Table 2.
Regression coefficients of the logistic regression model
| Intercept and risk factors | Regression coefficient | p value | OR | 95% CI of OR |
|---|---|---|---|---|
| Intercept | 11.0424 | 0.0003 | – | – |
| Age (years) | 0.0264 | 0.0004 | 1.026 | 1.0118–1.0419 |
| Lymphocytes (cells/µl) | − 0.8777 | 0.0029 | 0.414 | 0.2324–0.7376 |
| LDH (U/I) | 0.0044 | 0.0001 | 1.0044 | 1.0022–1.0066 |
| CRP (mg/dl) | 0.047 | 0.0013 | 1.048 | 1.0185–1.0784 |
| SpO2 (%) | − 0.1585 | < 0.0001 | 0.8537 | 0.8045–0.9059 |
The linear predictor can be calculated as 11.042 + 0.0264 × each year (age)—0.1585 × each unit of SpO2(%) − 0.877 × each lymphocyte (× 103 cells/µl) + 0.0044 × each unit of LDH + 0.047 × each unit of C-reactive protein (mg/dl). The predicted probability of respiratory failure can be calculated with the formula 1/(1 + exp [− linear predictor])
OR odds ratio; CI confidence interval; SpO2 peripheral pulse oximetry; LDH lactate dehydrogenase; CRP C-reactive protein