Fixed-effect negative binomial models for association between new case detection rate of leprosy and Bolsa Família Program and Family Health Program coverage, Brazil 2004–2011.

Data are risk ratio (95% CI) unless otherwise specified. TPC = target population coverage. ‡Cutoff taken from tertiles of the distribution of FHP coverage of the total population. *Cutoff is median value.The regression model to be estimated was as follows: Yit = αi + β1BFPit + β2FHPit + βnXnit + uitWhere Yit was the leprosy detection rate for the municipality i in year t, αi is the fixed effect for the municipality i that captures all unobserved time-invariant factors, BFPit is the Bolsa Familia Program coverage for the municipality i in the year t, FHPit the Family Health Program coverage for the municipality i in the year t, Xnit was the value of each n covariate of the model with in the municipality i in the year t, and uit was the error.Fixed-effect negative binomial models for association between new case detection rate of leprosy and Bolsa Família Program and Family Health Program coverage, Brazil 2004–2011.