Table 5.
Logistic regression analysis: predictive value for each outcome variable.
| Outcome variable | Odds ratio (95% CI) |
P value | |
|---|---|---|---|
| LOS in ICU a) | Cardiac pathology subgroup | 0.85 (0.71 to 1.03) | 0.12 |
| Baseline ScO2 | 0.95 (0.93 to 0.97) | < 0.001* | |
| Intraoperative TBBL ScO2 | 1.02 (1.01 to 1.03) | < 0.001* | |
| Duration of MV b) | Cardiac pathology subgroup | 0.85 (0.69 to 1.05) | 0.13 |
| Baseline ScO2 | 0.92 (0.90 to 0.95) | < 0.001* | |
| Intraoperative TBBL ScO2 | 1.02 (1.01 to 1.03) | 0.002* | |
| LOS in hospital c) | Baseline ScO2 | 0.93 (0.91 to 0.96) | < 0.001* |
| Intraoperative TBBL ScO2 | 1.02 (1.01 to 1.04) | < 0.001* | |
| Mortality within first 30 postoperative days d) | Baseline ScO2 | 0.94 (0.91 to 0.98) | 0.007** |
| Intraoperative TBBL ScO2 | 1.03 (1 to 1.05) | 0.02 | |
| STS-EACTS score 15 | 1.56 (0.95 to 2.53) | 0.08 |
a) 0.18 (Hosmer-Lemeshow), 0.06 (Cox & Snell), 0.9 (Nagelkerke). Model X2 (2) = 35, P < 0.001*
b) 0.23 (Hosmer-Lemeshow), 0.09 (Cox & Snell), 0.15 (Nagelkerke). Model X2 (2) = 54.2, P < 0.001*
c) 0,09 (Hosmer-Lemeshow), 0.08 (Cox & Snell), 0.13 (Nagelkerke). Model X2 (3) = 42.29, P < 0.001*
d) 0.09 (Hosmer-Lemeshow), 0.08 (Cox & Snell), 0.13 (Nagelkerke). Model X2 (3) = 42.29, P < 0.001*
* P is considered significant when <0.0125 (Bonferroni-correction)
** P is considered significant when <0.01 (Bonferroni-correction)
95% CI = 95% confidence interval; ICU = intensive care unit; LOS = length of stay; MV = mechanical ventilation; ScO2 = cerebral oxygen saturation; TBBL = time below baseline ScO2; P = significance value.