TABLE 1

Significant univariate and multivariate logistic regression analyses of predictors for 30-day mortality

Univariate#Multivariate
OR (95% CI)p-valueOR (95% CI)p-value
Age ≥65 years4.37 (3.08–6.20)<0.0014.37 (2.46–7.75)<0.001
Sex male1.34 (1.02–1.75)0.036
Influenza vaccine1.34 (0.97–1.84)0.074
Chronic cardiovascular disease1.74 (1.28–2.36)<0.001
Chronic renal disease2.26 (1.54–3.30)<0.001
Chronic liver disease1.85 (1.18–2.89)0.0073.40 (1.75–5.03)<0.001
Diabetes mellitus1.42 (1.06–1.92)0.021
Neurologic disease3.84 (2.94–5.03)<0.0013.35 (2.23–5.03)<0.001
Pleuritic pain0.42 (0.31–0.57)<0.001
Altered mental status3.92 (3.01–5.10)<0.0011.63 (1.10–2.4300.015
Creatinine ≥1.5 mg·dL−14.06 (3.14–5.23)<0.001
SaO2 <92%2.44 (1.80–3.30)<0.001
PaO2/FIO2 <2504.79 (3.55–6.46)<0.0013.24 (2.20–4.76)<0.001
CURB-65 risk class 3–56.12 (4.69–7.99)<0.001
PSI risk class IV–V10.91 (7.19–16.55)<0.001
Multilobar+<0.0010.006
 Bilateral2.71 (2.01–3.67)<0.0012.13 (1.33–3.41)0.002
 Unilateral multilobar1.61(1.10–2.37)0.0141.09 (0.64–1.86)0.75
 Localised11
Patterns of infiltrate§0.0150.046
 Alveolar11
 Interstitial1.39 (0.69–2.79)0.353.11 (1.24–7.78)0.015
 Mixed2.00 (1.23–3.26)0.0050.89 (0.43–1.84)0.75
Acute renal failure4.63 (3.58–6.00)<0.0012.90 (1.97–4.28)<0.001
Septic shock10.02 (7.38–13.61)<0.0016.30 (4.02–9.86)<0.001
Aetiologyƒ<0.001
 Unknown1
 Bacterial1.04 (0.77–1.39)0.80
 Respiratory virus or atypical bacterial0.34 (0.16–0.69)0.003
 Mixed1.86 (1.20–2.90)0.006
  • OR: odds ratio; SaO2: arterial oxygen saturation; PaO2: arterial oxygen tension; FIO2: inspiratory oxygen fraction; PaO2/FIO2; PaO2 and FIO2 ratio; CURB-65: confusion, urea, >7mmol·L−1 respiratory rate ≥30 breaths·min−1, blood pressure, 90 mmHg (systolic) ≤60 mmHg (diastolic), age ≥65 years; PSI: pneumonia severity index. #: the variables included in the univariate analysis were age, sex, smoking, alcohol consumption, previous antibiotic, influenza vaccine, pneumococcal vaccine, inhaled corticosteroids, systemic corticosteroids, chronic pulmonary disease, chronic cardiovascular disease, diabetes mellitus, neurological disease, chronic liver disease, pleuritic pain, altered mental status, creatinine, C-reactive protein level, white blood cell count, SaO2, PaO2/FIO2, CURB-65 risk class, PSI risk class, multilobar, pattern of infiltrate, septic shock, acute renal failure, and aetiology. : Hosmer–Lemeshow goodness-of-fit test, p=0.17. Internal validation of the final logistic regression model was conducted using bootstrapping with 1000 samples. The nine variables included in the model showed robust results, with small 95% CIs around the original coefficients. +: the p-value corresponds to the differences between the three groups (bilateral, unilateral multilobar, or localised). §: the p-value corresponds to the differences between the three groups (alveolar, interstitial, or mixed). ƒ: the p-value corresponds to the differences between the four groups (unknown, bacterial, respiratory virus or atypical bacterial, or mixed).