Table 2. Cox proportional hazards regression analysis results for overall mortality.
Univariate Cox regression | Multivariate Cox regression | ||||||||
---|---|---|---|---|---|---|---|---|---|
Model 1 | Model 2 | ||||||||
HR | 95% CI | P value | HR | 95% CI | P value | HR | 95% CI | P value | |
Age | 1.073 | 1.037–1.111 | <0.001 | 1.053 | 1.012–1.095 | 0.010 | - | - | - |
Age > 50 years | 6.626 | 1.988–22.086 | 0.002 | - | - | - | 4.507 | 1.289–15.755 | 0.018 |
Fever | 2.386 | 1.036–5.493 | 0.04 | - | - | - | - | - | - |
AIP | 24.385 | 11.065–53.742 | <0.001 | 30.119 | 11.547–78.562 | <0.001 | 14.433 | 6.076–34.286 | <0.001 |
Cyclophosphamide | 2.300 | 0.867–6.106 | 0.09 | - | - | - | - | - | - |
Hemoglobin | 0.809 | 0.647–1.010 | 0.06 | - | - | - | - | - | - |
ANC | 1.066 | 1.031–1.103 | <0.001 | - | - | - | - | - | - |
Neutrophilia | 2.232 | 1.012–4.926 | 0.047 | - | - | - | - | - | - |
ALC | 0.897 | 0.849–0.948 | <0.001 | - | - | - | - | - | - |
Lymphopenia | 3.948 | 1.759–8.864 | 0.001 | - | - | - | - | - | - |
Log-NLR | 2.570 | 1.651–4.002 | <0.001 | 2.216 | 1.176–3.843 | 0.01 | - | - | - |
High NLR | 7.640 | 2.880–20.267 | <0.001 | - | 5.201 | 1.923–14.068 | 0.001 | ||
CRP | 1.125 | 1.046–1.209 | 0.001 | - | - | - | - | - | - |
CRP > 0.5 mg/dL | 2.734 | 1.188–6.291 | 0.02 | - | - | - | - | - | - |
Protein | 0.388 | 0.246–0.613 | <0.001 | 0.441 | 0.232–0.838 | 0.01 | - | - | - |
Protein < 5.5 g/dL | 3.787 | 1.518–9.448 | 0.004 | - | 3.586 | 1.333–9.650 | 0.01 | ||
Albumin | 0.329 | 0.192–0.564 | <0.001 | - | - | - | - | - | - |
Albumin < 3.5 g/dL | 3.071 | 1.290–7.313 | 0.011 | - | - | - | - | - | - |
Log-CAR | 1.437 | 1.173–1.760 | <0.001 | - | - | - | - | - | - |
High CAR | 8.199 | 1.936–34.716 | 0.004 | - | - | - | - | - | - |
Creatinine | 3.758 | 1.543–9.154 | 0.004 | 4.375 | 1.840–10.401 | 0.001 | - | - | - |
ILD | 3.561 | 1.428–8.880 | 0.006 | - | - | - | - | - | - |
AE-ILD | 7.930 | 3.585–17.541 | <0.001 | - | - | - | - | - | - |
Malignancy | 1.089 | 0.410–2.888 | 0.86 | - | - | - | - | - | - |
All continuous variables were maintained in model 1. In model 2, continuous variables were converted to dichotomous variables using clinically relevant or the ROC-derived cut-points. Neutrophilia was defined as an ANC > 7500/mm3 and lymphopenia as an ALC < 1000/mm3. NLR or CAR values were dichotomized into high and low groups using a cut-off point of 4.775 and 0.0735, respectively.
Abbreviation: HR, hazard ratio; CI, confidence interval; AIP, acute interstitial pneumonia; ANC, absolute neutrophil count; ALC, absolute lymphocyte count; NLR, neutrophil-to-lymphocyte ratio; CRP, C-reactive protein; CAR, CRP-to-albumin ration; ILD, interstitial lung disease; AE-ILD, acute exacerbation of ILD.