia group(n=56)and non-anemia group(n=74)based on whether anemia occurred.The influencing factors of anemia in patients with multiple myeloma were analyzed by using multivariate Logistic regression,and a nomogram prediction model was constructed to predict the risk of anemia in patients with multiple myeloma according to the influencing factors.The receiver operating characteristic(ROC)curve was drawn,the area under the curve(AUC)was calculated,and the predictive value of nomogram prediction model for risk of anemia in multiple myeloma patients was evaluated.The reliability and accuracy of the nomogram prediction model were evaluated by calibration curve,and the clinical application value of the nomogram prediction model was evaluated by decision curve analysis(DCA).Result Multivariate Logistic regression showed that decreased platelet count,increased proportion of bone marrow plasma cells,increased blood creatinine levels,and increased interleukin-6(IL-6)were independent risk factors for anemia in patients with multiple myeloma(P<0.05).Based on this,a nomogram prediction model was constructed.ROC curve showed that the AUC of the nomogram prediction model for predicting the risk of anemia in multiple myeloma patients was 0.951,indicating a high predictive value.The calibration curve showed that the C-index of the nomogram prediction model for predicting the risk of anemia in multiple myeloma patients was 0.778(95%CI:0.741-0.815),indicating that the observed values of the nomogram prediction model had good agreement with the actual observed results for predicting the risk of anemia.The DCA curve showed that the prediction model had clinical significance in predicting the risk of anemia in patients with multiple myeloma,with a high clinical net benefit.Conclusion The risk of anemia in patients with multiple myeloma is closely related to decrease in platelet count,increase in the proportion of bone marrow plasma cells,and increase in blood creatinine levels and IL-6.The nomogram prediction model constructed based on these risk factors has high accuracy in predicting the risk of anemia in patients with multiple myeloma.