What does integration do to spending and quality?

Most of the time, vertical integration is expensive for patients and insurers. I published a review (Medical Care Research and Review 2018) evaluating this literature. Most studies concluded that integration increased prices and spending. At the same time, quality measures did not seem much affected. That is disappointing, because, in principle, hospitals and physicians working together should help to improve care coordination and reduce unwanted outcomes like hospital readmissions. There are, of course, exceptions where quality has improved, too.

But what about patient care?

My co-authors and I have found that integration spills over into all kinds of patient care issues:

• Integration can change not just the cost of treatment, but the treatment itself. Patients of integrated cardiologists were more likely to get hospital-based treatments like angioplasties (Health Affairs 2023).

• Hospital-integrated physicians could, in theory, better coordinate care for more clinically complex patients. Unfortunately, we did not find evidence that their patients were any more complex than those of independent physicians (forthcoming, American Journal of Managed Care). 

• Hospitals don’t appear able to or interested in acquiring the highest-quality physicians (Health Care Management Review 2024)

• Once integrated, physician productivity falls - they see fewer patients and provide fewer services (my job market paper - Health Services Research 2024)

• Patients of integrated cardiologists are more likely to get cardiac rehabilitation, which is a very valuable service that is widely underused (forthcoming, JAMA Network Open 2025). 

Any takeaways for public policy?

Yes.

• Re-imagine antitrust. Integration has been happening not just with outright practice acquisitions, but also with individual employment contracts between hospitals and physicians (Health Services Research 2022). Antitrust policy – traditionally oriented around acquisitions, not employment – would need re-invigorating to slow the pace of integration.

• Integration mediates performance under MIPS. Integrated physicians perform a little better on the Merit-based Incentive Payment System quality measures that favor bigger systems (e.g., quality that relates to EMR use), but integrated physicians do not perform better across-the-board on MIPS measures (Medical Care 2023).

• Upcoding is a concern. Integration raises the coded diagnostic severity of patients – that is, on paper, patients appear sicker than they were just before their doctor became integrated, likely because hospital diagnosis-coding services are more sophisticated (Health Economics 2022). Higher coded severity increases certain Medicare payments.

• Spending effects may differ by clinical context. There are some clinical settings - we studied stable angina, though there may be others - in which integration might be at least spending-neutral if not spending-reducing (Health Services Research 2025).

• Cardiac rehab is a bright spot for integration. There are cases in which integration offers benefits – for example, by increasing the chances of cardiac rehabilitation, which is a very valuable service that is widely underused (forthcoming, JAMA Network Open 2025). 

• Bundled payment could throw a curveball. Orthopedists in bundled payment (CMS Comprehensive Care for Joint Replacement Program) prescribed slightly fewer knee replacements. That effect disappeared, though, if the orthopedists were hospital-integrated (under review). All of this is important since Medicare's TEAMS model goes into effect in 2026.