Skip to main content

Table 3 Logistic regression onto hospital use (N = 432)

From: Drug use, homelessness and health: responding to the opioid overdose crisis with housing and harm reduction services

 

Unadjusted

Adjusted

 

OR

95% CI

P value

OR

95% CI

P value

Housing unstable (1 = yes)

2.409

1.619–3.586

0.000

2.035

1.289–3.214

0.002

Sex (1 = female)

0.937

0.627–1.400

0.753

0.055

0.006–0.474

0.008

Ethnicity (1 = indigenous)

1.319

0.873–1.963

0.193

1.221

0.757–1.968

0.413

Age

0.99

0.97–1.00

0.117

1.00

0.98–1.02

0.69

Location (1 = medicine hat or red deer)

0.686

0.468–1.007

0.054

0.669

0.429–1.046

0.078

Had an overdose (1 = yes)

3.965

2.584–6.083

 < 0.001

3.586

2.206–5.829

 < 0.001

How often: polydrug use (1 = 2–3 times/week or more)

1.801

1.223–2.663

0.003

1.108

0.675–1.819

0.684

How often: heavily influenced by drugs (1 = weekly or daily)

1.842

1.144–2.966

0.012

1.298

0.722–2.330

0.383

How often: irresistible longing to use (1 = weekly or daily)

1.519

1.023–2.257

0.038

0.753

0.418–1.359

0.347

How often: unable to stop use (1 = weekly or daily)

1.363

0.924–2.011

0.118

0.767

0.439–1.338

0.350

How often: neglected tasks due to use (1 = weekly or daily)

2.839

1.861–4.334

 < 0.001

2.187

1.284–3.726

0.004

How often: feel guilty due to drug use (1 = weekly or daily)

1.661

1.069–2.581

0.024

1.094

0.637–1.878

0.745

Others worry due to use (1 = yes)

2.365

1.292–4.329

0.005

1.728

0.850–3.511

0.131

Diagnosed addiction or mental health concern (1 = weekly or daily)

2.449

1.428–4.200

0.001

1.262

0.634–2.513

0.508

Inadequate access (1 = yes)

1.601

1.086–2.361

0.017

1.283

0.815–2.021

0.282

Inadequate amount (1 = yes)

1.204

0.818–1.773

0.346

1.000

0.638–1.569

0.999

Sex * diagnosed

1.331

0.879–2.015

0.177

22.433

2.466–204.064

0.006

How often: need to use in morning (1 = yes)

1.837

1.206–2.797

0.005*

Hurt due to use (1 = yes)

1.747

1.022–2.987

0.041*

Constant

   

0.001

  
  1. *Not statistically significant in final adjusted regression model