Table 2. Cox proportional hazard regression analysis for number of days a patient received oxygen supplementation.
| Parameter | Multivariate analysis | |
|---|---|---|
| HR (95% CI) | p valuea | |
| Age | ||
| 0–<3 months | - | 0.001 |
| 3–6 months | 3.76 (1.75–8.11) | |
| 6–<12 months | 2.94 (1.37–6.29) | |
| 12–<48 months | 3.81 (1.78–8.13) | |
| Gender | ||
| Female | - | 0.976 |
| Male | 1.01 (0.62–1.65) | |
| Underlying risk | ||
| No | - | 0.202 |
| Yes | 1.61 (0.79–3.31) | |
| Length of symptoms at intercept | ||
| ≤3 days | - | 0.389 |
| >3 days | 0.78 (0.44–1.37) | |
| PES3-total score (3 items; 1-unit increase) | 0.86 (0.75–0.99) | 0.029 |
N = 72, 2 patients were excluded as they did not have PES score on day 1 available. 1 patient was excluded due to missing age.
ap value was calculated by a likelihood ratio test.
A global chi-square test for the proportional hazard assumption showed no deviation (p = 0.447). A graphical review of possible time-dependent coefficients over time (plots of residuals for individual predictors) showed no deviation from the proportional hazard assumption.
Abbreviations: CI–Confidence Interval, HR–Hazard Ratio, PES–Physical Examination Scoring