Table 4 : Results of the ordinary least square regression model and the negative binomial regression model.


  OLS Model (AIC=1919) Negative Binomial Model (AIC=1531)

Effect p-value Least Square Adjusted
Mean
Standard
Error
p-value Least Square
Adjusted
Mean
Standard
Error
Parameter
Estimate
Standard
Error
Wald
Chi-Square

Intercept <0.0001 -- -- 0.0497 -- -- 0.33 0.17 3.85
Shift (Day) 0.0002 1.72 0.09 0.0001 1.63 0.08 0.44 0.11 15.04
Shift (Night) Ref 1.28 0.10 Ref 1.05 0.09 Ref Ref Ref
Not Difficult <0.0001 1.16 0.09 <0.0001 0.98 0.08 -0.58 0.12 24.95
Difficult Ref 1.84 0.11 Ref 1.75 0.09 Ref Ref Ref
Dehydrated (No/Unknown) <0.0001 1.10 0.07 <0.0001 0.98 0.07 -0.57 0.12 21.62
Dehydrated (Yes) Ref 1.90 0.13 Ref 1.74 0.10 Ref Ref Ref
Not Cooperative Child 0.0004 1.75 0.10 0.0003 1.63 0.08 0.44 0.12 12.95
Cooperative Child Ref 1.25 0.11 Ref 1.05 0.09 Ref Ref Ref
Nurse Experience (<1 Year) 0.0013 1.70 0.11 0.0012 1.56 0.09 0.36 0.11 10.43
Nurse Experience (1 Year+) Ref 1.30 0.09 Ref 1.10 0.08 Ref Ref Ref
OSBDM Score 0.0804 * * 0.2492 * * -0.0214 0.02 1.33
Dispersion --  --  --  -- -- -- 0.56 0.11 --

This table lists the results of the regression models for both the ordinary least squares (OLS) model and the negative binomial (NB) model.Parameter estimates
for the OLS model are not provided, as direct comparison of these values to the NB model is not appropriate. Instead, the model based adjusted means for each
factor are listed.

Brinkley et al.Journal of Medical Statistics and Informatics  2014 2:6DOI : 10.7243/2053-7662-2-6