Table 4.
Logistic regression modeling of factors associated with in-hospital death following ICU discharge
| Factor | Odds ratio (95% confidence interval) | P-value |
|---|---|---|
| Male | 1.30 (1.02-1.65) | 0.031 |
| SAPS II (per point) | 1.03 (1.02-1.03) | <0.001 |
| Discharge SOFA score (per point) | 1.19 (1.14-1.23) | <0.001 |
| DFLST order | 4.55 (3.34-6.20) | <0.001 |
| Diagnosis | 1 (reference) | - |
| Other diagnoses | ||
| Cardiovascular disease | 1.50 (1.04-2.16) | 0.029 |
| Respiratory disease | 1.53 (1.11-2.10) | 0.008 |
| Gastrointestinal | 1.83 (1.30-2.56) | <0.001 |
| Infection | 1.76 (1.20-2.56) | 0.003 |
| Discharge day | ||
| Monday | 1 (reference) | - |
| Tuesday | 0.99 (0.66-1.48) | 0.949 |
| Wednesday | 0.91 (0.60-1.36) | 0.631 |
| Thursday | 0.80 (0.53-1.21) | 0.296 |
| Friday | 1.45 (1.01-2.10) | 0.046 |
| Saturday | 1.26 (0.80-2.00) | 0.315 |
| Sunday | 1.41 (0.84-2.35) | 0.193 |
| Night discharge | 1.54 (1.12-2.11) | 0.008 |
The model (n = 5,992) had fair discrimination (area under receiver operator characteristic curve 0.771) and calibration (goodness of fit p = 0.4370). Decision to forego life sustaining therapy (DFLST)