Skip to main content

Table 4 Posterior means and correspondence p-values for parameter estimation with inclusion and exclusion of patient specific variance

From: Determinants of CD4 cell count change and time-to default from HAART; a comparison of separate and joint models

Effect

Model (2): With homogeneity of variance for patients

Model (5): With patient specific variance/heterogeneity

Posterior estimate

p-value

Posterior estimate

p-value

Intercept

4.350

0.004

4.130

0.002

Age

−0.006

0.003

− 0.036

0.005

Weight

0.001

0.082

0.001

0.082

Baseline CD4

0.005

0.008

0.002

0.008

Area (Reference = Urban)

 Rural

−0.002

0.064

−0.003

0.003

Marital Status (Reference = without partner)

 With partner

0.007

0.003

0.005

0.003

Sex (Reference = male)

    

 Female

0.034

0.004

0.023

0.003

Level of education (Reference = Tertiary)

 No educationl

0.001

0.076

0.001

0.076

 Primary education

0.017

0.044

0.017

0.034

 Secondary education

0.020

0.028

0.010

0.038

Level of Income (Reference = High income)

 Low income

−0.010

0.074

−0.020

0.064

 Middle income

−0.006

0.062

−0.016

0.052

Owner of Cell phone (Reference = With phone)

 Without cell phone

−0.001

0.004

−0.001

0.004

Level of adherence (Reference = Good adherence)

 Poor adhere

−0.523

0.004

−0.483

0.002

 Fair adhere

−0.452

0.001

−0.552

0.002

Level of disclosure the disease (Reference = yes)

 No

- 0.00195

0.0031

−0.1020

0.0031

WHO stages (Reference = Stage IV)

 Stage I

0.1511

0.0021

0.2511

0.0021

 Stage II

0.1567

0.0321

0.1567

0.0321

 Stage III

0.1381

0.2311

0.4381

0.0311

 Time

0.0210

0.0013

0.1210

0.0213

 Var (νi0)

0.8820

 

0.8620

 

 Var (νi1)

0.0210

 

0.0110

 

 Cov (νi0, νi1)

- 0.0542

 

- 0.0642

 

 Corr (νi0, νi1)

−0.8214

 

−0.7314

 

 DIC

234,763

212,393

  1. νi0 is the random of patient i effect and  νi1 is the random fraility term