Skip to main content
. 2008 May 15;12(3):R68. doi: 10.1186/cc6903

Table 3.

Results of final model multiple regression analysis total cost per stay as a 'dependent' outcome variable

Number = 764
Variance explained = 0.680
Unstandardised coefficients Standardised coefficients
Predictors B Standard error Beta t Significance

Period (before → after IMC) -0.077 0.054 -0.087 -1.420 (0.156)a
TISS-28 score on 1st day 0.826 0.152 0.207 5.443 (<0.001)a
Surgical (no → yes) 0.183 0.024 0.207 7.714 <0.001
LOS in the ICU 0.025 0.001 0.688 17.307 (<0.001)a
LOS in ward 0.016 0.002 0.616 9.940 (<0.001)a
Primary organ failure (reference category is metabolic)
1. Metabolic → cardiovascular 0.066 0.076 0.070 0.871 0.384
2. Metabolic → multitrauma 0.129 0.080 0.088 1.603 0.109
3. Metabolic → gastrointestinal 0.143 0.086 0.064 1.667 0.096
4. Metabolic → haematological 0.324 0.125 0.066 2.589 0.010
5. Metabolic → renal 0.141 0.093 0.050 1.522 0.129
6. Metabolic → neurologic/neurosurgical 0.120 0.076 0.116 1.580 0.115
7. Metabolic → respiratory 0.152 0.075 0.145 2.022 0.044
8. Metabolic → sepsis 0.081 0.085 0.039 0.955 0.340
TISS × period 0.470 0.178 0.184 2.634 0.009
LOS in the ICU × period -0.005 0.002 -0.135 -3.161 0.002
LOS in ward × TISS-28 -0.022 0.005 -0.289 -4.526 <0.001
Constant 3.220 0.086 37.259 <0.001

aP values within brackets should not be taken into account as the predictors involved are also present in an interaction effect. Predictors before the arrow have a score of 0, those after the arrow have a score of 1. In the test, 10log-transformed costs were analysed. ICU, intensive care unit, IMC, intermediate care unit; LOS, length of stay, TISS-28, Therapeutic Intervention Scoring System with a median value of 28.