One out of every five people discharged from US hospitals after suffering an acute stroke dies or is rehospitalized within 30 days after discharge, according to a study published in Stroke last year.
STUDIES have found wide variability in the hospital readmission rate for stroke patients.
But the study, led by cardiologist Gregg C. Fonarow, MD, and colleagues at the University of California, Los Angeles (UCLA), also found wide variability between hospitals in this readmission rate — a factor that may become very important as the Centers for Medicare and Medicaid Services (CMS) will soon add stroke to the conditions for which it calculates 30-day readmission rates. Hospitals with “excessive” rates of stroke readmissions may see their reimbursements for these additional hospitals stays cut back or denied altogether.
It seems obvious that interventions to identify the patients at highest risk of readmission, and to help improve their transition from the hospital back to a home setting, would be essential to help lower hospitals' readmission rates and improve stroke outcomes. There's just one problem, experts told Neurology Today: it's hard to know exactly who those patients are.
“In the national study published by [Dr.] Fonarow and colleagues, along with a cohort study done in Minnesota this year that had similar results [by Kamakshi Lakshminarayan, MD, PhD, and co-investigators] there was a wide variety of different diagnoses for which the patients were re-hospitalized,” said Michael Weinrich, MD, director of the National Center for Medical Rehabilitation Research. “It wasn't just recurrent stroke. In the Fonarow study, for example, only 14 percent of the re-hospitalizations were for recurrent stroke. About 11 percent of patients were readmitted for a cardiac problem. Some 11 percent had pneumonia. About half the diagnoses for which people were re-hospitalized had nothing to do with cardiovascular disease. That makes it much more challenging to identify what can help stop them coming back.”
“The last time I looked, over 300 studies had been carried out to try to predict readmissions after stroke. As far as I know, there are no predictive standardized models that are very accurate,” said Bruce H. Dobkin, MD, professor of neurology and director of the Neurologic Rehabilitation and Research Program at the Geffen School of Medicine at UCLA. “One problem is that the patients differ quite a bit.”
DR. MICHAEL WEINRICH: “Transitions of care are a problem in our system. Most patients do not go home, or to the next level of care, with their problems optimally managed and all of the information that they, their families and their primary caregivers need.”
Indeed, a review of studies from 2000-2010 in press by the federal Agency for Healthcare Research and Quality Evidence-based Practice Center in April confirmed that observation in an early draft of the report. “There were no transition of care interventions that consistently improved functional recovery after stroke or MI [myocardial infarction], and none seemed to consistently improve quality of life or psychosocial factors such as strain of care, anxiety, or depression,” the report authors wrote.
DR. BRUCE H. DOBKIN said he thinks hospitals and stroke centers can focus less on patient-specific factors and more on improving the transition process for all stroke patients. “The transfer to outpatient care requires a considerable amount of preparation, strategy and thought,” he said.
But, said Dr. Dobkin, there are, of course, some demographic factors that do appear to influence rehospitalization rates. “These include older age, particularly those over 80; people who weren't functioning well before their stroke; stroke severity, particularly difficulty with swallowing; and comorbidities, such as diabetes. But still, if I have 10 patients and some have all of those factors, some have a few, and some have none, it's still hard to anticipate based on that alone who's going to tend to be coming back.”
Dr. Dobkin thinks hospitals and stroke centers can focus less on patient-specific factors and more on improving the transition process for all stroke patients. “The transfer to outpatient care requires a considerable amount of preparation, strategy and thought,” he said.
Studies suggest that that process is drastically improved when stroke patients transition from hospital to home via an inpatient stroke rehab unit. For example, a 2006 study in Stroke by Anne Deutsch, PhD, RN, and colleagues, found that stroke patients who went through inpatient rehabilitation facilities before going home had much better overall outcomes compared to those who transitioned through skilled nursing facilities.
“But only about 20 percent of stroke patients in the US end up being transferred to an inpatient rehabilitation service for an average of 15-20 days of rehab,” Dr. Dobkin said.
Inpatient rehabilitation allows time to assess what ongoing problems a patient continues to have after their stroke and will need to cope with upon returning home. “It allows you to assess the patient's support, educate caregivers, provide equipment, modify medications and assess fall risk, and generally create a more optimal discharge plan for the patient,” Dr. Dobkin noted.
Whether from an inpatient rehab facility or just from the hospital in general, the US may be discharging its stroke patients far too quickly, said Mary Stuart, ScD, professor and director of the Health Administration and Policy Program at the University of Maryland, Baltimore County. “Today's ‘quicker and sicker’ discharge perspective, because of incentives for hospitals to get people out as quickly as possible, does indeed get people out fast, but readmission rates have also gone up dramatically. That's in large measure because of the frailty of the patient at that point of discharge. While the patient is arguably stable enough to go home, the follow-up is complicated and a lot can go wrong, either from their primary diagnosis of stroke, or a lot of other factors.”
Drs. Stuart and Weinrich have investigated stroke care and rehabilitation at hospitals in Switzerland and Italy. “They're keeping them longer than we are,” Dr. Stuart said. “In Switzerland, for example, we were surprised to find that they were discharging people from the acute care hospital at about the time when we were discharging from rehab. The patients then went into rehab for another period of time. They hold their patients in acute care until they think they'll get the greatest benefit from rehab, and keep them in rehab much longer as well.”
While the ability to compare costs of the two approaches was limited because of the different payment systems, Dr. Stuart noted that one long admission might be less costly than multiple shorter ones. “The early acute period of a hospital admission is the most expensive time, when someone is in a crisis,” she said. “The cost per day for keeping someone gets cheaper toward the end of admission, when you're watching them and keeping a close eye but not engaged in so many interventions.”
In addition to the length of stay in acute and rehabilitative care, the organization of services upon discharge is another key factor affecting readmission, according to Dr. Stuart.
“Transitions of care are a problem in our system,” said Dr. Weinrich. “Most patients do not go home, or to the next level of care, with their problems optimally managed and all of the information that they, their families and their primary caregivers need.”
One of the biggest problems is polypharmacy. Post-stroke patients frequently are sent home on multiple medications, some of which they may have been on prior to their hospitalization, and some of which may have been switched because of hospital formulary requirements. The confusing array of pills they must take can leave even family members with medical training at sea, said Dr. Stuart.
“Communication with an outpatient physician at the point of discharge is critical,” Dr. Dobkin said. “It's not enough to dictate a summary on discharge and hope it gets to primary care docs. The treating physician needs to talk to them directly, as well as give the patient specific information on what medications they are on, why some may have been changed, and what they will need to do after discharge.”
Dr. Dobkin suggests that hospitals should establish a system to give stroke patients and their families, upon discharge, one point of contact to help them sort through medications, physical therapy requirements, home health aides, safety equipment, and other needs. “They need to have one person that they can call to find out what's going on,” he says. “That often falls to the family physician, who is in the dark and doesn't have experience in ordering equipment or getting therapy.”
He's also excited about the use of new technologies to monitor stroke patients' recovery remotely. In the August issue of Stroke, Dr. Dobkin and his colleagues published a paper on the use algorithms for ankle accelerometer data to quantify the mobility of stroke patients in their home and community.
“This wasn't feasible even ten years ago, when accelerometers cost $4,000,” he said. “But now, I can get accelerometers that send information to a smartphone for $100. They can monitor how much, how far, and how fast a patient walks continuously, detect exercise, and send the information over the Internet, so a physician or researcher can see if patients are as active as expected or not. If they aren't, it's a warning signal that the person isn't doing well.”
Dr. Dobkin's research shows that information gathered this way has high reliability and validity as a real-world monitoring tool. “I'm guessing that within 3-5 years, Medicare will have identified classes of patients with particular diseases for which this kind of monitoring becomes one of the critical measures of risk for re-hospitalization,” he said.
There will probably have to be a lot of trial and error involved in improving the model for post-stroke care, Dr. Weinrich acknowledged. “In our current medical system, we really don't have a structure where there's clear accountability for this,” he said. “That's what CMS is trying to do in setting up accountable care organizations [ACOs]. Stroke would be a good pilot diagnosis for ACOs to develop new care models. It will also be useful to look at what's been done in other countries and in the VA system. But there's no doubt that we will have to try a number of different approaches.”
Fonarow GC, Smith EE, Schwamm LH, et al, for the Get with the Guidelines Steering Committee and Hospitals. Hospital-level variation in mortality and rehospitalization for medicare beneficiaries with acute ischemic stroke. 2011;42(1):159–66. E-pub 2010 Dec 16.