Skip to main content

Table 5 Poisson regression model (with interaction term between initial number of nodules and treatment arms)

From: Individuals living in an onchocerciasis focus and treated three-monthly with ivermectin develop fewer new onchocercal nodules than individuals treated annually

VariableSaturated modelStepwise model 
b (95% CI)P-valueb (95% CI)P-value
Increase in no. of new nodules for each additional initial nodulea
 150 µg/kg annually0.043 (0.014–0.071)0.0040.046 (0.019–0.073)0.001
 800 µg/kg annually0.067 (0.039–0.095)< 0.0010.070 (0.043–0.096)< 0.001
 150 µg/kg 3-monthly0.021 (− 0.004–0.047)0.1060.024 (− 0.000–0.048)0.052
 800 µg/kg 3-monthly0.026 (− 0.001–0.053)0.0630.029 (0.004–0.056)0.026
Age0.005 (− 0.00001–0.010)0.0510.006 (0.002–0.009)0.003
Weight− 0.001 (− 0.005–0.003)0.659  
CMFL in the village of residence
 LowRef.   
 Middle0.090 (− 0.154–0.333)0.470  
 High0.151 (− 0.072–0.375)0.185  
MFD
 0–30 mf/mgRef. Ref. 
 31–80 mf/mg0.089 (− 0.103–0.281)0.3640.105 (− 0.079–0.288)0.263
 81–170 mf/mg0.278 (0.095–0.461)0.0030.301 (0.128–0.474)0.001
 > 171 mf/mg0.371 (0.194–0.549)< 0.0010.394 (0.227–0.561)< 0.001
P of the likelihood-ratio test for the interaction term: 0.035
 P of the model< 0.001 < 0.001 
 AIC1649 1645 
 BIC1695 1678 
 Log likelihood− 813.5 − 814.6 
  1. aThe reference is defined as 150 µg/kg annually and 2 initial nodules
  2. Abbreviations: b, regression coefficient; 95% CI, 95% confidence interval; Ref., reference; CMFL, community microfilarial load; MFD, microfilarial density; AIC, Akaike’s information criterion; BIC, Bayesian information criterion