Skip to main content

Table 3 Baseline covariates* (before admission) associated with relapse of male and female PWUDs

From: Factors associated with relapse into drug use among male and female attendees of a three-month drug detoxification–rehabilitation programme in Dhaka, Bangladesh: a prospective cohort study

Covariate

Male PWUDs (N = 123)

Female PWUDs (N = 89)

Category

B

p

Adjusted RHR (95% CI)

B

p

Adjusted RHR (95% CI)

Socio-demographic factors

Had previously undergone drug treatment

   

0.66

0.03

1.94 (1.08−3.49)

Location where they live

      

  Stable housing§

     

1.00 (Ref.)

 Unstable housing†

   

0.89

0.01

2.44 (1.24−4.81)

Person(s) with whom they usually live

      

 Family member, relative

  

1.00 (Ref.)

   

 Spouse, sex partner

−0.10

0.85

0.91 (0.32−2.55)

   

 Non-PWUD friend, colleague

1.06

0.09

2.89 (0.84−9.90)

   

 PWUD friend, drug dealer

0.82

<0.01

2.27 (1.25−4.12)

   

 Alone

0.85

0.02

2.35 (1.18−4.68)

   

Income level

      

  More than the monthly average ‡

   

0.64

0.02

1.89 (1.13−3.17)

History of drug misuse (in the last six months before admission)

Type of drug used as the first choice

      

  Cannabis, sleeping pills, codeine containing cough syrup, alcohol

     

1.00 (Ref.)

  Heroin

   

1.29

0.01

3.62 (1.39−9.46)

  Buprenorphine, Pethidine

   

1.10

0.05

3.00 (1.02−8.81)

Sexual behaviour (in the last six months before admission)

Not had sex with non-commercial partner

0.82

0.01

2.27 (1.19−4.33)

   
  1. *Only the covariates with significance of at least a 5% level for association with relapse were analysed by backward stepwise Cox multivariate analyses.
  2. §Stable housing = house, working place, shop, office, club.
  3. †Unstable housing = street, abandoned place, park, roof-top, veranda, bus/train terminal.
  4. ‡Monthly average incomes are BDT 6144 for male, and BDT 11397 for female (1USD is equivalent to 76.20 BDT as of 31 October 2011).
  5. B: Net effect of independent variable. RHR relative hazards risk, CI confidence interval.