• Six Key Take-Aways from STAAR on Reducing Readmissions

    Dec 01, 2009

    The momentum building around hospital readmissions is evident in the pace of the phone conversation with Amy Boutwell, MD, MPP, codirector of the STate Action on Avoidable Rehospitalizations Initiative (STAAR).

    A multi-state project involving 53 hospitals, STAAR was launched by the Institute of Healthcare Improvement (IHI) in May 2009 with grant funding from The Commonwealth Fund. To date, participating hospitals have formed cross-continuum teams and submitted baseline 30-day readmission rates. Now they are busy determining how to improve the patient's transition from hospital to post-acute setting.

    The excitement in Boutwell's voice reflects the importance of STAAR's admirable goal: reduce statewide rehospitalization rates by 30 percent. As a physician and hospitalist, she is all too familiar with the revolving hospital door. Yet Boutwell breezes over the well-documented problem. She wants to talk about the significant progress that IHI is already making toward measuring and reducing readmissions-as well as identifying what this means financially to hospitals.

    Focus on All Readmissions-Not Just Heart Failure Readmits

    Until recently, efforts aimed at reducing rehospitalizations have primarily focused on specific disease groups. "The literature is replete with articles on heart failure readmission rates, psychiatric-related readmission rates, etc. And there have been some good efforts on reducing readmissions for certain disease categories," says Boutwell.

    Yet, as highlighted in an April New England Journal of Medicine study, the hospital-to-postacute transition is broken for everybody-not just patients with particular diseases. Almost 20 percent of Medicare patients hospitalized between 2003 and 2004 were readmitted to the hospital within 30 days. One year after discharge, 56 percent of these Medicare patients had either died or been readmitted (Jencks, S.F., et al, "Rehospitalizations Among Patients in the Medicare Fee-for-Service Program," NEJM, vol. 360, no. 14, April 2, 2009).

    "This analysis gave us our first appreciation that we are looking at a systemic problem," says Boutwell. "At IHI, we're trying to expand the view of readmissions and say, 'We're talking about improving transitions between care settings for everyone, despite their admitting or discharge diagnosis.'"

    Form a Cross-Continuum Team

    Hospitals might argue that it doesn't do much good to design an efficient discharge process if communication breaks down in the next care setting. Outpatient providers could say the same thing about breakdowns on the hospital side of the equation.

    "This is why IHI strongly believes that the senders and the receivers need to codesign the transition process together," says Boutwell.

    All the hospitals that participate in STAAR must invite members of the post-acute community onto their teams. Representatives from skilled nursing facilities, physician offices, home health agencies-as well as patients and families-have been invited onto the STAAR teams to look at the process of being transferred from the hospital to the next setting of care. Team members discuss what the sender needs to know, and what the receiver needs to know so that the providers can communicate more effectively and ensure safe transitions between care settings.

    "It's exciting work and it's work that we have not yet done as doctors and nurses," says Boutwell. "We tend to pay attention to the processes of care within our own walls and we haven't looked across organizational boundaries to figure out how to best communicate with our colleagues in other settings."

    Start Measuring Your All-Cause 30-Day Readmission Rate

    "All hospitals can and should be looking at their 30-day readmission rates," says Boutwell. She stresses the need to measure all-cause rates and not just
    disease-specific rates.

    Last summer, the Centers for Medicare and Medicaid Services (CMS) began posting 30-day readmission rates for three conditions-heart attack, pneumonia, and heart failure-on the Hospital Compare web site. While beneficial, these risk-adjusted data do not tell the whole story, says Boutwell.

    "There's a huge limitation in looking at readmissions through the lens of a particular condition," she says. "For instance, many heart failure patients are admitted for a host of diagnoses other than heart failure, and by limiting your field to heart failure readmissions you are really missing the larger picture."

    Another limitation: The CMS data only includes Medicare patients. "As a physician, I do not treat a 60 year old on private insurance any differently than a 70 year old on Medicare if they are both sick with heart failure."

    It is critical for hospitals to look at readmissions for their entire patient populations, says Boutwell. "If you are doing well on heart failure readmissions, then perhaps you need to broaden the scope to look at, 'Is this individual being readmitted for any condition? Is this patient in and out of the hospital with a high degree of frequency?' These are the clinically important questions to ask, and the answers can help us understand what is going on socially with the patient so we can ensure a safe transition to the next setting of care."

    Determine Your Baseline, Then Track Over Time

    The 53 hospitals in the STAAR project-in Massachusetts, Michigan, and Washington-recently submitted baseline 30-day all-cause readmission rates to IHI. Pulling and reporting the data was not difficult, says Boutwell. The data are readily available in hospital discharge data bases.

    However, confusion did arise over what data to actually pull. "A 30-day readmission rate seems like it would be a fairly straightforward measure. However, we've learned that many hospitals, vendors, and even departments within hospitals (for example, finance versus a clinical floor) have slightly different methodologies for thinking about their 30-day readmission rates."

    IHI learned that many hospitals have not actually tracked readmissions as a quality metric. Those that have often get the data from a vendor, and the vendor has specific rules for what medical conditions are included in the numerator and what medical conditions are included in the denominator.

    IHI developed recommendations to help STAAR hospitals determine a measurement strategy-but these recommendations are purposely flexible and nonprescriptive. Boutwell acknowledges that the IHI readmission measures are "crude rates" since every STAAR hospital may use somewhat different measurement specifications or methods.

    "What we will be comparing in the STAAR initiative is the progress that hospitals are making. Are they making month-to-month progress on that measure?" This goes back to what IHI considers the most important take-away about quality measurement: Measure your baseline and track over time.

    Focus First on Improving Current Processes

    IHI has identified four key improvements that should help STAAR hospital teams-as well as other providers-significantly reduce readmission rates.

    Improve the assessments of rehospitalized patients. Ask patients, their caregivers, and their primary providers (for example, primary care physician) an additional key question: "Why is the patient coming back this time?"

    Boutwell stresses that this is different than determining the patient's primary diagnosis or chief complaint. "The primary diagnosis may be heart failure. The chief complaint may be 'I'm short of breath.' But the reason why Mrs. Jones is back is because she lives with her daughter, and her daughter went out of town and Mrs. Jones ate high-sodium canned soup for a week, which made her breathless."

    Learning the story behind the patient's medical issues is key to effectively intervening in the cycle of recurrent rehospitalizations, says Boutwell. "This is very rich clinical information that many of us don't collect as a matter of routine care."

    Instead of providing education, document learning. "Right now, quality metrics are based on whether we provide the patient with, for example, heart failure information," says Boutwell.

    Nurses and physicians should limit patient education to a few critical teachings-and then ensure that the patient actually learns those essential lessons, she says. "For example, if Mrs. Smith learns about the sodium content in canned soup, we have made more progress than if we just hand her an information packet that she doesn't read."

    IHI recommends that nurses and physicians use the teach-back methodology, which involves giving patients critical information and then asking the patients to restate that information to ensure that they have learned what is being taught.

    Communicate in real time. "In many cases, we make decisions about whether a patient is ready for discharge early in the morning, and then the nurse tells the patient, 'You are going to be discharged at 1 p.m. today,' and the patient and family are surprised and unsettled."

    To avoid this confusion, IHI recommends keeping communication open with patients and families throughout hospitalization. In particular, patients and families need to know the clinical goals the medical team is looking at, and the milestones that predict when the patient will be discharged.

    The other part of real-time communication is getting the critical discharge information to the next provider at the time of discharge-not 48 hours or a week later. Essential discharge information may include vital signs, lab results, and current medications. The exact information that needs to be exchanged will vary among providers, depending on IT access and other factors, says Boutwell. "That's why senders and receivers need to codesign this transition process together."

    Ensure timely postacute follow up. Discharged patients are typically told to follow up with their physician in one to two weeks, says Boutwell. "However, by the time the patient picks up the phone to make an appointment and actually gets in to see the doctor, it could be a month later."

    IHI recommends providing high-risk discharged patients with some type of clinical touchpoint within 48 hours of discharge. Any one who has been in the hospital more than once in the past year is considered high risk by IHI. So is any patient who failed to learn critical teach-back lessons.

    All other hospital patients should receive a clinical touchpoint within five days of discharge.

    The clinical touchpoint does not have to be a physician visit, explains Boutwell. It could be follow-up phone call from the hospital or physician practice, a home health visit, or some other type of clinical contact.

    Hospitals and other providers may also consider using transition coaches or navigators to help coordinate patient care across care sites and to empower patients to become more responsible stewards of their own care. These approaches have been shown to successfully reduce rehospitalizations. (For more information, download the IHI paper Effective Interventions to Reduce Rehospitalizations: A Compendium of 15 Promising Interventions at www.ihi.org.)

    "We are big fans of these approaches," says Boutwell. "But these models require an additional professional to coach or navigate the patient. And those extra services are not currently paid for by Medicare or by many commercial payers. So, in STAAR, we are focusing on improving the care we are delivering already, as part of our normal course of work," says Boutwell.

    Simulate the Financial Impact of Reducing Readmissions

    Step one is for financial officers to look at their hospitalwide 30-day readmission rates, and to then look at the financial breakdown of those rehospitalized patients," says Boutwell.

    IHI has begun to do this with a number of CFOs around the country in hopes of determining how reducing readmissions will impact different types of hospitals in a range of markets.

    "We are learning that there are a variety of impacts," says Boutwell. "For an urban, tertiary, academic medical center, readmissions are costly no matter what because they contribute to emergency department (ED) overcrowding. Many of these patients are placed on observation status, which is an intensive staffing model that has unfavorable financial implications. So there are some medical centers where reducing 30-day readmissions might look financially favorable."

    The opposite may be true for hospitals that do not have ED capacity issues. "For some institutions, it looks like about 20 percent of their admissions are 30-day readmissions," says Boutwell. "Suburban or community hospitals in a saturated market probably won't be able to fill those beds. So there is a volume loss potential."

    Hospitals must also consider how much it will cost them to fix the broken transition process, says Boutwell. Incurred costs for hospitals can include adding additional quality improvement staff as well as added staff time spent faxing or emailing discharge summaries and making phone calls to patients post discharge.

    According to a July 28th Wall Street Journal article, Berkshire Medical Center in Massachusetts spends about $500,000 a year in salaries and other operating costs to run a heart failure clinic. The clinic has successfully reduced heart failure readmissions. Regular clinic visitors had a 30-day readmission rate of about 3 percent in 2008, compared to a national average of 24.5 percent among heart failure patients. However, financially, the hospital lost about 30 heart failure admissions in 2008-or about $225,000 in revenue (Winslow, R. and Goldstein, J., "Cutting Repeat Hospital Trips-Simple Idea, Hard to Pull Off," WSJ, July 28, 2009.)

    "That's a gap of $725,000, which for many hospitals in this climate, is a big economic disadvantage to doing the right thing," says Boutwell.

    Reducing readmissions in isolation of addressing payment and policy barriers is possible-but it could be very challenging for individual providers, says Boutwell. "It's very important to get hospital administrators, physicians, nurses, public policymakers, state leaders, and payers at the table to say, 'When is the time? When are we are going to be ready to pay for value over time and across settings?'"

    Care coordination across sites needs to recognized and compensated, she says. Plus, at the very least, a transition strategy needs to be determined that helps hospitals deal with the financial implications of losing volume due to fewer readmissions. "Otherwise, we are going to be faced with some hospitals that see this as a financial viability issue," says Boutwell. "And it will be hard to engage their hearts and minds in resolving this important issue."

    For more information about the STAAR initiative, visit www.ihi.org.

    Published in the December 2009 Leadership e-newsletter.