Reliability of the Dynamic Gait Index in People With Vestibular Disorders

Transcription

1 1528 Reliability of the Dynamic Gait Index in People With Vestibular Disorders Diane M. Wrisley, PhD, PT, NCS, Martha L. Walker, MS, PT, John L. Echternach, EdD, PT, ECS, FAPTA, Barry Strasnick, MD, FACS ABSTRACT. Wrisley DM, Walker ML, Echternach JL, Strasnick B. Reliability of the Dynamic Gait Index in people with vestibular disorders. Arch Phys Med Rehabil 2003;84: Objective: To examine the interrater reliability of the Dynamic Gait Index (DGI) when used with patients with vestibular disorders and with previously published instructions. Design: Correlational study. Setting: Outpatient physical therapy clinic. Participants: Subjects included 30 patients (age range, 27 88y) with vestibular disorders, who were referred for vestibular rehabilitation. Interventions: Subjects performance on the DGI was concurrently rated by 2 physical therapists experienced in vestibular rehabilitation to determine interrater reliability. Main Outcome Measures: Percentage agreement, statistics, and the ratio of subject variability to total variability were calculated for individual DGI items. Kappa statistics for individual items were averaged to yield a composite score of the DGI. Total DGI scores were evaluated for interrater reliability by using the Spearman rank-order correlation coefficient. Results: Interrater reliability of individual DGI items varied from poor to excellent based on values ( range, ). Composite values showed good overall interrater reliability (.64) of total DGI scores. The Spearman demonstrated excellent correlation (r.95) between total DGI scores given concurrently by the 2 raters. Conclusion: DGI total scores, administered by using the published instructions, showed moderate interrater reliability with subjects with vestibular disorders. The DGI should be used with caution in this population at this time, because of the lack of strong reliability. Key Words: Gait; Rehabilitation; Reliability and validity; Vestibular diseases by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation DYSFUNCTION OF THE vestibular system can result in disequilibrium manifested by ataxic gait and postural instability exacerbated by head turns, body turns, and situations From the Neurological Sciences Institute, Oregon Health and Sciences University, Portland, OR (Wrisley); School of Physical Therapy, Old Dominion University, Norfolk, VA (Walker, Echternach); Eastern Virginia Medical School, Norfolk, VA (Strasnick); and Bon Secours DePaul Medical Center Hearing and Balance Center, Norfolk, VA (Strasnick). Presented in part at the American Physical Therapy s Combined Section Meeting, February 2000, in New Orleans, LA. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Diane M. Wrisley, PhD, Neurological Sciences Institute, Oregon Health and Sciences University, West Campus Bldg 1, 505 NW 185th Ave, Beaverton, OR , wrisleyd@ohsu.edu /03/ $30.00/0 doi: /s (03) with absent or inaccurate sensory information. 1,2 Several clinical measures, developed to document both the quality of the movement and the temporospatial characteristics of gait, have been applied to the examination of gait in patients with vestibular disorders. 2-5 Of the gait analysis methods that have been used to assess mobility skills in patients with vestibular disorders, one was purely observational 2 and the others were primarily research tools that are not available to most clinicians. 3-5 None of the tools was specifically designed to assess the ambulation tasks that patients with vestibular disorders find difficult. The Dynamic Gait Index (DGI) was developed to assess dynamic postural stability in the older adult at risk for falling. 6 This functional gait scale consists of 8 tasks with varying demands, such as walking at different speeds, walking with head turns, ambulating over and around obstacles, ascending and descending stairs, and making quick turns. Each item is scored on a 4-level ordinal scale (table 1), with a maximum possible score, on the entire DGI, of 24. A score of 19 or less indicates an increased risk of falling in older adults. 7 Each item of the DGI provides simple instructions for performance of that particular item, with operational definitions for each of the possible grading options. However, no instructions are provided for general administration (eg, environment, equipment), and no decision rules are provided to help distinguish between 1 grading option and another. For example, the definition of good gait speed or minor disruption to smooth gait path may differ among clinicians, and, therefore, grading of the DGI may be inconsistent. Shumway-Cook et al 8 determined the reliability of the DGI by using a sample of 5 community-dwelling older adults with varying balance abilities. Five physical therapists, trained in the administration of the DGI, evaluated the subjects on the index. The developer of the test trained the therapists during a 1-hour training session in which they were instructed in her unpublished decision rules. Excellent interrater reliability (.96) was found by using the ratio of subject variability to total variability. Two therapists repeated the test 1 week later to determine test-retest reliability. Again, by using the ratio of subject variability to total variability, test-retest was also found to be excellent, at.98. The DGI, although not developed specifically for use with patients with vestibular disorders, includes gait activities that are of interest when evaluating patients with vestibular disorders, is easy to administer, and requires minimal equipment and space. The DGI has been shown to discriminate between patients with vestibular disorders who reported falls and those who did not report falls. 9 However, the reliability of the DGI, when used with the published grading criteria or when used with patients with vestibular disorders, has not been reported. If the DGI is shown to be reliable in subjects with vestibular disorders, it may provide an inexpensive clinical tool that measures dynamic postural stability in this patient population. The purpose of our study was to examine the interrater reliability of the DGI for patients with vestibular disorders by

2 RELIABILITY OF DYNAMIC GAIT INDEX, Wrisley 1529 Table 1: Dynamic Gait Index 6 1. Gait Level Surface. Instructions: Walk at your normal speed from here to the next mark (20 ) (3) Normal: Walks 20 ; no assistive devices, good speed, no evidence for imbalance, normal gait pattern. (2) Mild Impairment: Walks 20 ; uses assistive device, slower speed, mild gait deviations. (1) Moderate Impairment: Walks 20 ; slow speed, abnormal gait pattern, evidence for imbalance. (0) Severe Impairment: Cannot walk 20 without assistance, severe gait deviations or imbalance. 2. Change in Gait Speed. Instructions: Begin walking at your normal pace (for 5 ), when I tell you go, walk as fast as you can (for 5 ). When I tell you slow, walk as slowly as you can (for 5 ). (3) Normal: Able to smoothly change walking speed without loss of balance or gait deviation. Shows a significant difference in walking speeds between normal, fast, and slow speeds. (2) Mild Impairment: Is able to change speed but demonstrates mild gait deviations or no gait deviations, but unable to achieve a significant change in velocity, or uses an assistive device. (1) Moderate Impairment: Makes only minor adjustments to walking speed, or accomplishes a change in speed with significant gait deviations or changes speed but loses balance but is able to recover and continue walking. (0) Severe Impairment: Cannot change speeds, or loses balance and has to reach for wall or be caught. 3. Gait with Horizontal Head Turns. Instructions: Begin walking at your normal pace. When I tell you to look right, keep walking straight but turn your head to the right. Keep looking right until I tell you look left, then keep walking straight but turn your head to the left. Keep your head to the left until I tell you, look straight, then keep walking straight, but return your head to the center. (3) Normal: Performs head turns smoothly with no change in gait. (2) Mild Impairment: Performs head turns smoothly with slight change in gait velocity, ie, minor disruption to smooth gait path or uses walking aid. (1) Moderate Impairment: Performs head turns with moderate change in gait velocity, slows down, staggers but recovers, can continue to walk. (0) Severe Impairment: Performs task with severe disruption of gait, ie, staggers outside 15 path, loses balance, stops, reaches for wall. 4. Gait with Vertical Head Turns. Instructions: Begin walking at your normal pace. When I tell you to look up, keep walking straight, but tip your head and look up. Keep looking up until I tell you, look down. Then keep walking straight and turn your head down. Keep looking down until I tell you, look straight, then keep walking straight, but return your head to the center. (3) Normal: Performs head turns with no change in gait. (2) Mild Impairment: Performs task with slight change in gait velocity, ie, minor disruption to smooth gait path or uses walking aid. (1) Moderate Impairment: Performs task with moderate change in gait velocity, slows down, staggers but recovers, can continue to walk. (0) Severe Impairment: Performs task with severe disruption of gait, ie, staggers outside 15 path, loses balance, stops, reaches for wall. 5. Gait and Pivot Turn Instructions: Begin with walking at your normal pace. When I tell you, turn and stop, turn as quickly as you can to face the opposite direction and stop. (3) Normal: Pivot turns safely within 3 seconds and stops quickly with no loss of balance. (2) Mild Impairment: Pivot turns safely in 3 seconds and stops with no loss of balance. (1) Moderate Impairment: Turns slowly, requires verbal cueing, requires several small steps to catch balance following turn and stop. (0) Severe Impairment: Cannot turn safely, requires assistance to turn and stop. 6. Step over Obstacle. Instructions: Begin walking at your normal speed. When you come to the shoebox, step over it, not around it, and keep walking. (3) Normal: Is able to step over box without changing gait speed; no evidence for imbalance. (2) Mild Impairment: Is able to step over box, but must slow down and adjust steps to clear box safely. (1) Moderate Impairment: Is able to step over box but must stop, then step over. May require verbal cueing. (0) Severe Impairment: Cannot perform without assistance. 7. Step Around Obstacles. Instructions: Begin walking at your normal speed. When you come to the first cone (about 6 away), walk around the right side of it. When you come to the second cone (6 passed first cone), walk around it to the left. (3) Normal: Is able to walk around cones safely without changing gait speed; no evidence of imbalance. (2) Mild Impairment: Is able to step around both cones, but must slow down and adjust steps to clear cones. (1) Moderate Impairment: Is able to clear cones but must significantly slow speed to accomplish task or requires verbal cueing. (0) Severe Impairment: Unable to clear cones, walks into one or both cones, or requires physical assistance. 8. Steps Instruction: Walk up these stairs as you would at home (ie, using the rail if necessary). At the top turn around and walk down. (3) Normal: Alternating feet, no rail. (2) Mild Impairment: Alternating feet, must use rail. (1) Moderate Impairment: Two feet to a stair; must use rail. (0) Severe Impairment: Cannot do safely. Total Score (Score 19/24 indicates increased risk of fall). Reprinted with permission: Shumway-Cook A, Woollacott MH. Motor control: theory and practical applications. Baltimore: Williams & Wilkins; p 323 4, tbl

3 1530 RELIABILITY OF DYNAMIC GAIT INDEX, Wrisley using only the published instructions for the test without additional training of the test administrators. METHODS Participants The use of human subjects in this study was approved by Old Dominion University s Human Subjects Institutional Review Board. Subjects were recruited from patients referred for vestibular rehabilitation at Bon Secours DePaul Medical Center, in Norfolk, VA. All patients meeting the inclusion criteria between April and August 1998 were invited to participate. The inclusion criteria were that the patients be older than 18 years of age, have a Mini-Mental Status Examination score of more than 24, 10 and have the ability to give informed consent. Before referral for vestibular rehabilitation, each subject underwent a neuro-otologic examination and vestibular testing. Vestibular testing, consisting of electronystagmography, caloric testing, positional testing, and rotational chair testing, was completed at the physician s office as part of the diagnostic workup. Each subject was given a vestibular diagnosis by the neuro-otologist after vestibular testing. To provide information regarding the severity of the patients functional impairments, the subjects scores on the Dizziness Handicap Inventory 11 (DHI) and computerized dynamic posturography a (CDP) are provided. Subjects completed the DHI during the initial vestibular testing as a measure of self-perception of disability. The DHI is a 25-item test that was developed to evaluate a patient s perceived handicap because of dizziness. The test is scored on an ordinal scale of 0 to 4, with lower scores indicating less perceived handicap. Subjects completed CDP to assess the use of sensory interaction for balance. Procedures The 2 raters in our study were physical therapists, each with 14 years of experience in physical therapy (PT). One rater had 9 years and the second rater had 2 years of experience in the evaluation and PT intervention of patients with vestibular and balance dysfunction. Before the study, the raters individually reviewed the published instructions for the DGI and independently practiced administration of the test with several patients. No effort was made for the raters to agree on scoring criteria or on decision rules, because we wanted to determine reliability of the test as published. Rater 1 provided the standardized verbal instructions of the DGI and the guarding of all subjects. The subjects completed the DGI during either their first or second treatment session, depending on available time to administer the test and availability of raters. To assess interrater reliability, the DGI was scored concurrently by both raters once during the session. Raters were blind to each other s results. Each subject was given the same verbal instructions and proceeded to complete DGI tasks 1 (gait level surfaces) through 8 (walking up and down stairs) as well as he/she could (table 1). As the subject completed each task, he/she was given a rating that ranged from 0 (severe impairment) to 3 (normal) by the examiners. Data Analysis To evaluate for interrater reliability, the amount of agreement (percentage agreement) was determined for individual items and for the total score of the DGI, between raters. Because the score given for an individual item of the DGI is on an ordinal scale, the Cohen coefficient was calculated for each item. The values were averaged to give a composite reliability score, because values could not be calculated for Table 2: Descriptive Subject Information Age (y) (range, 27 88) Gender (n) Women, 23; men, 7 Diagnosis (n) Unilateral peripheral vestibular hypofunction: 18 Bilateral peripheral vestibular hypofunction: 1 Central vestibular dysfunction: 3 Cervicogenic dizziness: 5 Visual dependence: 3 DHI (range, 0 100) (range, 4 78) CDP (range, 0 100) (range, 26 83) Electronystagmography Normal: 13 (n) Unilateral reduced vestibular response: 13 Bilateral reduced vestibular response: 1 Directional preponderance: 1 Not tested: 2 Rotational vestibular testing (n) Normal: 16 Directional preponderance: 6 Abnormal phase: 1 Decreased gain: 4 Not tested: 3 NOTE. Values are mean standard deviation or as otherwise indicated. the composite scores because of the wide range of values. Studies have interpreted a value of greater than.80 as excellent, between.60 and.79 as good, between.40 and.59 as fair, and less than.40 as poor In addition to calculating values, the ratio of subject variability to total variability was calculated for individual and total DGI scores. This was done to allow for comparison with the reliability of the DGI as reported by Shumway-Cook et al. 8 Although not explicitly clear in the Shumway-Cook article, this was interpreted to mean the total variability and subject variability from an analysis of variance table. The Spearman rank-order correlation coefficient was calculated for DGI individual and total scores between raters, to evaluate systematic error caused by tester judgment. RESULTS Thirty subjects aged 27 to 88 years (mean, 61 17y) volunteered to participate in the study. The subjects included 23 women and 7 men. No subject who met the inclusion criteria was excluded from the study. Descriptive subject information is summarized in table 2. Vestibular diagnoses included unilateral vestibular hypofunction, bilateral vestibular hypofunction, cervicogenic dizziness, central vertigo, Meniere s disease, and postconcussive syndrome. Nineteen subjects had a peripheral vestibular diagnosis (unilateral and bilateral hypofunction), and 11 had a central vestibular diagnosis (cervicogenic dizziness, central vertigo, visual motion sensitivity). Scores on the DHI ranged from 4 to 78 out of 100 (mean, ). Twenty-six subjects completed the CDP, and composite CDP scores ranged from 28 to 83 (mean, 62 13). Fifteen patients showed abnormal CDP scores based on the normal values provided by the manufacturer (NeuroCom). The range of individual scores given by raters 1 and 2 are illustrated in figure 1. The total DGI scores ranged from 13 to 24, out of a possible 24 (fig 2). The median and the mode were 21. Shumway-Cook et al 7 defined a score of 19 or less out of 24 as indicative of increased risk of falling in community-

4 RELIABILITY OF DYNAMIC GAIT INDEX, Wrisley 1531 Fig 2. Error plots of total DGI scores for rater 1 and rater 2. Boxes represent mean item score; lines represent 2 SEM. Fig 1. Error plots of individual DGI scores for rater 1 and rater 2. Boxes represent mean item score; lines represent 2 standard error of mean (SEM). dwelling older adults. Most of our subjects showed total scores greater than 20, which indicates a minimal risk of falling using the criteria established by Shumway-Cook for older adults and the score validated by Whitney et al 9 for patients with vestibular disorders. 7 Because the scores on the DGI varied by no more than 1 grading level between raters or between tests on any individual DGI item, it was not necessary to weight. Although every effort was made to include subjects of varying balance abilities, there was little variability in subject performance on certain items. This may have yielded deceptively low estimates of reliability based on the statistic. For this reason, both and percentage agreement were calculated and reported. The amount of agreement between scores obtained when raters concurrently completed the DGI was calculated to determine interrater reliability. Percentage agreement, coefficients, Spearman rank-order correlation coefficient, and ratio of subject variability to total variability for interrater reliability of individual DGI items are listed in table 3. Interrater reliability of individual items varied from poor to excellent, based on coefficient values ( range, ). The percentage agreement of these items ranged from 73% to 97%. The ratio of subject variability to total variability ranged from.33 to Composite values and the ratio of subject variability to total Table 3: Interrater Percentage Agreement and Coefficient Values for Individual Items of DGI Gait Item No. % Agreement Value Interrater Reliability Spearman Correlation Coefficient Subject Variability/Total Variability 1. Self-paced gait Gait at various speeds Gait with horizontal head turns Gait with vertical head turns Pivot turn Step over obstacle Step around obstacle Stairs Composite

5 1532 RELIABILITY OF DYNAMIC GAIT INDEX, Wrisley variability showed moderate overall interrater reliability of total DGI scores (.64; subject variability/total variability.68). Spearman rank-order correlation coefficient showed excellent correlation between the total DGI scores of both raters (r.95, P.0001). DISCUSSION Our finding of interrater reliability of the DGI as.68 when measured as ratio of subject variability to total variability was lower than the reliability reported by Shumway-Cook et al. 7 Our findings may differ from Shumway-Cook s findings, because of the greater number of chronic balance problems of the older adults, homogeneity of the subjects, the differences in sample size in the previous study, or the decision rules that were provided to the therapists in the previous study but not available with the published version of the DGI. The measurement of reliability is the measure of the 3 potential sources of error: the subject or variable of interest, the rate, and the instrument. 15 Each of these sources may have contributed to the low values calculated for this study. Our study included only 30 subjects. The statistic becomes more useful with large sample sizes with greater variability. 12 Although data collection was performed during the first or second treatment session to obtain the greatest amount of variability possible, the variability was not as great as anticipated. Increasing the number of subjects included in the study or increasing the heterogeneity of the sample may yield values indicating higher estimates of reliability. The sample was considered representative of the population seen for vestibular rehabilitation. The scores on the DHI ranged from 4 to 78, with a mean score of 47. This indicates that patients perceive moderate disability from dizziness. Unfortunately, 6 of the 30 subjects scored either 23 or 24 on the DGI, indicating that there may be a ceiling effect in using this tool for patients with vestibular disorders. This ceiling effect probably contributed to the lack of variability seen in our data. Kappa coefficient values may be deceptively low with limited variability of data. 12 Item 6, ambulation around obstacles, showed poor interrater reliability (.35), even though agreement was 80%. There was little variability in subjects performance on this item, with subjects scoring mostly a 2 or a 3. Items 3 and 4, ambulation with head turns, showed the greatest variability, with subjects scoring from 0 to 3. These items showed fair reliability ( range,.57.58) despite a 73% agreement. The grading criteria for these 2 items, based on the definitions provided, were also the most difficult for the raters to agree on. Trends were seen in the direction of scoring between the 2 raters. Good correlations were found between the 2 raters scores although the percentage of actual agreement was low, which indicates that the scores varied in the same direction. For items 1 and 2, ambulation at various speeds, the first rater consistently scored the subjects performance higher than the second rater, whereas for items 3 and 4, ambulation with head turns, rater 1 scored the subjects consistently lower. The second rater never issued a score of 0 for any subject. These differences may have resulted from the viewing position of the rater, the interpretation of grading criteria, or an unwillingness to assign the lowest category. The first rater provided all verbal instructions and guarding of the subject. She stood lateral to and slightly behind the subjects, while the other rater stood approximately 1.2m (4ft) behind the subject. From their vantage points, the therapists may have seen different gait deviations. Both raters were experienced in the evaluation and treatment of patients with vestibular disorders. When the raters disagreed in scoring, the deviations from the agreement always occurred 1 level above or below the other s score on the DGI s ordinal scale. This indicates that, although there was not perfect agreement, gross differences in the interpretation of performance did not occur. The items with the lowest percentage agreement, items 3 and 4, had the least objective grading criteria, making it difficult to distinguish between 1 level of performance and another. With the 2 items that involve head turns, raters were asked to judge whether subjects displayed minor, moderate, or severe gait disturbances. These adjectives are briefly described qualitatively but lack objective criteria. Although it would be expected that there would be greater difficulty in distinguishing between the minor and moderate gait abnormality categories, the raters also showed difficulty distinguishing the normal category from the 3 abnormal categories. The definition provided in the grading criteria of normal might not have been sufficient. In item 1, normal is defined as walks 20 feet, no assistive device, good speed, no evidence for imbalance, normal gait pattern. 7 This requires a judgment by the rater as to what constitutes good speed, imbalance, and normal gait pattern. This may be based on the therapist s experience or on the age or the activity level of the subject. It has been repeatedly documented that older adults show slower gait velocities and increased gait deviations than younger adults Although ample kinematic data of normal gait pattern and gait speed at varying stages of the life span have been published, these data are not integrated into the grading criteria. The physical therapist s interpretation, based on personal knowledge and experience, of what is normal at different stages of the lifespan will determine his/her interpretation of this item. Although both therapists used the DGI frequently in clinical practice, most patients seen did not show severe gait deviations; therefore, criteria for the lower scores are less familiar. Rather than using less familiar grading criteria, the therapist could combine all abnormal performance into a familiar category. Walking with head turns requires interpretation of the amount of sway a subject exhibits as being mild, moderate, or severe. The current definitions are open to interpretation. Providing more objective criteria that incorporate published norms of gait speed and postural sway might increase the reliability of these items. It appears that the first rater used a more stringent interpretation of the DGI grading criteria. Reliability coefficients for currently used balance and gait assessments vary widely. Several scales (eg, the Berg Balance Scale, 19 timed up & go, 20 functional reach 21 ) have been shown to have excellent reliability. 21 Other balance and gait assessments (eg, the Performance Oriented Mobility Assessment, 22 the Modified Gait Abnormality Rating Scale 23 [GARS- M]) show moderate to good reliability of the total scores but have lower reliability of their individual scores. The reliability of the DGI found in our study was similar to the reliability of the GARS-M found by VanSwearingen et al. 23 They found scores of individual GARS-M items ranging from.49 to.68. However, they found higher values for individual GARS-M items when the raters were experienced physical therapists. Although the reliability of the DGI found in our study is similar to that of other clinical instruments, the moderate reliability is still cause for concern. A change of 4 points on the DGI has been considered a clinically significant change, 24 yet this difference may reflect measurement error based on the poor interrater reliability of several individual items on the DGI. One attempt has been made to modify the DGI by increasing the number of grading criteria to make them mutually exclusive and exhaustive. 25 Krishnan et al 25 found moderate to excellent intrarater reliability and poor to excellent interrater reliability

6 RELIABILITY OF DYNAMIC GAIT INDEX, Wrisley 1533 of individual DGI items by using videotaped sessions with 3 older adults with balance problems, with scores ranging from 0.0 to 1.0. Most clinical gait assessment tools have been developed to quantify specific gait deviations or to quantify fall risk in older adults. The DGI is the only clinical gait measure that has been validated for use with patients with vestibular disorders. 9 The number of items in the DGI that require head and body turns makes it a very appropriate test for use with patients with vestibular disorders and, therefore, worth modifying to improve the reliability and decrease the ceiling effect. The reliability of the DGI may be improved by providing (1) more objective grading criteria, (2) standardized instructions on administration and grading of the test, (3) training scenarios, and (4) items that more accurately assess the gait deficits seen in vestibular disorders. Based on this research, we have modified the DGI and are currently testing the reliability and validity of our modified DGI. CONCLUSION The DGI, using published instructions and decision rules, showed moderate interrater reliability with subjects with vestibular disorders. The DGI should be used with caution in this population at this time, because of the lack of strong reliability. Without sufficient reliability, the clinical significance of changes in scores of this functional gait assessment is unclear. Acknowledgments: We thank Marlene Kuntz, PT, for her assistance in data collection, Bon Secours DePaul Medical Center for its providing facilities and equipment; and Susan L. Whitney, PhD, PT, NCS, ATC, for her assistance in reviewing the manuscript. This research was completed as partial fulfillment of Dr. Wrisley s Postprofessional Master of Science Degree in Physical Therapy at Old Dominion University, Norfolk, VA. References 1. Herdman SJ. Assessment and treatment of balance disorders in the vestibular deficient patient. In: Duncan PW, editor. Balance: proceedings of the APTA Forum. Alexandria (VA): Am Phys Ther Assoc; p Borello-France DF, Whitney SL, Herdman SJ. Assessment of vestibular hypofunction. In: Herdman SJ, editor. Vestibular rehabilitation. Philadelphia: FA Davis; p Krebs DE, Gill-Body KM, Riley PO, Parker SW. Double-blind, placebo controlled trial of rehabilitation for bilateral vestibular hypofunction. Otolaryngol Head Neck Surg 1993;109: Ishikawa K, Edo M, Terada N, Okamoto Y, Togawa K. Gait analysis in patients with vertigo. Eur Arch Otorhinolaryngol 1993; 250: Kubo T, Kumakura H, Hirokawa Y, Yamamoto KI, Imai T, Hirasaki E. 3-D analysis of human locomotion before and after caloric stimulation. Acta Otolaryngol (Stockh) 1997;117: Shumway-Cook A, Woollacott M. Motor control: theory and practical applications. Baltimore: Williams & Wilkins; Shumway-Cook A, Baldwin M, Polissar NL, Gruber W. Predicting the probability for falls in community dwelling older adults. Phys Ther 1997;77: Shumway-Cook A, Gruber W, Baldwin M, Liao S. The effect of multidimensional exercises on balance, mobility and fall risk in community dwelling older adults. Phys Ther 1997;77: Whitney SL, Hudak MK, Marchetti GF. The dynamic gait index relates to self-reported fall history in individuals with vestibular dysfunction. J Vestib Res 2000;10: Folstein M, Anthony JC, Parshard I, Duffy B, Gruengerg EM. The meaning of cognitive impairment in the elderly. J Am Geriatr Soc 1985;33: Jacobson GP, Newman CW. The development of the dizziness handicap inventory. Arch Otolaryngol Head Neck Surg 1990;116: Portney LG, Watkins MP. Foundations of clinical research: applications to practice. East Norwalk (CT): Appleton & Lange; Haley SM, Osberg JS. Kappa coefficient calculation using multiple raters per subject: a special communication. Phys Ther 1989; 69: Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometerics 1977;33: Rothstein JM, Echternach JL. Primer on measurement: an introductory guide to measurement issues. Alexandria (VA): Am Phys Ther Assoc; Hageman PA, Blanke DJ. Comparison of gait of young women and elderly women. Phys Ther 1986;66: Winter DA, Patla AE, Frank JS, Walt SE. Biomechanical walking pattern changes in the fit and healthy elderly. Phys Ther 1990;70: Leiper CI, Craik RL. Relationship between physical activity and temporal-distance characteristics of walking in elderly women. Phys Ther 1991;71: Berg KO, Wood-Dauphinee S, Williams JI, Gayton D. Measuring balance in the elderly: preliminary development of an instrument. Physiother Can 1989;41: Podsiadlo D, Richardson S. The timed Up&Go : a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc 1991;39: Duncan PW, Weiner DK, Chandler J, Studenski S. Functional reach: a new clinical measure of balance. J Gerontol 1990;45: M Cipriany-Dacko LM, Innerst D, Johannsen J, Rude V. Inter-rater reliability of Tinetti balance scores in novice and experienced physical therapy clinicians. Arch Phys Med Rehabil 1997;78: VanSwearingen JM, Paschal KA, Bonino P, Yang J. The modified gait abnormality rating scale for recognizing the risk of recurrent falls in community dwelling elderly adults. Phys Ther 1996;76: Brown KE, Whitney SL, Wrisley DM, Furman JM. Physical therapy outcomes for persons with bilateral vestibular loss. Laryngoscope 2001;111: Krishnan LV, O Kane KS, Gill-Body KM. Reliability of a modified version of the Dynamic Gait Index a pilot study. Neurol Rep 2002;26:8-14. Supplier a. NeuroCom International, 9570 SE Lawnfield Rd, Clackamas, OR