• 
• 

Three Part Question

[In adults with confirmed or suspected Covid-19 infection attending the Emergency Department], do [computed tomography findings help] to [predict the requirement for hospital care]?

Clinical Scenario

A middle aged otherwise fit and well gentleman presents to your emergency department with symptoms highly suggestive of COVID-19 infection. He is streamed to be seen by you in the Amber area. He has been isolating at home for the last three days after developing myalgia and a fever. His oxygen saturations are 95% at rest on air and his respiration rate is 22 He is normally fit and well. Would a Computed Tomography (CT) scan help you to risk stratify him and decide whether admission is required?

Search Strategy

On 17 Apr 2020, Pubmed and EMBASE databases searched (via HDAS)
Title and Abstract: (covid OR coronavirus OR SARS-CoV-2) AND (prognos* OR score OR decision OR predic*)

Search Outcome

Pubmed 451 + EMBASE 290 results published since 01 Jan 2019
118 titles selected for abstract review, 76 selected for full text review
1 systematic review was identified relevant to the three part question

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Yuan M et al. March 2020 China	27 confirmed COVID-19 patients	Retrospective review of CT images	Mortality	80% of the mortality group had comorbidities. Higher CT score loosely correlates with mortality.	Small sample size. Not peer reviewed. No control for treatments
Xiong et al. March 2020 China	42 confirmed COVID-19 patients. (age 26-75)	Retrospective, single centre	Severity of COVID19 infection on CT. To determine relationship of CT findings with clinical parameters.	83% of patients exhibited a progressive CT features during the early stage from onset. Highest temperature and the severity of opacifications assessed on initial CT significantly related to progression of opacifications on follow-up CT. CRP, ESR and LDH showed significantly positive correlation with severity of pneumonia on initial CT.	Small sample size. Short follow up period.
Zhou et al. March 2020 China	62 confirmed COVID-19 patients	Retrospective	Identify CT features of early and late disease	Identification of ground glass opacifications and a single lesion on initial CT suggested early-phase disease. CT signs of aggravation and repair coexisted in advanced-phase disease.	Focuses on general CT findings, rather than prognosis. Small sample size.
Zhang R et al. April 2020 China	120 confirmed COVID-19 patients. Split into ‘common type’ and ‘severe type’ when common had 100% survival, severe had 77% survival.	Retrospective single centre	CT features of: common vs severe cases. Comparison of ‘common-type’ Symptomatic vs Asymptomatic and Febrile vs Afebrile	Severe patients commonly >20 y/o, comorbid and high LDH. No differences in CT findings between ‘common type’ symptomatic and asymptomatic. No difference in CT findings in ‘common type’ febrile vs afebrile. Diffuse consolidation, air bronchogram, old age, comorbidities, diarrhoea and high LDH may increase chance of severity.	Day at which survival was measured is unclear. No measurement comparison of severe disease.
Kunwei Li et al. March 2020 China	78 confirmed COVID-19 patients. Mild (24) common (46) severe-critical (8) types.	Retrospective single-centre	CT total severity score compared with clinical classification	CT Score was significantly higher in clinically severe group. High proportion of confirmed covid-19 cases with normal CT.	Not peer reviewed. Description of clinical classification differs between the abstract and the results section. Discharged patients only. Disproportionate group sizes.
Zhang X et al. March 2020 China	645 confirmed COVID-19 patients.	Retrospective observational.	Correlation between symptoms and biomarkers to radiograph findings (CT and CXR). Correlation between biomarkers and radiological findings to severe type.	Negative correlation between CT score and oxygenation index. Myalgia, shortness of breath, nausea and vomiting, lymphocytopenia, high serum creat and high CT score were predictive factors for severe type. Those with abnormal images were more likely to have higher bilirubin, CK and LDH and CRP.	Undisclosed number of hospitals. Not peer reviewed. All in-patients at time of study.

Comment(s)

There is a general consensus among the authors that CT findings correlate with severity of disease. No gold standard CT scoring system was identified, but the more abnormalities seen on the scan the poorer the prognosis for the patient. The definition of severe disease varied between papers. As expected, the presence of abnormal biomarkers positively correlates with the presence of abnormal CT findings. Interestingly, Kunwei L et al. demonstrated a significant proportion of their confirmed COVID-19 cases had a normal CT. However, the time between first symptoms and CT is not clear, and it is well-recognised that radiographic evidence of infection (in general) can lag behind symptoms. Zhang X et al. (2020) showed a negative correlation between CT score and oxygenation index, i.e. the more abnormalities on CT, the more oxygen required by that patient to maintain normal saturation levels. There is some evidence (Yang Z et al. 2020) for biochemical markers such as monocyte-lymphocyte ratio and homocysteine predicting progression of disease on chest CTs in positive patients. It is unclear how this correlates with disease severity, and this study only included ‘mild type’ COVID-19 positive patients. Therefore, it is of limited use when discussing progression of disease severity or predictors of mortality.

Clinical Bottom Line

Radiographic findings correlate with biochemical findings. Relevance of these findings to prognosis remains unclear.

References

1. Yuan M et al. Association of radiologic findings with mortality of patients infected with 2019 novel coronavirus in Wuhan, China
2. Xiong et al. Clinical and High-Resolution CT Features of the COVID-19 Infection Comparison of the Initial and Follow-up Changes
3. Zhou et al. CT Features of Coronavirus Disease 2019 (COVID-19) Pneumonia in 62 Patients in Wuhan, China
4. Zhang et al. CT features of SARS-CoV-2 pneumonia according to clinical presentation: a retrospective analysis of 120 consecutive patients from Wuhan city
5. Kunwei Li et al. CT image visual quantitative evaluation and clinical classification of coronavirus disease (COVID-19)
6. Zhang X et al. Epidemiological, clinical characteristics of cases of SARS-CoV-2 infection with abnormal imaging findings.