OP-ED

Finally Some Good News on Readmission Rates

Why readmission penalties are controversial

Penalizing hospitals for high readmission rates has been pretty controversial.  Critics of the program have argued that readmissions have little to do with what happens while the patient is in the hospital and are driven primarily by how sick or how poor the patient is.  Advocates of the readmissions program increasingly acknowledge that while readmissions may not reflect the quality of care that occurred within the hospital, someone should be accountable for what happens to patients after discharge, and hospitals are the logical choice.  While the controversy continues, there is little doubt that the metric is here to stay.  This October, the CMS Hospital Readmissions Reduction Program (HRRP) will increase its penalty on excess readmissions from 1% to 2% of total hospital reimbursement.

So far, CMS has focused on readmissions that occur after patients are discharged with one of three medical conditions—acute myocardial infarction, pneumonia, and congestive heart failure.  The data on the impact of the program are mixed:  while readmission rates appear to be dropping, the penalties seem to be targeted towards hospitals that care for some of the sickest patients (academic medical centers), poorest patients (safety-net hospitals) and for heart failure, some of the best hospitals (those with the lowest mortality rates).  No wonder the program has been controversial.

Why surgery may be different

In 2015, CMS extends the program to focus on surgical conditions, which provides an opportunity to think again about what readmissions measure, and what it might take to reduce preventable ones.  And if you think about it, surgery may be different.  Most patients who are admitted for Acute MI, CHF, and pneumonia are chronically ill and bounce in and out of the hospital, with any one hospitalization likely just an exacerbation of underlying chronic illness (especially true for pneumonia and heart failure).  Not so for surgery—at least not for the major surgeries.


Yes, many of these patients have chronic illnesses, but when a patient is admitted for surgery, he or she is relatively stable (except, of course, during emergency surgeries).  A hip replacement is not just an exacerbation of osteoarthritis but is meant to cure or alter that disease.  Patients should not need to be hospitalized again—at least not right away, if everything goes well.  But this is an empirical question – are surgical readmissions picking up something different than medical ones?

The evidence

In the September 19, 2013 issue of the New England Journal of Medicine, we publish our findings on the relationship between hospital surgical quality and readmission rates.  The quality of surgical care is deeply important—the average American undergoes 9 surgical procedures over the course his or her lifetime. We focused on six major surgeries.  The ones we chose are common, costly, and associated with substantial morbidity and mortality.  They also include a couple procedures that Medicare is thinking of including in the readmissions reduction program.  What we found was quite striking–the best hospitals, those with the highest volume and the lowest mortality rates, had much lower readmission rates.  Unlike medical conditions (where the relationship is often heading in the wrong direction) the story on the surgical side seems pretty different.  See figures below.

Surgical Readmissions Figure 1

 

Surgical Readmissions Figure 2Why would this be?  It’s likely because when people are readmitted after surgery, they are coming back for surgical complications.  Good hospitals don’t just have lower mortality rates, but they also likely recognize their complications and manage them early, before the patient is discharged.

Some complications, of course, don’t show up until after the patient is discharged.  Here too, good hospitals may manage things differently than poorer quality ones.  At high quality surgical hospitals, it may be that every patient is getting seen soon after discharge, and complications get recognized and treated in proactive ways that ward off rehospitalizations.  We don’t know what these best practices are, but whatever the mechanisms may be, it seems that some hospitals do a better job of this than others, which is no surprise.  The challenge is figuring out what these high quality hospitals are doing differently.

Some hospitals have intuitively figured this out.  At Brigham and Women’s Hospital in Boston, readmissions were recently added to weekly Morbidity and Mortality (M&M) conferences.  Each readmitted patient is carefully discussed and the causes for the readmission probed.  If the readmission was preventable, surgeons discuss what might have caused it – whether it was due to an intra-operative technical error, a systems error, or an error in management or communication in the post-operative course. By treating surgical readmissions as a quality measure with the same gravitas as a complication or mortality, the message is clear – it may not be possible to get perioperative mortality or complications to zero, but there has to be a ceaseless effort to get as close as possible.

Moving forward on surgical readmissions

The Medicare readmissions prevention policy has been controversial because there is a sense that we put the cart before the horse.  Before understanding what really drove readmissions or how we might prevent them, policymakers jumped in, feet first, into penalizing hospitals that had high readmission rates.  The intentions were good, but the data were lacking.  The evidence so far says readmission rates are falling, but the cost of the effort has been high.  We are penalizing those who care for the sickest and most vulnerable patients.

The data on surgical readmissions, on the other hand, is far more reassuring and reminds us why we need good data.  The evidence here is much clearer:  when we go after surgical readmissions, we are going after poorer quality care.  Right now CMS has proposed two surgical procedures—our data suggests that they can broaden their efforts to include many more.  Because the bottom line is if the readmissions reduction program can motivate hospitals to avoid these readmissions by improving surgical care, we can have better outcomes while saving money.  And that would be good news all around.

Ashish Jha, MD, MPH is the C. Boyden Gray Associate Professor of Health Policy and Management at the Harvard School of Public Health. He is also the Senior Editor-in-Chief for Healthcare: The Journal of Delivery Science and Innovation. Jha blogs at An Ounce of Evidence where this post originally appeared.

Thomas Tsai, MD, MPH is a general surgery resident at Brigham and Women’s Hospital and a research fellow at Harvard School of Public Health and the Institute for Technology Assessment at Massachusetts General Hospital.

Livongo’s Post Ad Banner 728*90

7
Leave a Reply

6 Comment threads
1 Thread replies
0 Followers
 
Most reacted comment
Hottest comment thread
6 Comment authors
Harlan KrumholzRobalan t falkoff, md, faafpsrLeah Binder Recent comment authors
newest oldest most voted
Harlan Krumholz
Guest
Harlan Krumholz

Agree completely that current measures can improve. We can do better in measurement and in quality. What bothers me is that many people are indicating that the readmission measures were premature. A patient-centric view might lead you to the conclusion that we have a crisis in our care. From the patient perspective the health care profession often views care in venues (inpatient, outpatient, SNF, etc) rather than as a holistic episode that the patient experiences. There is so much evidence, qualitative and quantitative, that we, as a profession, fail patients particularly at the point of transition from inpatient to outpatient… Read more »

Leah Binder
Guest

Beautifully said Harlan!

Rob
Guest
Rob

Isn’t it possible that the highest “quality” hospitals draw patients from all over the country(and the world) and that those patients when discharged return home, so that if they develop a late surgical complication the are readmitted to a local hospital and that these studies miss the true readmission rate? Perhaps patients that patronize these elite institutions are more self selected and therefore may be more likely to be compliant with discharge instructions? Isn’t it also true that an observational study like this, that lacks randomization or does not adequately control for case mix, severity, comorbidities, etc is very interesting… Read more »

alan t falkoff, md, faafp
Guest

Yes the focus has once again swung away from LOS in hospitals and back to gaming the system and maximizing reimbursement.

sr
Guest

I agree with Leah Binder in appreciating the work that B&W’s is doing. But having been a hospital consultant, I have to repeat what many of the practitioners used to say – ‘this policy isn’t going to work as intended, and the poor and sick are just going to get denied care’. I will quote from above – ‘before understanding what really drove readmissions or how we might prevent them, policymakers jumped in, feet first, into penalizing hospitals that had high readmission rates. The intentions were good, but the data were lacking’. The question I always wonder is – if… Read more »

Leah Binder
Guest

Another enlightening and important study from Ashish Jha. I agree we need more data to understand the causes of readmission and I worry as well that by using readmissions as a talisman we impose undeserved penalties on communities that are already struggling with the absence of resources and adequate social supports. At the same time, I believe hospitals should have a financial stake in reducing readmissions–or at the very least not have a financial stake in increasing readmissions (as they have historically). If nothing else, creating the right incentives around readmissions should prompt a cultural shift in hospitals to view… Read more »

Saeid Shahraz
Guest
Saeid Shahraz

These are striking results and congratulations on publishing the new findings. Combination of high volume and low mortality rate seems like a very reliable benchmark outcome for quality assessment in such studies. I just wonder what the results would be if you included 6000 plus hospitals (which is not technically possible I understand) instead of half of the number. The significant differences are uniformly less than 1%.