Table 3

Adjusted Odds of dying at home compared to odds of dying in other settings and 95% confidence intervals for patient characteristics from multivariable logistic regression analysis for subjects dying of cancer, heart disease, and cerebrovascular disease from 1980–2002.

Variables

b*

Wald Chi-square statistics

P

Odds Ratios

95% Confidence Interval


Intercept

-2.263

562139.321

< 0.0001

Year of death

1980–1984

1

1985–1989

-0.488

46346.315

< 0.0001

0.614

0.611–0.617

1990–1994

-0.798

119217.658

< 0.0001

0.450

0.448–0.452

1995–1999

-1.095

209436.362

< 0.0001

0.334

0.333–0.336

2000–2002

-1.334

209958.211

< 0.0001

0.263

0.262–0.265

Age

Under 65

1

65–74

0.263

8926.060

< 0.0001

1.300

1.293–1.307

75–84

0.777

97556.695

< 0.0001

2.174

2.164–2.185

85 or over

1.262

225213.075

< 0.0001

3.534

3.515–3.552

Gender

Female

1

Male

0.075

2204.343

< 0.0001

1.078

1.074–1.081

Cause of death

Cancer

1

Cerebrovascular

1.285

351250.089

< 0.0001

3.616

3.600–3.631

Heart disease

1.461

476921.234

< 0.0001

4.309

4.291–4.327

-2log likelihood

10113580

Chi-square model (df = 10)

1467394.7

P

< 0.0001

Overall rate of correct classification

74.0%


* Unstandardized logistic regression coefficients

Yang et al. BMC Palliative Care 2006 5:2   doi:10.1186/1472-684X-5-2

Open Data