Risk factors for outcomes of COVID-19 patients: an observational study of 795 572 patients in Russia (preprint)

Background Several factors that could affect survival and clinical outcomes of COVID-19 patients require larger studies and closer attention. Objective To investigate the impact of factors including whether COVID-19 was clinically or laboratory-diagnosed, influenza vaccination, former or current tuberculosis, HIV, and other comorbidities on the hospitalized patients' outcomes. Design Observational nationwide cohort study. Patients All subjects, regardless of age, admitted to 4,251 Russian hospitals indexed in the Federal Register of COVID-19 patients between March 26, 2020, and June 3, 2020. All included patients for which complete clinical data were available were divided into two cohorts, with laboratory- and clinically verified COVID-19. Measurements We analyzed patients' age and sex, COVID-19 ICD-10 code, the length of the hospital stay, and whether they required ICU treatment or invasive mechanical ventilation. The other variables for analysis were: verified diagnosis of pulmonary disease, cardiovascular disease, diseases of the endocrine system, cancer/malignancy, HIV, tuberculosis, and the data on influenza vaccination in the previous six months. Results This study enrolled 705,572 COVID-19 patients aged from 0 to 121 years, 50.4% females. 164,195 patients were excluded due to no confirmed COVID-19 (n=143,357) or insufficient and invalid clinical data (n=20,831). 541,377 participants were included in the study, 413,950 (76.5%) of them had laboratory-verified COVID-19, and 127,427 patients (23.5%) with the clinical verification. Influenza vaccination reduced the risk of transfer to the ICU (OR 0.76), mechanical ventilation requirement (OR 0.74), and the risk of death (HR 0.77). TB increased the mortality risk (HR 1.74) but reduced the likelihood of transfer to the ICU (OR 0.27). HIV comorbidity significantly increased the risks of transfer to the ICU (OR 2.46) and death (HR 1.60). Patients with the clinically verified COVID-19 had a shorter duration of hospital stay (HR 1.45) but a higher risk of mortality (HR 1.08) and the likelihood of being ventilated (OR 1.36). According to the previously published data, age, male sex, endocrine disorders, and cardiovascular diseases increased the length of hospital stay, the risk of death, and transfer to the ICU. Limitations The study did not include a control group of subjects with no COVID-19. Because of that, some of the identified factors could not be specific for COVID-19. Conclusions Influenza vaccination could reduce the severity of the hospitalized patients' clinical outcomes, including mortality, regardless of age, social, and economic group. The other factors considered in the study did not reduce the assessed risks, but we observed several non-trivial associations that may optimize the management of COVID-19 patients.

Diagnostic Accuracy of Computed Tomography for Identifying Hospitalization in Patients with Suspected COVID-19 (preprint)

The controversy of computed tomography (CT) use in COVID-19 screening is associated with ambiguous characteristics of chest CT as a diagnostic test. The reported values of CT sensitivity and specificity calculated using RT-PCR as a reference standard vary widely. The objective of this study was to reevaluate the diagnostic and prognostic value of CT using an alternative approach. This study included 973 symptomatic COVID-19 patients aged 42 $\pm$ 17 years, 56% females. We reviewed the disease dynamics between the initial and follow-up CT studies using a "CT0-4" grading system. Sensitivity and specificity were calculated as conditional probabilities that a patient's condition would improve or deteriorate relative to the initial CT study results. For the calculation of negative (NPV) and positive (PPV) predictive values, we estimated the COVID-19 prevalence in Moscow. We used several ARIMA and EST models with different parameters to fit the data on total cases of COVID-19 from March 6, 2020, to July 20, 2020, and forecast the incidence. The "CT0-4" grading scale demonstrated low sensitivity (28%) but high specificity (95%). The best statistical model for describing the pandemic in Moscow was ETS with multiplicative trend, error, and season type. According to our calculations, with the predicted prevalence of 2.1%, the values of NPV and PPV would be 98% and 10%, correspondingly. We associate the low sensitivity and PPV values with the small sample size of the patients with severe symptoms and non-optimal methodological setup for measuring these specific characteristics. The "CT0-4" grading scale was highly specific and predictive for identifying admissions to hospitals of COVID-19 patients. Despite the ambiguous accuracy, chest CT proved to be an effective practical tool for patient management during the pandemic, provided that the necessary infrastructure and human resources are available.