Table 4.
Parameter estimates of significant variables using negative binomial regressions on Lyme borreliosis infections (1998–2004)
| Non-spatial model | Spatial model | |
| Parameter | Estimate | Estimate |
| Intercept | -10.41*** | -11.08*** |
| propforest | 0.0225*** | 0.0254*** |
| income | 0.0001*** | 0.00004*** |
| housesep | 0.0084*** | 0.0098*** |
| roedeer | 0.0955*** | 0.0302 |
| Linear predictors in neigbourhood municipalities | 0.3675*** | |
| Degrees of freedom | 584 | 583 |
| Null deviance | 857.9 | 916.6 |
| Residual deviance | 642.4 | 642.3 |
| AIC | 3220.9 | 3182.9 |
| 2 × log-likelihood | -3208.9 | -3168.9 |
*** P-value < 0,0001