Table 3.
Base model (with no interaction term) and difference-in-difference (DID) model using the dichotomous pre-post variable on the effect of drug courts on overdose mortality in United States counties.
| Base model |
DID model |
|||||
|---|---|---|---|---|---|---|
| Variable | Coefficient | 95% CI | Z-score | Coefficient | 95% CI | Z-score |
| (Intercept) | 6.1 | (5.916, 6.402) | 49.7 | 6.0 | (5.787, 6.269) | 49.0 |
| Year Centered | 1.2 | (1.215, 1.279) | 75.3 | 1.2 | (1.228, 1.292) | 76.8 |
| Treatment | 2.9 | (2.649, 3.169) | 21.9 | 3.8 | (3.570, 4.116) | 27.6 |
| Pre/post | 0.7 | (0.498, 0.932) | 6.4 | 0.9 | (0.744, 1.178) | 8.6 |
| Treatment * Pre/post | −2.9 | (−3.478, −2.370) | −10.3 | |||
| Urban/rural classification | ||||||
| Metropolitan | Ref | Ref | ||||
| Rural | −0.1 | (−0.277, 0.075) | −1.1 | −0.1 | (−0.275, 0.077) | −1.1 |
| Urban | 0.0 | (−0.172, 0.134) | −0.2 | 0.0 | (−0.171, 0.135) | −0.2 |
| Number of OTP | 0.1 | (0.019, 0.147) | 2.5 | 0.1 | (0.032, 0.152) | 2.9 |
| Medicaid expansion | −2.7 | (−3.221, −2.211) | −10.5 | −2.6 | (−3.150, −2.144) | −10.3 |
| Northeast region | 0.5 | (0.305, 0.703) | 4.9 | 0.5 | (0.313, 0.707) | 5.0 |
| Return of presidential voting | ||||||
| Democrat | Ref | Ref | ||||
| Republican | 0.9 | (0.822, 1.144) | 11.9 | 0.9 | (0.781, 1.103) | 11.4 |
| Others | 1.0 | (0.679, 1.391) | 5.6 | 1.1 | (0.776, 1.478) | 6.2 |
| % of uninsured population | 0.1 | (0.051, 0.071) | 11.8 | 0.1 | (0.052, 0.072) | 12.0 |