Skip to main content

Table 3 Odds ratios in multilevel modelling for source of care (GLLAMM) (base category = attendance at private for profit; Data source - UNHS 2002/3 and 2005/6)

From: Changes in utilization of health services among poor and rural residents in Uganda: are reforms benefitting the poor?

No of Level 1 units = 2806      
No of level 2 units = 56      
  OR SE P value Confidence interval
PNFP relative to private for profit
Female 1.2245 .3207 .439 .7327 2.0462
Rural residence 1.4229 .3237 .121 .9109 2.2226
Within 5 km to health facility .9954 .1477 .976 .7442 1.3314
Vulnerable female headed Household .7287 .2319 .320 .3905 1.3597
Vulnerable elderly headed HH 1.3924 .2742 .093 .9466 2.0483
Quintile 2 .6735 .1539 .084 .4303 1.0541
Quintile 3 .9459 .2036 .796 .6203 1.4425
Quintile 4 .8714 .1973 .543 .5590 1.3584
Quintile 5 .7930 .2087 .379 .4733 1.3286
Time (2005/6 = 1)* 2.1485 .3348 .000 1.583 2.9160
Variances and covariances of random effects level 2 (districts)
var(1): 6468 (.2091) Confidence interval = (0.22-1.06)
Public facility relative to private for profit
Female* 1.3773 1.5008 .003 1.1124 1.7052
Rural residence* 1.3176 .1221 .003 1.0986 1.5801
Within 5 km from the health facility .9301 .0576 .242 .8238 1.0501
Vulnerable female headed household 1.084 .1428 .539 .8375 1.4039
Vulnerable elderly headed HH* 1.5383 .1316 .000 1.3007 1.8193
Quintile 2 .9287 .0860 .425 .7746 1.1136
Quintile 3 .8957 .0852 .247 .7433 1.0793
Quintile 4* .7470 .0745 .003 .6143 .9085
Quintile 5* .5479 .0630 .000 .4373 .6865
Time (2005/6 = 1)* 1.3092 .0837 .000 1.1549 1.4841
Variances and covariance's of random effects level 2 (districts)
var(1): .4277 (.1030) Confidence interval = 0.2217,0.6338
  1. *Statistically significant.
  2. UNHS = Uganda National Household Survey; GLLAMM = Generalized linear latent and mixed models; OR = Odds ratio; SE = Standard error