The Issue

To help keep costs down, improve quality of care, and increase health care providers’ accountability, Medicare has been experimenting with something called bundled payments. In one model hospitals are responsible for the patient’s entire “episode of care” — the care bundle — instead of Medicare paying piecemeal for each individual service or office visit. An episode may be defined as a surgery, for example, or treatment of a chronic medical condition. If the cost of care is lower than an historical benchmark, hospitals share in the savings; if the cost is higher, hospitals are responsible for a portion of the overage.

With support from the Commonwealth Fund, researchers looked at how participating hospitals performed in Medicare’s Bundled Payments for Care Improvement (BPCI) initiative for five medical conditions (congestive heart failure, pneumonia, chronic obstructive pulmonary disease, sepsis, and acute myocardial infarction) and compared them with matched control hospitals. The study covered the period 2013 to 2015.

$23,993 average Medicare payment per episode of care across the five conditions at BPCI hospitals, a drop from $24,280 at baseline. Control hospitals saw a similar decrease.

What the Study Found

No significant difference was found between the hospitals participating in Medicare’s bundled payment program and the control hospitals in terms of what Medicare paid for the care provided, patients’ clinical complexity, average length of hospital stay, emergency department use, hospital readmission, or mortality.

The Big Picture

A study of an earlier Medicare bundled care initiative for joint replacement showed that participating hospitals succeeded in lowering overall Medicare payments. Why were the results so different this time? Bundles for joint replacement differ from medical conditions in a few ways. For one, patients tended to be younger in the joint replacement study, with 30 percent of patients older than age 80 versus 55 percent in the study of medical conditions. For another, joint replacement patients have lower rates of poverty and disability: 13 percent of Medicare beneficiaries enrolled in the joint replacement study also had Medicaid, while 11 percent were disabled, compared to 25 percent and 21 percent, respectively, in the new study. These complexities may have rendered the BPCI intervention less effective.

In addition, BPCI hospitals may not have had sufficient ability to incentivize and influence care provided by the skilled-nursing and rehabilitation facilities where patients received postacute care. The authors note, however, that partnerships between hospitals and postacute care providers are becoming more common.

The Bottom Line

Hospitals participating in Medicare’s most recent bundled payment initiative did not have lower costs or other better outcomes compared with hospitals not participating. Bundling for medical conditions may require more time, new care strategies and partnerships, or additional incentives.