Use of the "Fast Evaluation of Mobility, Balance, and Fear" in elderly community dwellers: validity and reliability.

Key Words: Assessment, Balance, Falling, Function, Risk.

Falls are the leading cause of accidental death in the home,[1] and they are a contributing factor in 40% of the admissions to nursing homes.[2] Identifying elderly persons who are at risk for falling through the use of appropriate screening tools and referring elderly persons who are prone to falls for physical therapy for gait, balance, and strength deficits are important because this intervention appears to be effective in reducing the risk of falling.[3]

In spite of the social and medical consequences of falls and mobility restrictions for many older persons, primary care physicians do not always refer community-dwelling elderly clients for rehabilitation.[4] One reason for the lack of appropriate referral may be that the needs of older persons living in the community are not always clearly delineated by health care professionals. Identifying elderly community dwellers who are at risk for falling could be done by using a comprehensive screening tool that examines known risk factors, assesses physical performance, and evaluates the patient's fear of falling. The ideal tool would be easy to administer and would apply to persons with a wide range of medical conditions (eg, those with orthopedic or neurologic deficits).

Clinical tools that measure some aspect of balance or mobility in elderly people have received much attention in the literature.[5-18] Recent studies[19-24] have also addressed the influence of fear of falling on balance and mobility. None of the instruments described in the peer-reviewed literature, however, in our opinion, enable clinicians to integrate risk-factor assessment, evaluation of physical performance, and self-assessment of the factors that impair the performance of activities of daily living.

Clinical balance and mobility assessment tools that can be used in the home usually involve either a performance-oriented assessment of balance and mobility[7,912,14,25] or an assessment of the underlying mechanisms that might contribute to balance dysfunction.[5,6,26] Performance-oriented balance assessments[7,90,10,12,27] require people to perform various activities (eg, stand from a sitting position, turn while standing) while the therapist rates the level of performance based on a predetermined time or distance requirement[7,8,12] or determines a score based on a qualitative index of performance (eg, "normal," "adaptive," "abnormal").[10] In contrast to rating the performance of functional activities, the impairments underlying balance or mobility deficits can be evaluated by assessing the patterns of sensory dependence for balance derived from timed stance tests during distortion of the sensory environment (eg, the Clinical Test of Sensory Interaction on Balance [CTSIB]).[5,6]

Some performance-oriented balance assessments[12,23,29] as well as the CTSIB[6] are predictive of falls among elderly community dwellers. Tinetti et al[28] assessed the frequency of falls in 336 older persons living in the community (mean age=78.0 years, SD=5.1) through phone contacts every other month for a year. Thirty-two percent of the subjects (n=108) had fallen at least once during the study period. An increase in the number of abnormalities in balance or gait determined from a performance-oriented mobility assessment (eg, unsteady sitting, turning, or loss of balance following a nudge to the sternum) contributed to an increase in the relative risk of falling. Relative risk is the likelihood that someone with a balance or mobility deficit will fall compared with someone without the deficit. A relative risk of 1.0 means that the balance or mobility deficit does not increase the risk of falling. The relative risk was 1.0 with 0 to 2 abnormalities, 1.7 with 3 to 5 abnormalities, and 2.5 with 6 to 7 abnormalities. Berg et al(29) found similar results in a longitudinal study of a performance-oriented balance assessment with 113 elderly subjects (mean age=83.5 years, SD=5.3). They reported a relative risk of 2.7 for multiple falls over the next 12 months if subjects scored less than 45 points on the Berg Balance Scale. Duncan et al(12) reported that elderly men who were unable to reach 15.2 cm (6 in) were likely to have fallen two or more times within 6 months of testing. Twenty-six of 191 subjects in their cohort were classified as recurrent fallers, and the odds ratio adjusted for age, depression, and cognitive impairment was 8.07.[12] The odds ratio provides an estimate of relative risk. Di Fabio and Anacker[6] studied 47 elder persons (mean age=80.5 years, SD=9.0, range=65-96), using the CTSIB. Thirty-four percent of the subjects (n=16) fell at least twice within 6 months prior to data collection. For those subjects who scored below an average of 81 seconds during trials involving stance on a foam pad, the estimated relative risk of falling was reflected by an age-adjusted odds ratio of 8.67.

Although tests of physical performance and the underlying sensory interaction for balance are predictive of fall risk, the narrow focus of each of these tests limits the assessment of fall risk to unidimensional entities (eg, "physical performance," "sensory integration"). This limitation creates a problem for health care professionals assessing fall risk, because known risk factors for falling[3,28] or restricted mobility[20,30] are not measured. In addition, it is difficult to develop a comprehensive care program that targets the multiple causes of falls without a broad survey of risk factors. The use of several instruments to separately assess physical performance,[10,12,27] strength,[31,32] sensory systems,[33,34] and the influence of fear of falling on mobility[19-22,24] or other responses about performing mobility tasks--while a potential solution to the problem--is cumbersome and time consuming. A comprehensive screening tool developed by Arroyo and colleagues[35] was designed to address this problem by integrating risk-factor assessment, an evaluation of physical performance, and the patient's response to mobility performance.

Arroyo et al[35] introduced a tool referred to as the "Fast Evaluation of Mobility, Balance, and Fear" (FEMBAF) baseline questionnaire. The FEMBAF consists of three components: (1) an assessment of 22 factors that could place a person at risk for falling, (2) evaluation of the ability to complete 18 functional tasks, and (3) reports of fear, pain, mobility difficulty, and the perception of strength deficits for each of the 18 items in the performance-oriented assessment (Appendix). A preliminary analysis of fall risk among 241 elderly community dwellers (mean age=77.5 years, SD=7.9) was done using the FEMBAF in a case-controlled experimental design study.[35] Fifty-nine percent of the subjects (n=142) reported falling at least once in the year preceding the study. Arroyo and colleagues found that elderly persons with a previous history of falling had more complaints of fear, pain, lack of strength, and mobility difficulty during the 18-item performance-oriented balance assessment compared with elderly persons with no recent history of falling. Each activity in the performance-oriented assessment was scored on a three-point scale, and a maximum score of 54 indicated the best possible performance. Arroyo et al suggested that scores between 35 and 45 represented "moderate fall risk," whereas scores below 35 were proposed as the range of "severe fall risk."

The components of the FEMBAF are integrated to form a single tool that can be administered in about 15 minutes. The validity and reliability of measurements obtained with the FEMBAF, however, have not been reported. In addition, it is not known how the assessment of fall risk on the FEMBAF compares with other tests of mobility or sensory integration for balance.[6,28] The purpose of our study was to evaluate the concurrent validity and reliability of scores on the FEMBAF as a clinical index of functional ability and fall risk.

Method

Subjects

Participants were chosen sequentially from the referrals for home care services to a home health agency located in the Minneapolis-St Paul (Minn) area. Referrals were received from local clinics and hospitals. Once the referral for a physical therapy evaluation was received, the patients were evaluated at their residence. All patients meeting the following inclusion criteria were invited to participate: (1) over 65 years of age, (2) living at home or in a community-based assisted living facility, (3) ambulatory (with or without assistive device), (4) not enrolled in a hospice program, and (5) having a Folstein Mini-Mental State Examination[36] score greater than 20. Folstein and Folstein developed the Mini-Mental State Examination to evaluate the cognitive aspects of mental function.[36] This tool is suited to on-site use in patient's home. Standardization of the test on 206 people with and without cognitive impairment indicated that scores of 20 or less was found in patients with dementia, delirium, schizophrenia, or affective disorders and not in elderly people without mental disorders or people with neurosis and personality disorders.

Thirty-five of the 40 patients who were interviewed met the inclusion criteria. Five patients had Folstein Mini-Mental State Examination scores below 20 and were excluded from the study. We were seeking older persons without cognitive impairment. Elderly persons with impaired cognitive ability represent a different population that is already known to be at risk for falling and sustaining serious injury from falls.[37]

The characteristics of the patients who met the inclusion criteria and participated in the study are summarized in Table 1. Patients with a primary diagnosis of neurologic deficit included those with stroke (n=6), multiple sclerosis (n=1), and tumor or brain injury (n=2). The orthopedic category (Tab. 1) included patients with hip fracture (n=6), rib or humeral fracture (n=2), hip arthroplasty (n=2) or knee arthroplasty (n=3), and fractured vertebra secondary to osteoporosis (n=3). Each subject signed an informed consent form prior to the initiation of testing.

Table 1.
Characteristics of the Subjects

                      Age(y)

                   N    X     SD    Median   Range
All subjects      35   79.9   8.5    81.9    60-92

Gender
  Male             4   68.3   3.0    67.6    66-73
  Female          31   81.3   7.8    82.5    60-92

Living status
(alone)
  Yes             18   82.2   8.4    83.9    66-92
  No              17   77.4   8.1    81.0    60-88

Diagnosis
  Neurologic      9    74.9   7.1    74.2    66-84
  Orthopedic     16    80.1   8.1    82.3    68-91
  Weakness       10    82.8   8.9    83.8    60-92

                Folstein Mini-Mental State
                Examination(36)

                 X      SD    Median   Range
All subjects

Gender          29.2    1.2     30     25-30
  Male          30.0    0       30     0
  Female        29.1    1.2     30     25-30

Living status
(alone)
  Yes           29.2    1.4     30      25-30
  No            29.3    0.9     30      27-30

Diagnosis
  Neurologic    29.3    0.9     30      28-30
  Orthopedic    29.5    1.3     30      25-30
  Weakness      28.7    1.2     29      27-30

Raters and Reliability

All tests were administered by a single physical therapist who had 8 years of experience (7 years in the home care field). Testing of intrarater or interrater reliability using a test-retest design was not feasible because the subjects generally could not tolerate repeated examinations on the same day. Testing subjects on different days was not considered a viable option because of the potential effects of maturation. Interrater reliability, therefore, was assessed with the physical therapist and a physical therapist assistant (with 4 of 5 years of experience in home care). Five subjects were randomly selected from the sample. The physical therapist administered and scored the test while the physical therapist assistant observed and simultaneously scored the test during the same session. There was no discussion between raters during the evaluation. The testers were blind to the determination of fall risk derived from any of the tests that were given to each subject.

Outcome Assessment

The FEMBAF was used as the outcome measure, and the following components were assessed: (1) number of risk factors, (2) task completion and risk of falling, and (3) fear, pain, mobility, and strength.

Number of risk factors. The subjects were evaluated on 22 items, which were scored in a dichotomous fashion ("yes" or "no") (Appendix). All affirmative conditions were tallied, and this count provided a relative index of the number of risk factors that could contribute to falling. The risk-factor assessment was based on observation, patient report, and information in the medical chart.

Task completion and risk of falling. Each subject was then asked to perform 18 asks. Each task was scored according to the subject's ability to complete the task (3=task successfully completed without imbalance, 2=task initiated but unsteady or partially completed, 1=unable to perform or initiate task). The best possible score was 54. The assessment of fall risk suggested by Arroyo et al[35] was normal ([is greater than] 45), moderate fall risk (35-45), and severe fall risk ([is less than] 35).

Assessments of fear, pain, mobility, and strength. During each task, the subjects were asked whether fear (Are you fearful of falling?), pain (Does this movement hurt you?), difficulty moving (Is it difficult for you to get started and keep moving?), or lack of strength (Do you feel weak during the motion?) hindered task performance. Each affirmative answer was tallied as a "complaint" that potentially affected the subject's ability to complete the task. The number of "complaints" within each category (fear, pain, mobility, and strength) were evaluated as separate outcome variables.

Other Measures of Balance Ability

Three other measures of balance ability were used to evaluate the concurrent validity of the FEMBAF: (1) the balance subscale of the Tinetti Performance-Oriented Mobility Assessment[10] (B-POMA), (2) the CTSIB, and (3) the Timed Up and Go Test.[11]

B-POMA. This test consists of 13 tasks that are scored based on preestablished qualitative criteria.[10] The score for each task can be 2 (normal), 1 (adaptive), or 0 (abnormal). For example, the rating of a patient's response to a nudge on the sternum could be "steady, able to withstand pressure (normal)," "needs to move feet but able to maintain balance (adaptive)," or "begins to fall or needs assistance from examiner to maintain balance (abnormal)." Interrater reliability for aggregate scores on the gait and balance subscales of the Tinetti Performance-Oriented Mobility Assessment is r=.95.[38] Regarding validity, the B-POMA is highly predictive of falls and fall-related injuries in elderly community dwellers.[23,39,40] Five of the B-POMA tasks were the same as tasks that were included in the FEMBAF (Appendix). The B-POMA, however, had criteria developed specifically for each test item, whereas the FEMBAF used one rating system for all test items. Those items from the B-POMA that were identical to the FEMBAF, therefore, were scored twice (first using FEMBAF criteria for task completion to avoid a bias from exposure to the B-POMA ratings). That is, a single attempt at completing the task received two scores.

CTSIB. The CTSIB is a timed balance test that requires the patient to stand on a firm or compliant (foam) surface with eyes open, with eyes closed, or with the head inside a "visual dome."[5,6] There is a maximum score of 30 seconds per trial, 90 seconds per condition (summed across three trials), and 540 seconds for the composite score summed across all conditions and trials. The CTSIB is a reliable tool and provides valid measurements reflecting the sensory influences on postural control among elderly community dwellers.[5,6] Anacker and Di Fabio[5] reported a test-retest correlation for the CTSIB of r=.75, with 95% agreement of the composite score between sessions. The kappa (K) for the composite score was reported to be .77.[26] Di Fabio and Anacker[6] reported that the composite score for identifying fallers (cut-point=260 seconds) had a sensitivity of 44% and a specificity of 90%. When the average score of compliant-surface conditions was used as the boundary of normal/abnormal sensory integration, the sensitivity was 75% and the specificity was 65%. With an average score below 81 in the compliant-surface conditions, the estimated relative risk of falling was 8.67. We, therefore evaluated the FEMBAF against the CTSIB score averaged for the three compliant-surface stance conditions.

Timed Up and Go Test. A version of the Up and Go Test using qualitative descriptions of performance[41] was found to have weak concurrent validity.[5] We decided, therefore, to use the timed version of the test.[11] The Timed Up and Go Test requires a patient to stand up from sitting in a chair, walk 3 m, turn around, return to the chair, and resume a sitting position. Intrarater and interrater reliability have been reported as excellent (intraclass correlation coefficient=.99 for each type of reliability). Berg et al[25] demonstrated concurrent validity by correlating the Up and Go Test with the Berg Balance Scale (r= -.76) in a group of 31 elderly subjects living in residential care facilities. The Up and Go Test was the same as one item included in the FEMBAF performance-oriented assessment (Appendix). The Timed Up and Go Test, however, is a timed test, whereas the FEMBAF uses a three-point rating scale (described earlier) to evaluate task performance. The tester, therefore, scored the item twice, first using the FEMBAF criteria to avoid bias from exposure to the results of the Timed Up and Go Test.

Procedure

The testing was always done in the same order. The Folstein Mini-Mental Examination was given initially, followed by the FEMBAF. The Timed Up and Go Test and five items of the B-POMA were nested within the FEMBAF task-completion section as already described. Equipment needed for the FEMBAF were a chair with armrests, stairs, and cardboard 10 cm wide x 15 cm high.

The CTSIB was administered last. The testing was done on a hard surface of either linoleum or wood. The participants removed their shoes and assumed a posture of standing with malleoli touching, forefeet turned out 30 to 40 degrees, and arms crossed over the chest. Three conditions were timed in the following order: stance with eyes open, stance with eyes closed, and stance wearing a "visual dome." Each condition was then repeated during stance on high-density foam (7.62 x 50.8 x 50.8 cm, with a specific weight of 32.04 kg/[m.sup.3] and a compression of 31.75 kg). Compression is the amount of weight that will compress the pad to 75% of the original height. To assist in uniform foot placement during the compliant-surface stance conditions, an outline of feet was used on the foam to delineate proper foot position. Some participants had difficulty achieving this position because of posture changes secondary to cerebrovascular accident, so a posture as close to the one described was attempted. Stance was timed up to 30 seconds. If the subjects successfully completed the trial, they received 30 points (seconds) for each of the remaining trials. If the subjects moved the arms off the chest, took a step, flexed one or both knees, or moved heels or toes off the foam base, the timer was stopped and trials 2 and 3 were initiated.

Data Analysis

Description of risk factors. The number of patients with each risk factor (Appendix) was plotted for descriptive analysis.

Concurrent validity. Concurrent validity of each component of the FEMBAF was established by calculating Spearman rank-order correlation coefficients between scores from the FEMBAF and scores from the other measures of balance ability (B-POMA, CTSIB, and Timed Up and Go Test). The statistical significance of the correlation coefficients ([H.sub.0] r=0) was evaluated by converting the coefficient to a t statistic.[42] This procedure allowed us to test the null hypothesis of r=0 on a two-tailed t distribution with n-2 degrees of freedom. The smallest correlation coefficient that would still be significantly different from 0 was r=.35 (P[is less than] .05). All correlation coefficients equal to or greater than .35, therefore, were statistically significant.

Balance, living status, and diagnostic category. To determine whether the differences detected by the other three measures of balance ability within selected stratifications of the cohort were also detected by the FEMBAF, subjects were grouped according to living status (living alone or not alone) and diagnostic category (Tab. 1). A one-way analysis of covariance (ANCOVA) was done across each stratification for each outcome measure. The subjects' age was used as a covariate because a previous study[5] showed that stance duration on timed balance tests decreases (linearly) as age increases among nondisabled elderly community dwellers.

Descriptive comparison of fall risk. The score on the task-completion section of the FEMBAF was used to assign subjects to a fall-risk category. The three-level risk classification suggested by Arroyo et al[35] (ie, normal, moderate, severe) was collapsed to form a dichotomous variable (normal versus at risk) so that direct comparisons could be made with other balance assessments. Scores on the task-completion component of the FEMBAF greater than 45 were considered normal, and scores less than or equal to 45 were considered to indicate a risk for falling. This "cut-point" placed all subjects with moderate or severe fall risk (using Arroyo and colleagues' original classification[35]) into a "risk" category.

Tinetti et al[28] identified seven activities in their performance-oriented assessment that showed the greatest prevalence in their study group and reflected the highest relative risk of falling. They collapsed a three-point scale (ie, normal, adaptive, abnormal) to create a dichotomous assessment of performance on each activity (normal versus abnormal). Abnormalities on zero to two activities showed no relative risk of falling (1.0), whereas abnormalities on three to five activities had a relative risk of falling of 1.7. Three of the seven activities involved an assessment of gait and were not included in our study. We therefore used the remaining four activities (sitting down, stance on one leg, turning, and a nudge to the sternum) to estimate fall risk. A fall-risk index was estimated from the B-POMA by determining the number of abnormal responses and assigning risk categories as normal (zero to two abnor-malities) or at risk (three or four abnormalities).[28] For the CTSIB, it was previously determined that an average stance duration score of less than 81 seconds for three compliant-surface (foam) conditions distinguished fallers from nonfallers.[6] We therefore used this cut-point to classify subjects who were at risk for falling. The number of subjects classified as "at risk" and "not at risk" was plotted to provide a descriptive comparison of fall risk for each tool.

Reliability. Kappa coefficients were calculated to determine the chance-corrected percentage of agreement between raters for each test item. Kappa coefficients were averaged to provide a composite reliability coefficient for the risk-factor and task-completion components of the FEMBAF.

Results

Description of Risk Factors

The prevalence of risk factors obtained from the FEMBAF is summarized in Figure 1. Eighty-nine percent of the subjects (n=31) used assistive devices for ambulation, and 83% of the subjects (n=29) reported falling at least one time during the past year. In addition, 86% of the subjects (n=30) had pathology that was likely to induce falls, and 94% of the subjects (n=33) were taking medications that were potentially dangerous with regard to falls. Although 63% of the subjects (n=22) limited their activities to basic activities of daily living at home, only 23% of the subjects (n=8) reported that fear of falling was the limiting factor.

[Figure 1 ILLUSTRATION OMITTED]

Concurrent Validity

The correlations between the FEMBAF and the other measures of balance ability are summarized in Table 2. Higher scores on task completion, indicating greater proficiency, correlated with higher (better) scores on the B-POMA (Tab. 2). As the number of risk factors or the number of tasks performed poorly due to perceived lack of strength or mobility problems increased, the B-POMA score decreased, indicating a decrement in balance function (Tab. 2).

Table 2. Spearman Rank-Order Correlation Coefficients for Each Component of the Fast Evaluation of Mobility, Balance, and Fear[35] (FEMBAF) Versus the Other Measures of Balance Ability (N=35)

                                               Timed Up and
FEMBAF                 B-POMA(a)   CTSIB(b)   Go Test(11)

Risk factors           -.69(c)     -.46(c)      37(c)
Task completion         .91(c)      .54(c)    -.38(c)
Fear complaints        -.26-       -.32       -.02
Pain complaints        -.01        -.18        .01
Mobility complaints    -.58(c)     -.24        .60(c)
Strength complaints    -.84(c)     -.56(c)     .42(c)

(a) B-POMA=balance subscale of the Tinetti Performance-Oriented Mobility Assessment.(10)

(b) CTISB= Clinical Test of Sensory Interaction on Balance[5,6] (compliant-surface [foam] stance conditions only).

(c) p < .05.

Longer stance duration on the CTSIB (average of scores for the compliant-surface stance conditions) also had an association with better task-completion scores (Tab. 2). In contrast, shorter stance duration (indicating a decrement in balance function) was associated with a greater number of risk factors or an increase in the number of tasks performed poorly due to perceived lack of strength (Tab. 2).

The magnitude of the association between the scores from the Timed Up and Go Test and the scores from the FEMBAF was low, overall, compared with the magnitude of association between the other measures of balance ability and the FEMBAF, but several relationships still achieved statistical significance. There was a positive association between the Timed Up and Go Test score and the number of risk factors, mobility, and strength complaints (ie, longer transit time during the Timed Up and Go Test was associated with more risk factors or complaints). In addition, a low proficiency in the FEMBAF task completion was associated with a prolonged duration for completing the Timed Up and Go Test.

Balance, Living Status, and Diagnostic Category

FEMBAF and living status. When corrected for age, the number of strength deficits perceived to affect performance was lower for subjects who lived alone than for subjects who did not live alone (F=5.57; df=1,32; P=.03; Fig. 2). The task-completion scores were higher for subjects who lived alone than for subjects who did not live alone (F=3.86; df=1,32; P=.058; Fig. 2). There were no differences in the number of risk factors, fear, pain, or mobility complaints between subjects who lived alone and subjects who did not live alone (Tab. 3, Fig. 2).

[Figure 2 ILLUSTRATION OMITTED]

Table 3. Means and Standard Deviations of Outcome Variables by Living Status and Diagnostic Category

                                    Living Status

                               Not Alone        Alone

                              X         SD      X      SD

FEMBAF(a)
  Risk factors               11.3       3.4     9.1    5.0
  Task completion            34.2       6.4    39.3    6.8
  Fear complaints             4.2       3.8     3.9    4.0
  Pain complaints             1.7       3.6     1.8    2.6
  Mobility complaints         8.0       5.4     6.6    5.3
  Strength complaints        12.5       3.4     8.6    5.0

Other measures of balance
    ability
  B-POMA(b)                  16.5       6.2     19.5   4.6
  CTSIB(c)                   51.9       32.9    57.0  33.1
  Timed Up and Go
    Test,[11](d)             25.9       22.3    30.8  19.7

                                  Diagnostic Category

                                Neurologic      Orthopedic

                              X          SD     X       SD

FEMBAF(a)
  Risk factors               13.3       2.6     9.5     4.8
  Task completion            29.8       5.5     39.4    6.0
  Fear complaints             5.8       4.9     3.2     3.8
  Pain complaints             2.0       4.1     2.6     3.1
  Mobility complaints         9.4       6.4     7.8     4.9
  Strength complaints        15.1       2.9     8.5     4.2

Other measures of balance
    ability
  B-POMA(b)                  11.7        4.9    19.9     4.5
  CTSIB(c)                   29.0       30.1    61.5    31.2
  Timed Up and Go
    Test,[11](d)             32.3       32.0    24.6    14.3

                                Weakness

                             X          SD

FEMBAF(a)
  Risk factors               8.4        4.1
  Task completion           39.0        5.7
  Fear complaints            3.8        2.7
  Pain complaints            0.1        0.3
  Mobility complaints        4.6        4.2
  Strength complaints        9.5        3.8

Other measures of balance
    ability
  B-POMA(b)                  20.8        2.5
  CTSIB(c)                   66.4       23.8
  Timed Up and Go
    Test,[11](d)             30.8       20.0

(a) FEMBAF=Fast Evaluation of Mobility, Balance, and Fear.[35] Risk factors=number of risk factors tallied. Task completion=score on 18 functional tasks, with each task rated front 1 (worst performance) to 3 (best performance). Fear, pain, mobility, and strength complaints=number of complaints tallied during each functional task.

(b) B-POMA=balance subscale of the Tinetti Performance-Oriented Mobility Assessment.[10] Values are total scores on the B-POMA, with each of the 13 items rated from 0 (worst performance) to 2 (best performance)

(c) CTSIB=Clinical Test of Sensory Interaction on Balance,[5,6] measured as stance time (in seconds) averaged for three compliant-surface (foam) conditions.

(d) Timed Up and Go Test measured as time (in seconds) to stand from a sitting position, walk 3 m, and resume sitting in a chair.

Other measures of balance ability and living status. There were no differences in B-POMA, CTSIB, or Timed Up and Go Test scores between subjects who lived alone and subjects who did not live alone (Tab. 3, Fig. 3).

[Figure 3 ILLUSTRATION OMITTED]

FEMBAF and diagnostic category. When corrected for age, the outcomes across diagnostic categories showed that the number of risk factors (F=3.35; df=2,31; P=.048) and the number of perceived strength deficits (F=7.69; df=2,31; P [is less than] .001) were greatest for subjects with neurologic diagnoses compared with subjects with orthopedic conditions or generalized weakness (Fig. 4).

[Figure 4 ILLUSTRATION OMITTED]

The rate of task completion was lower in the neurologic category (F=7.51; df=2,31; P .002) than in all other diagnostic categories. There were no differences in the number of complaints of fear, pain, or mobility deficit across diagnostic categories.

Other measures of balance ability and diagnostic categories. The B-POMA scores were lower (F=15.14; df=2,31; P [is less than] .001) and stance duration during the CTSIB was shorter (F=7.79; df=2,31; P [is less than] .001) for subjects with neurologic conditions than for subjects with orthopedic conditions or subjects with generalized weakness (Fig. 5). There were no differences among diagnostic categories with respect to the Timed Up and Co Test scores.

[Figure 5 ILLUSTRATION OMITTED]

Descriptive Comparison of Fall Risk

The FEMBAF classified 31 of 35 subjects as being at risk for falling. The B-POMA and the CTSIB classified 15 and 22 subjects, respectively, as being at risk for falling (Fig. 6).

[Figure 6 ILLUSTRATION OMITTED]

Reliability

There was high interrater agreement on the determination of risk factors (mean [Kappa]=.95, SD=.15) and task completion (mean [Kappa]=.96, SD=.12).

Discussion

The FEMBAF appears to provide valid and reliable measurements of balance, mobility, and fall risk in a group of elderly community dwellers who did not have cognitive impairments. There were correlations between several components of the FEMBAF (number of risk factors, task completion, strength, and mobility complaints) and the other measures of balance ability (Tab. 2). The number of fear or pain complaints during task performance did not show an association with any of the other measures of balance ability (Tab. 2).

A general fear of falling was expressed by 37% of the cohort, but only 23% of the subjects indicated that fear limited their activities (Fig. 1). The disassociation of "fear of falling" from functional performance has been documented by Tinetti et al.[19] In a study of more than 1,000 elderly persons living in the community, Tinetti and colleagues found that fear of falling was not associated with impairments of higher-level physical or social functioning (eg, home repair, yard work, sports participation) and that fear of falling was only marginally associated with activities of daily living.

It was clear that multiple factors influence mobility proficiency and fall-avoidance behavior, because the number of risk factors identified by the FEMBAF had a relationship to the outcome on each of the other measures of balance ability (Tab. 2). These findings support the notion that multiple factors contribute to fall risk.[28,39] One implication of these findings is that the FEMBAF might be a useful screening tool because it accounts for the "additive" effects of multiple disabilities on falling.

The identification of modifiable risk factors is an important aspect of developing effective strategies for therapeutic interventions to improve mobility and prevent injurious falls.[3,43] Tinetti et al[3] described several modifiable risk factors, and many of these factors are "scored" on the FEMBAF (ie, postural hypotension; use of sedatives; impairments in balance, gait, and strength). The risk-factor component of the FEMBAF identified and provided a count of the factors that might contribute to falling, but fall risk was determined by the task-completion score on the FEMBAF (Appendix). More subjects were identified as being at risk for falling on the FEMBAF compared with the B-POMA or CTSIB (Fig. 6). One possible reason for these differences might be that the FEMBAF incorporates more challenging balance tasks (eg, jumping, climbing stairs, standing from a kneeling position) than does the B-POMA or the CTSIB. The "ceiling" effect that might be expected with activities that do not challenge balance, therefore, was minimized with the FEMBAF. When the cohort was stratified according to living status, there was a reduction in the number of strength complaints (Fig. 2) and the task-completion scores tended to be higher for subjects who lived alone than for subjects who did not live alone (Fig. 2). These findings suggest that independent living requires a high level of functional competence. There were no differences when outcomes on the other measures of balance ability were evaluated with respect to living status. The comparison measures, however, tended to show better scores for subjects who lived alone than for subjects who did not live alone (Fig. 3). The scores of the measures of balance ability indicating better performance for subjects who lived alone (Fig. 3), therefore, were in the same direction as the FEMBAF task-completion scores for this group of subjects (Fig. 2).

Elderly persons who return home from the hospital following inpatient treatment for a neurological deficit may have up to three times the risk for falling compared with elderly persons without neurological deficits living in the community.[44] With respect to diagnostic category, the FEMBAF, B-POMA, and CTSIB showed poorer outcomes for subjects with primarily neurologic dysfunction than for subjects with orthopedic-related disorders or generalized weakness (Figs. 4, 5). The consistency of findings from the FEMBAF and each of the other measures of balance ability across diagnostic categories provides additional support for the validity of the FEMBAF.

Limitations

This study was a preliminary demonstration of the usefulness of a new screening tool that can be used to identify risk factors and functional deficits in elderly persons living in the community. Whether this tool will help clinicians modify the care of clients living in the community in order to prevent injury due to falls remains to be determined. The level of disease severity of the subjects in our study required us to limit the rigor of the reliability test. The design for evaluating reliability was restricted to interrater reliability, and one of two raters was required to score the test strictly as an observer. Additional research is needed to show the predictive capacity of the FEMBAF. In addition, one rater did not interact with the person being measured, which eliminated a source of error that would be present when the instrument is normally used. A more complete description of reliability (eg, in the form of a test-retest design) was not feasible.

Conclusions

Concurrent validity of the measurements from the FEMBAF was evident from associations with each of the other measures of balance ability in the areas of risk-factor assessment, task completion, mobility, and strength complaints. Differences in performance across diagnostic categories were detected by the measures of balance ability as well as by the FEMBAF. The interrater reliability of the measures was excellent, with interrater chance-corrected agreement on the order of 95%. The FEMBAF may enable practitioners to identify patients who are at risk for falling or mobility dependence, and it provides a format for delineating risk factors that are known to respond to treatment.

References

[1] Lamb K, Miller J, Mernadez M. Falls in the elderly: causes and prevention. Orthopedic Nursing 1987;6:45-49.

[2] Kellogg International Work Group on the Prevention of Falls by the Elderly. Dan Med Bull. 1987;34 (suppl 4):1-24.

[3] Tinetti ME, Baker DI, McAvay G, et al. A multifactorial intervention to reduce the risk of falling among elderly people living in the community. N Engl J Med. 1994;331:821-827.

[4] Hoenig H, Mayer-Oaks SA, Siebens H, et al. Geriatric rehabilitation: What do physicians know about it and how should they use it.? J Am Geriatr Soc. 1994;42:341-347.

[5] Anacker SL, Di Fabio RP. Influence of sensory inputs on standing balance in community-dwelling elders with a recent history of falling. Phys Ther. 1992;72:575-581.

[6] Di Fabio RP, Anacker SL. Identifying fallers in community-living elders with a clinical test of sensory interaction for balance. Eur J Phys Med Rehabil. 1996;6:61-66.

[7] Rossiter-Fornoff JE, Wolf SL, Wolfson LI, et al. A cross-sectional study of the FICSIT common database static balance measures. J Gerontol. 1995;50A:M291-M297.

[8] Berg K, Wood-Dauphinee S, Williams JI. The balance scale: assessment with elderly residents and patients with acute stroke. Scand J Rehabil Med. 1995;27:27-36.

[9] Mahoney FI, Barthel DW. Functional evaluation: the Barthel index. Md Med J. 1965;14:61-65.

[10] Tinetti ME. Performance-oriented assessment of mobility problems in elderly patients. J Am Geriatric Soc. 1986;34:119-126.

[11] Podsiadlo D, Richardson S. The timed "up & go": a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc. 1991;39: 142-148.

[12] Duncan PW, Studenski S, Chandler J, Prescott B. Functional reach: predictive validity in a sample of elderly male veterans. J Gerontol. 1992;47:M93-M98.

[13] Gabell A, Simons MA. Balance coding. Physiotherapy. 1983;68:286-288.

[14] MacKnight C, Rockwood K. A hierarchical assessment of balance and mobility. Age Ageing. 1995;24:126-130.

[15] Di Fabio RP. Reliability and validity of functional assessment in patients with stroke. Journal of Neurologic Rehabilitation. 1990;4: 145-152.

[16] Harada N, ChiU V, Damron-Rodriquez J, et al. Screening for balance and mobility impairment in elderly individuals living in residential care facilities. Phys Ther. 1995;75:462-469.

[17] Hu MH, Woollacott MH. Multisensory training of standing balance in older adults, II: kinematic and electromyographic postural responses. J Gerontol. 1994;49:M62-M71.

[18] Wolfson LI, Whipple RH, Amerman P, et al. Stressing the postural response: a quantitative method for testing balance. J Am Geriatr Soc. 1986;34:845-850.

[19] Tinetti ME, Mendes de Leon CF, Doucette JT, Baker DI. Fear of falling and fall-related efficacy in relationship to functioning among community-living elders. J Gerontol. 1994;3:M140-M147.

[20] Howland J, Peterson EW, Levin WC, et al. Fear of falling among the community-dwelling elderly. Journal of Aging and Health. 1993;5:229-243.

[21] Powell LE, Myers AM. The activities-specific balance confidence (abc) scale. J Gerontol. 1995;50A:M28-M34.

[22] Tinetti ME, Richman D, Powell LE. Falls efficacy as a measure of fear of falling. J Gerontol. 1990;45:P239-P243.

[23] Maki BE, Holliday PJ, Topper AK. Fear of falling and postural performance in the elderly. J Gerontol. 1991;46:M123-M131.

[24] Arfken CL, Lach HW, Birge SJ, Miller JP. The prevalence and correlates of fear of falling in elderly persons living in the community. Am J Public Health. 1994;84:565-570.

[25] Berg KO, Maki BE, Williams JI, et al. Clinical and laboratory measures of postural balance in an elderly population. Arch Phys Med Rehabil. 1992;73;1073-1080.

[26] Di Fabio RP, Badke MB. Relationship of sensory organization to balance function in patients with hemiplegia. Phys Ther. 1990;70:542-548.

[2]7 Berg KO, Wood-Dauphinee SL, Williams JI, Gayton D. Measuring balance in the elderly: preliminary development of an instrument. Physiotherapy Canada. 1989;41:304-311.

[28] Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons living in the community. N Engl J Med. 1988;319:1701-1706.

[29] Berg KO, Wood-Dauphinee SL, Williams JI, Maki BE. Measuring balance in the elderly: validation of an instrument. Clan J Public Health. 1992;83:s7-s11.

[30] Gill TM, Williams CS, Tinetti ME. Assessing risk for the onset of functional dependence among older adults: the role of physical performance. J Am Geriatr Soc. 1995;43:603-609.

[31] Wolfson L, Judge J, Whipple R, King M. Strength is a major factor in balance, gait, and the occurrence of falls. J Gerontol. 1995;50A(Special Issue):64-67.

[32] Judge JO, King MB, Whipple R, et al. Dynamic balance in older persons: effects of reduced visual and proprioceptive input. J Gerontol. 1995;50A:M263-M270.

[33] Lord SR, Ward JA, Williams P, Anstey K. Physiological factors associated with falls in older community-dwelling women. J Am Geriatr Soc. 1994;42:1110-1117.

[34] Lord SR, Clark RD, Webster IW. Visual acuity and contrast sensitivity in relation to falls in an elderly population. Age Ageing. 1991;20: 175-181.

[35] Arroyo JF, Herrmann F, Saber H, et al. Fast evaluation test for mobility, balance, and fear: a new strategy for the screening of elderly fallers. Arthritis Rheum. 1994;37:S416. Abstract.

[36] Folstein MF, Folstein SE, McHugh PR. Mini-Mental State: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189-198.

[37] Nevitt MC, Cummings SR, Hudes ES. Risk factors for injurious falls: a prospective study. J Gerontol. 1991;46:M164-M170.

[38] Tinetti ME, Baker DI, Garrett PA, et al. Yale FICSIT: risk factor abatement strategy for fall prevention. J Am Geriatr Soc. 1993;41:315-320.

[39] Tinetti ME, Williams TF, Mayewski R. Fall risk index for elderly patients based on number of chronic disabilities. Am J Med. 1986;80: 429-434.

[40] Robbins AS, Rubenstein LZ, Josephson KR, et al. Predictors of falls among elderly people: results of two population-based studies. Arch Intern Med. 1989;149:1628-1633.

[41] Mathias S, Nayak USL, Isaacs B. Balance in elderly patients: the "Get-up and Go" Test. Arch Phys Med Rehabil. 1986;67:387-389.

[42] Glass GV, Stanley JC. Statistical Methods in Education and Psychology. Englewood Cliffs, NJ: Prentice-Hall; 1970:316.

[43] Overstall PW. Falls after strokes. BMJ. 1995;311:74-75.

[44] Forster A, Young J. Incidence and consequences of falls due to stroke: a systemic inquiry. BMJ. 1995;311:83-86.

RP Di Fabio, PhD, PT, is Professor, Program in Physical Therapy, Department of Physical Medicine and Rehabilitation, University of Minnesota, UMHC Box 388, 420 Delaware St SE, Minneapolis, MN 55455 (USA) ([e-mail]). Address all correspondence to Dr Di Fabio.

R Seay is a graduate student in the advanced master's degree program at the I University of Minnesota.

This study was approved by the University of Minnesota Human Subjects Committee.

This article was submitted August 30, 1996, and was accepted March 5, 1997