Abstracts

RETURN TO ABSTRACT LISTING


1521. Imaging Tests for the Diagnosis of COVID-19
Authors * Denotes Presenting Author
  1. Jean-Paul Salameh *; Queen's University; The Ottawa Hospital
  2. Mariska Leeflang; Amsterdam University Medical Centers
  3. Nayaar Islam; University of Ottawa
  4. Yemisi Takwoingi; University of Birmingham
  5. Daniël Korevaar; Amsterdam University Medical Centers
  6. Matthew McInnes; The Ottawa Hospital; University of Ottawa
Objective:
The diagnosis of infection by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) presents major challenges. These include understanding the value of signs and symptoms in diagnosing possible infection and assessing whether existing biochemical and imaging tests can identify infection and those needing critical care. Chest imaging tests could be used to assist with the diagnosis of suspected cases of the coronavirus disease 2019 (COVID-19) (1). We aimed to assess the diagnostic accuracy of chest imaging (computed tomography (CT), X-ray and ultrasound) in the evaluation of people suspected to have COVID-19.

Materials and Methods:
We searched the COVID-19 Living Evidence Database from the University of Bern, the Cochrane COVID-19 Study Register, and The Stephen B. Thacker CDC Library. We did not apply any language restrictions. We conducted searches up to May 5, 2020. We included studies of all designs that produce estimates of test accuracy or provide data from which estimates can be computed. We assessed the risk of bias and applicability concerns using the quality of diagnostic accuracy studies 2 (QUADAS-2) checklist and presented the results of estimated sensitivity and specificity, using paired forest plots, and summarized in tables. We used a bivariate model for meta-analysis where appropriate.

Results:
We included 84 studies categorized into 3 study designs: the first category included 71 studies with 6331 participants known to have COVID-19 at the time of recruitment; the second category included 10 studies with 1399 participants suspected of COVID-19, and the third category included 3 case-control studies with 549 cases and controls in total. Heterogeneity judged by visual assessment of the forest plots was high. Risk of bias in participant selection was high in 76 (90%) studies, and concern about the applicability of the index test was high in 74 (88%) studies. The pooled sensitivity of chest CT (65 studies, 5759 cases) and X-ray (9 studies, 682 cases) in confirmed cases were 93.1% (95%CI: 90.2-95.0) and 82.1% (95%CI: 62.5–92.7), respectively. The pooled sensitivity of ultrasound (2 studies, 32 cases) in confirmed cases was 96.9% (95%CI: 80.8; 99.6). The pooled sensitivity and specificity of chest CT in suspected cases (13 studies, 2346 participants) were 86.2% (95%CI: 71.9-93.8) and 18.1% (95%CI: 3.71-55.8), respectively.

Conclusion:
Included studies were at high risk of bias and limits confidence in the findings. Chest CT appears to be sensitive but not specific for the diagnosis of COVID-19 in suspected patients. Because of limited data, accuracy estimates of chest X-ray and ultrasound of the lungs are uncertain. Future diagnostic accuracy studies should recruit a representative sample of patients suspected of having COVID-19 and pre-define the criteria for positive imaging findings.