Bundled Payments For Medical Conditions Show No Savings

Bundled care, actually bundled payments refers to packaging hospitalization and the following 90 days of care into a single fee. CMS has successfully deployed these programs for knee and hip joint replacements generating significant savings which they share with the hospitals that have volunteered for the programs. A new study in the New England Journal of Medicine looks at bundled payments for medical conditions raising questions about how much we can generalize the results from surgical conditions.

The Study

The researchers obtained cost data on bundled payments to hospitals for five common medical conditions, heart failure, pneumonia, chronic obstructive pulmonary disease, sepsis and acute myocardial infarction. They utilized data for six month periods before and after hospitals began their bundled care participation (it is a voluntary demonstration project of CMS) comparing their costs to control hospitals matched for both hospital and patient characteristics. They calculated the difference in costs between the two periods for both sets of hospitals and the differences in the differences. The hospitals using bundled care demonstrated a cost savings of $286 or 1.1%, the control hospitals saved $398 or 1.6 % neither of which was statistically or financially significant. Readmissions and deaths also did not differ between bundled and control hospitals. But these savings were not close to the 10% or greater associated with bundled surgical care. 

So why the difference between surgical and medical care? To see the outline of the problem we need to consider how surgical bundling saves money.

Surgical vs. Medical Hospitalization

The researchers reference three widely cited articles about bundled payments for joint replacements. These programs demonstrated saving during hospitalization from two sources. First, the knee or hip implant cost was reduced – when implanting surgeons stand to make or lose money based on the choice of implant, they begin to pressure device manufacturers for better rates. Second, and this is probably more important, total joint replacement is an elective procedure, and the patients are stable and optimized before admission, and they all undergo surgery on the day of admission. This allows more standardized care so that everyone is getting out of bed the next day; everyone is getting the same physical therapy. If I can characterize patients as “inputs,” they are pretty uniform, amenable to assembly line care - as a result, surgeons reduced the length of stay in the hospital, often by half.

People do not recover from joint surgery in the three days that they are hospitalized, they require post discharge care for physical therapy, adjustment of medications, etc. This post-discharge care was another source of savings for the surgical bundles because they drastically reduced their use of two very costly post-discharge “destinations," in-patient rehabilitation and skilled nursing facilities  Instead they utilized more home health services achieving the same outcomes at less cost.

Medical patients hospitalized for these five conditions are not standardized like those undergoing joint replacement; entering the hospital with varying severity of illness. When the input varies this much assembly line “care pathways” don’t work as well. As a result, there were no decreases in length of stay. In fact, a careful look at the data suggests that heart failure and sepsis, two diseases that might require 24-48 hours to fully understand their underlying cause and adopt focused therapy showed the least cost savings. Pneumonia and myocardial infarction where cause and treatment can be readily identified within a few hours of admission showed greater savings. (The study is silent on this distinction and there are no statistical or financial data beyond a bar graph) 

Like their surgical counterparts, medical patients often require post discharge care. But again, the hospitals continued to use more expensive in-patient services; presumably because that is what was needed to properly care for their patients. As the authors point out in their discussion

…another possibility for the failure of BPCI [Bundled Payments for Care Improvement] hospitals to reduce allowed payments is a lack of ability to influence skilled nursing facilities, inpatient rehabilitation facilities, long-term care hospitals or home health agencies.

The surgical solution was to direct the care through home health agencies which might not be controlled but were certainly less expensive. With the continuing consolidation of healthcare, more health systems are able to integrate hospitalization and post-discharge care so this may be less of an issue moving forward.

Bundled payments seems to reduce surgical costs, but because of the nature of medical versus surgical disease, similar cost savings do not result from bundling medical care. In the same way, physicians personalize care for their patients and their problems; CMS will have to develop savings programs more tailored to the necessary care of medical diseases – one size does not fit all. Assembly line care may work well for surgical problems, where input and treatment can be standardized; but for many medical problems, this is just not possible.  

Source: Evaluation of Medicare’s Bundled Payments Initiative for Medical Conditions NEJM DOI: