Since October 2012, the Medicare program has penalized hospitals when too many patients in traditional Medicare are re-hospitalized within a month of discharge. This policy appears to be having unintended consequences for patients in Medicare and in the commercial market.

Hospitals with readmission rates above the national average now receive lower payments across-the-board from Medicare; the higher the rate of “excess” readmissions, the greater the penalty. The fines are intended to push hospitals to provide better care for their patients both in and out of the hospital, to avoid having patients return.

However, when patients do return to the hospital, the policy may unintentionally encourage hospitals to keep them “under observation” to circumvent the penalties. Although observation patients often receive care in a hospital’s inpatient unit, patients “under observation” are considered to have outpatient status under Medicare’s rules; they are not formally admitted to the hospital. It is a subtle difference, but one that may have significant financial consequences for some patients.

For example, as we described in another paper, treatment under observation could mean denial of coverage for necessary follow-up care in a skilled nursing facility, and huge medical bills. Why? Because, under other complex Medicare rules, beneficiaries must have spent at least three inpatient hospital nights, before the program will cover care provided in a skilled nursing facility; observation status doesn’t count toward those three nights.

Readmission Trends

Hospital readmissions have been declining for several years and this has intensified since Medicare began fining hospitals. The Centers for Medicare and Medicaid Services (CMS) has touted the recent decrease in readmission rates as evidence that hospitals are providing better quality care while better coordinating their efforts with other clinicians; and as proof that patients and their caregivers are leaving the hospital with clearer instructions on how to take their medications or where to get appropriate follow-up care. But, while physicians are readmitting fewer patients, they are also more likely to place their patients under observation. The rate of Medicare observation use almost doubled in the six years preceding implementation of Medicare’s readmission penalties and continued growing between 2012 and 2013.

Our findings suggest that these diverging trends may be related: In the new world of readmission penalties, some clinicians may be placing returning Medicare patients under observation rather than admitting them. Our independent analysis of Medicare data published by CMS revealed that the top 10 percent of hospitals with the largest drop (16 percent on average) in readmission rates between 2011 and 2012 also increased their use of observation status for Medicare patients returning within 30 days by an average 25 percent over the same time period (Figure 1).

Recent data indicate that Medicare’s readmission penalties may also be affecting some commercially insured patients. Patients not covered under Medicare do not impact a hospital’s readmission fine. But Medicare, as the country’s largest payer for health care services, often influences how hospitals care for all patients.

Using data from the OptumLabs™ Data Warehouse, which includes claims from a large private health insurance carrier, we found that hospitals that reduced readmissions within 30 days also increased their share of returning observation patients in private plans. The top third of hospitals with the largest six-year (2009-2014) reduction in 30 day readmissions (26 percent on average) increased their share of returning observation patients in private plans by an average of 45 percent (Figure 2). Much of that increase started in 2012, the same year that Medicare hospital readmission penalties began. To be sure, these findings apply to the health insurance carrier’s enrolled population and are not necessarily representative of the U.S. population. Nonetheless, the shifts are noteworthy.

Interestingly, similar trends in the use of observation stays were also seen in Veterans Hospitals where provider incentives and payment policy differ significantly from those of Medicare and private plans. A new article published by Brad Wright and colleagues in the October 2015 issue of Health Affairs shows that rates of observation stays in the Veteran’s Health Administration more than doubled between 2005 and 2013.

In addition to Medicare’s readmission penalties, there may well be other reasons for observation stays’ increasing popularity. For instance, as Wright and colleagues point out, veterans owe the hospital a significantly lower copay when they remain under observation ($50) than when they are admitted as inpatients ($1,184). The situation is similar for Medicare beneficiaries: nine out of 10 owed less for hospital services—not counting follow-up care—when they were “observed” than if they had been admitted. What’s more, some studies show that by placing certain patients under observation, physicians can save hospitals money while giving patients better and safer care. Finally, Medicare’s Recovery Audit program and similar initiatives by private-payers that carefully check the appropriateness of short inpatient stays, may be driving hospitals to observe patients more often.

Questions Remain

Our findings suggest that at least some hospitals are substituting observation status for inpatient readmissions, both for Medicare and privately insured patients. These trends raise a number of questions. For instance, do observation patients get the same quality of care as inpatients? Further, do drops in readmission rates truly mean that hospitals are providing better quality care? Or, as David Himmelstein and Steffie Woolhandler suggested in a recent Health Affairs Blog, is it merely that some hospitals are avoiding penalties by relabeling patients they previously would have readmitted as observation patients?

In fact, declining readmission rates may be a misleading measure of hospitals’ success in reducing medical complications or in coordinating patients’ care with other clinicians. By the same token, tying hospital payments to readmission rates may well be equivalent to allowing some hospitals to avoid financial penalties by simply relabeling patients rather than by improving patient care.

Figure 1


Figure 2