OR of block-level shootings in relation to alcohol outlet proximity
| Variable | Model 1 | Model 2 | Model 3 | Model 4 | ||||
| OR (95% CI) | P value | OR (95% CI) | P value | OR (95% CI) | P value | OR (95% CI) | P value | |
| Beer stores | ||||||||
| Same block | 1.7 (0.9 to 3.5) | 0.12 | 1.7 (0.9 to 3.5) | 0.12 | 1.7 (0.9 to 3.5) | 0.12 | 1.2 (0.6 to 2.5) | 0.62 |
| Within one block | 1.6** (1.2 to 2.2) | 0.005 | 1.6** (1.2 to 2.2) | 0.004 | 1.6** (1.2 to 2.2) | 0.004 | 1.5* (1.1, 2.1) | 0.02 |
| Within two blocks | 1.0 (0.8 to 1.2) | 0.86 | 1.0 (0.8 to 1.2) | 0.94 | 1.0 (0.8 to 1.2) | 0.95 | 1.0 (0.8 to 1.2) | 0.96 |
| Bars/restaurants | ||||||||
| Same block | 2.2*** (1.5 to 3.2) | <0.001 | 2.1*** (1.4 to 3.0) | <0.001 | 2.1*** (1.4 to 3.0) | <0.001 | 1.6* (1.1 to 2.4) | 0.02 |
| Within one block | 1.3** (1.1 to 1.6) | 0.009 | 1.3** (1.1 to 1.6) | 0.009 | 1.3** (1.1 to 1.6) | 0.008 | 1.2 (1.0 to 1.5) | 0.07 |
| Within two blocks | 0.9 (0.7 to 1.1) | 0.29 | 0.8 (0.7 to 1.1) | 0.34 | 0.9 (0.7 to 1.1) | 0.28 | 0.9 (0.7 to 1.1) | 0.33 |
| Black population | – | – | 2.0*** (1.5 to 2.7) | <0.001 | 1.9*** (1.4 to 2.6) | <0.001 | 1.9*** (1.4 to 2.6) | <0.001 |
| Hispanic population | – | – | 1.8** (1.3 to 2.7) | <0.01 | 1.4 (0.9 to 2.0) | 0.09 | 1.4 (1.0 to 2.0) | 0.09 |
| Unemployed population | – | – | 0.9 (0.87 to 1.2) | 0.54 | 1.0 (0.7 to 1.3) | 0.82 | 1.0 (0.7 to 1.3) | 0.75 |
| Income | – | – | 1.1 (0.9 to 1.4) | 0.24 | 1.1 (0.9 to 1.4) | 0.24 | 1.2 (0.9 to 1.4) | 0.19 |
| Street trees | – | – | 0.9 (0.8 to 1.1) | 0.43 | 0.9 (0.8 to 1.1) | 0.45 | 0.9 (0.7 to 1.1) | 0.32 |
| Vacant lots | – | – | 1.0 (0.9 to 1.2) | 0.85 | 1.0 (0.8 to 1.2) | 0.90 | 1.0 (0.8 to 1.2) | 0.86 |
| Drug activity | – | – | – | – | 1.6*** (1.3 to 1.9) | <0.001 | 1.6*** (1.3 to 2.0) | <0.001 |
| Commercial land use | ||||||||
| Same block | – | – | – | – | – | – | 1.6*** (1.2 to 2.0) | <0.001 |
| Within one block | – | – | – | – | – | – | 1.2 (0.9 to 1.7) | 0.12 |
| Within two blocks | – | – | – | – | – | – | 0.8 (0.6 to 1.0) | 0.07 |
*Significant at p<0.05, **significant at p<0.01, ***significant at p<0.001.
Note: all model estimates are adjusted for unit area using third-degree polynomial splines (not shown) and adjusted for spatial dependence using thin plate splines fit to the deviance residuals of the model in a first-stage fitting (not shown). To allow direct comparisons of model coefficients, continuous variables were mean centred and scaled by 2 SD prior to modelling.