Research Letter
July 30, 2024
Hospital Assets Before and After Private Equity Acquisition
Elizabeth Schrier, Hope E. M. Schwartz, David U. Himmelstein, et al
JAMA. 2024;332(8):669-671. doi:10.1001/jama.2024.13555
Private equity firms spent $505 billion on health care acquisitions between 2018 and 2023.1Financial infusions may augment resources for care.2,3 However, firms have sometimes sold acquired hospitals’ land and buildings, repaying investors with proceeds and burdening hospitals with rent payments for facilities they once owned.3
We assessed changes in hospitals’ capital assets after private equity acquisition.
Methods
We identified private equity–acquired US acute care hospitals and their acquisition years with the Private Equity Stakeholder Project Private Equity Hospital Tracker (eAppendix in Supplement 1), which we verified and augmented using a prior study,4 press reports, and Medicare cost reports. We excluded HCA Healthcare–owned hospitals, which prior studies suggest are atypical of private equity acquisitions.5
Using 2006 to 2021 Medicare cost reports, we assessed hospitals’ total capital assets (summing land, buildings, equipment, and health information technology), year last acquired, share of Medicaid inpatient discharges, rurality, teaching status, number of beds, and region. We excluded acquired hospitals with total assets outside the 95th and fifth percentiles during acquisition year (to minimize distortions from data-entry errors) and those acquired after 2019 (to ensure at least 2 years of postacquisition data).
We matched acquired hospitals to 10 nonacquired controls, using exact matching on year, region, and bed size category and nearest-neighbor matching on total capital assets during acquisition year (eAppendix in Supplement 1).4,6 We analyzed assets of acquired and control hospitals in 2 years before and after acquisition, using a difference-in-differences event-study specification. We used a linear mixed-effects model with year fixed effects and a random intercept term for each matched group to adjust for within-group correlation. The interaction term between time and acquisition status provides the difference-in-differences estimate. To facilitate verification of the parallel trends assumption and assess whether the 2-year postacquisition trends persisted, we also plotted and analyzed mean asset values of acquired and control hospitals for 10 years before and 5 years after acquisition.
Analyses were performed with R version 4.2.3 (R Foundation for Statistical Computing), with 2-tailed P < .05 statistically significant.
Results
Of 197 acquired hospitals, 156 met inclusion criteria and were matched with 1560 controls. Hospitals were acquired between 2010 and 2019, most commonly in 2018 (n = 67) and 2017 (n = 38). Acquired hospitals were mostly in the South (n = 87, 55.8%) and had 50 to 149 beds (n = 70, 44.9%). Compared with controls, fewer acquired hospitals were critical access (7.7% of acquired, 16.1% of controls; P = .008) or rural (0.6% of acquired, 4.2% of controls; P = .05). There were no significant differences in teaching hospital status (21.8% of acquired, 20.1% of controls; P = .68) or Medicaid share of discharges (11% for acquired, 10.3% for controls; P = .39).
Assets during acquisition year were $91 316 399 for acquired hospitals and $96 450 259 for controls (P = .49) (Figure). In the 2 years after acquisition, assets decreased by a mean (SD) of 15.0% (78.2%) for acquired hospitals and increased by 9.2% (36.3%) for controls, a 24.2% difference (95% CI, 11.6%-36.8%; P < .001), or from a mean (SD) of $87 705 471 ($76 823 078) in the 2 years before acquisition to $69 615 470 ($72 146 133) after (compared with $95 086 116 [$91 236 379] before and $102 495 629 [$100 050 927] after for controls, a difference-in-differences estimate of −$27 954 180; 95% CI, −$21 992 157 to −$33 916 202) (Table). After 2 years, 61.0% of acquired hospitals had reduced capital assets vs 15.5% of controls. Visual inspection of the graph of trends in assets in the 10 years before the acquisition year supported the parallel trends assumption, and data for 5 years after acquisition showed trends persisted and widened.
Discussion
After private equity acquisition, hospital assets decreased by 24% relative to that of controls during 2 years.
Study limitations included the small sample of acquired hospitals despite use of multiple sources to identify acquisitions. Private equity acquisitions are not required to be publicly reported, hampering identification. Closed hospitals were not included, possibly causing underestimation of capital losses. Many private equity acquisitions occurred recently and the drop-off in data at 4 and 5 years after acquisition precluded analysis of longer-term trends. Medicare cost reports may include errors, although they are audited. Many hospitals did not report components of total capital assets.
Private equity acquisitions appear to have depleted, rather than augmented, hospital assets. Although funds from asset drawdowns might be redeployed to enhance care or efficiency, previous studies suggest such effects may not occur.4 Financial outcome of private equity hospital acquisitions and effects on patient care require further study.
