Table 2.
Ability of FINDER to detect unsafe restaurants as compared to BASELINE rate and with subcategories of the baseline inspections, including complaint-based inspections that occurred in Chicago and routine inspections from both Chicago and Las Vegas
| FINDER n = 132 |
BASELINE n = 10,786 |
Odds ratioa [95% CI] | p-value | |
|---|---|---|---|---|
| Overall, number unsafe (%) | 69 (52.3%) | 2662 (24.7%) | 3.06 [2.14–4.35] | <0.001 |
| Risk level | ||||
| High, number unsafe (%) | 42 (50.0%) | 1909 (33.5%) | 1.98 [1.28–3.05] | 0.002 |
| Medium, number unsafe (%) | 23 (59.0%) | 536 (23.1%) | 5.50 [2.83–10.72] | <0.001 |
| Low, number unsafe (%) | 4 (44.4%) | 217 (7.9%) | 7.35 [1.79–30.13] | 0.006 |
| Comparison of FINDER to complaint-based inspections | ||||
|
FINDER n = 71 |
COMPLAINT n = 1291 |
|||
| Overall, number unsafe (%) | 37 (52.1%) | 508 (39.4%) | 1.68 [1.04–2.71] | 0.03 |
| Risk level | ||||
| High, number unsafe (%) | 27 (47.4%) | 374 (39.4%) | 1.38 [0.81–2.36] | 0.24 |
| Medium, number unsafe (%) | 9 (75.0%) | 115 (39.3%) | 4.64 [1.23–17.51] | 0.02 |
| Low, number unsafe (%) | 1 (50.0%) | 19 (38.8%) | 1.58 [0.09–26.78] | 0.75 |
| Comparison of FINDER to routine inspections | ||||
|
FINDER n = 132 |
ROUTINE n = 9495 |
|||
| Overall, number unsafe (%) | 69 (52.3%) | 2,154 (22.7%) | 3.16 [2.22–4.51] | <0.001 |
| Risk level | ||||
| High, number unsafe (%) | 42 (50.0%) | 1531 (32.2%) | 2.07 [1.35–3.20] | 0.001 |
| Medium, number unsafe (%) | 23 (59.0%) | 424 (20.9%) | 5.52 [2.84–10.76] | <0.001 |
| Low, number unsafe (%) | 4 (44.4%) | 199 (7.3%) | 7.65 [1.90–30.89] | 0.004 |
aOdds ratios from binomial logistic regressions