GAIT RESTORATION HAS LONG been recognized as a

Transcription

1 234 Ambulation After Stroke: How Important and Obtainable Is It and What Measures Appear Predictive? Susan E. Lord, MSc, Kathryn McPherson, PhD, Harry K. McNaughton, PhD, Lynn Rochester, PhD, Mark Weatherall, MBChB, FRACP ABSTRACT. Lord SE, McPherson K, McNaughton HK, Rochester L, Weatherall M. ambulation after stroke: how important and obtainable is it and what measures appear predictive? Arch Phys Med Rehabil 2004;85: Objectives: To assess how important community ambulation is to stroke survivors and to assess the relation between the level of community ambulation achieved and other aspects of mobility. Design: A multicenter observational survey. Setting: setting in New Zealand. Participants: One hundred fifteen stroke survivors living at home were referred from physical therapy (PT) services at 3 regional hospitals at the time of discharge and were assessed within 1 week after returning home. Another 15 people with stroke who did not require further PT when discharged were assessed within 2 weeks after they returned home to provide insight into community ambulation status for those without mobility impairment, as recognized by health professionals. Interventions: Not applicable. Main Outcome Measures: Self-reported levels of community ambulation ascertained by questionnaire, gait velocity (m/min), Functional Ambulation Categories (FAC) score, and Rivermead Mobility Index (RMI) score. Results: Mean gait velocity for the participants was 53.9m/ min (95% confidence interval [CI], ); mean treadmill distance was 165.5m (95% CI, ); median RMI score was 14; and median FAC score was 6. Mobility scores for the 15 people who did not require PT were within the normal range. Based on self-reported levels of ambulation, 19 (14.6%) participants were unable to leave the home unsupervised, 22 (16.9%) were walking as far as the letterbox, 10 (7.6%) were limited to walking within their immediate environment, and 79 (60.7%) could access shopping malls and/or places of interest. Participants with different levels of community ambulation showed a significant difference in gait velocity (P.001). The ability to get out and about in the community was considered to be either essential or very important by 97 subjects (74.6%). Conclusions: ambulation is a meaningful outcome after stroke. However, despite good mobility outcomes From the Department of Medicine, Wellington School of Medicine and Health Sciences, University of Otago, Otago, New Zealand (Lord, Weatherall); University of Southampton, Southampton, UK (McPherson); Medical Research Institute of New Zealand, Otago, New Zealand (McNaughton); and Northumbria University, Northumbria, UK (Rochester). Preliminary results presented at the 3rd World Congress in Neurological Rehabilitation, April 2002, Venice, Italy. Supported by the New Zealand Society of Physiotherapists. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the author(s) or on any organization with which the author(s) is/are associated. Reprint requests to Susan Elizabeth Lord, MSc, Dept of Medicine (Rehabilitation), Wellington School of Medicine and Health Sciences, PO Box 7343, Wellington South, New Zealand, sue.lord@ts.co.nz /04/ $30.00/0 doi: /j.apmr on standardized measures for this cohort of home-dwelling stroke survivors, nearly one third were not getting out unsupervised in the community. Furthermore, gait velocity may be a measure that discriminates between different categories of community ambulation. These findings may have implications for PT practice for people with mobility problems after stroke. Key Words: Cerebrovascular accident; Physical therapy; Rehabilitation; Walking by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation GAIT RESTORATION HAS LONG been recognized as a key goal in stroke rehabilitation, and considerable resources are spent to achieve it. In recent years, this emphasis has been extended to include the attainment of community ambulation as an important mobility and social outcome. ambulation has been broadly defined as locomotion outdoors to encompass activities such as visits to the supermarket, shopping mall, and bank; social outings; vacations; and pursuit of leisure activities. 1-3 Based on this definition, the criteria for the measurement of community ambulation have concentrated on mobility dimensions, in particular distance and temporal factors, because these are perceived to best reflect the dimensions of the task. 1,4 Perry et al 2 extended the criteria by classifying 147 people with stroke into 6 functional walking categories, 3 of which were related to community ambulation, based on gait velocity and the self-reported ability to ambulate in the community for selected activities. Hill et al 5 included the locomotion subscale score of the FIM instrument and the Functional Ambulation Categories (FAC) score in their criteria, along with gait velocity and distance. More recently, Patla and Shumway-Cook 6 have looked beyond distance and temporal factors to develop an operational definition of community mobility that considered 8 environmental dimensions: ambient conditions, terrain characteristics, external physical load, attentional demands, postural transition, traffic level, time constraints, and walking distance. This framework was explored further in a recent study 7 that highlighted the importance of 4 of these dimensions when people with and without mobility disability were compared as they ambulated in the community: temporal factors, postural transitions, physical load, and terrain. ambulation as a specific outcome has been the principle focus of 2 studies. 1,2 It was also considered as a separate outcome in several studies that investigated levels of mobility, community reintegration, and quality of life (QOL) for home-dwelling stroke survivors. Reduced levels of community ambulation and community participation, 5,8,9 decreased satisfaction with levels of outdoor mobility and community reintegration, and perceived difficulty in coping with public transportation and outdoor locomotion 13 have all been reported. In a qualitative study of 40 people with stroke, Pound et al 3 identified loss of independent ambulation, especially outdoors, as among the most debilitating aspects of the stroke sequelae. The main theme to emerge from the study

2 COMMUNITY AMBULATION AFTER STROKE, Lord 235 was frustration over not being able to leave the house, a feature that was also associated with mood disorder. However, results from these studies offer only limited insight into aspects of community ambulation for people with stroke, specifically its importance and how many people achieve it. The studies are compromised by small sample sizes, varying definitions of community ambulation, and limited approaches to measurement. The principal aim of this study was to assess the level of community ambulation and other aspects of mobility in a cohort of community-dwelling stroke survivors with motor impairment who had received physical therapy (PT) to improve their gait. Secondary aims were to identify the important features of a community environment for these people to further inform a definition of community ambulation and to investigate its importance. In addition, the results of this study allowed the estimation of effect size and power calculations for the measures that best characterized community ambulation for use in an ongoing multicenter, randomized, controlled trial in which approaches to improving community ambulation for people after stroke are being investigated. METHODS Participants In this multicenter, observational study, stroke survivors from 3 regional hospitals in New Zealand, which serve a combined population of approximately 800,000 people, were recruited between July 2000 and February The inclusion criteria were people who presented with a first-ever or recurrent stroke (MONICA definition) 14 and who had been ambulant in the community before their stroke. The participants were a convenient sample of people who had been discharged home after inpatient rehabilitation and who were recruited via referral from physical therapists in rehabilitation services. The first group of participants, who had subsequently attended PT with the goal of improving their general mobility and gait, were assessed within 1 week of discharge from PT services. This time frame was chosen to ensure that the assessment reflected what might be considered their optimal mobility and gait outcome. A second, smaller group of participants who did not attend PT as outpatients because their treating health professionals felt they were fit enough not to require it were assessed 2 weeks after discharge home, a time period that would allow a degree of reorientation to home and community life. The rationale for including this group was to test the assumption that community ambulation was both feasible and desirable for people who did not present with motor impairment and who were functioning at high levels after stroke. Each of the general hospitals serves a large geographic area, but because resources were limited for transporting subjects to the hospitals for assessment, participants living outside the city limits could not always be recruited into the study. All participants gave written consent to enter the study, which was approved by the relevant ethics committees for each center. Assessment Procedures Assessments were conducted in the PT department at each center by a senior physical therapist with expertise in neurologic rehabilitation and who was familiar with the clinical measures. The assessment took up to 1 hour, and the participants were invited to rest between each test item so that the effects of fatigue would be reduced. Measures The outcome measures used in this study included 4 standardized mobility measures and a questionnaire developed and pilot tested on several people with stroke before the study (appendix 1). The mobility measures were (1) gait velocity (m/min), measured with a stopwatch over a 10-m timed walk 15 ; (2) the FAC, a 6-point scale that classifies walking ability both indoors and outdoors 16 ; (3) the Rivermead Mobility Index 17 (RMI), a 15-item scale of functional mobility; and (4) gait endurance, measured by distance walked on the treadmill (participants were asked to walk up to 300m, to approximate the distance required for walking in suburban environments). These particular measures were selected because they have been shown to be reliable and valid for a stroke population. 15 The assessors were experienced physical therapists with expertise in neurologic rehabilitation who used the measures in their own clinical practice; the principal investigator (SEL) also conducted a training session with each assessor before the study. Interrater reliability has been shown to be good when users are trained and experienced, and therefore it was not formally tested in this study. 17,18 The questionnaire (appendix 1) was used to identify community ambulation by self-report according to the level of unsupervised mobility. Participants were allocated to 1 of 4 community ambulation groups based on this self-report: group 1, not ambulant outside the home; group 2, ambulant as far as the letterbox; group 3, ambulant in the immediate environment; and group 4, ambulant in a shopping center and/or places of special interest. The 4 categories were used to discern different levels of community ambulation and community participation and were based on the responses participants gave to question 2, which related to the types of places people liked to visit before they had a stroke. Statistical Analysis Frequency and descriptive statistics were examined for all variables. Relationships between the gait variables were investigated using the Spearman r. Analysis of variance (ANOVA) was used to examine mobility variables in relation to community ambulation. Responses to the open questions on the questionnaire were grouped into categories of similar items. The statistical package SPSS, version 9, a was used to analyze the results. RESULTS Sample Characteristics A total of 130 people with stroke participated in the study: 50 subjects from each of 2 centers and 30 subjects from the third center. Based on the findings of a community-based study of stroke in Auckland, New Zealand, this figure is likely to represent about one third of stroke survivors who present with a moderate or severe motor deficit 1 month after stroke and who live at home. 19 Table 1 describes the characteristics of the total sample and the amount of PT they received after their stroke. physical therapists visited participants in their homes if a home assessment was required at discharge, but intensive rehabilitation was not offered in the home environment. Participants receiving PT (n 15) were seen on average twice a week in the PT department the frequency of visits was more widely dispersed for those who received community input. The majority of the sample were of European descent, and 3 people identified themselves as Maori. Mobility Outcomes Mobility outcomes for the group as a whole were high for a stroke sample (table 2). Gait velocity for the 15 people who did not require PT was considerably higher than the overall mean,

3 236 COMMUNITY AMBULATION AFTER STROKE, Lord Table 1: Study Sample Characteristics n (%) Men 71 (54.5) First stroke 108 (83) PT 115 (88.4) Mean SD Range Age (y) Weeks since onset Inpatient PT (wk) Outpatient PT (wk) PT (wk) Abbreviation: SD, standard deviation. at 81.4m/min, which is within normal limits for the age group studied. 20 The overall mean gait velocity for the study sample was 56.2m/min, and 63 subjects (48.5%) achieved an RMI score of 14, which describes one s ability to go up and down 4 steps with no rail, using an aid if necessary. The highest score of 6 on the FAC was achieved by 94 subjects (72.3%) and indicates one s ability to walk independently on nonlevel and level surfaces, inclines, and stairs. Distances walked on the treadmill were low, even for those who did not require PT, with 58 (44.6%) able to walk the maximum 300m. Gait speed had a normal distribution and no ceiling effect, but this was not seen in the other mobility measures. A significant relationship existed for RMI, treadmill, and gait velocity, with values ranging from.54 to.76 (Spearman, P.01) (see table 3). Based on the self-reported levels of mobility, 19 participants (14.6%) were not walking outside the home (group 1), 22 participants (16.9%) were walking as far as the letterbox (group 2), 10 participants (7.6%) were walking in their immediate outside environment (eg, down the road, around the block) (group 3), and 79 participants (60.7%) were able to walk to shopping venues and/or places of interest (group 4). Of the participants who had not received PT, 13 were in group 4 and 2 were in group 2 because of their reported comorbidities. If participants who did not receive therapy were removed from the analysis and the 115 participants who received PT were considered separately, 20 participants (15.4%) were in group 2 and 66 (50.8%) were in group 4. When the 4 groups were considered with respect to gait velocity, RMI, distance on the treadmill, and the FAC (see table 2), each variable showed significance at P equal to.000 (1-way ANOVA). However, gait velocity appeared to be the Table 3: Spearman Between 3 Variables* in 130 People With Stroke Treadmill RMI Gait velocity (m/min) RMI.63 *Significant at P.01 (2-tailed). most discriminatory measure for the 4 levels of ambulation with confidence intervals that overlapped for groups 1 through 3 and intervals that were discrete for group 4 (see fig 1). A stick was the most common walking aid and was used by 46 (41.4%) of the participants who walked outside. To make some comparison with earlier findings, community ambulation was also considered in relation to earlier published research (table 4). Of the 87 (66.9%) participants who achieved the minimum threshold of 48m/min for community ambulation described previously, 2,5 15 (17.2%) were either not walking outside or walking only as far as the letterbox. Conversely, 13 people (10%) with gait velocities below 48m/min were able to walk to the shops and/or places of interest. The 23 participants (17.6%) who achieved a gait velocity of 79m/min, the threshold set by Lernier-Frankiel et al 1 for community ambulation, were all in group 4, the highest category of ambulators. Thirty-nine people (30%) achieved 300m of treadmill walking, a minimum gait velocity of 48m/min, and an FAC score of 6, which are the 3 physical parameters that Hill et al 5 have considered to be the minimum level of mobility required for community ambulation. Of this group, 35 (89.7%) were in group 4: walking to the shops or places of special interest. The Questionnaire The ability to get out and about in the community was considered essential by 53 participants (40.8%), very important by 44 (33.8%), important by 24 (18.5%), mildly important by 8 (6.2%), and not important by 1 (0.8%). Table 5 summarizes the responses to question 2, which provided the basis for a new and more complete definition of community ambulation: independent mobility outside the home, which includes the ability to confidently negotiate uneven terrain, private venues, shopping centers and other public venues. Shopping centers and shopping malls were reported most frequently and most often as the first choice of the 3 places participants reported having an interest in visiting. In response to question 5, which sought information on the effects of dependency on others for assistance, 29 people All Participants (N 130) Table 2: Mobility Outcomes for All Participants Group 1 (n 19) Group 2 (n 22) Group 3 (n 10) Group 4 (n 79) No PT (n 15) Mean m/min % CI Mean treadmill distance (m) % CI to 108.8* Median FAC score Mean FAC score (95% CI) 5.5 ( ) 4.2 ( ) 5.3 ( ) 5.5 ( ) 5.9 (5.9 6) ( ) Median RMI score Mean RMI score (95% CI) 12.4 ( ) 7.5 ( ) 11.1 ( ) 12.3 ( ) 14 ( ) 14.4 ( ) Abbreviation: CI, confidence interval. *The 95% CI for treadmill exceeds the lowest possible value as a result of the CI arithmetic process. The difference among groups 1 through 4 for m/min, treadmill, FAC score, and RMI score was significant (P.000 for all variables).

4 COMMUNITY AMBULATION AFTER STROKE, Lord 237 Table 5: Frequency of Responses (N 130) to the Question: Which Places Outside the Home Did You Like to Get to Before Your Stroke? Venues/Activities Frequency (%) Shopping centers/malls 83 (63.8) Visiting friends/families 65 (50) Organized social activities (eg, lawn bowling, church) 61 (46.9) Work 37 (28.4) Leisure activities/hobbies 33 (34) Beach 4 (3) Doctor 2 (1.5) NOTE. Respondents were asked to list a maximum of 3 types of places, in order of preference. Fig 1. Boxplot to describe the number of participants in each community level, based on gait velocity, showing median, interquartile range, and smallest and largest scores. Legend: 1.0, not walking; 2.0, as far as the letterbox; 3.0, immediate environment; 4.0, to malls and/or places of interest. (22.3%) did not require any assistance in accessing the community. Eighty-two participants (63%) reported dependency on carers, friends, and family members for assistance with transport, and 20 people (15.3%) reported the need for more general help, such as supervision with transfers, mobility, and assistance with equipment. Just under half of the participants (45%) commented that their dependency did not cause any particular problems for either themselves or for those assisting them, whereas 50.4% commented on their own feelings of dependency, especially as they were related to the loss of driving ability (43.5%). The difficulty associated with the use of public transportation was identified by participants who were dependent on others for transportation. These people cited lack of confidence, unfamiliarity, unreliability, poor proximity, and cost of taxis as negative factors. Table 4: Comparison of Ambulation Level Attained Compared With Thresholds Defined in Previous Research Group 3* (n 10) Group 4 (n 79) Perry et al 2 Most limited community walkers 25m/min 0 2 Limited community walkers 35m/min 2 11 walkers 48m/min 6 43 Hill et al 5 48m/min 6 43 Lernier-Frankiel et al 1 79m/min 0 23 *Group 3 walked in the immediate environment. Group 4 walked to shops and/or places of special interest. Two subjects walked below the threshold of 25m/min. DISCUSSION This study describes the level of ambulation for a cohort of home-dwelling people with stroke on discharge from PT services. Our findings suggest that although mobility outcomes were good overall, almost one third of the people were still unable to walk unsupervised in their communities. Of those who did achieve independent community ambulation, most were reluctant, or unable, to use public transportation and therefore had to rely on their carers if they were to go outside the home. Although the impact of these restrictions on QOL was not measured in this study, the importance people attached to the ability to leave the home was unequivocal. Our findings may not generalize to the stroke population because the sample was opportunistic and included only people discharged to a private home, which means they had comparatively good outcomes. They presented with cognitive and motivational levels sufficient for ongoing participation in a PT program. It is also important to note that study participants were assessed very soon after their final PT session and were, therefore, most probably at their peak physical and functional level a level that over time might not be sustained. 21,22 A description of community ambulation was developed based on the participants views of which environments matter most to them. Most previous definitions of community have been based on community settings perceived by researchers to be relevant to North American town and city-dwellers. Only 1 survey 1 has drawn on the views of people with stroke; that survey asked 30 people between 40 and 70 years of age about the 10 community sites to which they traveled most frequently. The destinations chosen were urban-based (eg, supermarket, drugstore, shopping mall), which is consistent with the highestranking activity reported in this study and higher than social outings and organized social activities. Furthermore, we tested definitions of community for a population that included not only city but also small town and rural dwellers in a country (New Zealand) where many people identify with nonurban environments. The definition of community ambulation derived from this study offers a standardized framework of earlier broad definitions of community ambulation. In this study, gait velocity was a useful and discriminative clinical measure. However, participants showed a relatively high level of mobility. For people with lower functional status, gait velocity is less discriminative than other clinical measures. 23,24 The emphasis placed on obtaining optimal gait velocity during rehabilitation, if outside walking dexterity is an agreed goal, appears to be justified as at least an important component of PT. However, the wide variation of gait speed within each group of ambulators and the achievement of independent community ambulation for some people despite low gait velocities adds weight to the argument that the skills required for the task are much more extensive and complex than gait speed alone. 6,7

5 238 COMMUNITY AMBULATION AFTER STROKE, Lord Our study results suggest that gait speed may be an appropriate primary endpoint in clinical trials for interventions to improve community mobility. However, of itself, it is not sufficient as a single measure of community ambulation and should be augmented by self-report and the use of measures of social integration and community participation. Endurance was less discriminating than gait velocity, perhaps because of the use of the treadmill to measure this variable. Few participants were familiar with the treadmill, and most were concerned about walking on it without an opportunity to practice the task. The treadmill was chosen because it was used as a criterion for community ambulation in a previous study 5 and because it was easier to standardize the testing procedure for the 3 centers than the 6-minute walking test, which has also been used to measure endurance in stroke survivors. 25 Although a recent study 7 indicated that endurance may be less important to successful community ambulation than other factors, it is plausible that the ability to walk a reasonable distance contributes to more satisfying ambulation. Aerobic fitness, which has been shown to be important after stroke, 26,27 has not been measured with respect to community ambulation. Further research is required to determine optimal levels of endurance and fitness for comfortable community ambulation. The ability outcomes for our sample exceeded outcomes in earlier studies, although it is difficult to comment on the comparability of the samples that were drawn from both inpatient and home-dwelling stroke populations. In the study by Lernier-Frankiel et al, 1 all 10 subjects were ambulating in the community, with a mean gait speed of 48.6m/min, although they were unable to cover the distances required for many of the places they wished to visit. In the study by Perry et al, 2 only 26 of 147 (17.6%) patients attained the highest level of community ambulation (full community integration) with a threshold gait speed of 48m/min, although 78 (53%) achieved at least some level of community ambulation. The study included people who had been discharged from hospitals for at least 6 weeks (discharge destination was not specified), for whom the onset of stroke was greater than 3 months, and who could walk at least 6m. The amount of PT given was not detailed. In the Hill audit, % of the 107 patients assessed were discharged home and the remainder were discharged to special accommodations or nursing homes, whereas our participants were all living at home. CONCLUSIONS Despite favorable mobility outcomes having been achieved for this home-dwelling cohort of people with stroke, questions remain as to why such outcomes are not necessarily linked to good levels of community ambulation. It may be that rehabilitation efforts need to be refined to achieve a level of mobility sufficient for these people to be adept community ambulators. Research is required to identify treatment approaches that maximize community ambulation based on the attributes required for the task, of which gait velocity may be one. Acknowledgments: We thank the research assistants who collected data at each of the centers: Jackie Pithie at Princess Margaret Hospital, Leanne Robinson at Waikato Hospital, and Nik Yarrall at Hutt Hospital, and the physical therapists at the 3 centers. APPENDIX 1: QUESTIONNAIRE Ambulation After Stroke: A Multicenter Survey 1. How important is it for you to be able to get out of the home? Not important Mildly important Important Very important Essential 2. Which places outside the home did you like to get to before your stroke? (Please list a maximum of 3 types of places, in order of preference.) 3. Are you able to get out and about, by yourself, without physical assistance or supervision from anyone? Outdoors (eg, as far as the letterbox) but no farther (go to question 5) Yes (Give up to 3 examples.) No (Go to question 5.) 4. Do you require special equipment to achieve this? (If yes, please state type of equipment, for example, wheelchair, scooter, type of walking aid.) 5. Does the assistance you require to get out and about cause any problems to you or your carers? (If yes, please identify.) 6. Do you have any other comments you would like to make regarding getting out of the home? References 1. Lernier-Frankiel M, Vargas S, Brown M, Krusell L, Schoneberger W. Functional community ambulation: what are your criteria? Clin Manag Phys Ther 1986;6(2): Perry J, Garrett M, Gronley J, Mulroy S. Classification of walking handicap in the stroke population. Stroke 1995;26: Pound P, Gompertz P, Ebrahim S. A patient-centred study of the consequences of stroke. Clin Rehabil 1998;12: Robinett C, Vondran M. Functional ambulation velocity and distance requirements in rural and urban communities. Phys Ther 1988;68: Hill K, Ellis P, Bernhardt J, Maggs P, Hull S. Balance and mobility outcomes for stroke patients: a comprehensive audit. Aust J Phys Ther 1997;43: Patla A, Shumway-Cook A. Dimensions of mobility: defining the complexity and difficulty associated with community mobility. J Aging Phys 1999;7: Shumway-Cook A, Patla A, Stewart A, Ferrucci L, Ciol M, Guralnik J. Environmental demands associated with community mobility in older adults with and without mobility disorders. Phys Ther 2002;82: Bethoux F, Calmels P, Gautheron V. Changes in the quality of life of hemiplegic stroke patients with time. Am J Phys Med Rehabil 1999;78: Wilkinson P, Wolfe C, Warburton F, et al. A long-term follow-up of stroke patients. Stroke 1997;28: Astrom M, Asplund K, Astrom T. Psychosocial function and life satisfaction after stroke. Stroke 1992;23:

6 COMMUNITY AMBULATION AFTER STROKE, Lord Mayo N, Wood-Dauphine S, Ahmed S, et al. Disablement following stroke. Disabil Rehabil 1999;21: Sabari J, Meisler J, Silver E. Rehabilitation in practice. Disabil Rehabil 2000;22: Grimby G, Andren E, Daving Y, Wright B. Dependence and perceived difficulty in daily activities in community-living stroke survivors 2 years after stroke. Stroke 1998;29: Asplund K, Tuomilehto J, Stegmayr B, Wester P, Tunstall-Pedoe H. Diagnostic criteria and quality control of the registration of stroke events in the MONICA project. Acta Med Scand Suppl 1998;728: Wade D. Measurement in neurological rehabilitation. Oxford: Oxford Univ Pr; Holden M, Gill K, Magliozzi M. Gait assessment for neurologically impaired patients. Phys Ther 1986;66: Collen F, Wade D, Robb G, Bradshaw C. The Rivermead Mobility Index: a further development of the Rivermead Motor Assessment. Int Disabil Stud 1991;13: Collen F, Wade D, Bradshaw C. Mobility after stroke: reliability of measures of impairment and disability. Int Disabil Stud 1990; 12: Bonita R, Beaglehole R. Recovery of motor function after stroke. Stroke 1988;19: Craik RL, Dutterer L. Spatial and temporal characteristics of foot fall patterns. In: Craik RL, Oatis CA, editors. Gait analysis theory and application. St Louis: Mosby; p Paolucci S, Grasso M, Antonucci G, et al. Mobility status after inpatient stroke rehabilitation: 1-year follow-up and prognostic factors. Arch Phys Med Rehabil 2001;82: Wade D, Collen F, Robb G, Warlow C. Physiotherapy intervention late after stroke and mobility. BMJ 1992;304: Richards CL, Malouin F, Dean C. Gait in stroke: assessment and rehabilitation. Clin Geriatr Med 1999;15: Salbach N, Mayo N, Higgins J, Ahmed S, Finch L, Richards C. Responsiveness and predictability of gait speed and other disability measures in stroke. Arch Phys Med Rehabil 2001;82: Dean C, Richards C, Malouin F. Task-related circuit training improves performance of locomotor tasks in chronic stroke: a randomized, controlled pilot trial. Arch Phys Med Rehabil 2000; 74: Macko RF, DeSouza CA, Tretter LD, et al. Treadmill aerobic exercise training reduces the energy expenditure and cardiovascular demands of hemiparetic gait in chronic stroke patients. A preliminary report. Stroke 1997;28: Teixeira-Salmela L, Olney S, Nadeau S, Brouwer B. Muscle strengthening and physical conditioning to reduce impairment and disability in chronic stroke survivors. Arch Phys Med Rehabil 1999;80: Supplier a. SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL