The Locomotor Experience Applied Post-Stroke (LEAPS) trial was most expensive trial NIH ever funded for stroke rehabilitation. Despite its expense and breadth (almost 5,000 stroke survivors were screened for this study and more than 400 included) there remains a cloud of confusion about the trial.
At its core, LEAPS was a comparison of partial weight supported treadmill training (PWSTT) and additional walking practice versus a home-exercise program. There are many clinicians who read various media accounts and came to the conclusion that the trial was a "negative." Headlines such as "Stroke Rehabilitation Study Results Surprise Researchers," "Walking Improves With Home Therapy Just as Well as Treadmill Training," and "In Stroke Recovery, Some Low-Tech and High-Tech Options are Equally Effective," didn't help.
The results of the study were published in the New England Journal of Medicine. Many clinicians who read that article came to a conclusion similar to that of Ralph Sacco, MD, president of the American Heart Association. His assertion was stark: "Although [the LEAPS researchers] hoped that this innovative approach to improving walking would be beneficial, they weren't able to show that."
And there was another issue that seemed to spring from the data and into headlines such as "At-Home Physical Therapy Safer Than High-Tech Stroke Rehabilitation." The last headline alludes to the fact that, as Duke Medicine News and Communications pointed out, "The trial showed that, when the locomotor training was used early, patients were at a higher risk for multiple and injurious falls."
For many in PT the question became, why mess with the equipment, extra set-up time, extra therapist training and added a risk of injury if outcomes are similar to standard care? In an effort to clear these questions up, we decided it was time to go the source. As you'll see in the following interview with one of the study's principal investigators, the full story is a bit more nuanced than "it didn't work."
Controlled Trials and Rehab Results
Katherine Sullivan, PT, PhD, FAHA is co-principal investigator of the LEAPS clinical trial, and associate professor in the Division of Biokinesiology and Physical Therapy at the University of Southern California.
 |
|
A patient uses partial weight treadmill training as studied in the LEAPS trial with the assistance of physical therapists.
|
Beyond a reasoned explanation of the data from Dr. Sullivan, we were also interested in a larger issue she emphasized when we requested the interview: Are randomized controlled trials (RTC), like the LEAPS trial, the best way to test rehab interventions? RCT, the same sort of trials that are done to allow for FDA approval of drugs, do not offer flexibility and the ability to change and modify treatment protocols based on incoming data.
But modifying treatment protocols based on incoming data is the hallmark of effective therapy. A mixed model research design may be a better way of proving efficacy of interventions. A mixed model approach combines quantitative (RCTs) and qualitative methods (usually things like interviews, group interviews and observations).
Authors: What did the LEAPS trial test?
Dr. Sullivan: The trial was designed to answer three clinical questions about physical therapy interventions for walking after stroke.
1. Is a walking program that includes PWSTT as a treatment modality more effective than progressive strength and balance exercises for improving walking speed and distance?
2. Does the timing (two- or six-months post-stroke) of the PWSTT walking program affect walking outcomes? Also, how does severity and the time since stroke interact with the PWSTT walking program influence outcomes? For example, do individuals with severe stroke perform better if an intense walking rehabilitation program is provided later, at the six-month time point, after stroke?
3. What is the optimal dose (12-, 24-, or 36-sessions) to achieve clinically meaningful changes in walking speed?
Authors: There are many clinicians who read various media accounts and came to the conclusion that the LEAPS trial was a "negative" trial. That is, PWSTT did not work. In fact PWSTT did work, but no better than physical therapy at home. For many in PT the question is: Why mess with the equipment, extra set up time and extra therapist training needed if outcomes are similar to standard care? We know therapists who are in a quandary because they encouraged their facilities to purchase PWSTT systems, and those facilities are now demanding justification for the purchase. What would you say to those therapists and those facilities?
Dr. Sullivan: First, it is important to understand that LEAPS compared two standardized, progressive physical therapy interventions provided by physical therapists. One intervention was a progressive strength and balance program provided early at two-months post-stroke. The other intervention was a progressive walking program that included PWSTT as a treatment modality provided early at two-months, or later at six-months post-stroke.
The progressive exercise program was delivered at home for design purposes, but was based on typical exercise provided by physical therapists in any setting. We did believe that the group that received PWSTT would have a better walking outcome then the progressive exercise program.
However, the LEAPS trial was the first large-scale, multi-site randomized clinical trial in stroke rehabilitation to determine if PWSTT worked across populations with severe walking problems. Also, no previous study had investigated whether a progressive strength and balance exercise program provided in the early recovery period post-stroke worked either. The LEAPS trial demonstrated that progressive, standardized physical therapy programs of sufficient duration were more effective than usual care, thus, both were proved to be effective.
Authors: As you point out, progressive strengthening and balance exercises improved gait. So one of the key conclusions of the LEAPS trial is something PTs/PTAs already knew: treatments that focus on strengthening, balance training and gait training work in improving outcomes. Essentially, physical therapy works. So therapists should take comfort in the fact that they treat patients in totality, addressing all issues. Everyone can agree that therapy should be comprehensive and PWSTT can be used as an adjunct to other proven therapies such as strengthening and balance training. In other words, PWSTT, in addition to strength and balance training are important components of effective physical therapy for patients with stroke. The LEAPS trial had the highest expenditure of any stroke recovery and rehabilitation study ever funded by the NIH. Was it worth it?
Dr. Sullivan: LEAPS was a bang for our taxpayer dollars. We showed that two standardized, progressive physical therapy interventions were effective in improving all measures after first time stroke, and that recovery is not limited to the first few months but extends throughout the year after a seriously disabling stroke. Also, gains made from two to six months were not lost but maintained at one year for those that received the early interventions.
Authors: The LEAPS trial was a randomized controlled trial (RTC). It's the same sort of trial that a medication would have to go through in to get FDA approval. Are RTCs the best way to test rehab treatment options in stroke?
Dr. Sullivan: We do not think that RCT trials, designed and interpreted based on requirements used for pharmaceutical studies, have a place in complex interventions such as rehabilitation RCTs. The most recent approaches to address health-related research for complex conditions are what are being termed the "mixed-model" research design. This approach allows for people with complex health conditions such as stroke or Alzheimer's disease to be enrolled in a clinical trial.
The clinician who provides the treatment starts with the standardize protocol; however, instead of dropping the patient due to non-compliance or an adverse event, the clinician can use the "logic of practice" to decide what the next best intervention should be. The non-responsive participant is transitioned to another treatment arm that may be more effective for his health condition.
Authors: The rigidity of clinical trials, where one-and only one-intervention is tested irrespective of shifting needs of the patient, does not reflect the way therapists actually do business. That is, "logic of practice" seems to be "practice sans logic." Therapists usually take the attitude, "Research is interesting, but we see little clinical application." The mixed model that infuses research into real world clinical environments (i.e., hospitals, outpatient clinics) may help bridge the "bench side" (clinical research) to "bedside" divide. Is there an algorithm or a decision tree within the next model research design?
Dr. Sullivan: Mixed methods are evolving from the most basic type which is a quantitative design combined with qualitative methods. For example, if the LEAPS RCT included a qualitative method such as structured interviews of patients and therapists during the trial. However, recently there have been examples of mixed-method quantitative trials where intervention protocols incorporate alternative interventions triggered by an adverse event or the clinical judgment of the practitioner. This type of mixed method design would be based on algorithms that are more similar to the situations that clinicians face each day in the treatment of individual patients.
Authors: During the LEAPS trial the PWSTT groups had a higher percentage of serious adverse events, recurrent stroke, fracture and hospitalization. What is your perspective on the risk of injury-especially falls-with regard to PWSTT?
Dr. Sullivan: The reported risks in the one-year paper were collapsed over the whole year; thus, the early and late PWSTT groups were combined and compared against HEP. One year results that are collapsed across a recovery year can be misleading since it does not take out the known factors that affect outcomes in the first year of recovery after stroke, such as the time from onset and stroke severity.
A paper to be published on the LEAPS falls incident makes the critical distinction between time and severity clear. Early (two to six months post-stroke), participants with greater severity had a higher fall rate if they received PWSTT. Fall rate was no different in the higher severity group that received HEP or the less severe participants who received HEP or PWSTT. However, the PWSTT group delayed to six months fell significantly less than all other groups since they were less mobile. Thus, rehab has the risk of greater adverse events such as falls or exercise induced conditions; but the benefit of greater recovery, more mobility and increased community reintegration.
My clinical interpretation would be to recommend that patients with walking velocity less than 0.40 m/s be provided a progressive strength and balance program prior to starting PWSTT. All patients with walking impairment <0.80 m/s should receive a progressive, standardized PT intervention, compared to little or nothing at all like our delayed PWSTT group. In other words, the factors that determined whether patients fell were how severe their stroke was and when they received treatment-not the partial body weight support itself.
Peter G. Levine is the director of SynapsTogether LLC, the author of Stronger After Stroke and The Stroke Recovery Blog, a research consultant for the Neuromotor Recovery and Rehabilitation Laboratory and a seminar instructor teaching plasticity-driving stroke recovery strategies. Contact: StrongerAfterStroke@yahoo.com. Mawunyo Gletsu is the lead physical therapist at AG-Rhodes Health and Rehabilitation and a PhD candidate in physical therapy examining partial body weight support gait training in people with Parkinson's disease at Nova Southeastern University.