“Mr. Obama, Tear Down These (Hospital) Walls”: Readmissions in Context

By  |  April 3, 2009 |  10 

I like readmissions.

Well, that didn’t come out quite right, did it?

What I mean is that I like focusing on readmissions as a potentially actionable quality measure. I believe that it’s possible to prevent many readmissions, thereby improving quality and lowering costs. And compared to mortality (the other hot outcome measure), the need for case-mix adjustment is a bit less critical, and there is no such thing as “a good readmission.”

I also like DRGs. Paying hospitals a fixed fee for a given diagnosis has created the only corner of sustainable capitation in our healthcare system, one that is otherwise awash in inappropriate expenditures driven by the dominant fee-for-service payment structure.

But the DRG system created a big black hole, and it is time to fill it. It’s called the post-discharge period. And one large part of the detritus emerging from that hole is readmissions.

You probably saw this week’s NEJM study by Stephen Jencks (a former Medicare official and now a Baltimore-based consultant), and my pals Mark Williams and Eric Coleman, of Northwestern and Colorado, respectively. The study found that 20% of Medicare patients are readmitted within a month of discharge, and one-third return within 90 days. Even more remarkably, by a year out more than half of patients (56%) discharged from an acute care hospital are re-hospitalized. The authors estimate that the cost of preventable readmissions was $17 billion in 2004 (the study year), which would make it more like $25 billion today.

Although the frequency and cost of readmissions were reasonably well known before this study, the paper was chock-full of interesting findings and observations:

  • Like so many things in healthcare, there was striking geographic variation in readmission rates – from a low of 13% in Idaho to 23% in Washington, D.C.
  • There were also variations by DRG, with the highest readmission rates in patients with heart failure, psychosis, vascular and cardiac surgery, and COPD – pointing the way toward targeted interventions.
  • More than half the patients readmitted within 30 days appeared not to have had an outpatient visit between hospital discharge and readmission, perhaps another target for intervention.
  • Most (70%) surgical patients who are readmitted come back for a medical diagnosis such as pneumonia or UTI.
  • Approximately 30% of readmitted patients come back to a different hospital, so hospitals will underestimate the extent of their readmission problem by looking solely at their own bounce-backs.

Now that we know that readmissions are frequent and costly, the next questions are: Are they preventable? Do they reflect poor quality care? And, if yes and yes, what clinical and policy maneuvers might help prevent them?

The Obama budget plan depends on figuring this out. The budget, which aims to save $300 billion (which used to seem like a lot of money) in Medicare/Medicaid costs over the next decade, includes a projected $26 billion in savings from “driving down hospital readmission rates for Medicare patients” – both from preventing the costs of the readmissions themselves and by lowering payments to hospitals with excessive readmission rates. 

The policy issues are worth thinking through a bit. Until this decade, DRGs gave hospitals an incentive to shorten length of stay and cut hospital costs for their Medicare patients, but provided no market signal to improve quality or safety. The incentive to shorten length of stay remains, but it is now accompanied by a variety of initiatives (error and quality reporting, pay for performance, Joint Commission visits) that create at least some inducement for hospitals to maximize value (quality divided by cost), and not just efficiency.

Up till now, this broader value incentive was focused, laser-like, on care delivered within the four walls of the hospital. Hospitals could do quite well managing length of stay and costs and (more recently) scoring well on their publicly reported quality measures and accreditation surveys. The closest the present system came to putting any skin in the post-discharge game was public reporting of the presence or absence of documentation of a discharge plan for adults with heart failure or kids with asthma. Not a very high performance bar.

The manifestations of this myopic focus on hospitalization as the unit of analysis can be seen in the paucity of attention that hospitals give post-discharge care. Studies have chronicled a litany of post-discharge disasters (summarized in this marvelous Primer on post-discharge adverse events by my colleague Sumant Ranji in our federal patient safety portal, AHRQ PSNet): nearly 20% of patients experience adverse events within 3 weeks of discharge; 40% of patients are discharged with tests pending, and many of these balls are dropped; 15% of discharged patients have a discrepancy in their medication lists; and only the rarest discharge summary finds its way to the desk of the primary care physician by the time a patient is seem for his or her first post-discharge visit.

In other words, when it comes to post-discharge care, we suck.

Despite powerful literature that shows that simple interventions – like post-discharge phone calls or the use of a transitions coach – can lead to impressive improvements in post-discharge care and decreased readmission and return-to-ED rates, few hospitals have put these interventions in place. Outside of integrated delivery systems like Kaiser Permanente or the VA, virtually no hospitals have electronically connected themselves to their referring physicians’ offices; everybody argues that Stark laws prevent them from making these hook-ups, but the lack of an incentive to improve post-discharge care has been the more important culprit.

Even hospitalist programs, ostensibly erected to improve value, tend to have a narrow focus on improving care within the building, too often taking an out-of-sight/out-of-mind attitude to their patients’ status after discharge. That said, I’m pleased that hospitalists have taken the national lead in studying post-discharge errors and building programs to prevent them.

As with most things in safety and quality, I’m not arguing that physicians, or hospital administrators for that matter, stop caring about their patients after they’re discharged. Of course they do. But this is 2009 – remember, we’ve discovered that improving quality and safety are largely systems properties, not driven by individual commitment or even individual skill. The question is not whether a hospitalist is skilled and empathic, or whether the patient was discharged at the correct, evidence-based time. It is this: are there robust and sustainable systems in place designed to maximize the safety of patients in the vulnerable post-discharge period? The sad answer is that there are very few such systems.

Accompanying the Jencks study, Harvard’s Arnie Epstein reviews the policy initiatives addressing readmissions – including those that are here today (publishing readmission rates on the Web) and those being actively discussed (financial penalties to hospitals with high readmission rates). But the Cool Kid on the Payment Block is “bundling” – aggregating  payments for doctors and hospitals for a period of time after an illness (an “episode of care”) in an effort to create accountable integrated entities that will improve care across the continuum (the entities somehow have to split up the spoils between hospitals, hospitalists, SNFs, primary care docs, specialists, care coordinators… Have fun with that). Epstein’s verdict: worthy of pilot studies, but “the likelihood that [bundling] will prove to be a successful model is still uncertain.”

Hospitals, of course, moan about all of this new pressure on readmissions – claiming, correctly, that they don’t control much of what happens when a patient leaves the building. “How can you blame us,” goes the lament, “if we can’t find a PCP for a patient, or the outpatient doc chooses to readmit the patient.” Some of this is doubtless true, and the potential for unfairness is real. But some of this bellyaching is a manifestation of learned helplessness, borne of having no incentive to pay any attention whatsoever to filling the post-discharge black hole.

An era is dawning in which hospitals will, for the first time, have to think of the post-discharge period as being, at least partly, their responsibility. Luckily, this is an area in which there are tools ready for the taking (for example, those developed by the Society of Hospital Medicine through its splendid Project Boost), and some early experience to learn from. Some of us, suspecting that this train was coming down the tracks, have been working on the discharge process for the last few years (my UCSF hospitalists have focused on this issue as our main quality initiative for the past year). Others will have to play catch-up ball.

Ultimately, hospitals will have to figure out ways to get a discharge summary in the hands of a PCP by the day after discharge (as opposed to the year after discharge, today’s sad state of affairs); to ensure that patients receive robust and understandable discharge instructions (not simply a check box on a form); and to provide, or facilitate the provision of, a follow-up phone call (or email or Tweet – whatever works!) and, for high risk patients, a post-discharge clinic visit, a discharge or transitions coach (as promoted by Eric Coleman’s “Care Transitions Program”), and/or a high risk case manager. This isn’t rocket science – all of these interventions make sense, are less expensive than an MRI or surgical robot, and are not that hard to implement. They simply take institutional will.

I, like you, don’t know where the money will come from for all of this. But we do know that readmissions are terribly expensive and just plain bad for patients. With unplanned readmission rates at 20% and higher, it is high time that we got to work on this problem. When it becomes less expensive to prevent a readmission than to neglect the post-discharge period and help contribute to one, someone will find the money to improve care.

They always do.


  1. Brad F April 4, 2009 at 12:56 am - Reply

    Hey Bob
    I was interested in learning more about the 20-40% of patients readmitted to a a neighboring (or distant) hospital. Perhaps Eric or Mark could comment, ie, anything unique about this cohort–either from an SES (“they did not get me home care”), geographic (big city, three hospitals clustered in one zip code), or quality perspective (“I disliked the facility”), etc. Not that they can drill down to this level with the data set, but they do a cursory looksie to see if there was something difft about this group. There are interesting associations lurking in all probability, and I have wanted to know for some time what causes these service interruptions and why such a big chunk of folks go elsewhere for care.

  2. Annie April 4, 2009 at 2:26 am - Reply

    What was the % who were originally d/c home vs. d/c to SNF, rehab or assisted living?

    It seems to me to be worth exploring extending hospital d/c to include an initial comprehensive home assessment by a qualified community/home health nurse (BSN educ or higher, case management and disease management expertise). The assessment would include care coordination, patient and family understanding of d/c instructions and self-care strategies with evidence of successful compliance. It would also include verification of f/u appointments with PCP and relevant specialists along with transportation for same.

    Again, looking at this from an exclusive medicine point of view misses the bigger picture: the patient’s ability to adapt, recover and progress in an unsupervised setting which requires mastery of post d/c instructions (such as how to maintain hydration, respiratory toilet, and complete ATB courses – to address the infection stat above), medications, care coordination, transportation, nutrition and a host of other requirements of which the physician most likely has little knowledge and interest (this is rightly the purview of professional nursing – no slight intended toward physicians).

    It might be useful to collaborate with the ANA on developing standards of nursing care and practice around post hospital d/c x 30 days which address patients’ transition form hospital level dependency to sufficient competence and independence in managing their care in an unsupervised setting.

  3. DZA April 4, 2009 at 12:49 pm - Reply

    in my humble (30 year) experience, pts are re-admitted largely because they have chronic progressive organ failure(s) that predictably decompensate. you could say that hospitalists have become the primary care physicians for the chronically ill. on any given day, i personally know about a third of the pts on our 150 pt service (i have been a hospitalist for a large teaching hospital for almost 9 years). i know many of them much better than their PCP (as you mention, many discharges have not even made it to f/u appts before the bounce). i believe only a fraction of these bounces are preventable with systems improvements. what is needed is overt rationing such that (for instance) progressive multiorgan failure=palliative care. as opposed to intensive care, a few weeks of chi-chi for the survivors, then death by complications on the Trach/PEG respiratory care ward a month later….

  4. ron kirshner April 4, 2009 at 2:43 pm - Reply

    Readmissions are obviously a complex problem but if what we are talking about is quality then maybe there is another aspect to this. If there is a 30% readmission rate then that means that 70% of our patients did ok with discharge. If the 70% group stayed longer because we were afraid of a readmission what would their fate have been?
    So back to the 30% readmission group. When they get readmitted with pneumonia or whatever do we really believe that if they stayed in the hospital this would not have happened anyway.
    Look, I am not saying that we can’t do better but I just don’t think a readmission is necessarily a bad thing or a failure .Hospitals are not great places to be in. We discharge patients as a trial to see how they do. Patients have easy access to providers. if they don’t do well we readmit. With this we have a 70% success rate.

    The suggestion that more visits or contact with the patient after discharge will bring down readmission I think is an illusion or at least has yet to be proven.
    If this is just a money issue then let’s fix that .I just don’t know that keeping patients longer (and this will certainly be the outcome) will be any better from a quality point of view than having planned or easy access readmits

  5. geriatricdoc April 8, 2009 at 10:03 pm - Reply

    I think that a problem encountered by the Japanese discharge planners would pique the readers interest.
    “Influence of superstition on the date of hospital discharge and medical cost in Japan: retrospective and descriptive study”

    Objectives: To determine the influence of superstition about Taian (a lucky day)-Butsumetsu (an unlucky day) on decision to leave hospital. To estimate the costs of the effect of this superstition

    Key messages
    • Belief in Taian-Butsumetsu, a superstition relating to the six day lunar calendar, is common among Japanese people
    • This study showed that the mean number of patients discharged on Taian (a lucky day) is the highest and that on Butsumetsu (an unlucky day) is the lowest
    • Patients discharged on Taian were older, were more likely to be female, and had longer hospital stays than those discharged on other days
    • The findings suggest that patients were extending their stay to leave hospital on Taian
    • This superstitious belief increased the cost of medical care in Japan

    BMJ 1998;317:1680-1683 ( 19 December )

  6. DZA April 8, 2009 at 11:42 pm - Reply

    geridoc-cool find. here’s an american version that is admittedly my availability heuristic…

    discharges cluster around the end of the work week, most particularly friday. this seems to coincide with the weekly rounder going off service (reduces the number of transfer of care notes required when going off weekday service, the number of pts left on service to be picked up by the new guy (frequently a moonlighter), maximizes 99239 billing going off service, and, in a statistical/stochastic kind of way, to the avg LOS of 4.x days bookended by the peak of admissions on sun/mon). and everybody knows discharges on sat thru monday am are a fraction of a typical weekday. which itself is due to a combination of reduced case management activity on weekends, rocks (some patients feel “safer” in the hospital for a “few more days” and weekends offer the perfect opportunity), and moonlighters “holding the fort” and failing to move pts thru the process. so friday is our Taian and the whole weekend (sat/sun) is our Butsumetsu!

  7. Happy Hospitalist April 9, 2009 at 4:59 pm - Reply

    Two Comments.

    1) I am surprised that the number of patients readmitted isn’t higher. Why? Because many patients I discharge should be hospice, but aren’t. They should be Do Not Transfer To Hospital status, but aren’t. The 50% readmission rate at one year, I suspect would be higher, if it didn’t exclude all the patients that died in that time frame. That’s what old sick people do. They die a natural death. A death we delay for a few months by admitting them to the hospital, again and again, just to make ourselves feel good about making the patient believe we are doing something for them.
    2) Many admissions are legal driven. Too afraid to send them home for a trial of outpatient care because of a fear, real or not, that patients may fail outpatient care and that 1/100,000 chance that they die.

  8. Pamela Wilson April 19, 2009 at 9:09 pm - Reply

    I came across your website as I was doing research. After reading the recent WSJ article “Why Quality Care is Dangerous” (4/8/09) and an AJN article on Diabetes Care Recommendations (Jan 09) I feel that quality care is injurious and will only penalize institituions and physicians. After all who wants to be penalized for taking action on a project that is likely to have poor results i.e. patients with serious medical issues. Why treat them at all if a penalty is involved? The Obama administration fails to see that the solution to the health care crises is PREVENTION. Hospitals cannot prevent readmission. They do not control the actions of a patient once they walk out the door. Nor will offering universal health care make people take better care of themselves. Prevention needs to begin in the schools and annual physician visits and education shoudl be a requirement of the privilege of having medical insurance. Human nature is human nature and I see so many clients I work with who could have had different outcomes with a little education earlier on in life. Pamela

  9. Stephen Soled May 12, 2009 at 5:56 pm - Reply

    Minimizing Avoidable Hospital Readmissions
    Virtually no one opposes hospital readmissions. The area of concern is avoidable hospital readmissions. According to the CMS, avoidable hospital readmissions are unnecessarily costing Medicare $12 billion/year.

    And there is a practical and inexpensive solution to this problem. It involves providing family caregivers with a means to check on their loved ones and ensure that they are doing the things to ensure their well-being. The solution utilizes a new device called the Wellness Wizard, which is a $149 electronic caregiver’s assistant. If anyone wants more information, a white paper is available. For a copy email me at [email protected]

  10. Bob Wachter July 10, 2009 at 2:05 pm - Reply

    OK, the fun begins. As of this week, CMS has begun posting 30-day readmission rates for pneumonia, CHF, and MI on its Hospital Compare website. They are treating this outcome like they do mortality outcomes — providing a fairly wide statistical berth, so the performance for the vast majority of hospitals will be listed as “no different from national rate.”

    This, of course, is just the beginning, the scaffolding for efforts to cut payments for hospitals with “excess” readmissions. Of all the cost-reduction maneuvers on the health reform table (taxing benefits, cutting disproportionate share payments…), this one seems like the most likely to stay in the final mix.

Leave A Comment

About the Author: Bob Wachter

Robert M. Wachter, MD is Professor and Interim Chairman of the Department of Medicine at the University of California, San Francisco, where he holds the Lynne and Marc Benioff Endowed Chair in Hospital Medicine. He is also Chief of the Division of Hospital Medicine. He has published 250 articles and 6 books in the fields of quality, safety, and health policy. He coined the term hospitalist” in a 1996 New England Journal of Medicine article and is past-president of the Society of Hospital Medicine. He is generally considered the academic leader of the hospitalist movement, the fastest growing specialty in the history of modern medicine. He is also a national leader in the fields of patient safety and healthcare quality. He is editor of AHRQ WebM&M, a case-based patient safety journal on the Web, and AHRQ Patient Safety Network, the leading federal patient safety portal. Together, the sites receive nearly one million unique visits each year. He received one of the 2004 John M. Eisenberg Awards, the nation’s top honor in patient safety and quality. He has been selected as one of the 50 most influential physician-executives in the U.S. by Modern Healthcare magazine for the past eight years, the only academic physician to achieve this distinction; in 2015 he was #1 on the list. He is a former chair of the American Board of Internal Medicine, and has served on the healthcare advisory boards of several companies, including Google. His 2015 book, The Digital Doctor: Hope, Hype, and Harm at the Dawn of Medicine’s Computer Age, was a New York Times science bestseller.


Related Posts

By  | June 26, 2018 |  2
JAMA just published the largest trial I have seen on a Hospital at Home (HAH) model to date and the first one out in the last few years. It comes from Mount Sinai in NYC–who have led the pack in this style of care if national presentations are the judge. They launched the program three […]
By  | April 29, 2013 |  15
Everybody hates curbside consults – the informal, “Hey, Joe, how would you treat asymptomatic pyuria in my 80-year-old nursing home patient?”-type questions that dominate those Doctor’s Lounge conversations that aren’t about sports, Wall Street, or ObamaCare. Consultants hate being asked clinical questions out of context; they know that they may give incorrect advice if the […]
By  | March 19, 2012 |  15
These days, I’d never consider trying a new restaurant or hotel without reading the on-line ratings on TripAdvisor or Yelp. I seldom even bother with professional restaurant or travel critics. Until recently, there was little patient-generated information about doctors, practices or hospitals to help inform patient decisions. But that is rapidly changing, and the results […]