Table 1.
Univariate and multivariable analysis of factors associated with moderate-to-large pleural effusion
| Variables | Missing values, n (%) | Absolute standardized differences | Odd ratio (95% confidence interval), p value by logistic regression | |
|---|---|---|---|---|
| Univariate | Multivariable | |||
| Age (per year) | 0 | 60.5 | 1.04 (1.02–1.06), p < 0.001 | 1.03 (1.01–1.05), p = 0.017 |
| SAPS II (per point) | 0 | 26.9 | 1.02 (1.0–1.03), p = 0.048 | I/NR |
| Mc Cabe class II (yes vs. no) | 0 | 51.8 | 4.7 (2.1–10.3), p < 0.001 | 4.2 (1.8–9.9), p = 0.001 |
| Cancer or hematological malignancy (yes vs. no) | 0 | 47 | 3.5 (1.7–7.2), p = 0.001 | NI |
| Cardiac disease (yes vs. no) | 0 | 55.1 | 3.3 (1.1–3.3), p < 0.001 | 2.2 (1.1–4.4), p = 0.02 |
| Left ventricle ejection fraction at cardiac ultrasound (%), | 44 (18%) | 52.9 | 0.96 (0.93–0.98), p < 0.001 | NI |
| Supra-ventricular arrhythmias (yes vs. no) | 0 | 35.8 | 2.3 (1.3–4.2), p = 0.007 | NI |
| Acute respiratory failure as cause of intubation (yes vs. no) | 0 | 31.9 | 1.9 (1.3–3.9), p = 0.02 | 1.8 (0.98–3.2), p = 0.059 |
| Dialysis (yes vs. no) | 0 | 31.5 | 2.5 (1.2–5.4), p = 0.02 | 2.0 (0.9–4.6), p = 0.088 |
| Serum Creatinine (per µmol/L) | 0 | 25.8 | 1.0 (0.99–1.00), p = 0.20 | NI |
| Septic shock (yes vs. no) | 0 | 25.4 | 1.7 (0.98–2.9), p = 0.06 | I/NR |
| ARDS (yes vs. no) | 0 | 22.2 | 1.7 (0.91–3.1), p = 0.098 | I/NR |
| Duration of MV before first SBT (per day) | 0 | 19 | 1.04 (0.99–1.1), p = 0.15 | NI |
SAPS II simplified acute physiology score, COPD chronic obstructive pulmonary disease, ARDS acute respiratory distress syndrome, SBT spontaneous breathing trial, NI not included, I/NR included, but not retained by the final model
Among related univariate factors, only the most statistically robust (yet clinically relevant) was entered into the regression model in order to minimize the effect of colinearity. The selection process was guided by consistency (less than 5% missing values) and maximal imbalances between groups (as estimated by absolute standardized differences) as follows: Mc Cabe class II was selected among Mc Cabe class II, cancer and hematological malignancy; dialysis was selected among creatininemia and dialysis; cardiac disease was selected among supra-ventricular arrhythmias, left ventricle ejection fraction and cardiac disease; ARDS was selected among duration of mechanical ventilation before the first spontaneous breathing trial and ARDS before inclusion. The multivariable model showed a good calibration as assessed by the Hosmer and Lemeshow goodness-of-fit test [χ2 (8 df) = 6.42, p = 0.60] and a fair discrimination as assessed by the receiver operating characteristics curve [area under the curve of 0.74 (0.67–0.80), p < 0.001]