Usefulness of High-Resolution Computed Tomography in Early Diagnosis of Patients with Suspected COVID-19

Curr Med Imaging. 2022;18(14):1510-1516. doi: 10.2174/1573405618666220606161924.

Abstract

Background: Diagnosis of coronavirus disease 2019 (COVID-19) is mainly based on molecular testing. General population studies have shown that chest Computed Tomography (CT) can also be useful.

Objective: The study aims to examine the usefulness of high-resolution chest CT for early diagnosis of patients with suspected COVID-19.

Design and setting: This is a cross-sectional study from May 1, 2020, to August 31, 2021, at the COVID Hospital, Mexico City.

Methods: This study examined the clinical, high-resolution chest CT imaging, and laboratory data of 160 patients who were suspected to have COVID-19. Patients with positive Reverse Transcription- Polymerase Chain Reaction (RT-PCR) testing and those with negative RT-PCR testing but clinical data compatible with COVID-19 and positive antibody testing were considered to have COVID-19 (positive). Sensitivity and specificity of CT for diagnosis of COVID-19 were calculated. p < 0.05 was considered significant.

Results: Median age of 160 study patients was 58 years. The proportion of patients with groundglass pattern was significantly higher in patients with COVID-19 than in those without COVID (65.1% versus 0%; P = 0.005). COVID-19 was ruled out in sixteen (11.1%). Only four of the 132 patients diagnosed with COVID-19 (3.0%) did not show CT alterations (p < 0.001). Sensitivity and specificity of CT for COVID-19 diagnosis were 96.7% and 42.8%, respectively.

Conclusions: Chest CT can identify patients with COVID-19, as characteristic disease patterns are observed on CT in the early disease stage.

Keywords: COVID-19; Diagnosis; SARS-CoV-2; computed tomography; sensitivity; specificity.

MeSH terms

  • COVID-19 Testing
  • COVID-19* / diagnostic imaging
  • Cross-Sectional Studies
  • Early Diagnosis
  • Humans
  • Middle Aged
  • SARS-CoV-2
  • Tomography, X-Ray Computed / methods