This Article Abstract Full Text (PDF) All Versions of this Article: 35/11/2537    most recent 01.STR.0000145200.02380.a3v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Bode, R. K. Articles by Mallinson, T. Search for Related Content PubMed PubMed Citation Articles by Bode, R. K. Articles by Mallinson, T. Related Collections Health policy and outcome research Exercise/exercise testing/rehabilitation Rehabilitation, Stroke

(Stroke. 2004;35:2537.)
© 2004 American Heart Association, Inc.

Original Contributions

Relative Importance of Rehabilitation Therapy Characteristics on Functional Outcomes for Persons With Stroke

From the Center for Rehabilitation Outcomes Research, Rehabilitation Institute of Chicago, and the Department of Physical Medicine and Rehabilitation and Institute for Health Services Research and Policy Studies, Feinberg School of Medicine, Northwestern University, Chicago, Ill.

Correspondence to Dr Rita K. Bode, Center for Rehabilitation Outcomes Research, Rehabilitation Institute of Chicago, 345 E Superior St, Room O-950, Chicago, IL 60611. E-mail r-bode{at}northwestern.edu

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Background and Purpose— The purpose of this study was to evaluate the relative importance of therapy focus, intensity, and length of stay on greater than expected functional gain, controlling for stroke severity.

Methods— This observational study included 198 first-stroke patients who were recruited from 8 in-patient rehabilitation facilities and 5 subacute programs. Stroke severity (motor, sensory and cognitive impairment) at admission was measured using an instrument combining all 3 aspects; self-care, mobility, and cognitive status at admission and discharge were measured with the Functional Independence Measure. Time spent by physical, occupational, and speech-language therapists on function- and impairment-focused activities were used to compute therapy intensity by discipline and type of activity. Residual change scores, estimated by regressing discharge on admission functional status, were modeled using patient and therapy characteristics.

Results— Controlling for the stroke severity, greater than expected gains in self-care were predicted by longer lengths of stay and more intensive function-focused occupational therapy, and greater than expected cognitive gains were predicted by longer stays alone. Predictors of residual change in mobility, however, differed by gender: greater than expected gains in mobility for men were predicted by longer lengths of stay and more intense function-focused physical therapy whereas, for women, they were predicted by stroke severity alone.

Conclusions— Unlike previous studies using raw functional gains, therapies accounted for a significant proportion of the variance in residual functional change. The results support studies suggesting that both content and amount of therapy are important aspects.

Key Words: rehabilitation • stroke • stroke outcome

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
The need to understand what types and intensities of rehabilitation therapies result in optimal and cost-effective outcomes has been a driving force behind research for >20 years.^1–3 Assuring that patients receive sufficient and cost-effective services has humanitarian, clinical, and policy motives but requires sophisticated modeling to distinguish rehabilitation benefits from natural recovery. Simple regression models use therapy intensity (dose) to predict functional status at discharge or functional gain (response). Evidence of a dose-response relationship using simple models has been equivocal.⁴ Modeling functional improvement is limited by the strong associations among admission, discharge, and gain measures.⁵ Because admission functional status is the strongest predictor of discharge functional status,^6–7 when it is included as a predictor of gain, it accounts for most of the variance in the outcome. Consequently, relationships between therapy or any other potential contributor and functional gain are masked. Using residual change scores (RCSs; the difference between actual and predicted discharge status) as the dependent variable removes the overwhelming effect of admission status on improvement. When modeling raw gain, we are able to identify individuals who gain the most (typically those with the lower function at admission who have the most room for improvement), but when modeling the residuals, we are able to identify persons who gain more or less than expected, regardless of their initial status. This enables us to better understand the impact of clinical variables on response to therapy for all persons. This article illustrates how modeling RCSs unmasks therapy characteristics that are related to more or less than expected gains.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Sample
The data came from 2 multicenter observational studies that examined rehabilitation outcomes in acute and subacute inpatient rehabilitation facilities. Each facility used its existing protocol to determine the types and amounts of therapy and length of stay (LOS) for each subject. The 8 acute facilities included 1 Veterans Affairs hospital; the others were private facilities, including not-for-profit and for-profit facilities. The 5 subacute facilities included 1 skilled nursing facility and 4 subacute units within acute and rehabilitation hospitals. Although a number of impairment groups were eligible for inclusion in each setting, only persons with stroke were studied in sufficient numbers across settings to be included in this study.

The inclusion criteria were: 18 years of age, first stroke, and multidisciplinary inpatient rehabilitation in either an acute or subacute setting. Data were collected in 2 phases: acute rehabilitation (1993 to 1998) and subacute rehabilitation (1996 to 2000). A total of 228 persons with stroke were enrolled in these studies; 30 had atypical lengths of stay (17 <1 week; 13 >8 weeks) and were excluded from this analysis.

Instruments and Instrumentation
Instruments are described in detail previously⁸ and are summarized here. As part of the acute rehabilitation study, lists of therapy activities by discipline were developed by an advisory group. Members represented the major clinical disciplines that provide services at inpatient rehabilitation hospitals as well as consumer and family representatives. A Delphi process⁹ was used to reach consensus on appropriate goals, activities, and barriers.

Twenty-five therapy activities each were identified by occupational therapists and physical therapists, and 21 activities were identified by speech language pathologists. Therapists used their clinical judgment to classify each activity and record the number of 15-minute units spent primarily in that activity. With clinician input, we aggregated time spent in the 71 therapy activities into 5 categories: evaluation and screening, function-focused activities, impairment-focused activities, discharge planning, and case management. The evaluation category included initial evaluation and screening activities, the discharge planning category included patient/caregiver education, home visits, and team/family conferences, and the case management category included documentation, and consultation with team members and with payors. Table 1 shows the classification of therapy activities into the 2 remaining categories: function-focused and impairment-focused activities. We summed the units within each therapy activity category by discipline.

Motor, sensory, and cognitive impairments were assessed to characterize stroke (impairment) severity.⁸ Ratings were calibrated using rating scale (Rasch) analysis.¹⁰ For ease of interpretation, calibrated values were transformed to a measure ranging from 0 to 100, with higher values representing less impairment. The Rasch estimate of reliability (0.86, interpreted similarly to Cronbach’s ) provided evidence of reliability of the measure for this sample, and the fit of the items to a single construct provided evidence of its validity in this application.

The Functional Independence Measure (FIM)¹¹ instrument was used to characterize patients’ activity restrictions.¹² Motor items were grouped by activities that are typically the focus of occupational therapy (OT; eating, grooming, bathing, upper and lower body dressing, toileting, and bowel and bladder management) and physical therapy (PT; bed/chair/wheelchair, toilet, and tub/shower transfer, walking, and stair climbing); a third item set (comprehension and expression, social integration, memory, and problem solving) characterized cognitive function.⁶ These 3 item sets were calibrated and produced excellent reliability estimates (0.93, 0.90, and 0.90, respectively). As in previous FIM calibrations,¹³ ratings from admission and discharge were cocalibrated for each scale to provide stable estimates of item difficulty, and separate person measures were produced for admission and discharge. Measures from the Rasch analysis were transformed to range from 0 to 100, with larger numbers indicating less activity restriction.

Six additional variables were used in preliminary modeling: age at admission, setting (acute or subacute), and 4 computed variables. LOS was computed by subtracting admission date from discharge date, but because of the skewed distribution of LOS, the natural log (Log^eLOS) was used in the analysis. The interval between stroke onset and admission (OAI) was computed by subtracting the admission date from the onset data (using the Log^e for the same reason). Intensity of function-focused activity by discipline was computed by dividing the number of therapy units within that category by the LOS, and intensity of impairment-focused activities was computed similarly. Intensities were multiplied by 15 to represent the average number of minutes per day.

Analysis
In each case for persons receiving therapy from those disciplines, OT intensity was used to model self-care improvement, PT intensity was used to model mobility improvement, and speech language pathology (SLP) intensity was used to model cognitive improvement. For comparison purposes, improvement was also modeled using function- and impairment-focused intensities for the disciplines combined. In modeling these improvements, we chose the discipline that was most related to the intended effect. Because little time was spent on evaluation, case management, and discharge planning, analyses were restricted to time spent in function- and impairment-focused activities. Correlations between the independent variables were used to evaluate collinearity among the predictor variables. Preliminary regression analyses were conducted to identify nonsignificant independent variables common across analyses to reduce the number of predictors in the models. These analyses identified age, Log^eOAI, and setting (acute or subacute) as nonsignificant predictors of residual gain and these variables were dropped. Because the setting effect was nonsignificant, data from both settings were combined for analyses.

Simple regression was used to predict discharge functional status from admission status and compute RCS. Results are reported in terms of percentage of variance in discharge status explained by admission status. A hierarchical model was then used to assess the influence of therapy characteristics (LOS; function- and impairment-focused intensity) beyond the effect of patient characteristics in predicting residual change. In the first model, only impairment at admission was entered. In the second model, gender and therapy characteristics were added to the model. Results are interpreted in terms of the effect that therapy characteristics have on higher or lower than expected gains. Lower than expected gains are reflected in negative regression coefficients, and higher than expected gains are reflected in positive regression coefficients, with higher than expected gains considered beneficial. Significant changes in R² values were used to determine significant improvements in the explanatory power of the model provided by the addition of the therapy characteristics. Multiple regression results are reported in terms of the percentage of variance in the RCS explained by the independent variables and the standardized regression coefficients (ß), t-statistics, and significance levels for each independent variable. A similar analysis was conducted using the RCS based on the impairment measure.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Sample Description
Table 2 summarizes the demographic characteristics for the sample. Correlations and summary statistics for impairment severity, intensity of function- and impairment-focused activities, and LOS for each sample are presented in Tables 3 and 4. Impairment severity is negatively related to LOS for all groups; that is, more severely impaired persons tended to have longer LOSs. For the group receiving PT, the intensity of function- and impairment-focused therapy is related (collinear) but not for the groups receiving OT or SLP, perhaps because it is harder to differentiate the 2 foci in PT activities. For those receiving OT and SLP, of interest is the negative relationship between impairment severity and function-focused intensity: more impaired persons tended to receive more function-focused therapy. The average OT intensity is higher for impairment-focused activities (t[df=184])=–2.30; P=0.02), but the average PT and SLP intensity is higher for function-focused activities (t[df=180]=5.30; P<0.001 and t[df=153]=8.34; P<0.001), respectively.

Prediction of Discharge Status and RCS
In the simple regression model, admission functional status significantly predicts discharge functional status: the adjusted R²s are: 0.628 for self-care, 0.517 for mobility, and 0.737 for cognition. Less variance is accounted for in predicting RCS than discharge status. Admission impairment severity accounts for little variance in the hierarchical regression model: 2% in self-care, 7% in mobility, and 4% in cognition. For self-care and mobility, the percentage of variance explained by therapy characteristics increases significantly to 24% for self-care (F_change[df=4178]=13.93; P<0.001) and 25% for mobility (F_change[df=4174]=11.41; P<0.001). In general, over and above impairment severity, longer stays and more intense function-focused therapy are associated with greater than expected gains in self-care and mobility. However, when therapy characteristics are added to the model for cognition, the increase in variance explained (7%) is not statistically significant (F_change[df=4147]=2.29; P=0.06), possibly because of the fact that so little variance was left to explain.

Although gender was not significant in the modeling of RCS in self-care and cognition, its effect was significant for mobility, necessitating the separate modeling of this change by gender. For self-care and cognition, impairment severity at admission and Log^eLOS are significant predictors of RCS for men and women. But for mobility, the variables that predict RCS differ by gender. Men who improve more than expected in mobility are those who have longer LOSs and receive more function-focused PT, whereas women who improve more than expected are those who are less impaired at admission. The intensity of function-focused therapy explains additional variance in terms of who improves more than expected in self-care, whereas the intensity of impairment-focused therapy does not. That is, those who improve more than expected in self-care are those who are less impaired at admission, have longer LOSs and receive a higher average intensity of function-focused OT. Those who improve more than expected in cognition are those who are less impaired at admission and have longer LOSs. The coefficients and test statistics for each model are presented in Table 5. The ß-coefficients show the relative importance of each of the predictors in the model in explaining RCS. Using discipline-specific function- and impairment-focused intensities explained slightly more variance in the self-care and mobility RCS: 24% and 25%, respectively (compared with 20% and 23% when using intensities combined across discipline) and the same in cognition RCS: 7%.

When we modeled impairment severity at discharge controlling for admission severity (results not reported here), the intensity of impairment-focused activities was not a significant predictor. This result supports the conclusion that more time spent in impairment-focused activities is not associated with greater than expected improvement, whether defined in terms of activity improvement or impairment reduction.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
RCSs were used to evaluate the relative importance of therapy focus, intensity, and LOS on greater than expected functional gain, controlling for stroke severity. Although there is growing evidence that suggests rehabilitation therapies have at least a modest effect on functional outcomes, it has not been established that simply increasing the amount of therapy improves outcomes, or for whom more or less therapy should be provided. Studies repeatedly identify the impact of patient severity, but identifying factors of rehabilitation therapy that are most strongly related to differences in patient outcomes are an important next step in the exploration of maximizing therapy benefits.

We compared the explanatory power of using combined therapy intensities across discipline versus discipline-specific therapy intensity and found that discipline-specific intensities did a slightly better job of showing a dose-response effect. Classifying therapy activities into function- and impairment-focused activities also provides more specific information on the kind of therapy that is related to better outcomes. Although previous research has not examined function- versus impairment-focused activities, these results support research that concluded that the content and timing of therapy may be more important than the amount.¹⁴ This study does not address therapy timing, but it provides useful information on the effect of therapy content on outcomes.

The results show that persons who gain more than expected are less severely impaired at admission; however, stroke severity at admission alone accounts for little variance in residual change. Therapy characteristics significantly increase the amount of variance explained by the predictive model. In the regression analysis, the contribution of therapy intensity (ß) is comparable to that of stroke severity and LOS, contrary to previous results in which it contributed little compared with initial status and LOS.⁶ Modeling RCS instead of raw gain thus provides stronger evidence of a dose-response relationship than found previously. As indicated by the nonsignificant setting effect in the preliminary regression analysis, the relationship between function-focused intensity and residual change did not differ across settings, suggesting that content and intensity of therapy maximizes outcomes rather than the settings in which they are delivered.

There are 2 possible explanations for the significant gender effect in mobility improvement. In our sample, women were more impaired and lower functioning than men at admission (data not shown), thus confounding the effect of gender on RCS. As a result, the type of therapy activities might differ for men and women. For more severely impaired persons in general, physical therapists might focus on bed mobility rather than locomotion, and because bed mobility is not assessed in the FIM, their mobility improvement would be underestimated. As in previous research,⁷ age and OAI are related to admission status (data not shown) and complicate the modeling of raw gain. When RCSs are modeled instead of raw gain, neither variable has an independent effect. Using this approach, we may be able to detect the effect on outcomes of other patient or therapy characteristics, such as comorbidities, that are relevant but collinear with initial status.

RCS Role in Understanding Therapy Effectiveness
Change or gain in function is a fundamental concern in rehabilitation. However, measuring change is fraught with methodological difficulties.¹⁵ Raw gain scores are typically modeled in examining the effectiveness of rehabilitation therapy, but drawbacks to using raw gains are obvious. For example, extreme scorers are likely to show more change when assessed the second time than people with more moderate initial scores because scores regress toward the mean.¹⁶ As a result, raw gains have a curvilinear relationship to initial status.⁷ RCSs have the advantage of having a zero correlation with initial status, thereby allowing analyses to concentrate on questions of efficacy.¹⁷ Therefore, the identification of patients who gain more than expected is not biased by initial status, and the relationship of therapy to outcome can be examined more directly.

RCSs have drawbacks. As with raw gain, they are less reliable than the estimates at each time point,¹⁵ and their interpretation is cumbersome. Because RCSs are relative, not absolute, a complete understanding of them requires knowledge of the representativeness of the sample, average group change, and whether higher or lower scores represent benefits.¹⁷ In this study, the acute and subacute samples are comparable in terms of demographic characteristics and functional status at admission to persons with stroke receiving rehabilitation services in facilities that participated in the Uniform Data Systems for Medical Rehabilitation (UDS) database during the time period covered by this study.^18–19 A comparison of average group change is not possible because UDS reports data on the total FIM motor scale, whereas this study used separate self-care and mobility subscales. However, the direction of beneficial change is clear; higher than expected change scores are interpreted as greater rehabilitation effectiveness.

Implications for Future Research
Using an alternate model of improvement, we have established that longer duration and greater intensity of some types of therapy are associated with greater than anticipated reductions in activity restriction. That is, it provides evidence of a dose-response effect in that more therapy is related to greater than expected gains. Using this approach along with theory-based definitions of rehabilitation treatments,²⁰ the role therapy plays in specific types of improvement can be examined more rigorously in future research.

These results have implications for the practice of rehabilitation under postacute care prospective payment (PPS). The amount of therapy per day in skilled nursing facilities has dropped significantly to low-moderate levels since 1996.²¹ Although the optimal duration and intensity of function-focused therapy is unknown, this study suggests that unchecked, changes such as those currently occurring in postacute care may occur at the expense of reduced activity restrictions for persons with stroke.

Study Limitations
Although this study was conducted at a number of sites offering rehabilitation services, it may not reflect the diverse acute and subacute rehabilitation programs across the nation. Although the facility refusal rate was low, self-selection may create bias. Also, data were collected before PPS was implemented at these sites, and therefore, LOSs were longer than are now common. In assessing the effect of SLP intensity, the measures used may have been less sensitive to aspects of stroke disability that were the focus of therapy, such as swallowing and aphasia. Finally, our post hoc classification of therapy activities may not have produced information that is as useful as a theoretically derived taxonomy of rehabilitation therapies.²²

Summary
Unlike previous studies using raw functional gains, therapy intensity accounted for a significant proportion of the variance in residual functional change. Differences were unmasked by examining function- and impairment-focused intensities separately. Results differ for residual change in self-care, mobility, and cognition, and for mobility, by gender. The results support studies suggesting that content and amount of therapy are important aspects.

	Acknowledgments

 
Funding was provided by the US Department of Education, National Institute on Disability and Rehabilitation Research to the Rehabilitation Institute Research Corporation for a field-initiated project titled Relation of Rehabilitation Intervention to Functional Outcome in Acute and Subacute Settings (grant H133G60135), and by a grant to the State University of New York at Buffalo, Rehabilitation Research and Training Center on Functional Assessment and Evaluation of Rehabilitation Outcomes (grant H133B300041). The American Occupational Therapy Foundation provided supplemental funding. We extend our thanks to the nurses and allied health professionals who provided the therapy activity information, the research assistants who collected the data, and the project coordinators who oversaw data collection at MidAmerica Rehabilitation Hospital, The Institute for Rehabilitation and Research, Swedish Covenant Hospital, Erie Country Medical Center, Episcopal Church Home, Long Beach Veterans Administration Hospital, Baylor Institute for Rehabilitation, Lutheran General Hospital, West Suburban Hospital, and the Rehabilitation Institute of Chicago.

Received April 2, 2004; revision received July 14, 2004; accepted July 22, 2004.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 
1. Feigenson JS. Stroke rehabilitation: effectiveness, benefits, and cost. Some practical considerations. Stroke. 1979; 10: 1–4.[Free Full Text]

2. Lincoln NB. Stroke rehabilitation. Curr Opin Neurol Neurosurg. 1992; 5: 677–681.[Medline] [Order article via Infotrieve]

3. Fuhrer MJ. Overview of clinical trials in medical rehabilitation: impetuses, challenges, and needed future directions. Am J Phys Med Rehabil. 2003; 82 (10 suppl): S8–S15.[CrossRef][Medline] [Order article via Infotrieve]

4. Cifu DX, Stewart DG. Factors affecting functional outcome after stroke: a critical review of rehabilitation interventions. Arch Phys Med Rehabil. 1999; 80 (5 suppl 1): S35–S39.[CrossRef][Medline] [Order article via Infotrieve]

5. Good DC. Stroke: promising neurorehabilitation interventions and steps toward testing them. Am J Phys Med Rehabil. 2003; 82 (suppl): S50–S57.[CrossRef][Medline] [Order article via Infotrieve]

6. Chen CC, Heinemann AW, Granger CV, Linn RT. Functional gains and therapy intensity during subacute rehabilitation: a study of 20 facilities. Arch Phys Med Rehabil. 2002; 83: 1514–1523.[CrossRef][Medline] [Order article via Infotrieve]

7. Johnston MV, Wood KD, Fiedler R. Characteristics of effective and efficient rehabilitation programs. Arch Phys Med Rehabil. 2003; 84: 410–418.[CrossRef][Medline] [Order article via Infotrieve]

8. Bode RK, Heinemann AW, Semik P, Mallinson T. Patterns of therapy activities for persons with stroke. Arch Phys Med Rehabil. In press.

9. Fink A, Kosecoff J, Chassin M, Brook RH. Consensus methods: characteristics and guidelines for use. Am J Public Health. 1984; 74: 979–983.[Abstract/Free Full Text]

10. Wright BD, Masters GN. Rating Scale Analysis: Rasch Measurement. Chicago, Ill: MESA Press; 1982.

11. Hamilton BB, Granger CV, Sherwin FS, Zielezny M, Tashman JS. A uniform national data system for medical rehabilitation. In Fuhrer MJ, ed. Rehabilitation Outcomes: Analysis and Measurement. Baltimore, Md: Brookes Publishing Co.; 1987.

12. World Health Organization. ICF: International Classification of Impairment, Activity and Participation. Geneva, Switzerland: World Health Organization; 2000.

13. Heinemann AW, Linacre JM, Wright BD, Hamilton BB, Granger CV. Measurement characteristics of the functional independence measure. Top Stroke Rehabil. 1994; 1: 1–15.

14. Wade DT, Skilbeck CE, Hewer RL, Wood VA. Therapy after stroke: amounts, determinants and effects. Int Rehabi Med. 1984; 6: 105–110.

15. Cronbach LJ, Furby L. How should we measure "change" –or should we? Psychol Bull. 1970; 74: 68–80.[CrossRef]

16. Frasure-Smith N, Lesperance F, Gravel G, Masson A, Juneau M, Talajic M, Bourassa MG. Social support, depression, and mortality during the first year after myocardial infarction. Circulation. 2000; 101: 1919–1924.[Abstract/Free Full Text]

17. Hauser-Cram P, Krauss MW. Measuring change in children and families. J Early Intervent. 1991; 15: 288–297.

18. Fiedler RC, Granger CV, Post LA. The Uniform Data System for Medical Rehabilitation: report of first admissions for 1998. Am J Phys Med Rehabil. 2000; 79: 87–92.[CrossRef][Medline] [Order article via Infotrieve]

19. Iwanenko W, Fiedler RC, Granger CV, Lee MK. The Uniform Data System for Medical Rehabilitation: report of first admissions to subacute rehabilitation for 1998. Am J Phys Med Rehabil. 2001; 80: 56–61.[CrossRef][Medline] [Order article via Infotrieve]

20. Whyte J, Hart T. It’s more than a black box; it’s a Russian doll. Am J Phys Med Rehabil. 2003; 82: 639–652.[CrossRef][Medline] [Order article via Infotrieve]

21. White C. Rehabilitation therapy in skilled nursing facilities: effects of Medicare’s new prospective payment system. Health Aff. 2003; 22: 214–223.[Abstract/Free Full Text]

22. Horn S. Using clinical practice improvement (CPI) methodology to peer into the black box. http://www.bu.edu/cre/webcast/index.html. Boston University Center for Measuring Rehabilitation Outcomes; March 2004.

This article has been cited by other articles:

				M. S. Nadar and J. McDowd 'Show me, don't tell me'; is this a good approach for rehabilitation? Clinical Rehabilitation, September 1, 2008; 22(9): 847 - 855. [Abstract] [PDF]

				I. G.L. van de Port, G. Kwakkel, I. van Wijk, and E. Lindeman Susceptibility to Deterioration of Mobility Long-Term After Stroke: A Prospective Cohort Study Stroke, January 1, 2006; 37(1): 167 - 171. [Abstract] [Full Text] [PDF]

				L. Tooth, K. McKenna, K. Goh, and P. Varghese Length of Stay, Discharge Destination, and Functional Improvement: Utility of the Australian National Subacute and Nonacute Patient Casemix Classification Stroke, July 1, 2005; 36(7): 1519 - 1525. [Abstract] [Full Text] [PDF]

				R. W. Teasell and L. Kalra What's New in Stroke Rehabilitation: Back to Basics Stroke, February 1, 2005; 36(2): 215 - 217. [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 35/11/2537    most recent 01.STR.0000145200.02380.a3v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Bode, R. K. Articles by Mallinson, T. Search for Related Content PubMed PubMed Citation Articles by Bode, R. K. Articles by Mallinson, T. Related Collections Health policy and outcome research Exercise/exercise testing/rehabilitation Rehabilitation, Stroke