Table 5.
Independent predictors of unfavorable outcome (ICU admission and/or death): Multivariable logistic regression model.
| Adjusted odds ratio (95% CI) | p value | |
|---|---|---|
| Age ≥ 70 years | 3.58 (1.83–6.99) | < 0.001 |
| SpO2 < 95% | 11.07 (5.34–22.97) | < 0.001 |
| Neutrophils > 7.5 × 103/µL | 3.67 (1.74–7.74) | 0.001 |
| LDH ≥ 300 U/L | 2.11 (1.04–4.31) | 0.04 |
| CRP ≥ 100 mg/L | 2.61 (1.32–5.14) | 0.01 |
| Viral load ≥ 7.35 log10 copies/mL (50th percentile) | 1.49 (.75–2.96) | 0.25 |
| Viral load ≥ 8.27 log10 copies/mL (2nd tertile) | 1.84 (.92–3.68) | 0.09 |
The final multivariable model was composed of five variables (therefore 17 events per variable) demonstrated as independent predictors of unfavorable outcome: Age ≥ 70 years, SpO2 < 95%, neutrophils > 7.5 × 103/µL, LDH ≥ 300 U/L, and CRP ≥ 100 mg/L). Such model reported a Beta Coefficient of -4.08 (standard error = 0.46), a Wald statistic of 78.72 (degrees of freedom = 1), and an overall apparent performance of 84.2% (sensitivity = 70.6%, specificity = 89.4%, PPV = 70.3%, NPV = 89.1%). The variables included were explanatory and contributed to giving the model an ability to explain roughly 52.1% of the variation of the outcome (Nagelkerke R2 value = 0.521). A higher nasopharyngeal viral load (above the second quartile or the second tertile) was not independently linked to an increased risk of ICU admission or death.