Table 2.
Differences in Patient Characteristics between Acquired and Control Hospitals, before and after Acquisition.*
| Characteristic | Mean in Control Hospitals† | Pretransaction Difference between Acquired and Control Hospitals (95% CI)‡ | Differential Change from Pretransaction Period to Third Post-Transaction Year (95% CI)§ |
|---|---|---|---|
| Age (yr) | 72.5 | −0.14 (−0.77 to 0.49) | −0.13 (−0.32 to −0.07) |
| Female sex (%) | 55.4 | 0.15 (−0.76 to 1.05) | −0.11 (−0.49 to 0.26) |
| Non-Hispanic white race or ethnic group (%)¶ | 75.2 | −1.90 (−4.76 to 0.96) | −0.33 (−0.72 to 0.06) |
| Dually enrolled in Medicaid (%) | 30.7 | 2.08 (0.15 to 4.00) | 0.27 (−0.53 to 1.08) |
| Disability as original reason for Medicare eligibility (%) | 33.6 | 1.17 (−0.62 to 2.95) | 0.27 (−0.27 to 0.80) |
| Chronic conditions (no.)‖ | 8.2 | 0.05 (−0.08 to 0.18) | −0.01 (−0.05 to 0.03) |
| DRG payment weight** | 1.6 | −0.02 (−0.04 to 0.01) | 0.00 (−0.02 to 0.02) |
Hospital-level observations were weighted according to count of hospital beds. Pretransaction differences and differential changes were adjusted for hospital, year, and state-by-year fixed effects. Pretransaction differences and differential changes in female sex, non-Hispanic white race or ethnic group, dual enrollment in Medicaid, and disability as the original reason for Medicare eligibility are in percentage points.
The mean of patient characteristics for control hospitals was calculated in 2011 (i.e., at the midpoint of the study period).
Pretransaction differences reflect the mean for acquired hospitals minus the mean for control hospitals. For example, the mean age of patients was 0.14 years lower in acquired hospitals than in control hospitals.
Negative values for differential change indicate that the change from the pretransaction period to the third post-transaction year for acquired hospitals was less than the change for control hospitals. For example, the mean age differentially decreased by 0.13 years in acquired hospitals relative to control hospitals.
Race and ethnic group were determined from Medicare enrollment files with the use of the Research Triangle Institute-modified race code.
The number of a patient’s chronic conditions was computed from data in the Chronic Conditions Data Warehouse.
Shown is the average diagnosis-related group (DRG) payment weight for the admissions at each hospital computed from claims.