Association of Hospital System Affiliation with Spending and Postoperative Outcomes: A Longitudinal Study of Hospital Mergers And Acquisitions From 2010 To 2018.

OBJECTIVE:To determine whether hospital system affiliation was associated with changes in surgical episode spending or postoperative outcomes. BACKGROUND:Over 70% of US hospitals are now part of a hospital system. The presumed benefits of hospital consolidation include concentrating volume and expertise, care integration, and investment in quality improvement. However, there is conflicting evidence as to whether expanding hospital systems are actually reducing health spending or improving quality. These observations call into question whether systems are leveraging their collective volume and experience to standardize care and maximize efficiencies. METHODS:The American Hospital Association Annual Survey was used to identify whether a hospital was part of a system and in which year a hospital joined the respective system. Using 100% Medicare claims data, we identified fee-for-service Medicare patients undergoing elective inpatient coronary artery bypass graft colon resection, lung resection, hip replacement, or knee replacement from 2010 to 2018. We used a difference-in-differences framework to evaluate hospital spending and outcomes before and after joining a system. The primary outcome was Medicare 30-day episode spending, with specific attention to the total episode payment, index hospitalization, and post-acute care components. Secondary outcomes included serious complications, 30-day mortality, and 30-day readmission. RESULTS:The cohort included 3,395,565 Medicare beneficiaries who underwent surgery between 2010 and 2018. Patients were treated at 3961 hospitals, of which 1097 (27.7%) were never in a system, 2262 (57.1%) were always in a system, and 602 (15.2%) joined a system during the study period. By 1 year after system affiliation, 30-day episode spending had decreased by $303 (95% CI: 63, 454, P=0.01), and after 5 years, 30-day episode spending decreased by $429 (95% CI: 5, 853, P=0.04). One year after system association, index hospitalization spending was not statistically different from before system affiliation ($-30, 95% CI: -160, 100, P=0.65). Conversely, 1 year after system association, postacute care spending decreased by $268 (95% CI: 107, 429, P<0.01) and remained lower for ≥5 years. There was no significant change in hospitals serious complications (-0.14, 95% CI: -0.40, 0.11, P=0.27), 30-day readmission (-0.14, 95% CI:-0.52, 0.25, P=0.48), or 30-day mortality (-0.08, 95% CI: -0.18, 0.03, P=0.17), 1 year after joining a system; similar patterns were observed at three and ≥5 years. CONCLUSIONS:system affiliation was associated with a small decrease in 30-day episode spending, driven by decreased spending in postacute care services. Notably, there was no difference in postoperative outcomes after system affiliation.