Archives of Physical Medicine and Rehabilitation
Volume 86, Issue 12, Supplement , Pages 1-7, December 2005

Opening the Black Box of Poststroke Rehabilitation: Stroke Rehabilitation Patients, Processes, and Outcomes

  • Gerben DeJong, PhD

      Affiliations

    • National Rehabilitation Hospital, Washington, DC
    • Department of Rehabilitation Medicine, Georgetown University, Washington, DC
    • Corresponding Author InformationReprint requests to Gerben DeJong, PhD, National Rehabilitation Hospital, 102 Irving St NW, Washington, DC 20010
    • ,
  • Susan D. Horn, PhD

      Affiliations

    • Institute for Clinical Outcomes Research, Salt Lake City, UT
    • ,
  • Brendan Conroy, MD

      Affiliations

    • National Rehabilitation Hospital, Washington, DC
    • ,
  • Diane Nichols, PT, NCS

      Affiliations

    • National Rehabilitation Hospital, Washington, DC
    • ,
  • Edward B. Healton, MD, MPH

      Affiliations

    • National Rehabilitation Hospital, Washington, DC
    • Department of Rehabilitation Medicine, Georgetown University, Washington, DC

Article Outline

Abstract 

DeJong G, Horn SD, Conroy B, Nichols D, Healton EB. Opening the black box of poststroke rehabilitation: stroke rehabilitation patients, processes, and outcomes.

This article introduces the journal’s supplement devoted to the methods and findings of the 7-site Post-Stroke Rehabilitation Outcomes Project (PSROP), a study designed to provide a very granular in-depth understanding of stroke rehabilitation practice and how practice is related to outcomes. The article summarizes current knowledge about the effectiveness of poststroke rehabilitation, outlines where the PSROP fits into the broader traditions of stroke rehabilitation outcomes research, underscores the study’s methodologic innovations, and summarizes the scope of the articles that follow.

Key Words:  Intervention studies , Rehabilitation , Stroke , Treatment outcome

 

THIS ARCHIVES SUPPLEMENT reports on the Post-Stroke Rehabilitation Outcomes Project (PSROP), a large, multicenter stroke rehabilitation study that entailed the collaboration of 7 hospital-based rehabilitation centers—6 in the United States and 1 in New Zealand. These 7 centers enrolled nearly 1400 stroke rehabilitation patients from 2001 to 2003. The study’s database (N=1291) provides an in-depth view of inpatient rehabilitation practice. This supplement reports on the motivation for the study, its methods, and findings across several dimensions of practice. This supplement also addresses important epistemologic issues in rehabilitation research that are raised by the methods and findings of the PSROP.

Stroke remains among the most compelling public health issues in the world today. In the United States alone, an estimated 700,000 people experience a new or recurrent stroke each year.1 Approximately one quarter of these people die, and a significant portion of the remainder survive with long-term disability. There are approximately 4.8 million stroke survivors in the population, and about 1.1 million of these report having functional limitations. In economic terms, the estimated direct and indirect costs of stroke are $56.8 billion per year, as of 2005.1, 2

Stroke survivors account for about 17% of all inpatient rehabilitation admissions. Although lengths of stay (LOSs) in rehabilitation settings have diminished considerably over the last few decades, rehabilitation remains an extended and labor-intensive affair that has seen few major breakthroughs. Much of what we do in stroke rehabilitation may be routine, but much also remains a trial-and-error matter that is difficult to characterize. Rehabilitation practitioners, it is said, customize their interventions to each individual patient. One result is that stroke rehabilitation practice varies from one patient to another and from one rehabilitation center to another and thus often lacks the standardization that is being demanded in other areas of medical practice, as evidenced by the development of practice guidelines and standardized protocols. In other words, stroke rehabilitation remains a “black box” of sorts. We have good ways of characterizing what goes into the black box (ie, the patient) and what comes out (ie, the patient) but little notion of how best to characterize what goes on inside the black box. Our failure to do so also limits our ability to know exactly what the active ingredients are in the rehabilitation process that are supposed to shape patient outcomes. This lack of specificity also limits the claims that providers and consumers can make of health plans and government to secure the financial resources needed for stroke rehabilitation.

Back to Article Outline

Current knowledge of the effectiveness of poststroke rehabilitation 

Stroke survivors constitute one of the largest consumer groups of postacute rehabilitation services in the American health care system.3 Among inpatient rehabilitation facilities (IRFs), industry data for 2004 indicate that the average Medicare reimbursement per day for a stroke survivor is about $1050 and that the average LOS is 17.3 days. In this age of continuous quality improvement, cost containment, reimbursement reduction, and the drive for evidence-based practice (EBP), rehabilitation providers are obligated to make sure what they are doing is clinically effective, cost efficient, and supported by data. Despite the large body of stroke rehabilitation research, the truth is that we do not know exactly how the $1050 per day is spent. Medicare and sundry health plans remain willing for the moment to provide the funding for stroke rehabilitation.

Of the more than 700,000 people who experience a stroke each year,4 about 300,000 to 400,000 will need some rehabilitation services.5 These stroke survivors will be assessed and given initial rehabilitation treatments while in acute care; will be screened by a representative of a rehabilitation facility, both for clinical need and financial support availability; and then will be transferred or discharged to one of any of the following: a free-standing rehabilitation hospital, a rehabilitation unit located in an acute care hospital, a skilled nursing facility (SNF) for subacute rehabilitation, a nursing home for residential accommodations and care, or to home for care by family and to receive rehabilitation services either at home or as an outpatient. If the survivor goes to a hospital-based rehabilitation center—now commonly referred to as an IRF, he/she will receive an ongoing therapeutic program that consists of round-the-clock rehabilitation nursing and physician coverage; daily physical therapy (PT), occupational therapy (OT), and speech-language pathology (SLP); and possibly additional services from psychology, social work, therapeutic recreation, vocational rehabilitation, and rehabilitation engineering staffs. In addition, stroke rehabilitation patients will have access to medical consultants of all possible types, specialist nurses such as those for skin and ostomy care, chaplains, family members, legal representatives, insurance company case managers, and research investigators and their assistants.

The interaction between each stroke survivor, his/her comorbidities, personal behaviors, and coping ability and all of these health care providers and family members is complex and highly specific—with each and all factors having a possible impact on a patient’s outcome. The interaction of the patient with the facility’s system of care comprises the process of care that heretofore has not been systematically disaggregated, measured, and evaluated to determine the most active ingredients that affect patient outcomes.

The following is a far-from-exhaustive review of some important research findings about stroke rehabilitation in IRFs. In 1982, Lind6 reviewed the 7 best studies on the effectiveness of inpatient stroke rehabilitation. The results of these observational studies were conflicting and were only weakly comparable because of variations in research methods. Three studies showed a positive effect as a result of rehabilitation, 3 studies showed no effect, and the seventh showed a negative effect. Twenty years later (2003), Teasell et al7 were unable to find substantially more depth or consistency in their review. They reviewed 272 randomized controlled trials (RCTs) but were unable to find even 2 RCTs confirming the efficacy of any particular treatment.7 In 2002, Langhorne et al8 observed that before the field of stroke rehabilitation can evolve into an evidence-based field of practice, the field must first establish a reliable evidence base. Ottenbacher and Jannell9 noted that most RCTs in stroke rehabilitation are too small or scientifically inadequate to provide reliable guidance in establishing EBP.

Langhorne coordinated the Cochrane Stroke Unit Trialists’ Collaboration (SUTC),10 a meta-analysis of RCTs that compared dedicated stroke units with conventional care units in several European countries. The meta-analysis included 19 trials and concluded that stroke units have superior immediate and 1-year outcomes, in terms of function and survival. One would want to jump immediately into the data, to drill down and see what it was about the stroke units that produced the superior outcomes, but this level of data was not captured by any of the studies included. The best Langhorne could accomplish was to define stroke units as “geographically distinct wards with dedicated stroke teams, who provide coordinated multidisciplinary rehabilitation, programmes of education and training in stroke, and specialization of medical and nursing staff”10—and that is the extent of it.

Another problem of existing research on stroke rehabilitation is scientific rigor, with relatively few studies achieving what is commonly referred to as level 1 evidence. Moreover, the subject matter, selection criteria, measures used, and variables used in each study are sufficiently variable, making comparisons between studies difficult at best. The SUTC study supported the superior outcomes of stroke units10; Price and Pandyan’s study11 of poststroke shoulder pain supported the use of functional electric stimulation (FES). It cannot be known, however, whether the SUTC units used FES to improve their outcomes by reducing shoulder pain. This noncompatibility and lack of a comprehensive database compromises generalizability of results. Of course, not every study should be fully compatible with all others.

There has been substantial progress in the associated fields of neuroscience, radiology, medicine, and pharmacology to address the issues related to stroke management in recent years. Tissue plasminogen activator treatment protocols are gradually becoming the national standard of care for the initial presentation of an acute stroke at emergency departments.12, 13 Deep vein thrombosis prophylaxis is now routine and includes combinations of Doppler screening and the use of various antithrombotic drugs and compression devices.14, 15, 16 Finding better methods to prevent initial and recurrent cerebrovascular accident remains an ongoing challenge for both the medical and research communities.17, 18, 19

In addition to published research, there are national databases that record various aspects of the inpatient rehabilitation stay. The Centers for Medicare and Medicaid Services (CMS), for example, requires that all IRFs acquire data on all their patients on admission and at discharge using the Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF-PAI).20 Moreover, accrediting agencies such as the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) and Commission on Accreditation of Rehabilitation Facilities require that IRFs acquire outcome data on patients. JCAHO, for example, under the auspices of its ORYX initiative, requires that IRFs capture data on LOS, FIM score change, and discharge destination for each patient.21 IRFs report these data mainly to 1 of 2 national data systems—to eRehabData.com of the American Medical Rehabilitation Providers Association, an industry trade association, or to the Uniform Data System for Medical Rehabilitation. The upside to these databases is that they bring greater uniformity to the acquisition of rehabilitation patient data and aid in making comparisons across facilities. The downside is that they lack the depth needed to effectively examine stroke rehabilitation practice in any detail—nor would we expect these databases to do so. These data sets are limited mainly to patient data captured at admission and discharge, and nearly everything that happens in between remains largely unknown—the proverbial black box of rehabilitation.

A few studies have begun the process of opening and examining rehabilitation’s black box.22, 23 The excellent recent article by Bode et al24 was a multicenter study looking at IRF-PAI data, billing data, and discharge data but was limited to a sample of 177 patients—indicating once again how difficult it has been to penetrate the black box.

The neuroplasticity thesis has also spawned new research that examines the efficacy of specific interventions. For example, there has been substantial research evaluating the applications of the constraint-induced movement theories of Taub et al,25, 26 and modified versions of the initial protocol appear promising.27, 28 These newer interventions, however, rarely are compared with existing interventions or other therapeutic approaches such as neurodevelopmental therapy (NDT) or proprioceptive neuromuscular facilitation (PNF). They are usually considered in isolation from the large bolus of other rehabilitation-related interventions. Research on mental and physical practice, applications of learning theory,29 task-specific training,30 and functional imaging have all contributed important concepts to the treatment of patients with stroke in the clinic today. They allow us, for example, to see activation patterns of the brain to help understand motor recovery.31, 32, 33, 34, 35 We hope that the knowledge gained will lead to the development of new training approaches. Other new technologies being tested are the use of virtual reality and robotics to aid in the recovery of lost function.36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46 However, the best practices of existing therapeutic approaches have yet to be ferreted out.

The spring 2003 issue of Topics in Stroke Rehabilitation contained several detailed evidence-based reviews on numerous outcome and efficacy studies in stroke rehabilitation. Studies were rated based on the number and quality of RCTs. For example, Teasell et al7 developed a list of clinical findings based on RCTs having strong level 1 evidence. These findings, however, are quite nonspecific. They suggest that stroke rehabilitation improves functional outcomes, but it is not known whether the physical therapists used NDT, PNF, body weight–supported (BWS) gait training, FES, or had the patient practice walking. Did the occupational therapists use FES, slings, shoulder taping, positioning, tone inhibitory techniques, shoulder injections, or some combination of these to achieve the greater intensity of therapy to improve functional outcome? Did the treatment of neglect include placing all items of interest on a patient’s left side, or were red markings placed on the left side of all objects, or was there simply “maximal cueing”? Did patients with greater functional improvements receive serotonin-specific reuptake inhibitor antidepressants, stimulants, atypical antipsychotic medications, combinations of these, or none of these? To date, research on current practice has been able to tell us little more than that “rehab is good.” Now it is time to drill down, to get to the nitty-gritty of inpatient stroke rehabilitation. What really happens, how often, and to what effect?

These kinds of questions are not answered easily using traditional clinical research methods such as RCTs unless one is prepared to apply an RCT to each of these variations of practice—a solution that is neither practical nor likely to occur given current limitations in rehabilitation research funding. Clearly, different methods must be found if we are to address the various combinations and permutations of practice, including methods that provide highly granular-level data and allow researchers to examine microprocesses such as the impact of shoulder-hand syndrome pain and its treatments on participation and progress in rehabilitation. The clinical practice improvement (CPI) method used in the PSROP addresses this need for more granular treatment data, as outlined in the next section.

The 1995 Agency for Health Care Policy and Research (AHCPR) Post-Stroke Rehabilitation Guideline on stroke rehabilitation provided a review47 of the best research available at the time and supplemented that review with expert consensus recommendations in those instances where the literature did not provide level 1 evidence. The guideline panel found very few level 1 studies. Eight years later, in 2003, the Veterans Health Administration (VHA) issued its own stroke rehabilitation guideline48 by significantly updating the work of the 1995 AHCPR guideline, taking into account the studies conducted in the intervening years. The PSROP database offers a rare opportunity to test the AHCPR and VHA guidelines by determining whether patients treated in keeping with the guidelines had better outcomes. A previous study of 288 stroke survivors at 11 VHA sites throughout the nation found that compliance with AHCPR guidelines was positively associated with outcomes.49, 50, 51 Because this study was conducted within the VHA, it remains uncertain whether the findings generalize to women stroke survivors as well as men. The chief limitation in using the guidelines developed to date as a point of departure for future research is their lack of specificity, which mirrors the underlying literature’s lack of specificity with regard to the exact nature and timing of rehabilitation therapies such as OT and PT, including their intensity, frequency, and duration—the very dimensions captured by the PSROP.

Back to Article Outline

The PSROP 

The PSROP began initially as one of several projects under the auspices of the Rehabilitation Research and Training Center on Measuring Rehabilitation Outcomes hosted at Boston University’s Sargent College and funded by the National Institute for Disability and Rehabilitation Research in 1999. The leadership team for the study was drawn from 2 organizations: the Institute for Clinical Outcomes Research in Salt Lake City, UT, and the National Rehabilitation Hospital’s (NRH’s) Center for Health and Disability Research in Washington, DC. As the scope of the PSROP increased, additional funding was provided by the NRH Neuroscience Center with a grant from the U.S. Army and Materiel Command. The National Blue Cross Blue Shield Association contributed to the acquisition of 6-month follow-up data from the NRH site. Various authors who contributed to this supplement did so under the auspices of their own funding sources in addition to those mentioned here.

The PSROP’s Principal Research Question 

The PSROP’s main research question is an enduring one: what impact does each stroke rehabilitation activity or intervention, both individually and collectively, have on patient outcomes on discharge, controlling for patient differences including medical and functional status on admission? This rather global question can be partitioned into a series of subsidiary questions, several of which are addressed to one degree or another in the articles represented in this supplement. Answering these questions required the acquisition of detailed in-depth data on patient characteristics, processes of care, and outcomes and the creation of a large relational database that is described more fully in Gassaway et al.52

Critical to the success of the PSROP has been the steadfast participation of the study’s 7 clinical sites in the design, data collection, and analysis phases of the project. The 7 sites participated vigorously and contributed far beyond the funding levels provided by the project. The 7 sites and their respective site directors are identified by Gassaway.52 The participation of front-line clinicians was especially important to the study’s attempts to characterize rehabilitation activities and to collect data documenting each stroke rehabilitation activity and intervention.

We distinguish between activity and intervention, a distinction underscored by the PSROP’s clinical contributors. An activity, to borrow examples from PT, might include bed mobility, sitting, gait or walking, and community mobility. An intervention, to use PT again, may include strength exercises, aerobic or conditioning exercises, electric stimulation, parallel bars, BWS gait training, and family education, to cite only a few of the 57 interventions coded in the study. At the risk of some oversimplification, there are 2 broad levels at which individual therapies can be analyzed: the activity level and intervention level (ie, the therapy intervention used to facilitate each therapy activity). This supplement is limited largely to the therapy activity level and not to the intervention level. We seek first to determine how participation in individual activities—in terms of timing, duration, frequency, and intensity—shape outcome. In subsequent work we want to determine how interventions within select activities shape outcomes.

The PSROP’s Methodologic Innovations in Rehabilitation Research 

We devote an entire supplement to the PSROP because of its scope and depth but also because it breaks new ground in rehabilitation research methods. One breakthrough has been the PSROP’s approach to characterizing the black box of stroke rehabilitation. To do so, it was first necessary to develop a taxonomy of stroke rehabilitation activities and interventions. It was never the intent of the PSROP to develop a stroke rehabilitation activity or intervention taxonomy, but investigators and collaborating clinicians determined that they could not go further if they did not have a working taxonomy of activities and interventions that used a common vocabulary and uniform methods of documenting stroke rehabilitation activities and interventions. The de facto taxonomy that evolved from the study has been outlined previously by DeJong et al.22 We do not present this taxonomy as a definitive one for stroke rehabilitation but believe that it serves as a working taxonomy that provides useful insights into how future and more formal stroke rehabilitation taxonomies might be developed.

The PSROP is a CPI-type of study that is essentially an observational cohort study with 3 added features. First, CPI studies systematically harness the collective wisdom of front-line practicing clinicians and use their insights in planning the study, defining the treatments to be evaluated, narrowing the hypotheses to be tested, developing the data collection instruments, and collecting and analyzing the data. Second, to control for patient differences, CPI studies capture the clinical complexity of each patient by using the Comprehensive Severity Index in addition to measuring functional status, a mainstay of rehabilitation studies. Third, CPI studies use detailed descriptors of rehabilitation processes made possible by the taxonomy of rehabilitation activities and interventions, as noted. Like many observational studies, CPI studies use multivariate analyses to identify the variables most associated with outcomes, but CPI’s distinguishing features, particularly the detailed characterization of activities and interventions, allow researchers to unravel relations that might not otherwise become apparent. A full-fledged CPI study includes a fourth feature: it ascertains the predictive validity of the findings by evaluating the outcomes that result when study findings are introduced into practice as part of a larger practice improvement strategy, a feature that also gives this genre of study its name—clinical practice improvement. The PSROP did not include this fourth feature.

A CPI study’s methodologic features also address some of the weaknesses found in RCTs. For more on the relative advantages of CPI and RCT studies, the reader is directed to a commentary by Horn et al.53

A central theme in clinical and health services research is the call for EBP, a call that is sometimes synonymous with a call for more RCTs in health care, including rehabilitation. Unfortunately, there may never be enough resources or time to address all the myriad forms of rehabilitation practice through randomized trials. There are no good shortcuts in rehabilitation research, but we do have to find a faster way of ascertaining what constitutes EBP in rehabilitation. Current methods for determining best practices are much too slow and too expensive. A strength of the CPI approach is its ability to uncover best practices more quickly than conventional methods, and such practices can later be vetted in validation studies or through controlled trials. A major challenge is knowing what therapeutic activities and interventions are truly ready for prime-time controlled studies. In the earnest quest for randomized studies, we risk wasting rehabilitation research resources on studies that may show no or minimal differences. Through the use of CPI-type studies, many promising therapeutic activities and interventions can be identified and unproductive activities and interventions weeded out in advance of such confirmatory studies.

The PSROP’s Limitations 

Every study has its limitations, and this study is no exception. First, the PSROP did not include data beyond discharge into the postrehabilitation period, except at 2 sites that had made independent efforts to follow up patients up to 6 months after their strokes. Hence, the PSROP can provide insight into the more immediate effects of stroke rehabilitation therapy but not into its long-term effects. The original level of funding simply did not permit the research team to probe beyond the rehabilitation episode, except in the 2 instances already noted.

Second, the study’s documentations of nursing activities and interventions are not as strong as those for the mainline rehabilitation therapies. The study was conducted during a period of serious nurse shortages that, in some instances, compromised the completeness of the nursing data, and thus these data are not reported in this supplement.

Third, as an observational cohort study, the PSROP focuses on the associations between various rehabilitation inputs and outcomes, not on causation of outcome. Nonetheless, as seen in subsequent articles, some findings and themes remain remarkably consistent across different patient subsets and therapy activities.

Fourth, there are other real or perceived limitations—for example, potential selection bias and other classic study limitations—although 1 hallmark of this study has been its ability to control for patient differences through the use of a detailed severity-of-illness adjuster that probes well beyond similar tools. These and other limitations are addressed in the supplement’s other articles.

Finally, the study’s unit of analysis was very much at the patient level and did not address major differences such as organizational milieu and interdisciplinary team coherence—although team conferences were considered a rehabilitation activity or intervention. The long-standing work by Strasser et al54 on rehabilitation team functioning has shown positive associations between various dimensions of teamness with patient outcomes. One could make the case that well-functioning teams result in better outcomes, because they organize care more efficiently at the patient level. They may also have an independent effect on outcomes because team culture may spill over onto therapist and patient mood and behaviors that affect the rigor of their participation. The PSROP did not capture this dimension of the rehabilitation experience.

Some Findings 

The PSROP offers several insights into the stroke rehabilitation process as we know it today. We want to share an insight or 2 that transcend the individual articles represented in this supplement.

An important finding is the large practice variations between facilities represented in the study. For example, we find enormous variations in the use of medications such as antidepressants, with no clear clinical indications for the observed variations. We rarely think of medication as a distinct rehabilitation intervention in the same way we think of the 3 therapies most closely identified with rehabilitation—namely, OT, PT, and SLP. Moreover, the management of affective disorders can greatly affect a patient’s ability to participate in these therapies. There is much room to identify best, or at least better, practice in this area.

The PSROP also examines the relative distribution of therapy activities within and between the 3 main rehabilitation therapies. PSROP investigators are struck, for example, by how little attention is given to community mobility and integration activities relative to other therapy activities. Future studies will need to determine how the neglect of these areas affects longer-term outcomes. Such studies are needed to inform providers, health plans, and other payers about the relative merits of these activities in fostering greater community independence and mobility after discharge.

One of the more compelling insights to emerge from the PSROP and the articles presented here is that earlier and more aggressive therapy is better, controlling for patient differences. In other words, starting therapy earlier is better than later, and moving patients on to higher-order and more difficult activities more quickly has a way of resolving some of the lower-order activities that rehabilitation providers sometimes focus on as necessary steps to more advanced activities. The earlier-is-better observation confirms many previous studies. The more-aggressive-is-better finding presents new opportunities to improve practice and presents new hypotheses for research. The case for the earlier-and-more-aggressive finding is also evident in some of the differential findings between the United States and New Zealand facilities. Compared with New Zealand facilities, U.S. facilities provide a more aggressive shorter-term program of rehabilitation with better outcomes, despite a more challenging case mix.

Back to Article Outline

Scope and organization of the supplement 

This supplement consists of 11 articles and 1 commentary, plus 2 commentaries by third parties. The second piece in this series is a commentary that raises fundamental epistemologic issues in rehabilitation research and identifies the methodologic genre of research from which the PSROP springs. In short, it provides the methodologic context that also enables the reader to understand better the contributions and limitations of the PSROP.

The third article52 serves as the common methods piece and baseline findings for the articles that follow. We chose to provide a separate baseline article describing the study’s methods and the study group’s principal characteristics because of the PSROP’s nontraditional approach and because the baseline article reduces the need for each subsequent article to repeat all the same background material with respect to study design and study group characteristics. Instead, each subsequent article’s methods section summarizes the study’s overall methods and focuses on those methods that are more specific to the topic of the article, especially if subsets of the study group were used and not the entire study group. We encourage readers of subsequent articles to have a basic acquaintance with this baseline article to more fully interpret the findings of individual articles that follow and to understand the limitations of the study. We also encourage readers to become acquainted with an earlier article22 published by the study team on stroke rehabilitation intervention taxonomies, as noted earlier.

A long-standing issue in rehabilitation is the timing of rehabilitation—timing from onset of the stroke and timing from acute hospital care to rehabilitation. The conventional wisdom has been that early rehabilitation results in better outcomes and that undue delays from acute care to rehabilitation result in further deconditioning and atrophy that limit participation in therapy or require a more prolonged rehabilitation process. The supplement’s fourth article examines this question and confirms much of the conventional wisdom on this topic.

The supplement’s fifth, sixth, and seventh articles examine and characterize the content of inpatient rehabilitation’s 3 main therapies—PT, OT, and SLP. Previous studies have been able to quantify the hours or minutes of therapy received over the course of an inpatient rehabilitation stay, but they provide little insight into the actual content of the 3 therapies with respect to specific activities and interventions and how much of each therapy activity or intervention was associated with higher levels of function and independence. The authors examine how participation in select therapy activities relates to progress in specific functional activities (eg, gait training, walking).

The authors of the 3 therapy articles do not attempt to show how participation in select activities relates to overall functional status at discharge and discharge disposition. This discussion is reserved for the supplement’s 11th article, in which we bring all the independent variables together in explaining the study’s observed outcomes. One of the lessons learned is that there is overlap of activities across the 3 therapies and that looking at these activities within the confines of the individual therapies in isolation from the other therapies provides an incomplete picture of the therapeutic encounter. The take-home message from this experience is that we cannot examine rehabilitation practice and therapy merely through the lens of individual therapy professions but need to look across professional domains to understand more fully how individual therapy activities and interventions relate to functional outcomes.

In addition to rehabilitation’s 3 core therapies, there are many other rehabilitation activities and interventions that the PSROP examined. We can report on only a few of these in this supplement. The supplement’s eighth article examines the use of neurotropic drug therapy. Many patients with stroke experience poststroke depression or other affective disorders that may slow their recovery, limit their participation in therapy, and diminish their outcomes. The authors believe that drug therapy is an understudied area of stroke rehabilitation that is ripe for more significant advances in rehabilitation practice and outcome.

Similar observations could be made about the role of nutrition in stroke rehabilitation outlined in the supplement’s ninth article. Many patients with stroke come to rehabilitation malnourished or inadequately hydrated either because of long-standing behaviors or because their new impairment may limit their ability to consume a more balanced diet. Nutrition is not commonly thought of as a rehabilitation intervention, but malnourished patients may lack the energy and mental focus needed to participate more effectively in rehabilitation therapy. In this area as well, we have observed considerable variation in practice and believe that a better understanding of the role of nutrition and malnutrition may help accelerate the rehabilitation process.

Midway through the PSROP, CMS implemented the long-awaited prospective payment system (PPS) for IRFs. The IRF-PPS presents important financial incentives that are likely to reshape provider behavior and rehabilitation practice. We believe that we cannot ignore this development and, coming midway through the PSROP data collection process in 2002, we have a singular opportunity to examine how the IRF-PPS may have altered the mix of stroke rehabilitation patients, the extent of therapy rendered, and the LOS. The supplement’s 10th article provides a before-and-after–PPS view of stroke rehabilitation in 3 of the 6 U.S. facilities that had significant numbers in both the pre- and post-PPS periods. This article, however, did not find striking short-term changes in stroke rehabilitation practice, as had been expected.

The supplement’s 11th article considers all the findings of the previous articles to help identify likely predictors of rehabilitation outcome in terms of functional status on discharge and discharge disposition. We chose not to examine these outcomes in previous articles except tangentially, in part because we believe that all patient and process variables cannot be evaluated independently of one another. Moreover, as noted earlier, the overlap in therapies across professional domains requires an integrated analysis that simply cannot be achieved by looking at the impact of therapies in isolation from one another.

The PSROP’s seventh site, the Wellington and Kenepuru hospitals in Wellington, NZ, provides an international dimension to stroke rehabilitation practice. The supplement’s 12th article examines how stroke rehabilitation practice in New Zealand is both similar to, and different from, stroke rehabilitation in the United States as represented by the 6 U.S. sites. We added New Zealand to the original U.S. cohort in the interest of locating additional variation in practice that might not be available in the United States. (The original PSROP study design included SNFs as well as IRFs. In 2000 and 2001, when the PSROP got underway, the SNF industry was experiencing considerable turmoil in the wake of changes resulting from the Balanced Budget Act of 1997, and surviving SNFs were difficult to recruit.) Unlike RCTs, which require very rigid adherence to practice protocols, CPI studies like the PSROP thrive on practice variation to help differentiate intervention effects that might otherwise be more difficult to identify when there are fewer practice differences.

Although the New Zealand site is the only non-U.S. site in the study, the PSROP is very similar to a 4-country, 5-site study on stroke rehabilitation currently underway in Europe, known as the Collaborative Evaluation of Rehabilitation in Stroke across Europe (CERISE). Sponsored by the European Commission, the study is being led by a team of investigators in Belgium at the Free University of Brussels and the Catholic University at Leuven. CERISE and PSROP investigators currently are examining ways to merge their 2 databases to provide a richer cross-national understanding of stroke rehabilitation practice and outcomes and to achieve a level of understanding that cannot be achieved by the 2 databases independently. Moreover, because the CERISE study ascertained 6-month outcomes, merging these 2 data sets will enable researchers to make more effective use of the 6-month outcome data obtained from 2 of the PSROP sites—one in the United States and the other in New Zealand.

Back to Article Outline

Conclusions 

Given the depth and scope of the PSROP database, there is a great deal more to be explored than is represented in this supplement. The research and findings presented here offer insights as to how we can understand practice variation and find best practices in stroke rehabilitation. The search for EBP begins with a better understanding of current practice. All too often, the quest for innovation ignores the gems that already exist in current practice and within the collective wisdom of rehabilitation practitioners. The PSROP offers 1 way in which these gems can be identified and disseminated into mainstream stroke rehabilitation practice.

Back to Article Outline

Acknowledgments 

We acknowledge the role of their collaborators at each of the clinical sites represented in the Post-Stroke Rehabilitation Outcomes Project: Brendan Conroy, MD (Stroke Recovery Program, National Rehabilitation Hospital, Washington, DC); Richard Zorowitz, MD (Department of Rehabilitation Medicine, University of Pennsylvania Medical Center, Philadelphia, PA); David Ryser, MD (Rehabilitation Department, LDS Hospital, Salt Lake City, UT); Jeffrey Teraoka, MD (Division of Physical Medicine and Rehabilitation, Stanford University, Palo Alto, CA); Frank Wong, MD, and LeeAnn Sims, RN (Rehabilitation Institute of Oregon, Legacy Health Systems, Portland, OR); Murray Brandstater, MD (Loma Linda University Medical Center, Loma Linda, CA); and Harry McNaughton, MD (Wellington and Kenepuru Hospitals, Wellington, NZ). We also acknowledge the role of Alan Jette, PhD (Rehabilitation Research and Training Center on Medical Rehabilitation Outcomes, Boston University, Boston, MA).

Back to Article Outline

References 

  1. American Heart Association . Heart disease and stroke statistics—2005 update . Dallas: AHA; 2005;
  2. American Stroke Association. Impact of stroke. Available at: http://www.strokeassociation.org/presenter.jhtml?identifier=1033. Accessed September 4, 2005.
  3. Liu K , Gage B , Kramer A . Appendix A. 1998 reports Medicare post-acute care (quality measurement final report) . Washington (DC): U.S. Department of Health and Human Services; 1998; Contract No. HHS-100-97-0010
  4. Broderick J , Brott T , Kothari R , et al.   The Greater Cincinnati/Northern Kentucky Stroke Study (preliminary first-ever and total incidence rates of stroke among blacks) . Stroke . 1998;29:415–421
  5. National Institute of Neurological Disorders and Stroke. Post-stroke rehabilitation fact sheet. Available at: http://www.ninds.nih.gov/disorders/stroke/poststrokerehab.htm. Accessed September 4, 2005.
  6. Lind K . A synthesis of studies on stroke rehabilitation . J Chronic Dis . 1982;35:133–149
  7. Teasell R , Foley N , Bhogal S , Speehley M . An evidence-based review of stroke rehabilitation . Top Stroke Rehabil . 2003;10:29–58
  8. Langhorne P , Legg L , Pollock A , Sellars C . Evidence based stroke rehabilitation . Age Aging . 2002;31(Suppl 3):17–20
  9. Ottenbacher KJ , Jannell S . The results of clinical trials in stroke rehabilitation research . Arch Neurol . 1993;50:37–44
  10. Organised inpatient (stroke unit) care for stroke. Stroke Unit Trialists’ Collaboration. Cochrane Database Syst Rev 2000;(2):CD000197.
  11. Price C, Pandyan D. Electrical stimulation for preventing and treating post-stroke shoulder pain. Cochrane Database Syst Rev 2000;(4):CD001698.
  12. Albers GW , Amarenco P , Easton JD , Sacco RL , Teal P . Antithrombotic and thrombolytic therapy for ischemic stroke (the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy) . Chest . 2004;126(3 Suppl):483S–512S
  13. Molina CA , Montaner J , Arenillas JF , Ribo M , Rubiera M , Alvarez-Sabin J . Differential pattern of tissue plasminogen activator-induced proximal middle cerebral artery recanalization among stroke subtypes . Stroke . 2004;35:486–490
  14. Kamran S , Downey D , Ruff R . Pneumatic sequential compression devices reduce risk of deep vein thrombosis in stroke patients . Neurology . 1998;50:1683–1688
  15. Pambianco G , Orchard T , Landau P . Deep vein thrombosis (prevention in stroke patients during rehabilitation) . Arch Phys Med Rehabil . 1995;76:324–330
  16. Desmukh M , Bisignani M , Landau P , Orchard T . Deep vein thrombosis in rehabilitating stroke patients (Incidence, risk factors and prophylaxis) . Am J Phys Med Rehabil . 1991;70:313–316
  17. Gorelick PB , Sacco RL , Smith DB , et al.   Prevention of a first stroke (a review of guidelines and a multidisciplinary consensus statement from the National Stroke Association) . JAMA . 1999;281:1112–1120
  18. Mohr JP , Thompson JL , Lazar RM , et al.   A comparison of warfarin and aspirin for the prevention of recurrent ischemic stroke . N Engl J Med . 2001;345:1444–1451
  19. Algra A, de Schryver EL, van Gijn J, Kappelle LJ, Koudstaal PJ. Oral anticoagulants versus antiplatelet therapy for preventing further vascular events after transient ischaemic attack or minor stroke of presumed arterial origin. Cochrane Database Syst Rev 2001;(4):CD001342
  20. Centers for Medicare and Medicaid Services. Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF-PAI). Available at: http://www.cms.hhs.gov/providers/irfpps/irfpai.asp. Accessed September 4, 2005.
  21. Joint Commission on Accreditation of Healthcare Organizations. Performance measurement in health care. Available at: http://www.jcaho.org/pms/. Accessed September 4, 2005.
  22. DeJong G , Horn SD , Gassaway JA , Slavin MD , Dijkers MP . Toward a taxonomy of rehabilitation interventions (using an inductive approach to examine the “black box” of rehabilitation) . Arch Phys Med Rehabil . 2004;85:678–686
  23. Ballinger C , Ashburn A , Low J , Roderick P . Unpacking the black box of therapy (a pilot study to describe occupational therapy and physiotherapy interventions for people with stroke) . Clin Rehabil . 1999;13:301–309
  24. Bode R , Heinemann A , Semik P , Mallinson T . Patterns of therapy activities across length of stay and impairment levels (peering inside the “black box” of inpatient stroke rehabilitation) . Arch Phys Med Rehabil . 2004;85:1901–1908
  25. Taub E , Morris DM . Constraint-induced movement therapy to enhance recovery after stroke . Curr Atheroscler Rep . 2001;3:279–286
  26. Taub E , Uswatte G , Morris DM . Improved motor recovery after stroke and massive cortical reorganization following Constraint-Induced Movement therapy . Phys Med Rehabil Clin N Am . 2003;14(1 Suppl):S77–S91
  27. Page SJ , Sisto S , Levine P , McGrath RE . Efficacy of modified constraint-induced therapy in chronic stroke (a single-blinded randomized controlled trial) . Arch Phys Med Rehabil . 2004;85:14–18
  28. Van der Lee JH , Wagenaar RC , Lankhorst GJ , Vogelaar TW , Deville WL , Bouter LM . Forced use of the upper extremity in chronic stroke patients (results from a single blind randomized clinical trial) . Stroke . 1999;30:2369–2375
  29. Winstein CJ , Rose DK , Tan SM , Lewthwaite R , Chui HC , Azen SP . A randomized controlled comparison of upper-extremity rehabilitation strategies in acute stroke (a pilot study of immediate and long-term outcomes) . Arch Phys Med Rehabil . 2004;85:620–628
  30. Carr J , Shepherd R . Movement science (foundations for physical therapy in rehabilitation) . 2nd ed.. Rockville: Aspen; 2000;
  31. Zemke AC , Heagerty PJ , Lee C , Cramer SC . Motor cortex organization after stroke is related to side of stroke and level of recovery . Stroke . 2003;34:e23–e28
  32. Ward NS , Brown MM , Thompson AJ , Frackowiak RS . Neural correlates of outcome after stroke (a cross-sectional fMRI study) . Brain . 2003;126(Pt 6):1430–1448
  33. Jackson PL , Lafleur MF , Malouin F , Richards CL , Doyon J . Functional cerebral reorganization following motor sequence learning through mental practice with motor imagery . Neuroimage . 2003;20:1171–1180
  34. Seidler RD , Noll DC , Thiers G . Feedforward and feedback processes in motor control . Neuroimage . 2004;22:1775–1783
  35. Shepherd RB . Exercise and training to optimize functional motor performance in stroke (driving neural reorganization?) . Neural Plast . 2001;8:121–129
  36. Dobkin BH , Firestine A , West M , Saremi K , Woods R . Ankle dorsiflexion as an fMRI paradigm to assay motor control for walking during rehabilitation . Neuroimage . 2004;23:370–381
  37. Rose FD , Attree EA , Brooks BM , Parlow DM , Penn PR , Ambihaipahan N . Training in virtual environments (transfer to real world tasks and equivalence to real task training) . Ergonomics . 2000;43:494–511
  38. Merians AS , Jack D , Boian R , et al.   Virtual reality–augmented rehabilitation for patients following stroke . Phys Ther . 2002;82:898–915
  39. Fasoli SE , Krebs HI , Stein J , Frontera WR , Hughes R , Hogan N . Robotic therapy for chronic motor impairments after stroke (follow-up results) . Arch Phys Med Rehabil . 2004;85:1106–1111
  40. Hesse S , Schulte-Tigges G , Konrad M , Bardeleben A , Werner C . Robot-assisted arm trainer for the passive and active practice of bilateral forearm and wrist movements in hemiparetic subjects . Arch Phys Med Rehabil . 2003;84:915–920
  41. Volpe BT , Krebs HI , Hogan N , Edelstein L , Diels C , Aisen M . A novel approach to stroke rehabilitation (robot-assisted sensorimotor stimulation) . Neurology . 2000;54:1938–1944
  42. Hesse S , Uhlenbrock D , Sarkodie-Gyan T . Gait pattern of severely disabled hemiparetic subjects on a new controlled gait trainer as compared to assisted treadmill walking with partial body weight support . Clin Rehabil . 1999;13:401–410
  43. Werner C , Von Frankenberg S , Treig T , Konrad M , Hesse S . Treadmill training with partial body-weight support and electromechanical gait trainer for restoration of gait in sub-acute stroke patients (a randomized crossover study) . Stroke . 2002;33:2895–2901
  44. Hesse S , Konrad M , Uhlenbrock D . Treadmill walking with partial body weight support versus floor walking in hemiparetic subjects . Arch Phys Med Rehabil . 1999;80:421–427
  45. Barbeau H , Visitin M . Optimal outcomes with body-weight support combined with treadmill training in stroke subjects . Arch Phys Med Rehabil . 2003;84:1458–1465
  46. Sullivan KJ , Knowlton BJ , Dobkin BH . Step training with body weight support (effect of treadmill speed and practice paradigms on post-stroke locomotor recovery) . Arch Phys Med Rehabil . 2002;83:683–691
  47. Gresham GE , Duncan PW , Stason WB , et al.   Post-stroke rehabilitation (assessment, referral and management) . Clinical practice guideline no. 16 . Rockville: U.S. Department of Health and Human Services, Public Health Services, Agency for Health Care Policy and Research; 1995; AHCPR Publication No. 95-0663
  48. Duncan PW , Zorowitz R , Bates B , et al.   Management of adult stroke rehabilitation care (a clinical practice guideline) . Stroke . 2005;36:e100–e143
  49. Duncan PW , Horner RD , Reker DM , et al.   Adherence to postacute rehabilitation guidelines is associated with functional recovery in stroke . Stroke . 2002;33:167–177
  50. Hoenig H , Duncan PW , Horner RD , et al.   Structure, process, and outcomes in stroke rehabilitation . Med Care . 2002;40:1036–1047
  51. Reker DM , Duncan PW , Horner RD , et al.   Postacute stroke guideline compliance is associated with greater patient satisfaction . Arch Phys Med Rehabil . 2002;83:750–756
  52. Gassaway J , Horn SD , DeJong G , Smout RJ , Clark C , James R . Applying the clinical practice improvement approach to stroke rehabilitation (methods used and baseline results) . Arch Phys Med Rehabil . 2005;86(12 Suppl 2): S16-33
  53. Horn SD , DeJong G , Ryser DK , Veazie PJ , Teraoka J . Another look at observational studies in rehabilitation research (going beyond the holy grail of the randomized controlled trial) . Arch Phys Med Rehabil . 2005;86(12 Suppl 2): S8-15
  54. Strasser DC , Falconer JA , Herrin JS , Bowen SE , Stevens AB , Uomoto J . Team functioning and patient outcomes in stroke rehabilitation . Arch Phys Med Rehabil . 2002;86:403–409

 Supported by the National Institute on Disability and Rehabilitation Research (grant no. H133B990005) and the U.S. Army & Materiel Command (cooperative agreement award no. DAMD17-02-2-0032). The views, opinions, and/or findings contained in this article are those of the author(s) and should not be construed as an official Department of the Army position, policy, or decision unless so designated by other documentation.No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated.

PII: S0003-9993(05)01259-1

doi:10.1016/j.apmr.2005.09.003

Archives of Physical Medicine and Rehabilitation
Volume 86, Issue 12, Supplement , Pages 1-7, December 2005

Article Tools