Table 2. Summary measures associated with the posterior distribution of parameters comprising the predictive structure of the selected Hurdle Logit-Poisson model.
| Binary portion of the model | I | Odds Ratioa | Pb | |
|---|---|---|---|---|
| mean | CI 95% | |||
| Intercept | 9.76 | (6.27; 15.27) | ||
| Number of dengue cases (lag = 1 week) | 14 | 9.39 | (5.25; 15.97) | 100.0% |
| EPI (lag = 4 weeks) | 0.29 | 1.20 | (1.09; 1.33) | 99.9% |
| Daytime temperature (lags = 6,7,8 weeks) | 2.5 | 1.37 | (1.20; 1.57) | 100.0% |
| Poisson portion of the model | I | Relative Risk a | P b | |
| mean | CI 95% | |||
| Intercept | 0.59 | (0.37; 0.94) | ||
| EDI (lags = 3,4,5 weeks) | 26 | 1.03 | (0.99; 1.07) | 90.9% |
| Poor population density | 1013 | 1.20 | (1.13; 1.28) | 100.0% |
aAssociated with increment I.
bP: Posterior probability of relative risk/odds ratio greater than 1, associated with increment I on the regressor.