基于定量CT构建的诺模图对重症新型冠状病毒肺炎的诊断价值

Establishment and evaluation of normogram in prediction of severe COVID-19 based on quantitative CT

目的探讨基于定量CT构建的诺模图对重症新型冠状病毒肺炎的诊断价值。方法回顾性分析117例新型冠状病毒肺炎(corona virus disease 2019,COVID⁃19)患者的临床和胸部CT资料。所有患者分为轻症组(82例)和重症组(35例)。对两组间差异有统计学意义的临床和定量CT指标进行多因素logistic回归分析,确定重症COVID⁃19相关的独立危险因素,构建诺模图,并通过ROC曲线分析、校准曲线及Hosmer⁃Lemeshow拟合优度检验进行模型验证。结果多因素logistic回归结果显示年龄(OR=1.155,95%CI:1.069~1.247)、淋巴细胞计数与白细胞计数比值(OR<0.001,95%CI:0~0.005)、LeV%(OR=1.136,95%CI:1.013~1.274)、MLeD(OR=1.009,95%CI:1.001~1.018)是重症COVID⁃19的影响因素。绘制诺模图,其ROC曲线下面积为0.969,校准曲线显示预测概率与实际概率符合度良好。Hosmer⁃Lemeshow拟合优度检验(χ^(2)=4.352,P=0.824)显示诺模图诊断重症COVID⁃19具有较好效能。结论基于定量CT构建的诺模图对于重症COVID⁃19的临床诊断具有较好的效能。

Objective To explore the efficiency of nomogram based on quantitative CT in prediction of severe corona virus disease 2019(COVID-19). Methods The clinical and chest CT data on 117 COVID-19 patients were retrospectively analyzed. All the patients were divided into a mild group(82 patients)and a severe group(35 patients). Multivariate Logistic regression analysis was performed on the clinical and quantitative CT indexes which had significant differences between the two groups to determine the independent risk factors associated with severe COVID-19. Nomogram was constructed for predicting severe COVID-19,and then verified by ROC analysis,calibration curve and Hosmer-Lemeshow goodness-of-fit test. Results Multivariate logistic regression showed that age,lymphocyte count-to-white blood cell ratio,LeV%,and MLeD were independent risk factors for severe COVID-19. The area under the curve of nomogram was 0.969. The calibration curve showed that the predicted probability was highly concordant with the actual probability. Hosmer-Lemeshow goodness-of-fit test(χ^(2)= 4.352,P = 0.824)showed that nomogram had higher efficiency in predicting severe COVID-19. Conclusions The nomogram based on quantitative CT has better efficiency and application value in predicting severe COVID-19.