Skip to main content

Table 3 Regression results of stratified models

From: Effects of health intervention programs and arsenic exposure on child mortality from acute lower respiratory infections in rural Bangladesh

  β SE p
Treatment area    
Negative binomial model    
 Arsenic (100 µg/L) 0.010 0.042 0.807
 High socioeconomic status −0.348 0.140 0.013
 Cost distance to clinica 0.125 0.147 0.396
 (Intercept) −4.777 0.195 0.000
 Log(theta) 0.674 0.559 0.228
Zero-inflated model    
 Arsenic (100 µg/L) −2.921 2.742 0.287
 High socioeconomic status 0.502 1.677 0.765
 Cost distance to Clinic1 1.945 1.488 0.191
 (Intercept) −3.214 1.405 0.022
N 3251   
Comparison area    
Negative binomial model    
 Arsenic (100 µg/L) 0.008 0.022 0.702
 High socioeconomic status −0.302 0.090 0.001
 Cost distance to clinica −0.151 0.058 0.010
 (Intercept) −4.306 0.084 0.000
 Log (theta) 2.807 1.815 0.122
Zero-inflated model    
 Arsenic (100 µg/L) −1.686 3.006 0.575
 High Socioeconomic Status −1.310 1.712 0.444
 Cost Distance to Clinica −5.137 3.003 0.087
 (Intercept) −8.242 2.693 0.002
N 3442   
  1. Zero-inflated negative binomial regression analysis of bari-level acute lower respiratory infection (ALRI) mortality rates in children under 2 years of age in Matlab, Bangladesh, 1989–1996 by treatment versus comparison area. Data source: icddr,b Matlab Health and Demographic Surveillance System
  2. aCentered to the mean and scaled by the standard deviation