Skip to main content
. 2022 Mar 24;59(3):2102538. doi: 10.1183/13993003.02538-2021

TABLE 5.

Logistic regression analysis to assess the relationship between demographic, epidemiological, clinical variables and mortality

Univariable analysis Multivariable analysis
OR (95% CI) p-value OR (95% CI) p-value
Age, years (10-year increase) 1.82 (1.58–2.09) <0.0001 1.93 (1.60–2.32) <0.0001
Male (yes versus no) 2.08 (1.16–3.71) 0.014 2.92 (1.38–6.16) 0.005
≥1 comorbidity (yes versus no) 6.01 (3.21–11.27) <0.0001
Diabetes mellitus (yes versus no) 2.69 (1.67–4.35) <0.0001
Cardiovascular disease (yes versus no) 5.12 (3.19–8.22) <0.0001
Chronic respiratory disease (yes versus no) 3.00 (1.74–5.18) <0.0001
HIV (yes versus no) 1.48 (0.77–2.87) 0.241
Chronic liver disease (yes versus no) 1.69 (0.82–3.46) 0.155
Chronic renal disease (yes versus no) 3.00 (1.74–5.18) <0.0001
Invasive ventilation (yes versus no) 25.18 (12.64–50.13) <0.0001 28.22 (1.37–64.39) <0.0001
Active TB (yes versus no) 1.5 (1.0–2.5) 0.069
Presence of key symptoms (yes versus no) 49.3 (19.7–123.9) <0.0001
Hospitalisation due to COVID-19 (yes versus no) 3.54 (1.95–6.41) <0.0001
Duration of hospitalisation (1-day increase) 0.98 (0.96–1.01) 0.072
Europe (yes versus no) 1.63 (1.04–2.57) 0.034

Multivariable model −2 log likelihood: 301.6, p<0.0001; percentage of cases correctly classified: 91%; area under the curve: 0.89 (0.86–0.91). TB: tuberculosis; COVID-19: coronavirus disease 2019.