Methods:
Adults (33-88 years old) with severe COVID-19 were recruited from the Department of Medicine of the EpiCURA Hospital (Hornu, Belgium; Ethics Committee: EpiCURA-2020-2303). The disease was confirmed through nasopharyngeal swab (RT-PCR). Patients were defined as severe COVID-19 if they required continuous care (oxygenotherapy, blood pressure monitoring) in internal medicine or intensive care units.
Patients with neurological disorder, chronic rhinosinusitis or history of nasal surgery prior the pandemic were excluded. Epidemiological and clinical data were collected at the hospital discharge. Details of the patient-reported outcome questionnaire used for data collection are reported in a previous study.3 Briefly: 1) olfactory and gustatory questions were based on the smell and taste component of the National Health and Nutrition Examination Survey; 2) symptoms were evaluated through a 4-point scale ranging from 0 (no symptom) to 4 (severe symptoms);3 3) nasosinusal symptoms were evaluated through the French sino-nasal outcome test-22 (SNOT-22).6Patients benefited from psychophysical olfactory evaluation through sniffin’stick tests (Medisense, Groningen, Netherlands): 16 pens were presented to patients every 30 seconds. The patient had to choose the adequate term describing the smell between 4 given options. The test was scored on a total of 16 points and allowed categorization into in 3 groups: normosmia (score between 12-16), hyposmia (score between 9-11) and anosmia (score <9).3 Moreover, the following hospitalization outcomes were recorded: duration of hospitalization (days); admission biology (D-dimer; hemoglobin; leucocyte count; lymphocyte count; CRP; creatitin; bilirubin; platelet count; LDH; Na+; K+; Cl-); 1-month serology (IgG) and chest computed tomography findings. Subjective and objective evaluations were made meanwhile.
The relationship between clinical and olfactory outcomes was analyzed through multiple linear regression between scale variables and through Mann-Whitney test and boxplot representation for groups versus scale variables (SPSS, v22,0; IBM-Corp, Armonk, NY, USA).