All models were refined through a backward stepwise process with variables retained if Likelihood ratio P-values were < 0.05 (Williams et al., 2013a). At each step of model building Omnibus tests identified if factors had a significant effect on the model fit and should be retained(P < 0.05) (Pallant, 2010). Model fit was assessed using the Hosmer-Lemeshow goodness of fit test (P > 0.05) (George and Mallery, 2010). The predictive ability of the model was examined through receiver operating characteristic (ROC) curve analysis (Reardon et al., 2012). All statistical analyses were conducted using Statistic Packages for Social Sciences (SPSS) version 21 (Chicago, IL, USA).