Objective: To construct and evaluate a diagnosis pathway (Xiangya pathway) for Corona Virus Disease 2019 (COVID-19). Methods: Consecutive subjects aged ≥12 years old who were screened for COVID-19 were included in Xiangya Hospital of Central South University from January 23 to February 3, 2020, and the subjects were further divided into the inception cohort and the validation cohort. The gender, age, onset time of disease of the subjects were recorded. The information of epidemiological history, fever, and the declined blood lymphocytes were collected as clinical indicators, CT scan was used to evaluate the possibility of COVID-19 and range of lung involvement. According to the current Chinese national standards, throat swabs of suspected cases were collected and the nucleic acid of COVID-19 was detected by reverse transcription-polymerase chain reaction (RT-PCR). The Xiangya pathway was constructed with multi-indexes, compared with clinical indicators, CT results and Chinese national standards, their effectiveness of detecting confirmed cases were verified in the inception and validation cohort. Results: A total of 382 consecutive adults who was screened for COVID-19 were included, and 261 cases were in the inception cohort and 121 cases were in the validation cohort. Among the 382 cases, 192 were males (50.3%) and 190 were females (49.7%), with a median age of 35 years (range: 15-92 years). There were 183 cases (47.9%) with epidemiological history, 275 cases (72.0%) with fever, 212 cases (55.5%) with decreased peripheral blood lymphocytes, 114 cases (29.8%) with positive CT findings, 43 cases (11.3%) with positive CT-COVID-19, and 30 cases (7.9%) with positive virus nucleic acid by throat swab. Compared with clinical indicators, the sensitivity and specificity of CT were 0.950 and 0.704, respectively. The accuracy of CT to make a definite diagnosis was higher than that of epidemiological history, fever, and declined blood lymphocyte count (0.809 vs 0.660, 0.532, 0.596, P=0.001, 0.002, 0.003, respectively). The sensitivity of this pathway and the pathway recommended by the Health Commission of China were both high (all were 1.000), while the specificity and accuracy of the Xiangya pathway were higher than the one recommended by the Health Commission (0.872 vs 0.765, 0.778 vs 0.592, both P<0.001). The CT-COVID-19 reduced the missed diagnosis rate caused by false negative of nucleic acid test (31 vs 64), with difference rate of 51.6%, and the positive rate of nucleic acid test was 64.5% (20/31). In validation cohort, the specificity and accuracy of the Xiangya pathway was 0.967, the positive rate of nucleic acid test was 76.9%(10/13). Conclusions: The Xiangya pathway can predict the nucleic acid test results of COVID-19, and can be applied as a reliable strategy to screen patients with suspected COVID-19 among people aged ≥12 years in areas other than Hubei during the epidemic period of COVID-19. The cohort size needs to be increased for further validation.
目的： 构建新型冠状病毒肺炎（COVID-19）的诊断流程（简称湘雅流程），评估其准确度。 方法： 纳入2020年1月23日至2月3日在中南大学湘雅医院排查COVID-19的≥12岁人群连续队列，按时间顺序分为试验库和验证库；记录该队列患者的性别、年龄、病程、流行病学史及发热，检测血常规及流感抗体。所有患者均行胸部CT，评价罹患COVID-19的可能性（第1类和第2类被认为CT-COVID-19阳性，可能为COVID-19），评估肺受累范围。按国家卫健委推荐诊疗方案（简称卫健委方案），采集疑似病例咽拭子标本，实时荧光定量聚合酶链式反应（RT-PCR）法检测新型冠状病毒核酸，在试验库中计算流行病学史、发热、血淋巴细胞降低及CT检出核酸阳性的效能，多指标构建COVID-19诊断的湘雅流程，检验湘雅流程及卫健委方案检出核酸阳性的效能，最后在验证库内验证湘雅流程的效能。 结果： 共纳入382例排查者，其中试验库261例、验证库121例；男192例（50.3%）、女190例（49.7%），中位年龄35岁（范围：15~92岁），有流行病学史183例（47.9%）、发热275例（72.0%）、血淋巴细胞降低212例（55.5%），胸部CT异常114例（29.8%），CT-COVID-19阳性43例（11.3%），咽拭子核酸阳性30例（7.9%）。胸部CT检出核酸阳性的灵敏度及特异度分别为0.950与0.704，准确度（0.809）均高于流行病学史（0.660）、发热（0.532）及血淋巴细胞降低（0.596）（P=0.001、0.002、0.003）。湘雅流程与卫健委方案均具有较高的灵敏度（均为1.000），但特异度、准确度均高于后者（0.872比0.765、0.778比0.592，均P<0.001）。湘雅流程需检测核酸的人数少于卫健委方案（31比64例），相差率为51.6%，核酸检测阳性率为64.5%（20/31）。在验证库中，湘雅流程筛选疑似病例的准确度为0.967，核酸检测阳性率为76.9%（10/13）。 结论： 湘雅流程能预判新型冠状病毒核酸的检测结果，在流行期内可作为湖北省外地区≥12岁人群筛选COVID-19疑似病例的参考，但还需扩大样本进一步验证。.
Keywords: Clinical protocols; Diagnosis; Pneumonia, viral; Program evaluation; Tomography, X-ray computed.