Authority, Vancouver, BC, Canada

Transcription

1 This article was downloaded by:[canadian Research Knowledge Network] On: 31 March 2008 Access Details: [subscription number ] Publisher: Informa Healthcare Informa Ltd Registered in England and Wales Registered Number: Registered office: Mortimer House, Mortimer Street, London W1T 3JH, UK Disability & Rehabilitation Publication details, including instructions for authors and subscription information: Measurement properties of the Activities-specific Balance Confidence Scale among individuals with stroke Erica M. Botner a ; William C. Miller abc ; Janice J. Eng ab a School of Rehabilitation Sciences, University of British Columbia, b GF Strong Rehabilitation Research Lab, c Centre for Clinical Epidemiology and Evaluation at the Vancouver Coastal Health Authority, Vancouver, BC, Canada Online Publication Date: 18 February 2005 To cite this Article: Botner, Erica M., Miller, William C. and Eng, Janice J. (2005) 'Measurement properties of the Activities-specific Balance Confidence Scale among individuals with stroke', Disability & Rehabilitation, 27:4, To link to this article: DOI: / URL: PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: This article maybe used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

2 Disability and Rehabilitation, 2005; 27(4): RESEARCH PAPER Measurement properties of the Activities-specific Balance Confidence Scale among individuals with stroke ERICA M. BOTNER 1, WILLIAM C. MILLER 1,2,3, & JANICE J. ENG 1,2 1 School of Rehabilitation Sciences, University of British Columbia, 2 GF Strong Rehabilitation Research Lab, and 3 Centre for Clinical Epidemiology and Evaluation at the Vancouver Coastal Health Authority, Vancouver, BC, Canada (Accepted date August 2004) Abstract Purpose: To examine the reliability and validity of the Activities-specific Balance Confidence (ABC) Scale among individuals with stroke. Method: Descriptive measurement study using a 4-week test retest design. Data were collected at a tertiary rehabilitation centre from a community-dwelling sample of ambulatory older adults who sustained one stroke at least 1 year prior to the study. During the first measurement session, the total sample (n = 77) completed the ABC as well as the Berg Balance Scale (BBS) and gait speed. The reliability sample (n = 24), a subset of the larger data set, completed the ABC again 4 weeks later. Results: Internal consistency of the ABC was 0.94 and test retest reliability was ICC = 0.85 (95% CI, 0.68, 0.93). A moderate significant positive, linear correlation with both the BBS (r = 0.36, P ) and gait speed (r = 0.48, P ) was observed. A factor analysis using principal component analysis indicated that the ABC measures two components of balance self-efficacy (component 1 = perceived low-risk activities; component 2 = perceived high-risk activities). Conclusions: The ABC has acceptable measurement properties as demonstrated in this sample of individuals with stroke. This study provides further support for the use of the scale. Keywords: Stroke, balance confidence, ABC, reliability, validity Introduction Confidence in performing specific tasks without falling is important to maintaining independent functioning among individuals with stroke. Balance impairments, gait impairments and a history of falls [1 5] are risk factors that may put individuals with stroke at a greater risk of developing a fear of falling than older adults. Fear of falling results in a persistent concern about falling that restricts participation in everyday activities, limits functioning and may cause a risk of further falls [6 10]. Balance confidence, also called falls efficacy, is the mediator between fear of falling and such consequences [11]. There is little information on the prevalence of fear of falling among individuals with stroke; however, three studies suggest that individuals with stroke experience lower confidence in performing various tasks without falling than older adults [12 14]. In addition, a study conducted by Friedman et al. [15] revealed that women with a history of stroke were at increased risk of developing a fear of falling. It is important to identify individuals with stroke who have or are at risk of developing low falls efficacy so that appropriate interventions can be provided. Several studies suggest that interventions such as balance and gait training, the use of hip pads, and social support systems can significantly improve balance confidence [16 21]. Specifically for individuals with stroke, Rodriquez et al. [21] demonstrated that post-acute training of gait in hemiparetic individuals resulted reduced fear of falling. Using reliable and valid measures is critical in such studies in order to limit the threat of instrumentation error. The Activities-specific Balance Confidence (ABC) scale [22 24] was first developed for older adults in 1995 in order to address the lack of variation in scores derived from previously existing measures of Correspondence: William C. Miller, PhD., O.T.(C) CCEE, VGH Research Pavilion, 828 West 10th Ave, Vancouver, BC, Canada V5Z 1L8. Tel: , ext Fax: bcmiller@telus.net ISSN print/issn online # 2005 Taylor & Francis Group Ltd DOI: /

3 balance confidence, such as the Falls Efficacy Scale (FES) [25]. It is based on Bandura s concept of selfefficacy, which is a belief in oneself and in one s perceived ability in performing a specific task and that a low self-efficacy can result in avoidance of tasks [26]. Thus, the ABC measures confidence in performing various tasks without falling [23]. The ABC was chosen for this study because it has a wider range of difficulty among its items than previous measures [22 24], potentially making it more responsive for use in populations with varying levels of function such as individuals with stroke. The ABC has been used in various studies [22 24, 27 34] and the reliability and construct validity have been assessed among various populations including older adults [22,23], and individuals with lower limb amputations [30]. The results of these studies have consistently demonstrated excellent internal consistency and test retest reliability. When examined in healthy older adults, Cronbach s a was 0.96 and test retest reliability was r = 0.92 (P ) [23]. Magnitudes were similar among individuals with lower limb amputations [30]. In addition, convergent and discriminant validity were examined in healthy adults and as expected a strong association was found with the Falls-Efficacy Scale, while a low correlation was found with the Positive and Negative Affectivity Scale [23]. Among individuals with lower limb amputations, strong associations were found with two different walk tests [30]. Despite the solid psychometric properties of the ABC among healthy older adults and individuals with lower limb amputations, reliability and validity have not been reported among individuals with stroke. Stroke is the number one cause of neurological disability and is one of the greatest risk factors for falls in older adults [35]. The incidence of falls has been reported to be as high as 73% of individuals with stroke falling within 6 months following hospital discharge to home with an average of 3.4 falls per person during this 6-month period [36]. Thus, the purpose of this study was to assess and report the reliability and validity of the ABC among individuals with stroke. Internal consistency, construct (convergent) validity and a factor analysis (factorial validity) of the ABC were assessed after the total sample completed the ABC and two validity measures during the first measurement session. Test-retest reliability was assessed after the reliability sample completed the ABC for the second time 4 weeks later. Method Design/participants Data from an existing convenience sample was used. The sample was recruited from the community Measuring balance confidence in stroke 157 through newspaper advertisements and community centres. Ethical approval was granted from both the university and hospital review boards. Inclusion criteria required participants to be older than 50 years of age, at least 1-year post stroke and to present with hemiparesis secondary to a single stroke. Participants were required to be able to provide informed consent, follow 1 and 2 step commands, and walk 10 m with rest intervals. Exclusion criteria included any musculoskeletal or neurological conditions in addition to stroke and a score of less than 24 on the Folstein Mini Mental exam [37]. During the first session, 77 participants (total sample) completed the ABC, the Berg Balance Scale (BBS) which is a measure of functional balance, and gait speed. Four weeks later, a sub-sample of 24 participants (reliability sample) completed the ABC a second time, to examine the test retest reliability. A 4-week time period between the two measurement sessions was chosen to reduce participant memory bias which was hopefully short enough that no real changes took place in the participant s balance confidence. Each participant filled out the ABC on his or her own. When necessary, individuals with severe language barriers, due to English as a second language, received help understanding the items in the scales. Of the 77 participants, 24 of the individuals (reliability sample) were part of an intervention study that ended 4 weeks prior to our first measurement session. No evidence of change in subject performance was observed for measures of BBS or gait speed during the time that our data were collected [38]; therefore, the sample was considered stable. The other 53 participants did not participate in any study previous to data collection. Measurement Activities-specific Balance Confidence (ABC) Scale. The ABC [22 24] is a 16-item self-report questionnaire that asks individuals to rate their balance confidence in performing specific ambulatory activities on a numerical rating scale (0 100). A score of zero represents no confidence, while a score of 100 represents complete confidence in performing the activity [24]. The overall ABC score for each participant is calculated by adding the individual item scores together and then dividing by the total number of items. Various studies [22 24] have shown the internal consistency, test retest reliability and validity of the ABC to be very strong among older adults. Myers et al. [22] demonstrated the discriminative properties of the ABC by suggesting that ABC scores lower than 50 indicate a low level of physical functioning, scores between 50 and 80 are indicative of a moderate level of physical functioning

4 158 E.M. Botner et al. and scores above 80 indicate high functioning among physically active older adults. Berg Balance Scale. The Berg Balance Scale (BBS) [39] was used as a validation measure for the ABC. The BBS consists of 14 tasks that require individuals to maintain their balance in positions and tasks of increasing difficulty. Items are graded along an ordinal scale ranging from 0 to 4 (with a possible total score of 56) based on either the amount of time an individual can hold a position or the amount of time to complete the given task. Fewer points are given if supervision, cueing or assistance is required during the task or if the time or distance requirements are not met. Full points are given if the participant is fully capable of performing the task. The scale has demonstrated strong test retest reliability and internal consistency among older adults [39] and the construct validity and responsiveness of the scale has been established in individuals following stroke [39 42]. Gait speed. Gait speed, a recognized and reliable indicator of gait performance [43] was also used as a validation tool for the ABC. Participants were asked to walk 8 m at a comfortable speed, wearing their own shoes and using their usual assistive devices (i.e., cane, ankle-foot orthosis). Self-paced gait speed was calculated from the mean of two walking trials. The cumulative distance and corresponding elapsed time were recorded using infrared emitting diodes (Northern Digital) which were attached to the foot. Gait speed was calculated using times from the middle section (i.e., approximately a 4-m section representative of constant gait speed) of the 8-m walk. Statistical analysis. Descriptive statistics were calculated using data collected from the total sample during the first measurement session. Internal consistency of the ABC and individual item consistency were examined by calculating Cronbach s a and observing any changes in a in the absence of each item. Construct validity (convergent) was measured by examining the relationship between ABC and both the BBS scale and gait speed using Spearman s correlation coefficient (r) because the distribution of ABC data were not normal (skewness = ( , kurtosis = ( ). Based on the results of previous research which found correlations between the ABC and functional measures in older adults between 0.3 and 0.5 [24], we hypothesized that our correlations would be at least that size. Cohen [44] suggests that a correlation of 0.1 is considered small, 0.3 is considered medium, and 0.5 is large. Factor analysis using principle component analysis with varimax rotation was conducted on the 16 balance self-efficacy items to detect the structure of relationships between the variables. Components were selected based on eigenvalues of 4 1 [45]. Relative test retest reliability was measured with intra-class correlation (ICC 1,1 ) and 95% confidence intervals were reported for all estimates. Coefficients and absolute reliability were measured by calculating the standard error of measurement (SEM) and by examining how the individual scores varied on repeated measurement using the Bland Altman Method [46]. Acceptable reliability (ICC) was set at 0.75 based on work by Anderson [47]. However, acceptable reliability for making decisions about individuals and treatments is often higher. For our purposes, we set it at 0.85 based on the magnitudes suggested by Weiner and Steward [48]. All data analyses were performed using SPSS for Windows version 11 with an a of 0.05 for significance. Results The average age of the 77 participants was years, of whom 71% were male, 57% had right side paresis, and the average stroke duration was years. There were no statistically significant differences between the reliability sub sample and the total sample. Participant characteristics are displayed in Table I. The mean ABC score for the total sample was Individual item mean scores ranged from 38.1 for confidence while standing on a chair to reach to 84.2 for confidence while getting in and out of a car (Table II). The items that the individuals in our population were least confident to perform without losing their balance included standing on a chair to reach, walking on an icy sidewalk, riding an escalator without using the rail and standing on tiptoes to reach. The average BBS score was with scores ranging from 34 to 54. The average gait speed was m/s, with speeds ranging from 0.06 to 0.74 m/s. Reliability Four-week test retest reliability on the ABC total score for the reliability sample demonstrated an ICC of 0.85 (95% CI ). Individual item test retest reliability ranged from 0.53 (95% CI ) for confidence walking up and down stairs to 0.93 (95% CI ) for confidence walking up and down a ramp. Other individual item ICCs are presented in Table II. Calculation of the SEM revealed the standard deviation of the measurement errors to be The Bland Altman Method [46] showed little difference in the total ABC mean scores between time one

5 Table I. Participant characteristics. Measuring balance confidence in stroke 159 Total sample Reliability sample (N = 77) (N = 24) Gender (M/F) 55/22 18/6 Mean age (years) Mean height (cm) Mean mass (kg) Mean time since stroke (years) Hemiparetic side (R/L/NA) 44/31/2 12/12/0 Mean Berg Balance Scale Score Mean gait speed (m/s) Type of stroke 40 Ischemic 12 Ischemic 25 Hemorrhagic 9 Hemorrhagic 12 Unknown 3 Unknown Table II. Values obtained for the items of the Activities-specific Balance Confidence (ABC) scale on the first measurement session and test retest reliability. On a Scale of 0 100%, how confident are you that you will not lose your balance or become unsteady when you...? Activity Mean (SD) Median Range ICC 95% CI A. Walk around the house 83.3 (18.6) B. Walk up and down stairs 76.3 (17.7) C. Pick up a slipper from the floor 74.1 (25.5) D. Reach at eye level 79.4 (20.9) E. Reach while standing on your tiptoes 57.3 (29.3) F. Stand on a chair to reach 38.1 (30.0) G. Sweep the floor 70.0 (28.7) H. Walk outside to nearby car 82.3 (19.4) I. Get in and out of a car 84.2 (18.3) J. Walk across a parking lot 78.8 (19.3) K. Walk up and down a ramp 71.9 (22.8) L. Walk in a crowded mall 72.9 (20.7) M. Walk in a crowd or get bumped 65.4 (21.4) N. Ride an escalator holding the rail 70.6 (26.0) O. Ride an escalator not holding the rail 46.3 (30.5) P. Walk on icy sidewalks 41.7 (28.7) Total ABC score 68.3 (17.5) (mean ) and time two (mean ) for the reliability sample. The mean difference was Figure 1 provides a visual representation of the individual responses, revealing that most of the data points fall within two standard deviations of the mean difference and of the SEM. This suggests that 95% of the time, the errors of measurement using this test will fall within this range [46]. Only one major outlier had a retest difference that was not within this range (difference in scores = 31.3). The data is equally distributed above and below the zero line (11 above and 13 below). Internal consistency (Chronbach s a) was 0.94 with a values ranging from 0.93 to 0.94 when each item was deleted. Out of interest, we calculated test retest reliability without the outlier and found a substantial improvement in both the ICC which became 0.92 ( ) and in the mean difference between time one and time two which reduced to Validity There was a moderate positive, linear correlation between the ABC total score and the BBS score using Spearman s correlation coefficient (r = 0.36, P ). The relationship between the ABC total score and gait speed was also a moderate, positive, linear correlation (r = 0.48, P ). A scatter plot of these relationships can be found in Figures 2 and 3, respectively. Factor analysis Table III reports the results of the factor analysis. In the initial factor extraction, two components with eigenvalues greater than 1 emerged, and they were retained for rotation. Rotation using Varimax with Kaiser normalization was performed. The results showed modest correlations between the two factors. Component loadings were deemed significant if the

6 160 E.M. Botner et al. Figure 1. Bland Altman plot of difference in ABC scores between time 1 and time 2 versus average ABC Scores from time 1 and time 2. Figure 2. Scatterplot comparing the Activity-specific Balance Confidence (ABC) Scale and the Berg Balance Scale (BBS) Scores. Figure 3. Scatterplot comparing the Activity-specific Balance Confidence (ABC) Scale and gait speed scores. Table III. Rotate component matrix. Extraction method: principal component analysis; rotation method: varimax with Kaiser normalization. Items Factor 1 PLRA Factor 2 PHRA Mean Score on ABC H. Walk outside to nearby car I. Get in and out of a car A. Walk around house J. Walk across parking lot D. Reach at eye level C. Pick up slipper from floor L. Walk in a crowded mall K. Walk up and down ramp B. Walk up and down stairs O. Ride escalator with no rail P. Walk on icy sidewalks F. Stand on chair to reach N. Ride escalator with rail M. Walk in a crowd/get bumped E. Reach while on tiptoes G. Sweep the floor magnitude of the variable loading was for the variable. Given this criterion, nine items loaded on the first component, and six on the second component. One item, confidence in sweeping floors, loaded almost equally on both components. Together, components 1 and 2 accounted for 68.6% of the variance in the set of variables (Factor 1 = 55.7%; Factor 2 = 12.9%). Discussion This is the first known published study to examine the measurement properties of the ABC among PLRA = Perceived Low Risk Activities; PHRA = Perceived High Risk Activities individuals with stroke. To our knowledge, it is also the first study to perform a factor analysis using principal component analysis in order to detect structure of relationship among the items in the scale. Using the ABC cut points derived by Myers et al. [22] to discriminate between different levels of physical functioning, 20% of our participants were classified as having a low level of physical functioning, 54% had a moderate level of physical functioning and 26% were highly functioning [22].

7 The coefficient a was very high, similar to that seen in previous studies [23,29], demonstrating that each participant s responses are consistent across scale items. The fact that a changed minimally when each item was deleted, indicates strong item consistency, and that none of the items decreased the scale stability. Therefore, the scale is internally consistent. Nunnally [49] suggests that a should be above 0.70, but probably not higher than 0.90 to ensure that there is no item redundancy. Despite the fact that our a magnitude is slightly higher that 0.90, it is comparable to that seen in previous studies [23, 30]. According to Streiner and Norman [45], a scores are dependent on both the magnitude of the correlations among the items as well as the number of items in the scale. In addition, examination of the internal consistency and factor structure of the ABC in this study provide evidence that item redundancy is not a concern. Based on magnitudes of acceptable reliability presented by Anderson [47] and Weiner and Steward [48], our estimates of test retest reliability of the ABC among individuals with stroke are considered excellent for description of groups; and are acceptable for making decisions about individuals and treatments. Observations from the Bland Altman plots revealed that the data were equally distributed above and below the zero line suggesting minimal bias with repeat testing. In addition, only one out of the 24 participants in the reliability sample fell outside of two standard deviations of the mean difference, suggesting good agreement and limited test retest variation [46]. Although several of the individual items demonstrated ICCs that appear to be low, it is important to focus on the reliability of the scale as a whole [45]. It is plausible that there is simply more random error in the activities that have demonstrated lower relative reliability because they are activities that may not be performed on a daily basis by the participants. The individual whose ABC score dramatically changed was low functioning according to his results on both the BBS and gait speed. The characteristics of the participant did not provide any clues as to his change in confidence; therefore, it is possible that the change could have been due either to a measurement error or an actual change in balance confidence even though the participant demonstrated no significant changes in any of the functional measures (BBS or gait speed). Out of interest, we calculated the reliability with and without this participant and as one would expect, the reliability of the scales increased. The correlations between the ABC and the functional performance measures (BBS and gait speed) were significant; however, the coefficient values were not particularly strong. This is consistent with previous findings and appropriate given that while indicative of actual performance, perceptions Measuring balance confidence in stroke 161 rarely predict the full variance of an actual skill or performance [24]. The factor analysis using a principal component analysis revealed two distinct components of balance self-efficacy represented in the 16 items of the ABC. Based on the item content, component labels were assigned. Component 1 (perceived low-risk activities) included the items that had the highest scores (over 70) indicating high levels of confidence in performing the tasks without losing balance, whereas component 2 (perceived high-risk activities) included the items with the lower confidence scores (under 70). The results of the factor analysis indicate that the ABC appears to be measuring two separate components. The fact that the two components were distinguishable by item mean score ( + 70) indicates that the ABC can separate out items with different levels of perceived risk (Table II). More than 80% of our participants scored between 40 and 80 on the ABC, indicating that there was minimal ceiling or floor effect in this sample. A limitation of our study should be noted. Our sample was drawn from an existing convenience sample. This sample consisted of community dwelling, ambulatory individuals with stroke and cannot be generalized to all individuals with stroke. However, the majority of individuals with stroke do regain walking ability [50]. Although the Berg Balance Scale is considered a fall-risk factor [51], an important follow-up study would be to examine the relationship between balance confidence and the number of actual falls in order to provide further evidence for the validity of the ABC. Conclusions A psychometrically sound measure of balance confidence, such as the ABC, may help clinicians understand the effects of fear of falling and low falls efficacy among individuals with stroke. The results provide further support for the measurement properties of the ABC and suggest that it is a useful tool in identifying individuals who have low confidence in performing everyday basic tasks without losing balance so that appropriate interventions can be made. Acknowledgements This project was funded by a New Investigator salary support award (JJE) and Post-Doctoral awards (WCM) from both the Canadian Institute of Health Research and Michael Smith Foundation for Health Research. Also funded by operating grants from CIHR (MOP-57862) and the Heart and Stroke Foundation of BC and Yukon (JJE). The authors wish to acknowledge Daniel Marigold, Kelly Chu,

8 162 E.M. Botner et al. Catherine Donnelly, Craig Tokuno and the rest of the GF Strong Rehab Research Lab for their help and support during data collection References 1. Niam S, Cheung W, Sullivan PE, Kent S, Gu X. Balance and physical impairments after stroke. Archives of Physical Medicine and Rehabilitation 1999;80: Mayo NE, Korner-Bitensky N, Kaizer F. Relationship between response time and falls among stroke patients undergoing physical rehabilitation. International Journal of Rehabilitation Research 1990;13: Nyberg L, Gustafson Y. Using the Downton Index to predict those prone to falls in stroke rehabilitation. Stroke 1996;27(10): Overstall P. Falls after stroke. British Medical Journal 1995;311: Sackley CM. Falls, sway, and symmetry of weight bearing after stroke. International Disability Studies 1991;13: Tinetti ME, Richman D, Powell L. Falls efficacy as a measure of fear of falling. Journal of Gerontology: Psychological Sciences 1990;45(6): Tinetti ME, Powell L. Fear of falling and low self-efficacy: A cause of dependence in the elderly. Journal of Gerontology 1993;48: Vellas BJ, Wayne SJ, Romero LJ, Baumgartner RN, Garry PJ. Fear of falling and restriction of mobility in elderly fallers. Age and Ageing 1997;26: Chandler JM, Duncan PW, Sanders L, Studenski S. The fear of falling syndrome: Relationship to falls, physical performance, and activities of daily living in frail older persons. Topics in Geriatrics and Rehabilitation 1996;11(3): Walker Peterson E. Fear of falling. Part I: Prevalence, intensity and implications. Home Community Health (Special Interest Section Newsletter) 2001;8(4): Li F, McAuley E, Fisher JK, et al. Self efficacy as a mediator between fear of falling and functional ability in the elderly. Journal of Aging and Health 2002;14(4): Hellstrom K, Lindmark B. Fear of falling in patients with stroke: A reliability study. Clinical Rehabilitation 1999;13: Hellstrom K, Nilsson L, Fugl Meyer AR. Relationship of confidence in task performance with balance and motor function after stroke. Physiotherapy Theory Practice 2001; 17: Hellstrom K, Lindmark B, Fugl Meyer A. The Falls-Efficacy Scale, Swedish version: Does it reflect clinically meaningful changes after stroke? Disability and Rehabilitation 2002;24 (9): Friedman SM, Munoz B, West SK, Rubin GS, Fried LP. Falls and fear of falling: Which comes first? A longitudinal prediction model suggests strategies for primary and secondary prevention. Journal of the American Geriatrics Society 2002;50: Cameron ID, Stafford B, Cumming RG, et al. Hip protectors improve falls self-efficacy. Age and Ageing 2000;29: Walls BS. A dangerous secret: I had a fall. OT Practice 1999;4(11): Riley G, Holding D. Tackling a fear of falling using graded exposure. Physiotherapy 2000;86(3): Walker PE. Help patients recover from the fear of falling. Hospital Case Management 2000;8(3): Gentleman B, Malozemoff W. Falls and feelings: Description of a psychosocial group nursing intervention. Journal of Gerontology and Nursing 2001;27(10): Rodriquez AA, Black PO, Kile KA, Sherman J, Stellberg B, McCormick J, Roszkowski J, Swiggum E. Gait training efficacy using a home-based practice model in chronic hemiplegia. Archives of Physical Medicine and Rehabilitation 1996;77: Myers AM, Fletcher PC, Myers AH, Sherk W. Discriminative and evaluative properties of the Activities-specific Balance Confidence (ABC) scale. Journal of Gerontology: Medical Sciences 1998;53A(4): M Powell LE, Myers AM. The Activities-specific Balance Confidence (ABC) scale. Journal of Gerontology: Medical Sciences 1995;50A(1): M Myers AM, Powell LE, Maki BE, Holliday PJ, Brawley LR, Sherk W. Psychological indicators of balance confidence: Relationship to actual and perceived abilities. Journal of Gerontology: Medical Sciences 1996;51A(1): M Tinetti ME, Richman D, Powell L. Falls efficacy as a measure of fear of falling. Journal of Gerontology: Psychological Sciences 1990;45(6): Bandura A. Self-efficacy mechanism in human agency. American Psychologist 1982;37: Petrella RJ, Payne M, Myers A, Overand T, Chesworth B. Physcial function and fear of falling after hip fracture rehabilitation in the elderly. American Journal of Physical Medicine and Rehabilitation 2000;79(2): Miller WC, Death BA, Speechley M, Koval J. The influence of falling, fear of falling, and balance confidence on prosthetic mobility and social activity among individuals with a lower extremity amputation. Archives of Physical Medicine and Rehabilitation 2001;82: Miller WC, Speechley M, Death BA. Balance confidence among people with lower-limb amputations. Physical Therapy 2002;82(9): Miller WC, Death BA, Speechley M. Psychometric properties of the Activities-specific Balance Confidence scale. Archives of Physical Medicine and Rehabilitation 2003;85(5): Hoeppner SC, Rimmer JH. Postural balance and self reported exercise in older adults. Adapted Physical Activities Quarterly 2000;17: Furzhong L, McAuley E, Fisher JK, Harmer P, Chaumeton N, Wilson NL. Self-efficacy as a mediator between fear of falling and functional ability in the elderly. Journal of Aging and Health 2002;14(4): Velozo CA, Peterson EW. Developing meaningful fear of falling measures for community dwelling elderly. American Journal of Physical Medicine and Rehabilitation 2001;80(9): Whitney SL, Hudak MT, Marchetti GF. The Activitesspecific Balance Confidence scale and the Dizziness Handicap Inventory: A comparison. Journal of Vestibular Research 1999;9: Jorgensen L, Engstad T, Jacobsen BK. Higher incidence of falls in long-term stroke survivors than in population controls: depressive symptoms predict falls after stroke. Stroke 2002;33: Forster A, Young J. Incidence and consequences of falls due to stroke: a systematic inquiry. British Medical Journal 1995; 311: Folstein MF, Folstein SE, McHugh PR. Mini-mental state : A practical method for grading the cognitive state of patients for the clinician. Journal of Psychology Research 1975;12: Eng JJ, Chu K, Kim CM, Dawson AS, Carswell A, Hepburn K. A community-based group exercise program for persons with chronic stroke. Medicine and Science in Sports and Exercise 2003;35(8):

9 39. Berg K, Wood-Dauphinee S, Williams JI, Gayton D. Measuring balance in the elderly: preliminary development of an instrument. Physiotherapy Canada 1989;41(6): Berg K, Wood-Dauphinee S, Williams JI. The Balance Scale: reliability assessment with elderly residents and patients with an acute stroke. Scandinavian Journal of Rehabilitation Medicine 1995;27(1): Stevensson TJ, Garland SJ. Standing balance during internally produced perturbations in subjects with hemiplegia: Validation of the balance scale. Archives of Physical Medicine and Rehabilitation 1996;77: Wood-Dauphinee SL, Berg K, Bravo G, Williams JI. The balance scale: Responsiveness to clinically meaningful changes. Canadian Journal of Rehabilitation 1997;10: Richards CL, Malouin F, Dumas F, et al Gait velocity as an outcome measure of locomoter recovery after stroke. In: Craik R, Oatic C, editors. Gait analysis: Theory and applications. St. Louis, MO: Mosby, pp Cohen J Statistical power analysis for the behavioral sciences. 2nd edn. Hillside, NJ: Lawrence Erlbaum Association. Measuring balance confidence in stroke Streiner DL, Norman GR Health measurement scales. A practical guide to their development and use. 2nd ed. Oxford: Oxford Medical Publications. 46. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1: Anderson EM. Criteria for assessing the tools of disability outcomes research. Archives of Physical Medicine and Rehabilitation 2000;12(2): S Weiner EA, Steward BJ Assessing individuals. Boston, MA: Little Brown. 49. Nunally JC Jr Psychometric theory. 2nd ed. New York: McGraw-Hill. 50. Feigenson J. Stroke rehabilitation: Outcome studies and guidelines for alternative levels of care. Stroke 1981;12: Berg KO, Wood-Dauphinee SL, Williams, JI, Maki, B. Measuring balance in the elderly: Validation of an instrument. Canadian Journal of Public Health 1992;83(2): S7 11.