Table 1.

Summary of (a) logistic GLMM (individual-level probability that a child became reinfected after praziquantel treatment) and (b) negative binomial GLMM (egg count of reinfected children) aiming to identify habitat-related, snail–habitat, and snail predictors of human urogenital schistosomiasis burden

Models are numbered and ordered by their BIC rank. Models within 10 ∆BIC of the top model are marked in gray in the dBIC column. dMSE is indicated, with top models by MSE marked in gray in the dMSE column. Also shown are marginal (i.e., associated with fixed effects) and conditional (i.e., associated with fixed plus random effects) R². The remaining columns describe which variables were included in each model. Habitat-related variables contain only information about habitat. Snail variables contain only information about snails. Snail–habitat values are obtained using information about both habitat and snails (SI Appendix, Table S1).