More to Quality Care Than Rehospitalization Rates

Abstract

Some years ago, my nursing home administrator and I met with the chief executive officer of the local hospital, whose idea was to start directing hospital patients to “preferred” nursing homes. “But, but, but,” I sputtered, “you have to give people choice! That’s illegal!” He replied, “Dan, you’re a fool” (a phrase all too frequently said to me, and unfortunately in many cases appropriately so). It turns out I really didn’t understand — he was talking about informed choice, the hospital’s plan to identify the best homes in the area and partner with them, thereby improving quality and reducing cost. To this end, the hospital formed a Transition of Care committee with the local nursing facilities. Initially we worked on how to improve communication, particularly with the emergency department. But with changes in health care, data availability, and financial pressures, our committee’s mission morphed into how to reduce rehospitalization rates. The nursing homes can’t say that they weren’t warned about what was coming — a few months ago, the hospital released their first set of “tiered” skilled nursing facilities. To almost no one’s surprise, the only data point used for what determined the preferred facilities was … yes, rehospitalization rates. At the roll-out meeting, I spoke up with my concerns about this, and over the past few months many of the problems have become clearer. This is an important issue — if it hasn’t occurred in your region yet, there’s a good likelihood that it will or at least will start to be discussed. To make matters worse for my locale, the hospital’s accountable care organization (ACO) is thinking about using the exact same metrics that the hospital has started to use. Which brings us to our questions: Are rehospitalization rates the sole indicator of quality care in a nursing home? What are some of the problems if you assume they are? What are some of the implications — perhaps some of them unintended — for nursing homes, physicians, patients, and families? And is there anything that can be done about all this? There is no doubt that financial pressures are affecting our industry. We are all judged on data. We may lose money with high rehospitalization rates and with bundled care and ACO models. There also is no question that rehospitalization is not the only measure of quality patient care, but the hospitals may not understand this. Perhaps more importantly, they may not want to understand it — if they save money by demanding a rate below, say, 22%, what do they care what the nursing home thinks? The same principles may apply to the ACO, although other factors might affect a “population health” model that takes other factors into account for costs (such as resource utilization, medication usage, or number of consultants). Already I have seen quality measures that are applicable to young, healthy patients being used in our population. AMDA — the Society for Post-Acute and Long-Term Care Medicine has appropriately recognized the importance of identifying quality in the PA/LTC world, but it’s a difficult nut to crack. Let’s review some of the issues. A focus on rehospitalization sends a message to nursing home staff to keep patients in the SNF as long as possible, perhaps inappropriately so. This is a point strongly made in the Interventions to Reduce Acute Care Transfers (INTERACT) materials. The financial truth, sad to say, is that if a sick patient dies in the nursing home before being sent to the emergency department, that helps the home’s rehospitalization data. The hospital may develop a palliative care team. While stressing that it isn’t “hospice,” the reality is that patients are clearly being driven toward hospice/do-not-hospitalize status, even when it may not be appropriate. End-of-life care is complicated, with many gray areas, and there are definitely staff (physicians included) who can only see black or white. Just because patients are receiving palliative care does not categorically mean they should not be rehospitalized, but that may be the way they get treated. The SNFs that take complicated patients are unfairly discriminated against. A fair assumption can be made that very sick patients are more likely to be readmitted to the hospital than those not as sick. Let’s say an SNF gets 10 highly complicated admissions; despite hospital-level care at the SNF, three of these patients are appropriately returned to the hospital. Another home gets 10 “easy” patients, of whom two are readmitted to the hospital inappropriately. Which home looks better in the data? The data do not show the difference between a potential avoidable hospitalization (PAH) and an unavoidable one. To go back to our previous example, using the current methods the first SNF has a rehospitalization rate of 30%, and the second has a 20% rate. But if you look at it from an adjusted PAH perspective, the first SNF drops down to 0%. Obviously the numbers matter. A smaller number of admissions to the SNF can make a huge difference, and the SNF needs a larger denominator if the patients are high risk. But if the SNF isn’t preferred, it gets fewer admissions — making it far more difficult to achieve the hospital-driven benchmark. A nursing home in a hospital’s 9-1-1 catchment area may not be treated fairly. When the hospital sends a patient to an SNF that directs its emergencies to a different hospital, that data may not be correctly captured. The hospital claims this is seen on the PointRight system (a national data set that is supposed to take into account every hospital that a patient is sent to), but our local homes believe that may not be the case. So apples are not being compared with apples. Hospitals should consider the local SNFs as their partners — they should work with them closely and meet frequently. Once a hospital has developed its SNF preference tiers, who exactly is presenting the information to the patients? Either social workers or case managers are likely informing patients and families of their choices. Do all the SNFs involved feel comfortable with that? How do we know what they are telling the patients and families? For instance, one physician claimed that the hospital’s social worker told a hospitalized Medicare patient that the non-preferred SNF, in addition to not meeting the desired rehospitalization rate, “may not accept your insurance” — which along with not being true, might also be unethical to say. The social workers and case managers work for the hospitals, so they share the hospital’s inherent biases. They also may not know all there is to know about the range of available facilities, even though they have the power to influence referrals. Should patients even be referred to specific facilities? Who assumes the responsibility of having made that choice? For years, I wouldn’t provide specific recommendations to my patients because that would make me responsible for a bad outcome or bad care. Instead I would tell the families to look at the homes themselves and take into consideration many factors, such as their opinion after visiting, how close it was for the family to visit, its community reputation, the available reviews of prior stays and survey results, and the doctors who attended the residents. With the hospital now using only one measure for recommendations — readmissions — none of those other factors matter. Now, if the hospital refers to a home where the family is unhappy, the hospital’s to blame. If the family chooses a different home and has a good outcome, the hospital was wrong. If the family chooses a different home and there’s a bad outcome, the family must deal with the guilt of going against the hospital’s recommendation. What’s more, every facility not on the preferred list is being subtly accused of giving poor care by the hospital, which isn’t a good message to send to the community. How often the data are assessed can also be a major problem point. If the SNF has even one bad month, which can occur for a variety of reasons either preventable or not, their rehospitalization rate may go above the “accepted” level. They are then dropped off the preferred list, and boom, financial peril ensues. In general this is an example of many people not understanding data — one fixed data point does not establish a trend. The implications of all this can be very profound. For hospitals, one of the risks is that a non-preferred home may not only start sending its patients to a different hospital, but also stop using the other ancillary services that the hospital has been providing (such as home care, labs, hospice, etc.). When a resident’s family asks why they aren’t being sent to the hospital next door, all the SNF has to tell them is that the hospital now sends its patients to facilities many miles away, without respecting the family’s wishes. It takes a careful look at the big picture to see whether that’s a risk worth taking for the hospital, both financially and for its reputation. What can be done about all this? I’m not sure I have the answer, but I have some suggestions: Become involved. Regardless of your discipline, communicate with your hospitals so they know your concerns and know your goal is to provide excellent care. Work together for solutions. In today’s changing medical field, one should proactively address these issues. Encourage your local hospital to work collaboratively with SNFs in their catchment area. Personally I think hospitals should consider the local SNFs as their partners — they should work with them closely and meet frequently. Referring patients elsewhere has too many drawbacks in my opinion. Discuss the issue with all providers. Nursing home administrators may not know the issues and potential solutions from the health care providers’ perspective, and vice versa. Talk to the providers in the hospital. As hospitalists become more common, they may not realize what’s going on in our PA/LTC world. The primary care practitioners in the community may also be unaware. Each group’s involvement may encourage discussion so the patient and family are making a true informed choice. Make sure you are giving the best possible care. Organizations have been working for years to determine who the highest risk patients are. Look at your process to focus your efforts on this highest risk group. Luke Kim, MD, and colleagues at the Cleveland Clinic Health System recently validated a measuring instrument to predict 30-day all-cause readmission rates of SNF residents to hospitals, which may be of use (J Am Med Dir Assoc 2016;17:863.e15–863.e18). Join or form a larger group. Many areas have long-term care alliances and consortiums. There is power in numbers. Identify and implement best practices. If you are a member of a larger group or if the hospital is running a committee involving nursing homes, find out what the best nursing homes are doing to prevent PAHs and also look into how to replicate what they do. Always remember that we can ask the Society to help and we can bring up the issue in our state affiliates (the Society or others). Larger organizations working together is a helpful option as well. This is why we belong to organizations — to call attention to important issues and seek help. Again, never forget that there is power in numbers. The Board Room is an editorial written by members of Caring’s Editorial Advisory Board. Dr. Haimowitz is a multi-facility medical director and physician in Levittown, PA, and a member of the board. The views he expresses are his own and not necessarily those of the Society or any other entity.