Keep the Bar Raised on Reducing Hospital Readmissions

Last week the head of the American Hospital Association sent a letter to Donald Berwick, director of the Center for Medicare and Medicaid Services, stressing how “crucial” it is that the agency consider the racial and ethnic backgrounds of patients when it determines how well hospitals are doing on preventing readmissions within 30 days of discharge.

Why so crucial? The Affordable Care Act includes a hospital readmission reduction program (HRRP) that uses incentives and starting in October 2012, penalties to encourage hospitals to enact better follow-up and other procedures that reduce preventable readmissions among Medicare patients. For Medicare patients aged 65 and older, about 19% percent of all hospital stays were readmissions within 30 days, according to a new statistical brief from the Agency for Healthcare Research and Quality. Some 2.3 million rehospitalizations a year racked up more than $17 billion in annual Medicare costs in 2008, and the Medicare Payment Advisory Commission (MedPac) reported that expenditures for “potentially preventable rehospitalizations” were as high as $12 billion in 2005 alone. It’s a significant—yet avoidable—expense that drives up the ever-rising cost of Medicare.

Under the HRRP, those hospitals that report higher than expected 30-day readmission rates for patients who had been hospitalized with heart attacks, heart failure and pneumonia could see their Medicare reimbursements decreased by up to 1% the first year, up to a maximum of 3% in 2015 with an expanded list of relevant health conditions. Hospitals already submit information about readmissions as part of Medicare’s “pay-for-reporting” program and these rates are then published on CMS’s Hospital Compare site. By adding incentives and penalties to this benchmark, the goal of HRRP is to drive quality of care improvements while saving Medicare $710 million each year.

Rich Umbdenstock, the AHA president and author of the letter, takes the position that the 10% of hospitals serving a larger proportion of black and non-English-speaking patients will be unfairly penalized in this pay-for-quality program. He argues that a recent study in the Journal of the American Medical Association (JAMA), for example, shows “that blacks have a higher likelihood of being readmitted to the hospital than do whites. In addition,” the letter continues, “the JAMA article concluded that hospitals serving a disproportionately large number of minorities have higher readmission rates. Proper accounting for these racial disparities is crucial as the Department of Health and Human Services (HHS) implements the Hospital Readmissions Reduction Program (HRRP)…”

The AHA letter acknowledges that the health agency already uses risk adjustment measures that account for disparities in patient populations served by individual hospitals. These factors include adjustments for severity of the underlying medical condition and whether patients suffer from multiple diseases as well as demographic information like the age and gender of patients. 

But, HHS is not considering “patients’ race or life circumstances [the letter specifically mentions limited English proficiency], which can have just as great an impact on health outcomes,” asserts the hospital group. “As it implements the HRRP, HHS [Department of Health and Human Services]should incorporate these additional characteristics into its risk-adjustment methodology both to comply with the law and to avoid penalizing the very providers who are trying to eliminate racial disparities in health care.”

This is where the logic gets a little twisted. The AHA is correct that the authors of the JAMA paper did find racial disparities in hospital readmission rates: Black Medicare patients had 13 percent higher odds of 30-day readmission than white patients (24.8 percent vs. 22.6 percent) and patients discharged from minority-serving hospitals had 23 percent higher odds of readmission than patients from non-minority-serving hospitals (25.5 percent vs. 22.0 percent).

In general, the researchers found that white patients at non-minority serving hospitals had the lowest rate of readmission while black patients at minority-serving hospitals had the highest rate of readmission. But interestingly, white Medicare patients who were treated at minority-serving hospitals still had 23% higher odds than their counterparts at non-minority hospitals of being readmitted. From this and other findings, the JAMA authors concluded that “the association of readmission rates with the site of care was consistently greater than the association with race, suggesting that racial disparities in readmissions are, at least in part, a systems problem—the hospital at which a patient receives care appears to be at least as important as his/her race…”

Clearly, this is a complicated problem whose resolution will need to involve taking into account regional variations and socioeconomic disparities. Preventing hospital readmission is going to require a different approach at a small community hospital in a wealthy suburb, for example, as compared to a large urban medical center with a high proportion of poor, minority Medicare patients. First of all, in the original JAMA study cited by the hospital association—which looked at over 3 million Medicare hospital discharges from 2006 to 2008—only 8.7% involved black patients and only 10% of the hospitals were classified as “minority-serving.” The authors note that care for minorities is concentrated in just a few facilities. As to why these minority-serving hospitals have higher readmission rates—irrespective of who is released from them—the researchers and commentators can only speculate. “[E]ither our measures of financial stress are inadequate or…the higher readmission rates among these hospitals are due to other factors, such as a failure to prioritize quality or inadequate focus on transitions of care and coordination of care,” they write.

Other studies have provided more insight into the disparities. The AHA letter seems to suggest that a “patient’s behavior” once he or she is released from the hospital—unhealthy lifestyle choices, poor social support or mental illness, for example—is a significant factor. The hospitals, they say, will be unfairly penalized for the dire circumstances many of their patients find themselves in. I think this is missing the point of the readmission reduction program: the goal is precisely to address just these issues at the system level. Rather than calling for race and ethnicity to be part of the equation in determining readmission target rates, perhaps the AHA should be calling for assurances that HHS will help minority-serving hospitals devise specific programs that will ameliorate the readmission disparity. One finding: patients who contract infections while in hospital are more likely to be readmitted. Therefore, making sure minority-serving facilities implement simple safety programs (hand-washing, checklists, etc.) to reduce hospital-acquired infections after surgery or at the site of central lines could be an affordable way to reduce readmission rates.

The fact is, we actually know quite a bit about how to reduce a hospital’s readmission rate, whatever the racial/ethnic mix of its patients. A critical mass of studies has found that the most important factor in preventing hospital readmissions is improving follow-up care after discharge. This includes devising a comprehensive discharge plan that facilitates communication between doctors who treated a patient in the hospital and those who will see him as an out-patient as well as a link to community services to help monitor those who are particularly vulnerable to ending up back in the hospital. Disease management programs can also be effective in reducing rehospitalization. The chart below offers some key strategies.

Reduce readmit

The focus on  “seamless care” is not currently the norm at most hospitals. A 2009 study in the New England Journal of Medicine of Medicare beneficiaries who were rehospitalized within 30 days of discharge, found that more than half had not visited a physician’s office between the time of discharge and being readmitted. This fact “is of great concern and suggests a considerable opportunity for improvement,” write the NEJM authors. “Hospitals and physicians may need to collaborate to improve the promptness and reliability of follow-up care.”

A letter from Mahesh Krishnamurthy, a geriatrician and clinical assistant professor at Drexel University that was published in the Annals of Internal Medicine last year, gives even more details of the lack of continuity of care: “Approximately 41% of patients discharged from hospital have a test result pending. In two-thirds of these cases the MDs involved were unaware of the results. Of these pending tests, 9.4% required potential urgent action.” Krishnamurthy continues, “One fourth of all discharged patients require additional work up. 33% of these follow up outpatient tests are not completed.”

The health reform law has specific provisions to help hospitals reduce their chance of being penalized. First of all, the ACA sets aside $500 million for a readmissions-reduction Medicare pilot program that started earlier this year. Secondly, the emphasis on increasing the number of primary caregivers—whether they be doctors or nurse practitioners—will be another key to reducing hospital readmissions. Increased funding for community health centers, another provision included in the ACA will also encourage better access to conveniently-located primary care. 

In a recent study, Brian Jack, associate professor at Boston University School of Medicine found that adding one family physician per 1,000 residents at the county level, or 100 per 100,000, could reduce hospital readmission costs by $579 million per year, or 83 percent of the ACA target.

The Boston researchers combined figures from the Hospital Compare database with a set of data that measures physicians per population at the county level. “Using these data, we found that 30-day readmission rates for (pneumonia, heart attack and heart failure) decrease as the number of family physicians increases,” writes Jack. “Conversely, increased numbers of physicians in all other major specialties, including general internal medicine, is associated with increased risk of readmission,” he added.

Another important change is the new emphasis on accountable care organizations—groups of providers that include hospitals, primary care doctors, specialists as well as nurses and other caregivers—that will be paid a set fee to follow a patient through all aspects of an illness. Theoretically, fragile patients leaving the hospital will receive follow-up telephone calls and other communications to remind them of tests and doctor visits and to monitor their care for problems before they become serious enough to require hospital readmission. It will be in the accountable care organizations financial interest to keep patients out of hospitals, not bouncing back and forth accruing charges.

It’s become increasingly clear that hospitals will not have to reinvent the wheel when they begin undertaking readmission reduction strategies. Groups like the Society of Hospital Medicine and their Project Boost (Better Outcomes for Older Adults Through Safe Transistions) and the Institute for Healthcare Improvement’s STAAR program (State Action on Avoidable Hospital Readmissions, an initiative funded by the Commonwealth Fund) have already helped dozens of hospitals serving a wide range of patients and geographic locales use grants to reduce Medicare readmission rates.

In the end, the health reform law is introducing many provisions that will fundamentally change the profit motive for providers; linking monetary reward more closely to quality over sheer utilization. When CMS rewards hospitals for reducing readmissions and penalizes them for excess rehospitalizations they are removing deep-seated incentives providers have had in the past to fill empty beds.

Throwing up our hands in resignation and giving minority-serving hospitals a break on having to meet quality standards is ultimately a disservice to those patients who seek care from these facilities. In fact, hospitals that struggle with high Medicare readmission rates are precisely the ones that need incentives to provide better care coordination and will benefit the most from improved follow-up as well as new community health centers and accountable care organizations. It is important that CMS, as well as private groups like IHI and Commonwealth continue to help fund efforts to institute proven strategies in reducing readmissions at the lowest performing hospitals. Only then can we begin making progress on eliminating preventable disparities in health care.

9 thoughts on “Keep the Bar Raised on Reducing Hospital Readmissions

  1. Readmissions have largely been a result of the way hospitals get reimbursed by Medicare, which occurs by diagnosis and does not relate to the length of stay. I can tell you there is clearly an incentive to get that patient out of the hospital as fast as possible since hospitals only lose money the longer the patient is hospitalized. It is clear to me there are many times that the patients course can still take a turn for the worse and this is happening with greater frequency due to these quick discharges.
    The question no one seems to be asking in this debate is if this is such a bad thing. If I discharge 100 patients from the hospital, having stabilized them and discharged them as quickly as possible, and 10 of them end up readmitted, is this any better than keeping them longer in the hospital and having longer admission rates? The resources used are really the issue, and of course whether any of these patients suffered significant morbidity or mortality as the result of the readmission.
    As an example, I had a patient who underwent cardiac cath with stent placement. He did well and was discharged from the hospital the day after the cath in good condition. 3 days later, he returned with a large hematoma at the cath site and with lightheadedness and dizziness. Could anyone have done anything different to prevent this? I am not sure. Should we keep everyone in the hospital for 3 days after cardiac cath to prevent this rare readmission? Doesn’t sound like a smart idea to me. Yet this will count against the hospital as a readmission and they will be dinged as such.
    Also, what if the patient does not return for follow up in the proper time as instructed, despite the best efforts of the hospital to assure that he/she understands the importance of this? Is the hospital to get penalized for this failure of compliance? What about when the follow up health care team has no affiliation to the admitting hospital?
    Right now, this all seems like a very clumsy way for Medicare to incentivise better behavior by health care organizations that I can see resulting in better post- hospitalization care and more emphasis on the part of hospitals to assure the patinet is attended to after discharge, but Medicare is going to have to work hard to get this one right. This will be a difficult task with all the variables we have to deal with in treating a diverse patient population.

  2. Keith-
    Concerning your patient readmitted with the hematoma, you ask whether anyone could have doe something to prevent this? If the answer is no, then it doesn’t matter if your hospital is “dinged” beacause all hospitals will be dinged at the same rate – the cost of readmission is included in the DRG for the original admission if a hospital has an average readmission rate. If the answer is yes, then something could have been done and it is correct to penalize the hospital.
    Is it such a bad thing if you discharge 100 patients from the hospital and 10 are readmitted? It might be if there is significantly greater morbidity due to having discharged these patients rather than keeping them longer. If you happen to be good at identifying the patients who are most likely to be readmitted and don’t discharge them, the system rewards you (less cost for the same payment). If you are average at identifying these patients, it costs you nothing on average. If you are worse than average at identifying these patients, then it does indeed cost you (and rightfully so).

  3. There is no doubt that blogs have great importance and relevance in health care. Issues such as this are of interest public for millions of people in United States and in the world.

  4. Certainly a worthy read. Your book, Money-Driven…” addresses the money issue surrounding expense of keeping the alive people who have no hope of “meaningful recovery”. What’s happening on the politician front regarding this issue.
    Thanks

  5. My father-in-law (age 97 at the time) underwent gall bladder surgery at a large teaching hospital. He did well, despite endless dithering before surgery was done. The day after he came home, he was re-hospitalized locally with congestive heart failure (this happened after a previous surgery as well.) As that problem was resolving, a large abscess at the surgical site burst out and drained profusely. He had to be readmitted to the teaching hospital. Because of the CHF, he was placed on a cardiac floor. Several days later, with nothing done about the abscess (interventional radiology was supposedly required to correct the problem) the new interns, having not read his primary diagnosis of post-surgical wound infection, decided to send him home. I, a retired nurse fluent in medical terminology, had to argue rather vehemently with the discharge nurse that his primary admitting condition had not been addressed. She was adamant that “the order had been written” and that was that. I won by stating that we would simply take him to the ER entrance for readmission.
    And that would have “dinged” the teaching hospital TWICE under the new rules. Needless to say, we don’t plan to go there again. Dad survived for another year.

  6. Jackie–
    This is Naomi’s post, so
    I’m just weighing in as a reader.
    When your father-in-law was 97 did HE really want to go through all of the hospitalizations and treatments that you describe?
    Some 97-year-olds would want to, but this is relatively rare.
    In their late 90′s, many people would prefer to avoid hospitals, surgeries, etc, and spend their remaining time at home, surrounded by loved ones, with pallcative care specialists or hospice workers who can insure that they are not in pain.
    When they are in their 80s, many (certainly not
    all) Americans have made their peace with dying.
    Again, this is not what everyone wants. Some people really want to fight, as hard as they can, to live as long as they can.
    I believe that people have a right to do that. And palliative care specialists will help them.
    One goal fo palliative care is to help people die their own way– “in character.” As one pallative care expert explains: “Some people have always been fighters.” This is who they are. The palliative care specialist will help these patients fight to the end.
    “Some people have always lived in denial,” my source added. And a pallaitve care specialist will also help him die in character. This may mean that the patient and pallative care specialist never use the “D”(Death) word.
    But I also believe that the majority of older patients (over 80) don’t want to go through all that you describe . .
    Your father-in-laws may have been one of those who truly wanted to fight to the finish, or who just was not willing to accept the idea of dying.
    I am certainly not questioning his choice.
    I’m just suggesting that, as we go forward, we all should make sure that we’re asking the dying patient what he/she wants.

  7. Maggie,
    Even with younger patients issues like the one Jackie described can be a problem. In the case of her dad, the overriding issue wasn’t the CHF . . . it was the untreated post op infection. Even assuming her dad was a hospice patient, he received horrible care due to the fact that a) the interns didn’t treat the admitting diagnosis (why he didn’t get septic is a miracle of sorts), and b) the discharge nurse failed to be a patient advocate. She should have listened to Jackie, called the interns and told them “this patient’s primary admitting diagnosis has not been addressed and he is not ready for discharge.” She didn’t do that. She just wanted to clear a task off her list for the day. Pathetic.
    But getting back on topic . . .
    Naomi’s article raises some good points. Quite frankly, I find the AHA’s stance to be a cop out . . . a way to dodge the requirements of the law. What they’re really saying is, “We don’t want to be penalized for not taking proper care of patients who are poor and have no resources.”
    When a patient is admitted to the hospital, the hospital has an obligation to give proper care. To prevent readmission, this means teaching and follow up.
    If the patient doesn’t speak English, invest in Spanish speaking staff who can translate (I mean nurses, not housekeepers).
    If economics is the problem, put a focus on preventative care . . . now that the PPACA is putting an emphasis on that, they should be able to get reimbursed for keeping patients OUT of the hospital.
    The old models of “treat ‘em and street ‘em” aren’t going to work in the new environment. It is time for hospitals to modernize their thinking and adapt. Cutting them slack using race as an excuse just gives them permission not to change.

  8. As a nurse with unusually broad experience particularly in hospitals that serve a predominately African American population. I could not agree more with Naomi’s position. The “well, our population is sicker” excuse is so tired but never seems to go away. My experience tells me that there is a significant reason for readmissions that is being missed in this conversation, and actually in any discussion about the subject I have ever read.
    Hospitals don’t admit patients , doctors do. What I have seen in each hospital I worked for is a relatively small group of physicians whose practices are largely chronically ill, elderly blacks, many in nursing homes. These physicians maintain a constant census in the hospital – whether the patients need to be there or not. Often I would review a chart not finding anything particularly wrong, so I would ask them why they were in the hospital. The usual answer would be,”my doctor told me to come in”. It is not at all unusual to have patients spend a week in the hospital every month, year after year. They don’t look any different when they leave or when they return. Some of these patient actually exhaust their Medicare inpatient benefit, but hospitals are so dependent on these physicians for admissions that it is ignored. Consider the primary care physician who wants to follow their patient in the hospital, is it economically efficient to make the trip for one patient? No, ten? Definitely. As Dr. Atul Gawande said, they will feed at this trough till someone takes it away. I assert that, by far, readmissions are NOT due to premature discharges or incomplete care, they are due to doctors who practice primary care in the acute facility – because the system pays them better that way. CMS should look at that the data behind my assertion before it responds to the AHA.

  9. Panacea–
    I, too, consider the AHA’s stance to be a “cop out.”
    They are talking about the patients who most need follow-up, and they are saying “We can’t follow up on these patients; it’s just too hard!”
    Not true.
    In NYC, Brooklyn’s Maimonides hospital provides care for an extraordinary number of new immigrants from many nations, as well as conservative Jewish families who have lived in Maimonide’s neighorhood for years. The hospital also serves well-educated Brooklynites who recognize just how good Maimonides is. (My daughter is having her baby there in August.)
    Maimonides has translators who speak a great many languages. (I cannot remember the exact number.)
    Maimonides invests in translators rather than putting money into hotel-like amentiies,
    This can be done, and is essential if a hospital wants to provide patient-centered care.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>