[Sorting and detection of COVID-19 by low-dose thoracic CT scan in patients consulting the radiology department of Fann hospital (Dakar-Senegal)].

Introduction: COVID-19 has spread rapidly since its emergence in China and is currently a global health issue. Its definitive diagnosis is made by PCR on nasopharyngeal swabs. However, this diagnostic test has low sensitivity with delayed results. Hence, thoracic computed tomography represents an interesting alternative. The aims of this study were to assess the frequency of computed tomography (CT) lesions suggestive of COVID-19 and to compare the results of CT and PCR test.

Methods: a prospective study carried out over15 working days and involved 47 patients. These patients were recruited based on the presence of at least 2 clinical signs of COVID-19. Chest CT without contrast according to the "LOW-DOSE" protocol was performed. A PCR test on nasopharyngeal swabs was done in patients with signs suggestive of COVID on CT. A serological test was performed in case of a discrepancy between the CT and PCR results.

Results: thoracic CT was abnormal in 38 patients and normal in 9 patients. Lesions suggestive of COVID-19 have been identified in 32 patients. Two patients had lesions of non-specific pneumonia. Tuberculosis lesions were visualized in 3 patients. One patient had lesions of interstitial pneumonia. The mean DLP was 59 mGy.cm with extremes of 25 and 95 mGy.cm. Ground-glass opacity was present in 100% of COVID-19 suspects on CT. The results of the PCR test were the same than CT in 12 patients. The positive predictive value for CT was 37.5%. In 20 patients with COVID lesions on CT, the PCR test was negative with a false positive rate of 62.5%. In the patients with negative PCR test, 4 had a serological test for COVID-19 and this test was positive in 3.

Conclusion: low-dose chest CT can reduce radiation exposure in COVID-19 patients who are at risk of cumulative dose due to repetitive exam. CT can identify lesions suggestive of COVID-19. It also enables the triage of patients by identifying other diagnoses.