Multivariable logistic regression analysis for in-hospital mortality with the individual quality performance measures instead of the full versus incomplete compliance (in ED patients with severe sepsis and septic shock)
| Variable | Adjusted OR (95% CI) |
| PIRO 0–8 | Reference |
| PIRO 9–17 | 2.34 (0.88 to 6.30) |
| PIRO ≥18 | 6.42 (2.11 to 19.56) |
| Admission to ICU/MCU | 3.65 (1.85 to 7.18) |
| Treatment at academic medical centre | 0.91 (0.45 to 1.83) |
| Correct suspected source of infection | 0.45 (0.3 to 0.7) |
| No unanticipated transfer | 0.12 (0.05 to 0.30) |
| Lactate measured within 6 hours | 0.78 (0.19 to 3.31) |
| Blood cultures drawn before antibiotics administration | 0.93 (0.22 to 3.98) |
| Antibiotics administration within 3 hours | 1.39 (0.47 to 4.15) |
| Mean arterial pressure ≥65 mm Hg within 6 hours | 0.87 (0.37 to 2.04) |
| Adequate fluid resuscitation when systolic blood pressure <90 or lactate >4 mmoL/L | 0.80 (0.37 to 2.04) |
| Appropriate antibiotics administered in ED | 0.68 (0.35 to 1.30) |
For exact definition of quality performance measures, see text. Treatment in the University Medical Centre was compared with treatment in an urban hospital. The quality performance measure ‘appropriate ICU consultation in ED’ was excluded from the model because of collinearity with ICU/MCU admission.
For all variables, adjusted ORs with 95% CI are shown.
The Hosmer-Lemeshow goodness-of-fit test p value was 0.492, n=353.
ICU, intensive care unit; MCU, medium care unit; PIRO, Predisposition, Infection, Response and Organ failure illness severity score.