Table 4. Estimates of Association Between Reforms Following the Patient Protection and Affordable Care Act and Total Hospital Spending for All Diagnoses From Alternative Specifications.
| Specification | DID | Generalized synthetic control, acute care vs critical access hospitals | |
|---|---|---|---|
| Targeted vs untargeted diagnosesa | Acute care vs critical access hospitalsb | ||
| No price standardization | |||
| No. | 6 551 547 | 7 634 242 | 7 566 671 |
| Treatment effect estimate (95% CI), $ | −418 (−494 to −343) | −1947 (−2042 to −1852) | −1978 (−2279 to −1683) |
| No adjustment for coded risk | |||
| No. | 6 551 547 | 7 634 242 | 7 566 671 |
| Treatment effect estimate (95% CI), $ | −398 (−459 to −337) | −1781 (−1859 to −1704) | −1214 (−1458 to −946) |
| Hospitals in regions with highest tercile of Medicare Advantage penetration | |||
| No. | 2 190 772 | 2 562 193 | 2 476 099 |
| Treatment effect estimate (95% CI), $ | −382 (−524 to −240) | −1562 (−1774 to −1350) | −929 (−1272 to −591) |
| Hospitals in regions with lowest tercile of Medicare Advantage penetration | |||
| No. | 1 890 385 | 2 268 390 | 2 254 501 |
| Treatment effect estimate (95% CI), $ | −485 (−616 to −354) | −2164 (−2298 to −2030) | −1433 (−1898 to −1064) |
| Winsorized payment | |||
| No. | 6 551 547 | 7 634 242 | 7 566 671 |
| Treatment effect estimate (95% CI), $ | −470 (−511 to −430) | −1413 (−1473 to −1353) | −995 (−1189 to −793) |
Abbreviation: DID, difference-in-differences.
a
All targeted vs untargeted diagnoses models were adjusted for age, race, sex, hierarchical condition category score, clinical classifications software, hospital profit status, proportion of Medicare days, and proportion of Medicaid days.
b
All acute care vs critical access hospitals DID models were adjusted for age, race, sex, hierarchical condition category score, diagnosis-related group weight, hospital profit status, proportion of Medicare days, and proportion of Medicaid days.