Computerized tomography (CT) scans of the chest are not a suitable independent screening tool for coronavirus-19 (COVID-19); use of only CT could lead to misdiagnosis and potential infection risk, according to findings published in European Radiology. However, CT images are valuable for assessing clinical severity and guiding treatment of COVID-19 in conjunction with clinical information.
This retrospective single-center study on patients with COVID-19 in Zhuhai, China, from January 18 to February 7, 2020, divided participants (N=78) into 4 categories based on the diagnosis and treatment plan for COVID-19 issued by the Chinese National Health Commission. Categories were (1) mild type with minimal clinical symptoms without pneumonia on imaging; (2) common type with respiratory tract, fever, and evidence of pneumonia in imaging; (3) severe type with respiratory distress, respiratory rate ≥30 breaths/min in resting state, oxygen saturation ≤93%, ratio of arterial partial pressure of oxygen to fraction of inspired oxygen of ≤300 mm/Hg; and (4) critical type with respiratory failure requiring mechanical ventilation, and having other organ failure and shock requiring treatment in the intensive care unit.
Among the 78 patients included in the study, 24 (30.8%) were categorized as mild, 46 (59.0%) as common, 6 (7.7%) as severe, and 2 (2.6%) as critical; all participants in the cohort were discharged from hospital after a mean of 20±7 days (range, 9-45 days). There was a high consistency of CT visual quantitative analysis between the 2 observers (intragroup correlation coefficient, 0.976; 95% CI, 0.962-0.985).
Findings on CT imaging demonstrated that in 87.5% of cases with evidence of pneumonia had peripheral distribution, 80.4% had ground-glass opacities, 76.8% had mixed ground-glass opacities, 73.2% had air bronchograms, 53.6% had fibrotic lesions, no cavitation was seen, 44.6% had interlobular septal thickening within the lesions, 32.1% had peribronchovascular distribution, 21.4% had consolidation, and 8.9% had pleural effusion. No centrilobular nodules or lymphadenopathy was found.
Further, all 5 lobes were involved in the severe-critical type while the lower lobes were usually involved in the common type (87.0%). Compared with the severe-critical type, the common type had a lower incidence of right upper lobe and middle lobe involvement (P = .016; P = .006, respectively).
The mean total severity score (TSS) in the critical group (TSS, 8-18; median 10, P25 9, P75 15.25) was significantly higher than that in the common group (TSS, 1-11; median, 5; P25, 2.75; P75, 6.25; P <.001). Receiver operating characteristic analysis showed the TSS area under the curve for diagnosing critical participants was 0.918 (95% CI, 0.843-0.994). The TSS cutoff of 7.5 had 100% specificity and 82.6% sensitivity. Among the 78 confirmed patients COVID-19 (positive real-time reverse-transcriptase polymerase chain reaction test for severe acute respiratory syndrome coronavirus 2), 24 (30.8%) with mild type virus had normal chest CT scans.
Study authors concluded that “screening for COVID19 with chest CT alone can lead to misdiagnosis in some patients, which would lead to a potential infection risk, so CT was not suitable as an independent screening tool. Visual quantitative analysis based on CT images has high consistency and high diagnostic ability, which can reflect clinical classification; it is expected to accurately assess the clinical severity of COVID-19 and guide the clinical treatment by combining with the clinical information.”
Li K, Fang Y, Li W, et al. CT image visual quantitative evaluation and clinical classification of coronavirus disease (COVID-19) [published online March 25, 2020] Eur Radiol. doi:10.1007/s00330-020-06817-6