Skip to main content
. 2014 May 15;14:64. doi: 10.1186/1471-2318-14-64

Table 1.

Logistic regression analysis predicting 5-year mortality

 
 
Univariate logistic regression
Multivariate
    OR (IC 95% ) p R 2 (%) AUC AUC (*)
Covariates of adjustment
 
 
 
 
Age
(Continuous)
1.17(1.06,1.28)
0.001
14.6
0.70(0.59,0.81)
 
Sex
Female
Reference
 
 
 
 
 
Male
1.17(0.52,2.65)
0.708
0.2
0.52(0.40,0.64)
0.79(0.71,0.88)
Cognitive
No
Reference
 
 
 
 
Impairment
Mild
2.02(0.66,6.24)
 
 
 
 
 
Mod/severe
7.58(2.10,27.4)
0.007
15.3
0.69(0.59,0.79)
 
Barthel
4
Reference
0.056
11.2
0.72(0.62,0.82)
 
3
2.14(0.70,6.58)
2
2.18(0.55,8.68)
1
7.14(1.54,33.15)
Comorbidity indices
 
 
 
 
ChI
1
Reference
 
 
 
 
 
2
1.92(0.66,5.61)
0.144
7.0
0.64(0.53,0.75)
0.81(0.73,0.89)
 
3
2.30(0.67,7.90)
 
 
 
 
 
4
4.03(1.06,15.31)
 
 
 
 
DC
1
Reference
 
 
 
 
 
2
5.76(1.17,28.24)
0.106
8.1
0.58(0.45,0.70)
0.80(0.72,0.88)
 
3
2.38(0.84,6.75)
 
 
 
 
 
4
1.92(0.66,5.56)
 
 
 
 
CIRS
1
Reference
 
 
 
 
 
2
1.23(0.40,3.78)
0.601
2.4
0.54(0.42,0.66)
0.80(0.71,0.88)
 
3
0.97(0.34,2.80)
 
 
 
 
 
4
2.25(0.60,8.46)
 
 
 
 
ICED
1
Reference
 
 
 
 
 
2
0.70(0.21,2.34)
0.344
4.5
0.56(0.45,0.67)
0.79(0.71,0.87)
 
3
0.75(0.26,2.12)
 
 
 
 
 
4
2.17(0.58,8.20)
 
 
 
 
GIC
1
Reference
 
 
 
 
 
2
0.50(0.15,1.67)
0.012
16.1
0.66(0.56,0.76)
0.79(0.71,0.88)
 
3
1.04(0.27,4.01)
 
 
 
 
 
4
4.62(0.96,22.09)
 
 
 
 
Prognostic indices
 
 
 
 
 
BISEP
1
Reference
 
 
 
 
 
2
1.41(0.52,3.85)
0.010
17.2
0.73(0.63,0.82)
0.81(0.73,0.89)
 
3
6.92(1.29,37.05)
 
 
 
 
 
4
10.0(1.91,52.47)
 
 
 
 
PI
1
Reference
 
 
 
 
 
2
3.10(0.83,11.66)
0.005
20.9
0.72(0.62,0.83)
0.80(0.71,0.88)
 
3
5.85(1.52,22.57)
 
 
 
 
  4 18.00(4.01,80.71)        

BISEP: Burden of Illness Score for Elderly Persons; ChI: Charlson Index; CIRS: Cumulative Illness Rating Scale; DC: Disease Count; ICED: Index of Coexistent Disease; GIC: Geriatric Index of Comorbidity; PI: Prognostic Index *Quartile ranges do not aply to ICED and GIC, as these indices are predefined into 4 classes

(*) The multivariate AUC for the “covariates of adjustment” refers to the model including only age, sex, Barthel and cog impairment, whereas the multivariate AUC given for each index is the one obtained from the multivariate model including the formers and each one of the indices.