PEOPLE WITH STROKE often have difficulties changing

Transcription

1 2156 ORIGINAL ARTICLE The Four Square Step Test is a Feasible and Valid Clinical Test of Dynamic Standing Balance for Use in Ambulant People Poststroke Jannette M. Blennerhassett, PhD, Victoria M. Jayalath, BAppSci (PT) ABSTRACT. Blennerhassett JM, Jayalath VM. The Four Square Step Test is a feasible and valid clinical test of dynamic standing balance for use in ambulant people poststroke. Arch Phys Med Rehabil 2008;89: Objective: To examine if the Four Square Step Test (FSST), a previously reported clinical test of dynamic standing balance, which involves stepping over obstacles and turning, was a feasible and valid test, and sensitive to change during stroke rehabilitation. Design: Prospective observational cohort study over a 4-week duration. Setting: Rehabilitation hospital. Participants: People with stroke (N 37) who could walk at least 50m with minimal assistance were recruited consecutively when attending physical therapy during rehabilitation. Interventions: Not applicable. Main Outcome Measures: Dynamic standing balance was examined at 2 weekly intervals using 2 clinical tests: the FSST and the Step Test. Falls events were monitored using a falls diary and by an audit of medical histories. Results: Strong agreement was observed between performance scores for the FSST and Step Test obtained within the same testing session (intraclass correlation coefficient 3,k,.94.99). A moderate to strong inverse relationship (Spearman.73 to.86) was observed between the FSST and Step Test scores at each assessment. Scores from both tests revealed significant improvements in dynamic balance across the 4-week period (P ). Five of the participants reported falls during the study. These 5 people had low scores for both clinical tests and difficulty clearing their foot when stepping over objects in the FSST. Conclusions: The FSST is a feasible and valid test of dynamic standing balance that is sensitive to change during stroke rehabilitation. Key Words: Accidental falls; Outcome and assessment (health care); Rehabilitation; Stroke; Walking by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation From the Physiotherapy Department, Austin Health: Royal Talbot Rehabilitation Centre, Kew (Blennerhassett); and Physiotherapy Department, Donvale Rehabilitation Hospital, Donvale (Jayalath), Victoria, Australia. 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 authors or on any organization with which the authors are associated. Correspondence to Jannette M. Blennerhassett, PhD, Physiotherapy Department, Austin Health: Royal Talbot Rehabilitation Centre, 1 Yarra Blvd, Kew, Victoria, 3101, Australia, Jannette.Blennerhassett@austin.org.au. Reprints are not available from the author /08/ $34.00/0 doi: /j.apmr PEOPLE WITH STROKE often have difficulties changing direction and stepping to negotiate obstacles when walking. 1,2 This can compromise safe walking within the hospital, home, and community, 3 and can increase the likelihood of falls. 4 The incidence of falls among stroke survivors is estimated to be high with 73% of stroke survivors having a fall within 6 months postonset, and 60% reporting a fall shortly after discharge from hospital. 3-5 In addition to the risk of injury, a fall can lead to fear of falling and reduced confidence with mobility. 4 This could lead to a vicious cycle where the person becomes less active, and at risk of decline in physical capacity and further falls. 4 After stroke, impaired dynamic standing balance can limit the ability to walk safely and increase the risk of having a fall. 1,3,6 Because balance control is adapted to suit the task and environment, and involves many factors, 7 clinicians examine dynamic standing balance using a range of clinical tests. 8,9 Tests that include tasks that are relevant for daily life, such as stepping over objects and turning, and also have the potential to identify people at risk of falling, would contribute valuable information to clinical decision-making during stroke rehabilitation. The FSST 8 is a clinical test of dynamic standing balance that may be suitable for this purpose. The FSST was designed to examine ability to step over small objects and change direction within a clinical setting, and requires minimal space, equipment, and time. The FSST was found to be reliable 8,10 and capable of discriminating between nonfallers and fallers in older adults, 8 transtibial amputees, 11 and people with vestibular dysfunction. 10 However, it is not known whether the FSST is sensitive to change over time. 8,10 The FSST also seems advantageous to use with people with stroke because performance during the test requires use of visual perception and cognitive skills in addition to physical abilities 8,12,13 that are relevant to walking indoors and outdoors. However, the FSST has not been validated for use with people poststroke. The purpose of this study was to evaluate the clinical utility of the FSST during stroke rehabilitation. The specific research questions examined were: (1) Is the FSST feasible to use in people with stroke? (2) Is the FSST sensitive to change across a 4-week period of rehabilitation? (3) Can concurrent validity of the FSST be shown? To explore this, we examined if performance on the FSST poststroke was related to that described by another clinical test of dynamic standing balance, the Step Test. 14 The Step Test was selected because it also involves rapid stepping and has been established for use with CI ES FSST ICC List of Abbreviations confidence interval effect size Four Square Step Test intraclass correlation

2 A TEST OF DYNAMIC STANDING BALANCE POSTSTROKE, Blennerhassett 2157 people with stroke. 9 (4) Does the FSST have potential to provide clinically relevant information when assessing falls risk in people with stroke? METHODS Participants The participants were a consecutive sample of people with stroke who met the study criteria and who attended physical therapy at Austin Health: Royal Talbot Rehabilitation Centre, a public facility that provides therapy to inpatients and outpatients in Melbourne, Australia. All potential candidates were informed about the study and then invited to participate. No person declined to be involved in the study. To be eligible, participants needed to be able to walk at least 50m with minimal assistance, as indicated by a score of 4 to 7 on the walking item of the FIM instrument. 15 People were not excluded if they had language, visual, or cognitive impairments. Thirty-seven people participated in the study (table 1). Because some participants were discharged from rehabilitation during the study, data are available for 28 participants at 2 weeks and 20 participants at 4 weeks. The study was approved by the Austin Health Human Research Ethics Committee. Procedure The study used 2 clinical tests (the FSST 8 and the Step Test 14 ) to examine dynamic standing balance in people with stroke across a 4-week period of rehabilitation. The FSST and Step Test were conducted concurrently on 3 occasions (ie, at baseline, 2 weeks, and 4 weeks). A physical therapist who was not involved in the participant s rehabilitation conducted the tests. Where possible, the same physical therapist performed all testing for each participant. To control for potential order effects within testing, participants who were assigned a study code with an even number commenced the test with the FSST. Those assigned with an odd number were tested first with the Step Test. After testing, the physical therapist s and participant s impressions about the safety and clinical use of each test were noted. During the 4-week study, each participant was asked to complete a falls diary. Falls events for all participants were also monitored by the treating physical therapist and an audit of the medical history. Table 1: Details of Participants at Commencement of the Study Variable Detail Number (n) 37 Sex (n, men:women) 25:12 Age (y, mean [range]) 53 (23 75) Time since stroke (d, median [range]) 66 (9 1094) Affected side (n) Right 14 Left 16 Bilateral 7 Scores for FIM walking item (n) Rehabilitation status (n) Inpatient 24 Outpatient 13 Outcome Measures Dynamic standing balance was measured using 2 clinical tests: the FSST 8 and the Step Test. 14 Each test was demonstrated and then conducted using standardized instructions and procedures, as detailed below. Participants were permitted 1 practice trial before trials were scored. The Step Test has been found to be reliable 14 and sensitive to change during stroke rehabilitation. 9 During the Step Test, the person was instructed to maintain their balance on 1 leg, while placing the whole of the opposite foot on and off a small step (7.5cm in height, positioned 5cm directly in front) repeatedly for 15 seconds. A trained physical therapist closely supervised each trial. No gait aid was permitted during testing. For each trial, the score recorded was the number of steps completed in 15 seconds, or before the person required assistance to maintain their balance. People who could not perform the test received a score of 0. Each participant was scored for 2 trials of the Step Test for both the right and left side. The FSST has been found to be a reliable and valid test of dynamic standing balance for use in people who are older, 8 have transtibial amputation, 11 or have vestibular dysfunction, 10 as mentioned previously. During the FSST, the person sequentially steps over 4-single point walking canes placed in a cross configuration on the ground. Initially, the person steps around the 4 squares in a clockwise direction, then reverses in a counterclockwise direction. The instructions supplied by Dite and Temple 8 were slightly modified (as shown in italics below) to reduce complexity and emphasize the need for safety. The instructions were: Try to complete the sequence as fast and as safely as possible without touching the canes. Both feet need to make contact with the floor in each square. Participants that used a walking cane were allowed to use it during testing. During demonstration of the FSST, the examiner faced forward during the entire stepping sequence. However, participants often chose a different strategy, such as turning to face each cane to ensure that they avoided touching the canes. Such strategies were permitted during testing, and noted accordingly. A trained physical therapist closely supervised all trials. At times, a second person was assigned to operate the stopwatch so that the physical therapist could provide assistance to prevent a fall. The FSST was scored by the time taken to complete the task (ie, from initial contact of first step, to initial contact of final step). Four attempts at the FSST were permitted to complete 2 successful trials. Unsuccessful trials were when the participants did not complete the sequence successfully, lost their balance and required assistance, or made contact with the canes. Unsuccessful trials were noted but the time taken was not recorded. At the end of each testing session, the participant and physical therapist were asked to record if they preferred to use: the FSST; the Step Test; either test; or neither test to evaluate dynamic standing balance in the clinical setting. Reasons for preferences were also recorded. Data Analysis The level of agreement between the 2 repeated scores obtained for each test during the 3 assessment points was analyzed using ICC. Although both the FSST and Step Tests yielded ratio data, nonparametric statistical analyses were performed to address the remaining research questions. The reason for this was that the data were not normally distributed and relationships observed between scores for each test were not linear in nature. Differences between first and second test scores were examined for potential practice effects using the Wilcoxon signed-rank test. Similarly, differences arising from

3 2158 A TEST OF DYNAMIC STANDING BALANCE POSTSTROKE, Blennerhassett the order in which the FSST and Step Tests were conducted were examined using the Mann-Whitney U test. As no order effects were found (see Results), performance for each test was summarized by the average score for the 2 test trials. The only exception to this was when only 1 successful score was obtained for the FSST. The available data obtained across the 4-week period were then summarized for descriptive purposes. The relationship between performance on the FSST and Step Test at each assessment point were inspected visually using scatterplots and analyzed statistically using Spearman. For participants who were not discharged from rehabilitation and were tested at 2 and 4 weeks, differences in test performance over time were statistically analysed using the Wilcoxon signed-rank test. We have also presented ESs and 95% CIs for the mean differences in paired data to help interpret the clinical change recorded by the tests. The ESs calculated were interpreted using the guidelines proposed by Cohen 16 (ie, d.2 [small]; d.5 [medium]; and d.8 [large]). All analyses were performed using SPSS a version RESULTS Agreement Between Scores for Repeated Trials Performance scores for the FSST and Step Test across the 4-week period are summarized for the available sample in table 2. There was excellent agreement between the scores obtained during the 2 repeated trials for both tests (ICC 3,k.94.99) at each assessment point. In addition, no significant difference was found between the first and second successful trial score for either test (Z 1.40 to.21, P.16.84), indicating that a practice effect was not evident. The order in which tests were conducted did not significantly influence group performance (U.04 to.94, P.36.97). Concurrent Validity for the FSST and Step Test We observed a moderate to strong inverse relationship between FSST and Step Test performance scores across the 3 assessments (table 3). The scores for the FSST and Step Test at the initial assessment are also presented in figure 1. That figure shows that people who required more time to complete the FSST also had lower scores for the Step Test. However, the relationship observed was not linear. Figure 1 also displays performance scores for the stroke sample relative to the fifth percentile score for community-dwelling healthy older adults reported for the FSST (ie, 65y of age) 8 and Step Test (ie, Table 3: Relationship Observed Between FSST and Step Test Performance Tests Examined Initial 2 Weeks 4 Weeks Step test: right and left stance.86*.92*.96* FSST and step test right stance.86*.78*.81* FSST and step test left stance.78*.73*.84* NOTE. Spearman correlation coefficients are reported. *P y of age). 14 For the FSST and Step Test respectively, only 15 (40%) and 21 (57%) of the 37 people with stroke achieved scores equivalent to healthy older adults at the initial assessment. For the 20 participants examined at 4 weeks, the number who achieved scores equivalent to healthy older people was 7 (35%) for the FSST, and 12 (60%) for the Step Test. Change Over Time Participants who had repeated test scores, (ie, not discharged or received a FSST score at the earlier test) were observed to improve significantly for both tests across the 4-week interval. For instance, participants took less time to complete the FSST and performed a higher number of repetitions for the Step Test after 2 or 4 weeks rehabilitation, as anticipated. Nevertheless, the ESs calculated for both tests of dynamic standing balance were found to be small (table 4). The only exception was a medium ES found for the change between baseline and 4 weeks for the right Step Test. Details of the statistical significance, and the observed ES and 95% CI for the paired differences in test scores are provided in table 4. Feasibility for Clinical Use The physical therapists reported that both tests took less than 5 minutes to complete and were practical for use in the clinical setting. No fall or injury occurred during testing. Nevertheless, many participants found the FSST difficult to perform, and needed close supervision to be tested by trained staff. For example, between 40% and 62% of participants had unsuccessful trials for the FSST at the 3 assessment points. The main reasons recorded for not being able to complete the FSST were difficulty maintaining balance and trouble clearing the canes. Only 2 of the 37 participants had unsuccessful trials due to difficulty following or remembering the test instructions. At the Table 2: Summary of FSST and Step Test Performance at the 3 Assessment Points Test Test Occasion Initial (N 37) 2 Weeks (n 28) 4 Weeks (n 20) Step Test (rep/15s) Left stance Mean SD Range Right stance Mean SD Range FSST (s) Mean SD Range Participants with unsuccessful trials n (%) 23 (62%) 11 (39%) 8 (40%) Participants unable to be scored n (%) 5 (14%) 3 (11%) 3 (15%) Abbreviation: rep, repetition.

4 A TEST OF DYNAMIC STANDING BALANCE POSTSTROKE, Blennerhassett 2159 Fig 1. Comparison of FSST and Step Test scores at the initial assessment (n 32). The dashed lines indicate the fifth percentile score (ie, poorest value) documented for community dwelling older people for the FSST 8 and Step Test. 14 initial test, 5 participants failed all 4 attempts of the FSST and a score could not be obtained. The FSST remained difficult for those 5 participants, and 3 of those people continued to fail all trials at the 2- and 4-week assessment points. These people also found the Step Test difficult and as a consequence had low Step Test scores for both right and left stance (ie, scores 0 4). Therapists and Patients Impressions of the Tests In general, the physical therapists indicated that both tests were clinically useful. No apparent bias was reported for either test. For instance, the physical therapists preferences across the 3 occasions of testing are as follows: 17% to 20% for the Step Test; 27% to 35% for the FSST; 30% to 42% either test; 0% to 3% neither test; and 11% to 16% not recorded. The reason given for specific preferences appeared to be based on the information gained from observing the participant s performance, and ease or safety concerns when conducting the tests. For instance, the Step Test was described as easier and safer to perform than the FSST for people who had marked balance dysfunction and leg weakness, or were impulsive. One therapist commented that the repeated action examined during the Step Test also provided a means to carefully observe the coordination of movement for the stepping leg, and ability to balance on the stance leg. Preferences were given to the FSST when the physical therapist felt they gained relevant information about the person s ability to plan, step, and change directions to clear the obstacles. Of the 37 participants, people tended to prefer the FSST (38%) because they felt that the test was highly relevant for daily life, and examined skills that were challenging to perform. Eight percent of participants preferred the Step Test because they felt it was easier to perform than the FSST, and reported greater confidence during testing. No preference for either test was given by 21% of participants, but 3% did not like being tested on either test, because both were perceived as challenging. Preferences were not recorded for 30% of participants. Participants Who Fell and Their Performance on the Clinical Tests Only 7 (18%) of the 37 participants completed the falls diary during the study period. However, monitoring by the treating physical therapist indicated that 5 participants had a fall during the 4-week period. Interestingly, 4 of the 5 fallers were outpatients and the only inpatient had the fall at home when on weekend leave. The medical audit concurred that no participant had a fall when in the hospital setting over the 4-week period. For each fall, the person reported that they tripped over an object or caught their foot on the ground when rushing or changing direction. All of the 5 fallers had unsuccessful trials during the FSST, and 2 could not achieve a FSST score at the initial test. At the initial assessment, the 3 fallers who achieved a FSST score took between 15.6 to 27.7 seconds to complete the test. The lower score for the Step Test at the initial assessment for the 5 fallers were 0, 3.5, 4.5, 7, and 11. Table 4: Change Observed for FSST and Step Test Scores During Stroke Rehabilitation Study Interval Test Value Baseline to 2 Weeks 2 Weeks to 4 Weeks Baseline to 4 Weeks FSST n 24 n 17 n 16 P *.01* d % CI (s) 0.4 to to to 10.7 Step Test n 28 n 20 n 20 Left stance P d % CI (rep/15s) 2.3 to to to 1.9 Right stance P d % CI (rep/15s) 3.3 to to to 2.1 NOTE. Values reported are number in subset analysed (n), Cohen s d ES, 95% CIs for the mean difference in paired scores. Abbreviation: Rep, repetitions. *P.05. P.01.

5 2160 A TEST OF DYNAMIC STANDING BALANCE POSTSTROKE, Blennerhassett DISCUSSION The FSST was found to be a feasible and valid clinical test of dynamic standing balance to use with people with stroke who are at least ambulant with minimal assistance. The FSST was also able to detect change in dynamic standing balance over a 2- or 4-week period of stroke rehabilitation. Nevertheless, the ESs calculated for change in FSST scores were found to be small. Although the FSST was challenging for some participants, physical therapists who were trained to provide physical assistance were able to conduct the test without incident. Moreover, the FSST offers a standardized means to observe how people with stroke plan and organize movement to step over small obstacles and change directions, which may assist to make clinical decisions about balance control to plan therapy and discharge from hospital. The strong relationship observed between performance for the FSST and Step Test poststroke demonstrates that the FSST has concurrent validity as a test of dynamic standing balance. In addition, the FSST and Step Test could both detect change during stroke rehabilitation, and the ESs calculated for those changes were found to be similar but relatively small for both tests. Both tests were reported to be quick and practical to use in a clinical setting, and could be administered without incident by physical therapists. The Step Test was described as easier than the FSST to use with people who were impulsive or required assistance for safety. These types of difficulties often contributed to the unsuccessful trials of the FSST, and the inability to achieve a FSST score in 15% of the participants. However, both the therapist and patient felt that important clinical information about safety and dynamic standing balance control was gained from observing performance on the FSST. Nevertheless, the inherent risks observed when conducting a test that involves obstacle negotiation and turning in people with stroke reinforces that the FSST should only be administered by staff with adequate training, as recommended by Dite and Temple. 8 Observation of FSST performance highlighted that people with stroke who can ambulate require more time to step safely over objects, and yet often have difficulties maintaining balance and avoiding contact with a small obstacle. These difficulties are likely to have an impact on the person s capability to walk independently within the hospital, home, and community. 3,17,18 The observed performance for FSST poststroke contrasts with that described for community-dwelling older adults, especially when considering that unsuccessful trials of the FSST were not reported for older people who were classified as a multiple fallers. 8 At baseline in our study, 62% of participants with stroke had unsuccessful attempts of the FSST and only 41% achieved FSST scores equivalent to those of communitydwelling older adults. Although participants with stroke made significant improvements for FSST during the study, the ability to step over objects safely remained problematic. For instance, at 4 weeks, 40% of participants continued to record unsuccessful FSST trials, and 65% achieved FSST scores equivalent to those of healthy older people. However, these comparisons using published data for community-dwelling people over 65 years of age 8 are likely to underestimate the true extent of balance limitation because 81% of our sample were under 65 years of age, and younger people are reported to require less time to complete the FSST. 10 Normative data for FSST performance in people under 65 years are therefore needed to permit age-appropriate comparisons during clinical assessment. Clinically, it was interesting to observe that people who reported a fall during the study took more than 15 seconds to complete the FSST, or had difficulty clearing the obstacles during the test. The participants also indicated that their fall occurred when catching the foot while stepping, consistent with the difficulties observed during the FSST. This type of mechanism for falling is often reported after stroke. 2,3,17,18 It has been suggested that tests of dynamic standing balance that involve cognitive and physical processes, and are conducted under challenging situations may be advantageous when assessing balance limitation and risk of a fall after stroke. 13 Indeed, the FSST was designed to be a challenging clinical test of dynamic standing balance in that participants were asked to rapidly step and change directions to negotiate obstacles in a specific sequence. 8,11 Performance on the FSST is therefore likely to involve planning, cognitive, and movement processes. This combination of processes may have contributed to the finding that the FSST provided a stronger prediction of falls risk in community dwelling people over 65 years of age than the other clinical tests, such as the Timed Up & Go test, Step Test, and Functional Reach. 8 Moreover, a score of 15 seconds or more on the FSST was found to have an 86% chance of positively predicting a history of falls in community-dwelling older adults. 8 Interestingly, the 5 participants with stroke who reported a fall during this study, and could achieve a FSST score, took 15 seconds or more to complete the FSST. These 5 participants with stroke also scored 11 or less on the Step Test, which was a cut-off score found to offer a 63% chance of positively predicting falls history in older people. 8 In addition, all but 1 of the 5 stroke participants who reported a fall during this study scored 7 or less on the Step Test, which has been suggested as a score that may identify people with stroke at risk of recurrent falls after discharge from hospital. 6 Clinically, our preliminary findings suggest that a person with a stroke who fails an attempt of the FSST or takes 15 seconds or more to complete the FSST, or who scores 7 or less on the Step Test may be at a higher risk of falling. However, these clinical observations require more investigation before drawing conclusions about the value of the FSST in predicting people at risk of having a fall poststroke. Nevertheless, our observations support that this issue receives further investigation. Study Limitations There are several limitations that require consideration when interpreting the findings of this study. The sample size was small and was recruited by convenience from 1 rehabilitation setting. The available data were further restricted because patients were discharged from rehabilitation during the study. There was low compliance with maintaining a falls diary and most information about falls events was obtained from selfreports. Although an audit concurred that no participant had a fall while supervised in hospital, the reliability of recording of falls events within the hospital system is unknown. The period during which falls were monitored was relatively short, especially when incidence of falls is more commonly surveyed over a 6-5,6,8,11 to 12-month 19 period. The study did not investigate underlying impairments and processes that may have contributed to balance dysfunction and risk of falling within the sample. Effectiveness of treatment and management strategies to improve safety in mobility were also not evaluated. CONCLUSIONS The FSST appears to be a feasible and valid test to examine how people with stroke plan and organize movement to walk over small obstacles and turn within a confined space. Our findings support that clinicians who are trained to provide physical supervision are able to use the FSST during stroke

6 A TEST OF DYNAMIC STANDING BALANCE POSTSTROKE, Blennerhassett 2161 rehabilitation. We recommend that the person with stroke be able to walk at least 50m with minimal assistance before the FSST is considered. The FSST could be used in conjunction with other clinical assessments to make clinical decisions regarding dynamic standing balance and to monitor change over time. In particular, our preliminary data suggest that people who require more than 15 seconds or who fail a trial of the FSST may be at risk of having a fall. Acknowledgments: We thank all participants for their involvement in the study; the physical therapists at Austin Health-Royal Talbot Rehabilitation Centre who recruited and tested participants; and the undergraduate physical therapy students from The University of Melbourne who assisted in data management for the study. References 1. Said C, Goldie P, Patla A, Sparrow W, Martin K. Obstacle crossing in subjects with stroke. Arch Phys Med Rehabil 1999; 80: Said CM, Goldie PA, Culham E, Sparrow WA, Patla AE, Morris ME. Control of lead and trail limbs during obstacle crossing following stroke. Phys Ther 2005;85: Hyndman D, Ashburn A, Stack E. Fall events among people with stroke living in the community: circumstances of falls and characteristics of fallers. Arch Phys Med Rehabil 2002;83: Mackintosh SF, Hill K, Dodd KJ, Goldie P, Culham E. Falls and injury prevention should be part of every stroke rehabilitation plan. Clin Rehabil 2005;19: Yates JS, Lai SM, Duncan PW, Studenski S. Falls in communitydwelling stroke survivors: an accumulated impairments model. J Rehab Res Dev 2002;39: Mackintosh SF, Hill KD, Dodd KJ, Goldie PA, Culham EG. Balance scores and a history of falls in hospital predict recurrent falls in the 6 months following stroke rehabilitation. Arch Phys Med Rehabil 2006;87: Huxham FE, Goldie PA, Patla AE. Theoretical considerations in balance assessment. Aust J Physio 2001;47: Dite W, Temple V. A clinical test of stepping and change of direction to identify multiple falling older adults. Arch Phys Med Rehabil 2002;83: Hill K, Ellis P, Bernhardt J, Maggs P, Hull S. Balance and mobility outcomes for stroke patients: a comprehensive audit. Aust J Physiol 1997;43: Whitney S, Marchetti G, Morris L, Sparto P. The reliability and validity of the Four Square Step Test for people with balance deficits secondary to a vestibular disorder. Arch Phys Med Rehabil 2007;88: Dite W, Connor H, Curtis H. Clinical identification of multiple fall risk early after unilateral transtibial amputation. Arch Phys Med Rehabil 2007;88: Lamb SE, Ferrucci L, Volapto S, Fried LP, Guralnik JM. Risk factors for falling in home-dwelling older women with stroke. The women s health and aging study. Stroke 2003;34: Hyndman D, Ashburn A, Yardley L, Stack E. Interference between balance, gait and cognitive task performance among people with stroke living in the community. Disabil Rehab 2006;28: Hill K, Bernhardt J, McGann A, Maltese D, Berkovits D. A new test of dynamic standing balance for stroke patients: reliability, validity and comparison with healthy elderly. Physiol Can 1996; 48: Hamilton B, Granger C. Disability outcomes following inpatient rehabilitation for stroke. Phys Ther 1994;74: Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. New York: Academic Pr; Said CM, Goldie PA, Patla AE, Sparrow WA. Effect of stroke on step characteristics of obstacle crossing. Arch Phys Med Rehabil 2001;82: Mackintosh SF, Goldie P, Hill K. Falls incidence and factors associated with falling in older, community-dwelling, chronic stroke survivors ( 1 year after stroke) and matched controls. Aging Clin Exp Res 2005;17: Andersson AG, Kamwendo K, Seiger A, Appelros P. How to identify potential fallers in a stroke unit: validity indexes of 4 test methods. J Rehab Med 2006;38: Supplier a. SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL