Department of Pulmonary Diseases, Uludağ University Faculty of Medicine, Bursa, Türkiye.
Department of Biostatistics, Uludağ University Faculty of Medicine, Bursa, Türkiye.
Tuberk Toraks. 2023 Dec;71(4):325-334. doi: 10.5578/tt.20239601.
In a resource-constrained situation, a clinical risk stratification system can assist in identifying individuals who are at higher risk and should be tested for COVID-19. This study aims to find a predictive scoring model to estimate the COVID-19 diagnosis."
Patients who applied to the emergency pandemic clinic between April 2020 and March 2021 were enrolled in this retrospective study. At admission, demographic characteristics, symptoms, comorbid diseases, chest computed tomography (CT), and laboratory findings were all recorded. Development and validation datasets were created. The scoring system was performed using the coefficients of the odds ratios obtained from the multivariable logistic regression analysis."
Among 1187 patients admitted to the hospital, the median age was 58 years old (22-96), and 52.7% were male. In a multivariable analysis, typical radiological findings (OR= 8.47, CI= 5.48-13.10, p< 0.001) and dyspnea (OR= 2.85, CI= 1.71-4.74, p< 0.001) were found to be the two important risk actors for COVID-19 diagnosis, followed by myalgia (OR= 1.80, CI= 1.08- 2.99, p= 0.023), cough (OR= 1.65, CI= 1.16-2.26, p= 0.006) and fatigue symptoms (OR= 1.57, CI= 1.06-2.30, p= 0.023). In our scoring system, dyspnea was scored as 2 points, cough as 1 point, fatigue as 1 point, myalgia as 1 point, and typical radiological findings were scored as 5 points. This scoring system had a sensitivity of 71% and a specificity of 76.3% for a cut-off value of >2, with a total score of 10 (p< 0.001).
The predictive scoring system could accurately predict the diagnosis of COVID-19 infection, which gave clinicians a theoretical basis for devising immediate treatment options. An evaluation of the predictive efficacy of the scoring system necessitates a multi-center investigation.
在资源有限的情况下，临床风险分层系统可以帮助识别高风险人群，并对其进行 COVID-19 检测。本研究旨在寻找一种预测评分模型来估计 COVID-19 诊断。
本回顾性研究纳入了 2020 年 4 月至 2021 年 3 月期间在急诊大流行诊所就诊的患者。入院时记录了人口统计学特征、症状、合并症、胸部计算机断层扫描（CT）和实验室检查结果。建立了评分系统并进行了验证。使用多变量逻辑回归分析得到的比值比系数进行评分系统分析。
在 1187 名住院患者中，中位年龄为 58 岁（22-96 岁），52.7%为男性。多变量分析发现，典型影像学表现（OR=8.47，95%CI=5.48-13.10，p<0.001）和呼吸困难（OR=2.85，95%CI=1.71-4.74，p<0.001）是 COVID-19 诊断的两个重要危险因素，其次是肌痛（OR=1.80，95%CI=1.08-2.99，p=0.023）、咳嗽（OR=1.65，95%CI=1.16-2.26，p=0.006）和乏力症状（OR=1.57，95%CI=1.06-2.30，p=0.023）。在我们的评分系统中，呼吸困难得 2 分，咳嗽得 1 分，乏力得 1 分，肌痛得 1 分，典型影像学表现得 5 分。该评分系统的截断值>2 时，其灵敏度为 71%，特异性为 76.3%，总分为 10 分（p<0.001）。
预测评分系统可以准确预测 COVID-19 感染的诊断，为临床医生制定即时治疗方案提供了理论依据。需要进行多中心研究评估该评分系统的预测效果。
