Skip to main content
. 2014 Jan 10;14:6. doi: 10.1186/1472-6874-14-6

Table 3.

Structured multiphase logistic regression model of demographics, causes of injury and injury episodes associated with multiple injuries (n = 199)

 
Multiple injuries
  OR a 95% CI b p
Phase I (Nagelkerke R square = 3.3%, Hosmer and Lemeshow Test p = .99)
Age
 
 
.418
≤ 29
1.0
 
 
30-49
.6
(.3, 1.3)
 
≥ 50
.6
(.2, 1.5)
 
Race
 
 
.098
Chinese
1.0
 
 
Non-Chinese
2.1
(.9, 4.9)
 
Phase II (Nagelkerke R square = 5.5%, Hosmer and Lemeshow Test p = 1.0)
Relationship with perpetrator
 
 
.041*
Married spouse
1.0
 
 
Cohabitant
3.3
(1.4, 7.8)
 
Boyfriend but non-cohabitant
1.1
(.1, 17.0)
 
Ex-husband / ex-boyfriend
2.1
(.4, 11.9)
 
Phase III (Nagelkerke R square = 18.2%, Hosmer and Lemeshow Test p = .23)
Episode of abuse
 
 
.894
1 st
1.0
 
 
2 nd
1.5
(.5, 4.1)
 
Multiple
1.2
(.5, 2.8)
 
Undisclosed
1.3
(.5, 3.9)
 
Causes of injury c
 
 
 
Punching with a fist
4.0
(2.1, 7.5)
<0.001*
Slapping
1.4
(.6, 3.2)
.377
Pushing, shoving or grabbing
1.6
(.7, 3.7)
.251
Weapon use
1.2
(.6, 2.5)
.545
Kicking
1.2
(.5, 3.1)
.670
Arm or hair twisting
1.0
(.3, 2.9)
.976
Slamming against a wall
1.2
(.4, 3.7)
.749
Throwing things
3.1
(.6, 15.0)
.167
Admission time
 
 
.747
Day-time
1.0
 
 
Night-time 1.1 (.6, 2.0)  

Note: aOR refers to odds ratio; bCI refers to confidence interval; Causes of injury; cThe comparators are “no punching with a fist”, “no slapping”, “no pushing, shoving or grabbing”, “no weapon use”, “no kicking”, “no arm or hair twisting”, “no slamming against a wall”, “no throwing things” respectively; *p < 0.05; Tolerance > 0.9 and VIF < 1.1.