Table 1.
Year |
2005 |
2006 |
2007 |
2008 |
2009 |
P§ |
p-value trend# |
|
Patient (n) |
38 |
31 |
27 |
27 |
18 |
|
|
|
Gender (m/f) |
20/18 |
16/15 |
14/13 |
18/9 |
11/7 |
0.728 |
0.289 |
|
Age (years)* |
68.6 ±15.9 |
73.8 ±9.5 |
68.0 ±13.5 |
67.5 ±13.9 |
65.4 ±11.6 |
0.196 |
0.125 |
|
SAPS II * |
45.0 ±17.7 |
43.4 ±17.9 |
42.5 ±19.8 |
44.3 ±19.0 |
37.0 ±10.9 |
0.603 |
0.177 |
|
Apache II * |
20.0 ±8.8 |
20.5 ±8.8 |
18.2 ±8.7 |
17.9 ±9.7 |
15.6 ±7.2 |
0.327 |
0.041 |
|
Effective cost weight * |
14.8 ±13.6 |
13.3 ±11.4 |
17.2 ±14.0 |
14.5 ±13.6 |
19.2±18.5 |
0.623 |
0.256 |
|
|
Average change per year |
|||||||
Mean |
95% CI |
p |
||||||
ICU stay (days) |
26.0 ± 1.9 |
19.5 ± 2.2 |
18.5 ± 2.3 |
16.3 ± 2.3 |
15.8 ± 2.8 |
−2.7 |
−4.1 to −1.3 |
< 0.001 |
Ventilation (hours) |
571.4 ± 41.5 |
455.0 ± 47.3 |
404.1 ± 49.2 |
362.0 ± 49.9 |
467.9 ± 60.9 |
−42.0 |
−72.6 to −11.4 |
0.008 |
Antibiotic use (days) |
14.3 ± 1.2 |
12.0 ± 1.3 |
13.8 ± 1.4 |
12.1 ± 1.4 |
9.0 ± 1.7 |
−1.0 |
−1.9 to −0.2 |
0.02 |
Antibiotic costs (EUR/patient) |
407.6 ± 57.2 |
420.7 ± 66.7 |
426.6 ± 67.7 |
393.2 ± 68.5 |
373.9 ± 83.7 |
−14.3 |
−55.7 to 27.1 |
0.495 |
ICU reinfection rate (proportion) |
0.31 ± 0.07 |
0.37 ± 0.08 |
0.13 ± 0.06 |
0.12 ± 0.06 |
0.08 ± 0.06 |
−35.1% |
−53.9% to −8.5% |
0.014 |
28-day mortality (proportion) | 0.54 ± 0.06 | 0.51 ± 0.07 | 0.55 ± 0.08 | 0.48 ± 0.07 | 0.34 ± 0.01 | −22.4% | −44.3% to 8.1% | 0.133 |
Demographic and clinical data:
* Mean ± standard deviation.
§ F test of ANOVA (continuous characteristic) or Wald chi square test (categorical characteristic) of logistic regression model.
# Linear polynomial contrasts (continuous characteristic) or Cochran-Armitage trend tests (categorical characteristic).
SAPS II, Simplified Acute Physiology Score II.
APACHE II, Acute Physiology And Chronic Health Evaluation Score II.
Primary outcome parameters:
Adjusted Means ± standard error and average yearly change of outcome parameters, with 95 confidence limits (CI) and p values of trend tests from ANCOVA or logistic regression models with covariates gender, age, SAPS II, APACHE II and effective cost weight.