David F. Torchiana, MD, chairman and CEO of the Massachusetts General Physicians Organization, proposes a new payment system that bundles hospital readmissions and discharge care into a single payment:
One of the more effective health policy interventions in recent history was the introduction by Medicare, in the 1980s, of payment by diagnosis-related groups, or DRGs. Before DRGs, hospitals were paid on a cost-plus basis, essentially a blank check to do more. Under a DRG system, hospitals have an incentive to shorten the number of days patients stay in the hospital because the payment amount is fixed for each diagnosis, encouraging hospitals to manage costs and length of stay. Over the next decade, hospital days in the U.S. fell by nearly half; they remain among the lowest per capita in the western world.
In recent years, it has become increasingly apparent that many hospital patients end up being readmitted, which limits the cost savings that DRG payment might otherwise generate.
One study, conducted by Stephen Jencks, MD and colleagues using Medicare data from 2003-2004, showed that 19.6% of Medicare patients returned to the hospital within 30 days of leaving. In fairness, about ten percent of these returns were planned – for chemotherapy or insertion of a stent, for example – but the rest were readmitted for unplanned reasons that may have been preventable with the right intervention after discharge. Half of the readmitted patients had no evidence of an outpatient physician visit between the time of discharge and readmission.
Studying readmissions in detail leads to some interesting observations, some of which may seem counterintuitive. First of all more patients are readmitted after long hospital stays than after short hospital stays. You also are more likely to be readmitted if you are discharged to a rehab hospital or nursing home than you are if you go from the hospital to home. That’s because long stay patients and patients who go to a post-acute facility after a hospital stay are more seriously ill and more likely to be readmitted.
Readmission is not especially easy to keep track of either – in our fragmented system up to half of the patients that are readmitted after hospitalizations for conditions like cardiac surgery are readmitted to a new hospital rather than the one that did the surgery. There also is a lot of variation in readmission rates. Recently, CMS has taken this on as a quality measure and reports readmission rates for some diagnoses, like heart failure or acute myocardial infarction, on their Hospital Quality Compare web site.
At Partners, we have been studying our readmissions and have found that most of the ones that happen within thirty days actually occur early, in the first week or two.
More than half of our readmission volume is Medicare patients even though they are a smaller share of admissions overall because older patients tend to be sicker and have fewer social supports than younger patients.
Clinicians across the country are testing different ideas for addressing this problem, from better discharge planning to supportive palliative care, improved coordination with home health services and scheduling follow-up MD visits before the patient leaves the hospital. But we have little concrete evidence to guide us toward interventions that work. Recently, an article in the New England Journal of Medicine cast doubt on the value of rigorous discharge planning as a solution, showing little correlation between the extent of discharge planning and actual readmission rates.
Other studies have achieved good results with better care coordination and early scheduled follow-up. Partners’ has joined a Commonwealth Fund/IHI effort called STARR – State Action on Avoidable Hospitalizations – which has the goal of reducing 30-day re-hospitalizations by 30 percent. The focus is on the transition out of the hospital and coordination between the acute hospital and any follow-on care as an outpatient.
A new question is whether a DRG-like payment that includes hospital costs and the post-hospital discharge interval in the fixed payment – commonly referred to as a bundled payment – could create the appropriate financial incentive to address this issue more effectively. The short answer is that it might and it may not take long to find out if it does given upcoming changes being considered in both state and federal payment policies.
Today the payment system doesn’t cover the cost of any extra effort soon after discharge and reduced readmissions actually reduce hospital revenue – not a great formula for action. A payment system that bundles readmissions and post-discharge care into a single payment makes sense and could both save money and improve care. It’s likely to be complicated to implement but it is worth trying and deserves our support.
This program aired on January 15, 2010. The audio for this program is not available.