基于实验室指标建立的COVID-19与非COVID-19患者鉴别诊断模型

A differential diagnosis model for patients with COVID-19 and non-COVID-19 pneumonia based on laboratory indicators

目的 回顾性分析新型冠状病毒肺炎(COVID-19)患者和同期非COVID-19患者的临床、影像学和实验室指标特征,找寻鉴别诊断指标。方法 选择2020年1月-2020年4月医院收治的COVID-19患者66例和同期非COVID-19患者40例作为研究对象,统计人口学资料、临床症状、基础疾病、影像学特征、住院时间、入院时实验室指标。结果 COVID-19患者白细胞(WBC)、白蛋白(ALB)和前白蛋白(PALB)均低于非COVID-19患者,而住院时间、谷草转氨酶(AST)、国际标准化比值(INR)、纤维蛋白原(Fbg)、乳酸脱氢酶(LDH)、肿瘤特异性生长因子(TSGF)和铁蛋白(Ferritin)均高于非COVID-19患者。COVID-19患者具有更高频率的支气管充气征、网格状改变、病变肺叶数及病变肺段数,以及更低频率的小叶中心结节影。COVID-19患者住院时间与患者年龄正相关,与LYM和ALB呈负相关性。此外,同时合并糖尿病和高血压的患者住院时间多于其他患者。受试者诊断曲线(ROC)显示,WBC、TSGF、LDH和Ferritin在鉴别COVID-19和非COVID-19患者时曲线下面积均>75%。多因素Logistic回归分析表明TSGF、LDH及Ferritin是鉴别COVID-19与非COVID-19的独立因素,三指标联合诊断曲线下面积高达0.9181。结论 普通型COVID-19和非COVID-19患者存在部分临床特征、影像学特征和实验室指标的差异。其中TSGF、LDH和Ferritin联合诊断模型或可成为鉴别普通型COVID-19和非COVID-19患者的有效指标。

OBJECTIVE To retrospectively analyze the clinical characteristics, imaging features and laboratory indexes of the patients with COVID-19 and non-COVID-19 so as to seek for differential diagnosis indexes. METHODS A total of 66 patients with COVID-19 and 40 non-COVID-19 patients were recruited as study subjects who were treated in the hospital from Jan 2020 to Apr 2020. The demographic data, clinical symptoms, underlying diseases, imaging features, length of hospital stay and laboratory test indexes at the admission were statistically analyzed. RESULTS The white blood cell(WBC),albumin(ALB) and prealbumin(PALB) of the COVID-19 patients were remarkably lower than those of the non-COVID-19 patients;while the length of hospital stay, aspartate aminotransferase(AST), international normalized ratio(INR), fibrinogen(Fbg), lactate dehydrogenase(LDH), tumor specific growth factor(TSGF) and ferritin(Ferritin) of the COVID-19 group were remarkably higher than those of the non-COVID-19 group. The COVID-19 patients had a higher frequency of air bronchogram, reticular pattern, number of affected lobes and number of affected segments, but a lower frequency of centrilobular nodules than did the non-COVID-19 patients. The length of hospital stay of the COVID-19 patients was positively correlated with the age but was negatively correlated with LYM and ALB, and the length of hospital stay of the patients complicated with diabetes mellitus and hypertension was longer than the patients with other complications. Receiver operating characteristic(ROC) curve analysis showed that the areas under curves of WBC, TSGF, LDH and Ferritin were more than 75% in distinguishing between COVID-19 and non-COVID-19. Multivariate logistic regression analysis showed that TSGF, LDH and Ferritin were the independent factors for distinguishing between COVID-19 and non-COVID-19, and the area under curve of the joint detection of the three indexes was 0.9181. CONCLUSION The ordinary COVID-19 patients and non-COVID-19 patients vary in some clinical characteristic, imaging features and clinical laboratory indexes. The joint diagnosis model of TSGF, LDH and Ferritin may be used as an effective indicator for distinguishing between ordinary COVID-19 and non-COVID-19.