Olfactory evaluation in hospitalised and self-isolated patients with COVID-19: a single-centre experience on 55 cases.

BACKGROUND: Smell loss is a common symptom of COVID-19 infection. Majority of the studies that evaluated olfactory impairment in COVID-19 used questionnaires (subjective smell evaluations) and did not compare the results with objective or semiobjective measures of smell. We performed smell testing in hospitalised and self-isolated patients with COVID-19 and control participants. METHODS: Fifty-five COVID-19 and 44 control participants underwent smell testing, using Burghart Sniffin' Sticks 'Screening 12 Test'. Participants also rated their smelling capability on the numerical scale. Differences between groups and correlation between smell loss and time from acute onset of symptoms were tested, as well as correlation between results of smell test and subjective assessment of smell. RESULTS: Hospitalised patients with COVID-19 correctly determined 6.5/12 odorants compared with 10/12 in the self-isolated and 11/12 in the control group (p<0.001). Hyposmia or anosmia were present in 87.5% of hospitalised and 29.0% of self-isolated patients (p<0.001). The correlation between subjective self-assessment and results of smell testing was non-significant in both groups of patients with COVID-19, while there was a moderate positive correlation (p=0.001, Spearman's correlation coefficient=0.499) in control participants. CONCLUSION: Contrary to some previous reports suggesting that the presence of olfactory loss may predict milder course of disease, our study found that a vast majority of hospitalised patients with COVID-19 had prominent olfactory impairment. The absence of correlation between self-rated and objective smell evaluation in patients with COVID-19 indicates that subjective smell assessment is unreliable.

COVID-19 diagnosis by routine blood tests using machine learning.

Physicians taking care of patients with COVID-19 have described different changes in routine blood parameters. However, these changes hinder them from performing COVID-19 diagnoses. We constructed a machine learning model for COVID-19 diagnosis that was based and cross-validated on the routine blood tests of 5333 patients with various bacterial and viral infections, and 160 COVID-19-positive patients. We selected the operational ROC point at a sensitivity of 81.9% and a specificity of 97.9%. The cross-validated AUC was 0.97. The five most useful routine blood parameters for COVID-19 diagnosis according to the feature importance scoring of the XGBoost algorithm were: MCHC, eosinophil count, albumin, INR, and prothrombin activity percentage. t-SNE visualization showed that the blood parameters of the patients with a severe COVID-19 course are more like the parameters of a bacterial than a viral infection. The reported diagnostic accuracy is at least comparable and probably complementary to RT-PCR and chest CT studies. Patients with fever, cough, myalgia, and other symptoms can now have initial routine blood tests assessed by our diagnostic tool. All patients with a positive COVID-19 prediction would then undergo standard RT-PCR studies to confirm the diagnosis. We believe that our results represent a significant contribution to improvements in COVID-19 diagnosis.