By Stephen F. Jencks, M.D., M.P.H.
From the Editors: The recent publication by the Centers for Medicare and Medicaid Services of hospital readmission rates creates a unique opportunity to further understand health care system performance. It comes at a time when health care organizations are engaged in efforts to redesign care to provide the best and most efficient services. The data on readmissions should thus be a key element in strategic planning by hospital leaders and care teams. Stephen F. Jencks, M.D., M.P.H., an expert in this area, explains his views on what the data mean--and how to interpret them for effective quality improvement.
If you are a hospital Chief Medical Officer or Chief Quality Officer, the CMS data on rehospitalization may raise more questions than they answer and leave real uncertainty as to what your next step should be. Some pointers may be helpful:
1. The CMS information uses a sophisticated model to correct differences in severity of illness between patients in different hospitals and hence to judge whether the hospital’s performance is better or worse than would be expected for patients with similar characteristics. This model is quite conservative in that it finds a very large fraction of hospitals to have rates that are neither better nor worse than expected. Small hospitals are particularly unlikely to be identified as better or worse. Since heart attack, heart failure, and pneumonia rehospitalization rates are fairly strongly correlated, looking at them together may be useful. For example, if all three rates are appreciably higher than expected, something is probably going on, even if no individual rate is significantly higher than expected.
2. It is also important for hospital leadership to keep in mind that the average rehospitalization rate, with which a hospital’s performance is compared, is generally not very good, and being average should not be a source of complacency, much less pride. It may be helpful to look at hospitals that are significantly above average to see what they are achieving; their performance level can be considered a plausible target and perhaps a benchmark.
3. The CMS rates reflect data averaged over time and take some time after the data set closes to get published. The result is that tracking improvement through this publication will be very difficult and lead to an improvement cycle so long as to be unhelpful.
4. Replicating the CMS results is essentially impossible for an individual hospital because it cannot know how many of its patients are rehospitalized elsewhere and it does not have the ambulatory claims data that are needed for the CMS risk-adjustment model.
5. The data reported here are limited to the outcome of rehospitalization within 30 days. They provide no information on processes of care and their adequacy. For example, it is impossible to ascertain whether the patient/family went home understanding what medications to take and able to obtain them, understanding what danger signs require action and who should be called, or having an appointment for follow-up and the means to get to it.
6. What a hospital can do is to examine the records of patients that it has discharged and who were later readmitted to the same hospital. These patients present an extremely rich source of information because hospital staff can ask the patient and family what led to the rehospitalization and take immediate steps to remedy any failures in the discharge process and in post-hospital care. These interviews are extremely important and valuable because they often reveal that education and arrangements documented in the record were ineffective from the patient and family perspective.
7. Above all, these data should be seen as one element in a very complex puzzle that invites a hospital to examine aspects of performance that it may never have known about but that are vital to its patients.
The mission of The Commonwealth Fund is to promote a high performance health care system. The Fund carries out this mandate by supporting independent research on health care issues and making grants to improve health care practice and policy. Support for this research was provided by The Commonwealth Fund. The views presented here are those of the authors and not necessarily those of The Commonwealth Fund or its directors, officers, or staff.