
Patrick M. Colletti
Professor of Radiology, University of Southern California
Section Editor for Cardiopulmonary Imaging, AJR
Published April 2, 2020
At the time of this writing, the American Journal of Roentgenology (AJR) has received more than 100 manuscripts describing imaging in patients infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Thus far, 18 articles and letters have been published online, open-access and ahead-of-print, in the AJR Coronavirus Disease (COVID-19) Collection.
Ultimately, what should a radiology department do during an infectious disease outbreak? Cheng and colleagues from Singapore General Hospital presented an approach to COVID-19 safety for imagers based on their experience with severe acute respiratory syndrome (SARS) in 2003. They list important actions to be carried over from that experience to protect and optimize radiology department operation:

- Share information so that all team members understand moment-to-moment changes in risks and resources needed to safely manage patients.
- Personal protection equipment must be made available and properly donned and duffed.
- Potentially infected patients must be identified and isolated.
- Ideally, dedicated CT scanners will be identified and managed for high-risk patients.
- Physical security and access control with proper signage must be assured (Fig. 1).
- Alternate decentralized work areas should be identified.
- Interventional radiology procedures should be modified for safety and efficiency.
- Radiologists must rapidly report potential COVID-19 findings electronically and by telephone conversation, when appropriate.
So, what’s going on in Singapore today? As of April 14, 2020, there were just over 1,300 active cases of COVID-19 in the country. Of these, 1,287 patients were hospitalized but in stable condition, while 28 were listed as critical. Singapore recorded its 10th death from COVID-19 on April 14.
Similarly, Hosseiny et al. compared the clinical and imaging findings of COVID-19 with those of two previous coronavirus infections: SARS and Middle East respiratory syndrome (MERS). There are similarities, but there are differences, too. Clinical signs and symptoms of COVID-19 include fever, dyspnea, and dry cough. Complaints of sore throat and diarrhea are less common in most reported cases, though there is substantial variation in presentation other than fever. Typical COVID-19 findings on CT include multilobar ground-glass opacities (GGOs) often with consolidation. Normal CT, as seen in perhaps 15%–20% of scans, does not exclude SARS-CoV-2 infection. As expected, consolidation is an indicator for poor prognosis. Pulmonary fibrotic changes after recovery are less well-described.

Open Access COVID-19 Resources
ARRS is committed to providing all radiologists with open access to the latest imaging research on COVID-19 to help understand the imaging features associated with coronavirus.
In Wuhan, Hubei, China, Han and colleagues described early clinical and CT manifestations of COVID-19 pneumonia. Clinical manifestations in the 108 patients they studied were fever in 94 (87%), dry cough in 65 (60%), and fatigue in 42 (39%) patients. Laboratory findings included normal WBC count in 97 (90%), normal or reduced lymphocytes in 65 (60%), and high-sensitivity C-reactive protein elevation in 107 (99%) patients. CT distribution included one lobe in 38 (35%), two or three lobes in 24 (22%), and four or five lobes in 46 (43%) scans. Most lesions were peripheral (97 [90%]) and patchy (93 [86%]). GGOs were seen in 65 (60%) scans, with consolidation in 44 (41%) scans. The size of opacities varied from less than 1 cm (10 [9%]) to more than 3 cm (56 [52%]). Vascular thickening was noted in 86 (80%), the “crazy-paving” pattern was found in 43 (40%), air bronchograms were seen in 52 (48%), and the halo sign appeared in 69 (64%) CT scans.
Zhou and colleagues, also in Wuhan, described their findings in 62 patients with COVID-19 pneumonia. They emphasized GGOs and bronchial distortion as signs of COVID-19. Again, as of today, Wuhan seems to be doing well. China has lifted its 76-day lockdown, and the city is reemerging from the coronavirus crisis. From various news reports, you can see that the citizens of Wuhan are wearing protective masks—some of them better than the masks that we have in the United States.
Meanwhile, in Shanghai, China’s most populous city, Cheng and colleagues pointed out that frontline physicians and radiologists should consider the diverse imaging presentations of COVID-19. A reverse transcription–polymerase chain reaction (RT-PCR) test remains necessary for patients with uncertain imaging findings, and testing is crucial for control of the outbreak—especially during the early period, when patients’ exposure history may be unknown.
Back in Hubei Province, Li and Xia from Tongji Hospital reported that, from their early experience, CT had a low rate of missed diagnosis of COVID-19 (3.9%, 2/51) and thus, “may be a standard method for the diagnosis of COVID-19 based on CT features.” The co-authors explained further: “Rapid diagnosis can lead to early control of potential transmission. With CT diagnosis of viral pneumonia, patients with suspected disease can be isolated and treated in time so that the management of patients will be optimized, especially for the hospitals or communities lacking nucleic acid testing kits.” They concluded, however, that “for the identification of specific viruses, CT is still limited,” also noting that “it is valuable for radiologists to recognize that the CT findings of COVID-19 overlap with the CT findings of diseases caused by other viruses.”
From Hunan, China, Zhao et al. reported on the relationship between chest CT findings and clinical conditions of COVID-19 pneumonia in a multicenter study of 101 patients retrospectively collected from four institutions. Most patients, 70%, were 21–50 years old, and 5% of the patients had family outbreaks. Fever was the onset symptom for 78% of patients. Fourteen patients in the emergency group were older than those in the nonemergency group. Most patients with COVID-19 pneumonia had GGOs (87 [86%]) or mixed GGOs and consolidation (65 [64%]), vascular enlargement (72 [71%]), and traction bronchiectasis (53 [52%]). Lesions were more likely peripheral (88 [87%]) and bilateral (83 [82%]) and lower lung predominant (55 [54%]) and multifocal (55 [54%]).
Salehi and colleagues published a nice systematic review of imaging findings in 919 patients with COVID-19. They concluded that although the majority of COVID-19 mortalities occur among patients with acute respiratory distress syndrome in the ICU, “in a patient population with low pretest probability of [SARS-CoV-2] infection, the typical imaging features should be interpreted with caution.”
One of the most unique papers AJR has published came from Wuhan. Liu et al. authored a preliminary analysis of the pregnancy and perinatal outcomes of women with COVID-19 pneumonia. Of the 15 pregnant women with chest CT-documented COVID-19, 11 had successful deliveries (10 cesarean, one vaginal) and four were still pregnant (three in the second trimester, one in the third) at the time of publication. Importantly, there were no abortions, neonatal asphyxias, neonatal deaths, stillbirths, or neonatal SARS-CoV-2 infections in any of the newborns. More recently, some papers have confirmed early-onset infection in neonates born to mothers with COVID-19, but mother-to-child transfer was not seen in this initial study of 15 patients.
Huang et al. in Wuhu, China analyzed 25 patients with RT-PCR-documented COVID-19. CT scores were rated 0–35 based on extent and intensity of lung involvement. Data were separated into two groups, based on time from symptom onset to diagnosis and treatment: group 1 was patients for whom this interval was less than or equal to 3 days and group 2 was those for whom the interval was greater than 3 days). CT scores were plotted against time, and after analyzing the resulting curves, the mean peak CT score was 10 and 16 for group 1 and 2, respectively, and the mean time to disease resolution was 6 and 13 days, respectively. The last CT scores were lower for group 1 than for group 2 (p = 0.025), which led to the conclusion that timely diagnosis and treatments are keys to providing a better prognosis for patients with COVID-19.
In early encounters with COVID-19 pneumonia, typical chest CT findings created the impression that CT could successfully screen for infected patients (Fig. 2).

A, Initial unenhanced axial chest CT image 4 days after admission shows multiple bilateral subsegmental peripheral patchy and ground-glass opacities with obscure boundaries and mainly subpleural distribution. Neither pleural effusion nor enlarged mediastinal lymph nodes are seen.
B and C, Unenhanced axial CT images 6 (B) and 12 (C) days after admission and after initiation of treatment shows enlarged lesions; lesions of both lungs are diffuse and patchy compared with previous CT images (A).
D, Unenhanced axial CT image 22 days after admission shows lesions in both lungs have absorbed gradually after treatment, and subpleural line can be seen.
On occasion, CT imaging showed asymptomatic opacities while RT-PCR testing was negative. As experiences with less-enriched COVID-19 cohorts were encountered, we learned that CT was considerably less efficient at detecting the many asymptomatic patients with COVID-19, especially compared with nucleic acid testing.
Logically, asymptomatic community members do not require RT-PCR testing unless there has been a known or potential exposure to COVID-19. CT is best reserved in planning therapy on selected patients with symptomatic COVID-19, or if doctors have reasonable suspicion that RT-PCR is falsely negative.
Of course, whereas the findings of CT lung opacities typical for COVID-19 may appear to be statistically reliable in the early stages of a pandemic, alternative diagnoses, including other infections and inflammatory conditions, cannot be readily excluded by image pattern alone.
As the newest article in AJR’s Coronavirus Disease Collection by Raptis and colleagues makes clear, “the radiology literature on COVID-19 has consisted of limited retrospective studies that do not substantiate the use of CT as a diagnostic test for COVID-19.”
The opinions expressed in InPractice magazine are those of the author(s); they do not necessarily reflect the viewpoint or position of the editors, reviewers, or publisher.