Table 3. Changes in Quality Measures for Long-Stay Nursing Home Residents After PE Firm Acquisition Compared With For-Profit Nursing Homes Without PE Firm Ownership Examined in Secondary Analysesa.
| Minimum data set quality measure | Pooled sample, 2012-2018, No. (%)b | Preacquisition period (2012) | Postacquisition period (2018) | Differential change | Relative change, %c | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| All | PE | For-profit | Unadjusted difference | PE | Non-PE | Unadjusted difference | Unadjusted (95% CI) | P value | Adjusted (95% CI) | P value | |||
| Antipsychotic medications (n = 230 687) | 49 598 (21.5) | 22.1 | 20.8 | 22.2 | −1.4 | 16.8 | 19.3 | −2.5 | −1.1 (−4.8 to 2.4) | .53 | −0.2 (−1.7 to 1.4) | .83 | −0.9 |
| Pressure ulcer (n = 278 188) | 13 631 (4.9) | 3.6 | 3.8 | 3.6 | 0.2 | 5.0 | 5.6 | −0.6 | 0.8 (−2.4 to 0.7) | .26 | 0.5 (−0.4 to 1.3) | .30 | 13.5 |
| Severe pain (n = 217 284) | 16 731 (7.7) | 9.4 | 9.1 | 9.4 | −0.3 | 4.9 | 5.3 | −0.4 | −0.1 (−2.5 to 2.3) | .94 | 0.2 (−1.1 to 1.4) | .79 | −0.2 |
Abbreviation: PE, private equity.
Linear regressions were used for estimation. All models included the following covariates: age group (65-69, 70-74, 75-79, 80-84, and ≥85 years), race and ethnicity (Black, White, and other non-White race [Asian, Hispanic, North American Native, and other]), sex, dual eligibility for Medicare and Medicaid, indicators for 66 chronic and disabling conditions used for risk adjustment (see eTable 2 in the Supplement for a list of the chronic conditions), activities of daily living score at initial assessment (range, 1-28, where a higher score indicates a greater need for assistance with activities of daily living), and severe cognitive impairment (scores >3 on the 4-point Cognitive Function Scale). Nursing home characteristics included occupancy rate, an indicator for multifacility chain affiliation, total number of beds, and terciles of the distributions of the percentage of patients covered by Medicare and the percentage covered by Medicaid. Other covariates included fixed effects for quarter, year, nursing home, Hospital Referral Region, and Hospital Referral Region interaction with year. Standard errors were adjusted for clustering at the level of the nursing home.
The pooled sample consists of all resident observations from 2012 to 2018.
Relative changes were derived from the sample by dividing the adjusted estimates for all outcomes by the unadjusted mean of the outcomes in the preacquisition period (2012).