| Post Period (vs. Pre Period) |
---|
Sample | Outcome | DID LevelA | P-Value | DID SlopeB | P-Value |
---|
Non-Hispanic White1 | Any Hospitalization | −0.45% | 0.28 | −0.04% | 0.22 |
Any ED Visit | −0.21% | 0.69 | −0.01% | 0.89 |
Non-Hispanic Black22 | Any Hospitalization | −0.70% | 0.37 | 0.01% | 0.87 |
Any ED Visit | −1.53% | 0.10 | −0.00% | 0.99 |
Hispanic3 | Any Hospitalization | −1.10% | 0.14 | −0.01% | 0.84 |
Any ED Visit | −0.96% | 0.33 | 0.02% | 0.73 |
- Notes: Linear regression used for utilization outcomes. Sample is person-months from 2010 to 2015. Regression covariates include Group (treatment vs. comparator); a linear monthly time trend, indicators and splines for the post period (2012–2015); and the interactions between these variables. Other covariates included sex, age, indicator for SSI status, 18 comorbidity indicators, and state fixed effects. Repeated measures adjusted for using Huber-White robust standard errors, clustering at the person level. 1. N = 205,371 2. N = 90,506 3. N = 64,528
- A. Difference between treatment vs. comparator in the discontinuity (change in level) for the post period, measured using the interaction between Group & an indicator variable for the post period
- B. Difference between treatment vs. comparator in the change in slope for the post period, measured using the interaction between Group & a spline variable for the post period