Skip to main content

Table 5 Results of random parameter logit model subgroup analysis for HIV status of respondent

From: Developing a patient-centered community-based model for management of multi-drug resistant tuberculosis in Uganda: a discrete choice experiment

Attributes and levelsa

HIV Negative

HIV Positive

P value†

 

Coefficient

SD

Relative importance

Coefficient

SD

Relative importance

 

DOT provider

  

34.2

  

24.2

 

 Family (reference)

–

–

 

–

–

  

 Community health work

1.22

1.78

 

0.95

2.07

 

0.356

 Expert client

0.89

1.50

 

0.84

2.37

 

0.741

Location

  

27.5

  

43.0

 

 Workplace (reference)

 Home

0.98

1.11

 

1.69

1.26

 

0.042

Support

  

38.4

  

32.8

 

SMS (reference)

 Phone call reminders

0.76

0.56

 

0.56

0.01

 

0.668

 Travel vouchers

1.37

0.47

 

1.29

0.53

 

0.582

  1. aDummy coded attributes (coefficient of the reference category is constrained to be 0)
  2. CI Confidence interval, SD standard deviation for preference heterogeneity (random component of the model coefficients)
  3. Number of observations = 1512 (HIV negative), 960 (HIV positive)
  4. † P value for indicating statistical difference in preference weights of attribute levels by HIV status