- Email: [email protected]
Use of the Berg Balance Scale to Predict Independent Gait After Stroke: A Study of an Inpatient Population in Japan
PM R 7 (2015) 392-399
www.pmrjournal.org
Original Research
Use of the Berg Balance Scale to Predict Independent Gait After Stroke: A Study of an Inpatient Population in Japan Hyuma Makizako, PhD, PT, Norihito Kabe, MSc, PT, Asami Takano, ST, Kanako Isobe, CW
Abstract Objective: To examine whether the Berg Balance Scale (BBS) score on admission and after 1 month can be used to predict the Functional Independence Measure (FIM) walking level 6 or 7 (modified independent and independent gait respectively) after 3 months and to determine the optimal BBS cut-off score for predicting independent gait among inpatients with first stroke in a rehabilitation facility. Design: Retrospective cohort study. Setting: Inpatient rehabilitation facility (Kaifukuki). Participants: Two hundred fifty-one consecutive patients with first stroke and nonindependent gait on admission. Outcome Measures: BBS on admission and after 1 month and FIM on admission and after 3 months. Results: Area under the receiver operating characteristic curves for predicting independent gait after 3 months were 0.81 (95% confidence interval [CI], 0.75-0.88) for BBS score on admission; the optimal cut-off score being 13 (63% sensitivity; 90% specificity). Achieving greater cut-off score on admission was significantly associated with subsequent independent gait (odds ratio, 9.7; 95% CI, 4.38-21.36; P < .001). A subanalysis of patients with poor balance on admission (BBS score <13; n ¼ 191) showed area under the receiver operating characteristic curves for predicting independent gait after 3 months of 0.88 (95% CI, 0.80-0.95) for BBS scores 1 month after admission. In this subgroup, BBS scores at 1 month of 27 were significantly associated with subsequent independent gait (odds ratio, 21.6, 95% CI, 6.40-73.20; P < .001). Conclusions: Admission BBS scores predicted FIM walking level 6 or 7, which denotes modified independent or independent gait, after 3 months, the optimal cut-off for BBS scores being 13 among inpatients with first stroke in a rehabilitation facility. Patients with poor balance on admission whose BBS scores had improved to 27 at 1 month after admission were likely to achieve modified independent or independent gait within 3 months after admission.
Introduction One of the most important goals of rehabilitation for patients with stroke is improvement in gait, ideally to the point of achieving independent ambulation without requiring physical assistance or supervision by another person [1]. Predictors of recovery of this ability in patients with stroke reportedly include age [2], activities of daily living scores [2,3], differences in muscle strength between the lower extremities [4], and balance [3]. In particular, initial walking speed is the most commonly used indicator of the eventual ambulatory ability of patients with stroke [5-7]. It is important to ascertain whether eventual ambulatory ability can be improved in those poststroke
inpatients in rehabilitation facilities who need assistance to walk on admission. It is difficult, however, to predict on the basis of admission assessments whether inpatients poststroke with severe limitations in walking on admission will reach functional independence. In the Post-Stroke Rehabilitation Outcomes Project, on admission to a rehabilitation facility, more than 60% of inpatients poststroke who received a 6-hour block of physical therapy reportedly had severe limitations in walking, with scores of 1 or 2 on the walk Functional Independence Measure (FIM) item [8]. Thus, identification of predictors of independent gait would allow refinement of admission assessments of inpatients poststroke with severe limitations in walking.
1934-1482/$ - see front matter ª 2015 by the American Academy of Physical Medicine and Rehabilitation http://dx.doi.org/10.1016/j.pmrj.2015.01.009
H. Makizako et al. / PM R 7 (2015) 392-399
The Berg Balance Scale (BBS) has been identified as the most commonly used balance assessment tool across the continuum of stroke rehabilitation [9]. Balance is one of the key components that determines eventual ambulatory ability after stroke [10], and the BBS score predicts most of the variation in this ability among patients admitted with strokes to rehabilitation facilities [3]. In a prospective study, BBS scores of 20 on admission to an inpatient rehabilitation facility indicated that persons with strokes were highly likely to achieve only household ambulation speeds (<0.4 m/s) by the time of their discharge [3]. These findings suggest that balance performance based on BBS scores can predict future gait speed in inpatients poststroke. In addition to gait speed, walking ability with or without physical assistance or supervision is important for discharge planning. In addition, another prospective study has reported that balance performance predicts the mobility of patients with stroke at discharge [11]; however, in this study the duration of hospitalization varied and the length of observation was not defined. To our knowledge, no studies have yet reported optimal cut-off balance scores for predicting independent gait among poststroke patients who, on admission assessment, cannot walk without physical assistance or supervision from another person. A better understanding of the ability of admission balance performance to predict attainment of independent gait within an established rehabilitation period (eg, 3 months) would provide useful information about inpatients with strokes in rehabilitation facilities. We decided that examining changes in BBS score at 1 month after admission would provide the best predictor of gait independence in these patients because functional outcomes, including balance, can reportedly improve within this time frame in inpatients with stroke in early rehabilitation [12]. Therefore, we were interested in investigating the ability of changes in BBS scores after 1 month to predict such as outcomes in the patients with the lowest initial scores. Because of the medical insurance system in Japan, rehabilitation generally occurs in dedicated convalescent rehabilitation wards that are known as the “Kaifukuki.” In 2000, these new interdisciplinary postacute “Kaifukuki” rehabilitation wards were incorporated into the Japanese medical insurance system, and it was anticipated that these changes in national insurance policies would affect the amount and organization of rehabilitation interventions and improve in patient outcomes [13]. In these wards, convalescent inpatients with stroke undergo rehabilitation programs provided by physical therapists, occupational therapists, speech therapists, and clinical nurses. These patients and their families are given comprehensive monthly rehabilitation plans that include information about achieved and planned goals; rehabilitative approaches to achieving the planned goals; discharge planning; and social
393
resources necessary for discharge home [13]. Therefore, it is important to predict whether inpatients with stroke who need physical assistance or supervision from another person to walk on admission will achieve independent gait and approximately how long this will take. The aims of this study were to determine both whether BBS scores predict a modified independent or independent gait (FIM walking level 6 or 7) after 3 months and the optimal BBS cut-off score for such predictions among inpatients of a Kaifukuki rehabilitation facility. We postulated that balance measured by BBS on admission assessment would predict gait ability after 3 months. In addition, we postulated that, among inpatients with poor balance on admission, changes in BBS scores within 1 month might help to predict improvement in ambulatory ability after 3 months. Because balance performance can improve rapidly and dramatically even when balance is poor on admission, we hypothesized that improvement in 1 month BBS scores would predict independent gait. Methods Patients Data for this retrospective cohort study were obtained from records stored in the Funabashi Rehabilitation Hospital database (patients between April 2009 and September 2012). This study included 251 inpatients who have had a stroke with a mean age of 68.4 years (SD 11.1). The inclusion criteria were (1) aged 40 years or older; (2) first stroke (physicians-determined based on magnetic resonance imaging findings and/or history obtained); (3) diagnosis of cerebral hemorrhage (CH) or cerebral infarction; (4) independent before onset; (5) walk item FIM score 5 on admission; and (6) underwent the inpatient rehabilitation program of the Funabashi Rehabilitation Hospital for more than 3 months. Exclusion criteria were diagnoses of subarachnoid hemorrhage and ability to walk independently on admission assessment (walk item FIM score 6). The study was approved by the Ethics Committees of the Funabashi Rehabilitation Hospital. Study Variables Gait independence had been assessed using the relevant subitem of the FIM motor scale [14]. The FIM motor scale is a 13-item instrument with 4 subscales: self-care (eating, grooming, bathing, dressing upper body, dressing lower body, and toileting), sphincter control (bladder and bowel management), transfers (in and out of bed, toilet and tub/shower), and locomotion (walking and stair climbing). Ratings for each item ranged from 1 (total assistance) to 7 (complete independence). Excellent internal consistency of FIM (Cronbach alpha >0.9) has been confirmed in a previous
394
Berg Balance Scale for Gait After Stroke
study [15]. In this study, all patients exhibited their scores on the subitem concerning walking ranged from 1 (total assistance) to 5 (supervision) at admission. Gait independence at 3 months had been assessed by the FIM walking score 6 (6 ¼ modified independence or 7 ¼ complete independence) [16]. Complete independence (score 7) means that subject can safely walk a minimum of 150 ft (50 m) without assistive devices. Modified independence (score 6) means that the subject can walk a minimum of 150 ft (50 m) without supervision but with the support of a brace (orthosis) or leg prosthesis, specially modified shoes, cane, crutches, or walkerette; takes more than a reasonable time; or there are safety considerations [8]. FIM scores 1-5 are determined by the level of physical assistance required for walking, whereas scores 6 and 7 denote independent walking [17]. Thus, there are 2 gross score classifications: dependent (helper required: scores 1-5) and independent (no helper required: score 6-7) [18]. Patients’ ability to balance on admission and 1 month later have been evaluated previously by physical therapists using the BBS [19]. The BBS is a 14-item test that uses a 4-point ordinal scale to quantify performance on tasks such as standing unsupported, standing with eyes open or closed, and turning 360 . Possible total scores range from 0 to 56, with greater scores denoting better balance. Patients’ cognitive function and aphasia had been assessed by speech therapists. Patients without aphasia were considered to have cognitive impairment if their precognitive scores were 23/30 on the Mini-Mental State Examination [20], whereas for patients with aphasia, lower age-adjusted cutoff values of Raven’s Coloured Progressive Matrices Scores [21] were used. In addition, lower limb paresis had been assessed by physical or occupational therapists using the Brunnstrom motor recovery stage [22]. Rehabilitation All patients in this study had undergone rehabilitation programs administered by physical therapists, occupational therapists, and speech therapists. Rehabilitation programs were based on a comprehensive approach and included physical therapy, occupational therapy, and speech therapy. Patients were provided 6-9 units (1 unit: 20 minutes) per day as necessary. Statistical Analysis Patients were classified as independent (n ¼ 65) or nonindependent (n ¼ 186) based on their ambulatory ability 3 months after admission. Means, SDs, and proportions were calculated. Kolmogorov-Smirnov tests confirmed that all continuous variables followed a normal distribution. Admission characteristics of patients were compared between the 2 groups with a
Student unpaired t-tests for continuous variables and c2 tests for categorical variables to determine whether there were any significant differences between the groups in these characteristics. Receiver operating characteristic (ROC) curves were used to determine the cutoff points for admission BBS scores that best discriminated between patients who did and did not achieve independent gait during the 3month follow-up period. In the subanalysis using ROC curves, BBS scores from a second assessment 1 month after admission of the patients (n ¼ 191) who had had poor balance on admission (below the cutoff BBS score at admission assessment for predicting independent gait) were assessed to determine optimal cutoff points of this second value for predicting independent gait after 3 months. The cutoff value that equally balanced high sensitivity and specificity was defined using the Youden Index [23] and the 3-month outcomes grouped into 2 categories: below and above this cutoff point. Multiple logistic regression analyses were performed to identify the relationships between independent gait and cutoff points for BBS at admission assessment among all patients and 1-month assessments among patients who had had poor balance on admission (below the cutoff BBS score for predicting independent gait on admission assessment). The covariates assessed by multiple logistic regression analyses included age, sex, type of stroke (cerebral hemorrhage or not), days between stroke onset and admission, Brunnstrom motor recovery stage (lower limb stage III or other), cognitive impairment (23/30 on the Mini-Mental State Examination for patients without aphasia or lower ageadjusted cutoff values of Raven’s Coloured Progressive Matrices Scores for patients with aphasia), and aphasia. Relevant clinical characteristics and factors that potentially affect the future functional status of inpatients poststroke were used as covariates [10,24-28]. Predictive equations were generated [29], and odds ratios for cut-off BBS scores were calculated. Statistical analyses were performed using SPSS for Windows, version 19.0 (IBM Corp, Armonk, NY). A P value of < .05 indicated statistical significance. Results Of the 251 patients in this study, 65 (25.8%) regained independent gait by 3 months after admission. Table 1 summarizes relevant patient characteristics and admission measurements according to independent (n ¼ 65) and nonindependent (n ¼ 186) groups as classified by ambulatory ability after 3 months. Patients in the independent group (63.3 11.7 years) were significantly younger than those in the nonindependent group (70.2 10.4 years) (P < .001). At admission assessment, the independent group had significantly better FIM total (independent group ¼ 67.0 17.9 points, nonindependent group ¼ 40.3 18.3 points; P < .001) and
H. Makizako et al. / PM R 7 (2015) 392-399
395
Table 1 Patient characteristics at admission Gait Ability After 3 Months Age, y, SD Gender, female/male Type of stroke, cerebral hemorrhage/cerebral infarction Hemisphere of stroke, left/right/other* Aphasia, yes/no Cognitive impairment, yes/no† Days between stroke onset and admission, SD Brunnstrom stage, n (%) Stage I Stage II Stage III Stage IV Stage V Stage VI Stage III Functional Independent Measure, SD Berg Balance Scale, SD
Total (n ¼ 251)
Independent (n ¼ 65)
Not Independent (n ¼ 186)
P Value
68.4 11.1 103/148 133/118 108/122/21 76/175 174/77 32.0 12.7
63.3 11.7 22/43 35/30 29/31/5 17/48 28/37 30.4 12.7
70.2 10.4 81/105 98/88 79/91/16 59/125 146/40 32.6 12.7
<.001 .171 .872 .945 .400 <.001 .232
16 (6.4) 78 (31.1) 83 (33.1) 24 (9.6) 31 (12.4) 19 (7.6) 157 (62.5) 47.2 21.6 9.0 12.5
0 (0.0) 10 (15.4) 29 (44.6) 9 (13.8) 5 (7.7) 12 (18.5) 55 (84.6) 67.0 17.9 19.3 14.6
16 (8.6) 68 (36.6) 54 (29.0) 15 (8.1) 26 (14.0) 7 (3.8) 102 (54.8) 40.3 18.3 5.4 9.3
<.001 <.001 <.001
Values are mean SD, n (%), or as otherwise indicated. * Other: bilateral strokes and cerebellar/brainstem damage. † Mini-Mental State Examination score 23/30 for patients without aphasia or lower age-adjusted cutoff values of Raven’s Coloured Progressive Matrices scores for patients with aphasia.
BBS scores (independent group ¼ 19.3 14.6 points, nonindependent group ¼ 5.4 9.3 points; P < .001) than the nonindependent group. There were statistically significant between-group differences in the data from admission assessments concerning prevalence of cognitive impairment (independent group ¼ 43.1%, nonindependent group ¼ 78.5%; P < .001) and Brunnstrom motor recovery stage ( Stage III) (independent group ¼ 40.0%, nonindependent group ¼ 25.8%; P ¼ .040). However, there were no statistically significant between-group differences in sex, type of stroke (CH or cerebral infarction), hemisphere of stroke (left, right, or other), aphasia, or days between stroke onset and admission. Table 2 shows cutoff values of BBS for predicting independent gait after 3 months among all patients. Area under the ROC curves for predicting independent gait after 3 months was 0.81 (95% confidence interval [CI], 0.75-0.88) for BBS scores on admission and its optimal cut-off was 13 (63% sensitivity and 90% specificity) (Figure 1A). Of 60 patients with BBS scores 13 on admission, 41 (68.3%) achieved independent gait after 3 months, whereas of 191 patients with BBS scores <13 on admission, only 24 (12.6%) achieved
independent gait after 3 months (Table 3). In the subanalysis of patients (n ¼ 191) with poor balance on admission (BBS score <13), 24 (12.6%) patients regained independent gait by 3 months after admission and area under the ROC curves for predicting independent gait after 3 months was 0.88 (95% CI, 0.80-0.95) for BBS scores 1 month after admission with a cut-off of 27 (75% sensitivity and 88% specificity) (Figure 1B). In this subanalysis, of 38 patients whose BBS score 27 at 1-month assessment, 18 (47.7%) achieved independent gait after 3 months, but of 153 patients whose BBS score <27 at 1-month assessment, only 6 (3.9%) achieved independent gait after 3 months (Table 3). Multiple logistic regression analyses using the factors of age, sex, type of stroke (CH or not), days between stroke onset and admission, Brunnstrom motor recovery stage (lower limb Stage III or other), cognitive impairment, and aphasia as covariates showed that BBS score better than cutoff points of 13 at admission was significantly associated with independent gait (odds ratio, 9.7; 95% CI, 4.38-21.36; P < .001) (Table 3). The final logistic regression equation derived from these data is as follows:
Table 2 The optimal cutoff value for gait independent after 3 months determined from the ROC curve Assessment Time of BBS BBS at admission BBS after 1-month
Analysis Data
Cutoff Value
Sensitivity
Specificity
Area under the ROC Curves (95% CI)
All participants, n ¼ 251 Poor balance function at admission (Admission BBS <13 points), n ¼ 191
13 27
0.63 0.75
0.90 0.88
0.81 (0.75-0.88) 0.88 (0.80-0.95)
ROC ¼ receiver operating characteristic; BBS ¼ Berg Balance Scale; CI ¼ confidence interval.
Berg Balance Scale for Gait After Stroke
396
A
independent gait (odds ratio, 21.6, 95% CI, 6.40-73.20; P < .001) (Table 3). The final logistic regression equation for the subgroup of patients with poor balance on admission is as follows: Logit P (gait independence after 3 months, FIM walking score 6) ¼ 1.981 þ 3.075 (BBS after 1-month 27) 0.044 (age) 0.681 (female) 0.369 (cerebral hemorrhage) 0.031 (days between stroke onset and admission) 0.137 (Brunnstrom stage III) 0.460 (cognitive impairment) 1.161 (aphasia).
1.0
Sensitivity
0.8 0.6
0.4 0.2 0.0
0.0
B
0.2
0.4
0.6
0.8
1.0
1-Specificity 1.0
Sensitivity
0.8 0.6 0.4 0.2 0.0 0.0
0.2
0.4
0.6
1-Specificity
0.8
1.0
Figure 1. Receiver operating characteristic curve analysis of Berg Balance Scale (BBS) score and optimal cut-off at admission of all patients (n ¼ 251) (A) and cut-off BBS score at 1-month assessment of patients with poor admission balance (BBS score <13) for gait independent after 3-month (n ¼ 191) (B).
Logit P (gait independence after 3 months, FIM walking score 6) ¼ 1.992 þ 2.269 (BBS at admission 13) 0.050 (age) 0.398 (female) þ 0.005 (cerebral hemorrhage) 0.011 (days between stroke onset and admission) þ 0.569 (Brunnstrom stage III) 0.904 (cognitive impairment) 0.192 (aphasia). According to the subanalysis of patients with poor balance on admission (n ¼ 191; BBS score <13), BBS score better than the cutoff point of 27 points at the 1 month assessment was significantly associated with
Discussion Our findings on data on inpatients with stroke of a Japanese Kaifukuki rehabilitation ward who needed assistance with or supervision of gait at admission suggest that balance as assessed by BBS scores can predict achievement of complete or modified independence, that is, the ability to walk a minimum of 150 ft (50 m) without supervision within 3 months of admission. ROC curves demonstrated that inpatients with BBS scores of 13 on admission assessment were highly likely to achieve independent gait within 3 months. The sensitivity and specificity of this cutoff point were 63% and 90%, respectively. In addition, of the subgroup of inpatients with below cutoff BBS scores (<13) on admission, those who had improved their BBS score to 27 at the 1 month after admission were highly likely to achieve independent gait by 3 months after admission. The sensitivity and specificity of the BBS score cutoff of 27 at a second assessment (1 month of the subgroup of patients with poor balance on admission [BBS score <13]) were 75% and 88%, respectively. Because FIM scores 1-5 are almost completely driven by the level of physical assistance required for walking, but scores 6 and 7 denote independent walking, we determined gait independence as functionally meaningful when the score for the subitem of the FIM concerning walking was 6 points (modified and complete independence) during the 3-month follow-up. Researchers have used different definitions for gait
Table 3 ORs for the association of cutoff value of BBS with gait independence after 3 months according to logistic regression models
All participants (n ¼ 251) BBS at admission (<13) BBS at admission (13) Poor balance function at admission† (n ¼ 191) BBS after 1 month (<27) BBS after 1 month (27)
No. Patients, n (%)
No. Gait-Independent Patients After 3 Months (%)
Crude OR (95% CI)
P Value
Adjusted OR* (95% CI)
P Value
191 (76.1) 60 (23.9)
24 (12.6) 41 (68.3)
Reference 15.0 (7.52-30.00)
<.001
Reference 9.7 (4.38-21.36)
<.001
153 (80.1) 38 (19.9)
6 (3.9) 18 (47.4)
Reference 22.1 (7.83-62.09)
<.001
Reference 21.6 (6.40-73.20)
<.001
OR ¼ odds ratio; BBS ¼ Berg Balance Scale; CI ¼ confidence interval. * Adjusted for age, gender, type of stroke (cerebral hemorrhage or not), and days between stroke onset and admission, Brunnstrom stage (III or other), cognitive impairment, and aphasia. † Subanalysis of patients (n ¼ 191) with poor balance on admission (BBS score <13).
H. Makizako et al. / PM R 7 (2015) 392-399
improvement or independent gait: some have determined gait improvement by changes in gait speed [3,6] or changes in scores for gait-related measurements, rather than by FIM scores [2,10]. Although more studies are needed for the various recovery phases and over different durations of follow-up our findings may provide useful information regarding inpatients of poststroke rehabilitation facilities, especially those in subacute phases who are without independent gait on admission. Our finding that BBS scores can predict independent gait is consistent with previous studies suggesting that balance is a key function in determining future ambulatory ability of poststroke inpatients [3,11,24]. Improvement in balance and gait of patients with stroke play important roles in increasing their physical activity [30] and activities in the community [31], improving their quality of life [32], and reducing the burden on caregivers [33]. One aspect of ambulatory ability at discharge, gait speed, reportedly correlates with admission gait speed. In addition, admission gait speed predicts discharge destination and length of stay of patients with stroke [7,34]. As expected, it is not possible to predict independent gait on the basis of admission gait speed among inpatients who need assistance with gait on admission. Even patients with the most severely limited admission mobility, however, reportedly spend more than one fourth of their physical therapy time in gait-related activities [8]. Therefore, predicting gait ability among inpatients who have had a stroke and do not have independent gait on admission assessment could provide important information for selecting follow-up strategies. In the present study, BBS scores at 1 month predict independent gait among those patients who needed physical assistance or supervision to walk on admission (BBS scores 13). Thus, our findings provide evidence that ability to balance on admission may predict 3-month ambulatory ability in all inpatients with stroke and that results of a 1-month assessment of balance predict 3-month ambulatory ability in the subgroup with BBS scores 13 on admission. The limitations of this way of determining independence, however, should be recognized: such patients may not be able to access the community independently. Advantages of the present study include establishing the usefulness of admission BBS scores and changes in BBS score 1 month after admission for predicting independent gait by 3 months poststroke. The percentages of patients whose gait improved in our study, however, seem lower than that in previous studies [2,35]. For instance, Kollen et al. reported that, by 6 months poststroke, 62% of patients had regained independent gait [2]. Conversely, in this study, only 26% (n ¼ 65) had regained independent gait by 3 months after admission. There are several possible explanations for these differences. This study included only patients who have
397
had a stroke and needed assistance with or supervision of gait on admission, and we assessed independence of gait at 3 rather than 6 months. Because there are differences between published studies in the means of admission assessment after stroke onset and interventions implemented in rehabilitation facilities, it may not be possible to accurately and directly compare the findings of this study with other published data. For instance, patients in the cited previous studies were assessed within 14 days of stroke onset [2,10], whereas the mean number of days between stroke onset and admission to our unit was 32.0 days in the present study. The shorter duration of follow-up and differences in relevant patient characteristics (eg, motor recovery stage) may in part be responsible for the lower percentage of patients who regained independent gait in our study. Interventions implemented at the Kaifukuki rehabilitation facility take longer than the rehabilitation programs used in traditional or general facilities in Japan. Therefore, there are no relevant data on gait improvement among inpatients with stroke with which to compare the programs of these facilities. Thus, more studies of the various recovery phases (acute, subacute, and chronic) and facilities over different follow-up times (3 months, 6 months, and longer) are needed. Although all inpatients with stroke underwent admission assessment within 1 week of admission to a rehabilitation unit and no differences between outcome groups in days between stroke onset and admission were identified, the interval between stroke onset and admission did vary. In addition, rehabilitation programs were provided individually and were not homogeneous. Therefore, the amount of time spent on gait activities during physical therapy was not standardized. Although our findings suggest that balance as assessed by optimal cutoff BBS scores is a predictor of independent gait within 3 months of admission, we did not assess other factors that potentially affect gait performance, such as muscle strength [36], trunk control [37], homonymous hemianopia [16], and urinary incontinence [38]. Although in the current study we determined the proportion of people who had attained independent gait at 3 months and the optimal admission BBS cut-off score for these patients, we did not ascertain the time to achieve independence. These are important limitations of our study; future studies are needed to clarify this. In this study, FIM was used as an outcome measure of gait independence. In addition, we used Brunnstrom stage to assess patients’ motor recovery. Most published studies concerning stroke rehabilitation have used other assessments such as the Barthel index, timed walk assessment, FugleMeyer assessment, and Ashworth scale [39]. We considered the FIM preferable for assessment of gait independence because it is scored by observation of patients’ activities of daily living associated with walking performance (eg, walking speed)
398
Berg Balance Scale for Gait After Stroke
[3,7]. The assessment of other specific gait variables (eg, gait speed and gait distance), however, would have provided additional objective information concerning gait and likely increased the sensitivity of predicting independent gait among patients in poststroke rehabilitation facilities. Although we recognize these limitations, our findings may provide useful information regarding inpatients of poststroke rehabilitation facilities, especially those in subacute phases and without independent gait on admission to our Kaifukuki wards. Specifically, BBS scores on admission and changes in these during the first month after admission would be useful for predicting gait dependence and refining approaches to rehabilitation of subacute inpatients with strokes. Knowing after assessment at 1 month that a poststroke inpatient with severe limitations of walking on admission is not likely to reach modified independent and independent gait within 3 months from admission assessments would assist clinicians to plan rehabilitation programs within rehabilitation facilities (Kaifukuki) and patients’ environments after discharge. Conclusions Our findings indicate that BBS scores on admission predict independent gait after 3 months, the optimal cut-off score being 13 among inpatients of Kaifukuki rehabilitation ward. Even among patients with poor balance on admission, those who show improvement of BBS scores to 27 by 1 month after admission are significantly more likely to achieve independent gait within 3 months of admission. Future studies with longer follow-up periods (eg, 6 months, on discharge, and after living in the community) and examining a combination of predictors including modifiable factors potentially affecting achievement of independent gait are needed. Acknowledgments We thank the staff and administrators at the Funabashi Municipal Rehabilitation Hospital and Ms. A. Morita (speech therapist) and Dr. M. Ishikawa, Hatsudai Rehabilitation Hospital, for their support for this study. References 1. Bohannon RW, Horton MG, Wikholm JB. Importance of four variables of walking to patients with stroke. Int J Rehabil Res 1991;14: 246-250. 2. Kollen B, Kwakkel G, Lindeman E. Longitudinal robustness of variables predicting independent gait following severe middle cerebral artery stroke: A prospective cohort study. Clin Rehabil 2006; 20:262-268. 3. Bland MD, Sturmoski A, Whitson M, et al. Prediction of discharge walking ability from initial assessment in a stroke inpatient rehabilitation facility population. Arch Phys Med Rehabil 2012; 93:1441-1447.
4. Kluding P, Gajewski B. Lower-extremity strength differences predict activity limitations in people with chronic stroke. Phys Ther 2009;89:73-81. 5. Schmid A, Duncan PW, Studenski S, et al. Improvements in speedbased gait classifications are meaningful. Stroke 2007;38: 2096-2100. 6. Goldie PA, Matyas TA, Kinsella GJ, Galea MP, Evans OM, Bach TM. Prediction of gait velocity in ambulatory stroke patients during rehabilitation. Arch Phys Med Rehabil 1999;80:415-420. 7. Kuys SS, Bew PG, Lynch MR, Morrison G, Brauer SG. Measures of activity limitation on admission to rehabilitation after stroke predict walking speed at discharge: An observational study. Aust J Physiother 2009;55:265-268. 8. Latham NK, Jette DU, Slavin M, Richards LG, Procino A, Smout RJ, et al. Physical therapy during stroke rehabilitation for people with different walking abilities. Arch Phys Med Rehabil 2005; 86:S41-S50. 9. Blum L, Korner-Bitensky N. Usefulness of the Berg Balance Scale in stroke rehabilitation: A systematic review. Phys Ther 2008; 88:559-566. 10. Kollen B, van de Port I, Lindeman E, Twisk J, Kwakkel G. Predicting improvement in gait after stroke: A longitudinal prospective study. Stroke 2005;36:2676-2680. 11. Tsang YL, Mak MK. Sit-and-reach test can predict mobility of patients recovering from acute stroke. Arch Phys Med Rehabil 2004; 85:94-98. 12. Veerbeek JM, van Wegen E, van Peppen R, et al. What is the evidence for physical therapy poststroke? A systematic review and meta-analysis. PLoS One 2014;9:e87987. 13. Miyai I, Sonoda S, Nagai S, et al. Results of new policies for inpatient rehabilitation coverage in Japan. Neurorehabil Neural Repair 2011;25:540-547. 14. Keith RA, Granger CV, Hamilton BB, Sherwin FS. The functional independence measure: A new tool for rehabilitation. Adv Clin Rehabil 1987;1:6-18. 15. Dodds TA, Martin DP, Stolov WC, Deyo RA. A validation of the functional independence measurement and its performance among rehabilitation inpatients. Arch Phys Med Rehabil 1993; 74:531-536. 16. Patel AT, Duncan PW, Lai SM, Studenski S. The relation between impairments and functional outcomes poststroke. Arch Phys Med Rehabil 2000;81:1357-1363. 17. Ditunno JF Jr, Burns AS, Marino RJ. Neurological and functional capacity outcome measures: Essential to spinal cord injury clinical trials. J Rehabil Res Dev 2005;42:35-41. 18. White DK, Wilson JC, Keysor JJ. Measures of adult general functional status: SF-36 Physical Functioning Subscale (PF-10), Health Assessment Questionnaire (HAQ), Modified Health Assessment Questionnaire (MHAQ), Katz Index of Independence in activities of daily living, Functional Independence Measure (FIM), and Osteoarthritis-Function-Computer Adaptive Test (OA-FunctionCAT). Arthritis Care Res (Hoboken) 2011;63:S297-S307. 19. Berg KO, Maki BE, Williams JI, Holliday PJ, Wood-Dauphinee SL. Clinical and laboratory measures of postural balance in an elderly population. Arch Phys Med Rehabil 1992;73:1073-1080. 20. 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. 21. Raven J, Court J, Raven J. Raven’s Colored Progressive Matrices. Oxford: Oxford Psychologists Press; 1995. 22. Brunnstrom S. Movement therapy in hemiplegia: A neurophysiological approach. Hagerstown, MD: Harper & Row; 1970. 23. Schisterman EF, Perkins NJ, Liu A, Bondell H. Optimal cut-point and its corresponding Youden Index to discriminate individuals using pooled blood samples. Epidemiology 2005;16:73-81. 24. Kwakkel G, Wagenaar RC, Kollen BJ, Lankhorst GJ. Predicting disability in strokeda critical review of the literature. Age Ageing 1996;25:479-489.
H. Makizako et al. / PM R 7 (2015) 392-399 25. Reid JM, Gubitz GJ, Dai D, et al. Predicting functional outcome after stroke by modelling baseline clinical and CT variables. Age Ageing 2010;39:360-366. 26. Wandel A, Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Prediction of walking function in stroke patients with initial lower extremity paralysis: The Copenhagen Stroke Study. Arch Phys Med Rehabil 2000;81:736-738. 27. Dong Y, Slavin MJ, Chan BP, et al. Cognitive screening improves the predictive value of stroke severity scores for functional outcome 3-6 months after mild stroke and transient ischaemic attack: An observational study. BMJ Open 2013;3:e003105. 28. Yavuzer G, Kucukdeveci A, Arasil T, Elhan A. Rehabilitation of stroke patients: clinical profile and functional outcome. Am J Phys Med Rehabil 2001;80:250-255. 29. Park SW, Wolf SL, Blanton S, Winstein C, Nichols-Larsen DS. The EXCITE Trial: Predicting a clinically meaningful motor activity log outcome. Neurorehabil Neural Repair 2008;22:486-493. 30. Michael KM, Allen JK, Macko RF. Reduced ambulatory activity after stroke: the role of balance, gait, and cardiovascular fitness. Arch Phys Med Rehabil 2005;86:1552-1556. 31. Alzahrani M, Dean C, Ada L. Relationship between walking performance and types of community-based activities in people with stroke: An observational study. Rev Bras Fisioter 2011; 15:45-51.
399
32. Samsa GP, Matchar DB. How strong is the relationship between functional status and quality of life among persons with stroke? J Rehabil Res Dev 2004;41:279-282. 33. Vincent C, Desrosiers J, Landreville P, Demers L. Burden of caregivers of people with stroke: Evolution and predictors. Cerebrovasc Dis 2009;27:456-464. 34. Rabadi MH, Blau A. Admission ambulation velocity predicts length of stay and discharge disposition following stroke in an acute rehabilitation hospital. Neurorehabil Neural Repair 2005;19:20-26. 35. Wade DT, Wood VA, Heller A, Maggs J, Langton Hewer R. Walking after stroke. Measurement and recovery over the first 3 months. Scand J Rehabil Med 1987;19:25-30. 36. Veerbeek JM, Van Wegen EE, Harmeling-Van der Wel BC, Kwakkel G. Is accurate prediction of gait in nonambulatory stroke patients possible within 72 hours poststroke? The EPOS study. Neurorehabil Neural Repair 2011;25:268-274. 37. Duarte E, Marco E, Muniesa JM, Belmonte R, Aguilar JJ, Escalada F. Early detection of non-ambulatory survivors six months after stroke. NeuroRehabilitation 2010;26:317-323. 38. Barer DH. Continence after stroke: useful predictor or goal of therapy? Age Ageing 1989;18:183-191. 39. Salter KL, Teasell RW, Foley NC, Jutai JW. Outcome assessment in randomized controlled trials of stroke rehabilitation. Am J Phys Med Rehabil 2007;86:1007-1012.
Disclosure H.M. Department of Functional Activation, Center for Gerontology and Social Science, National Center for Geriatrics and Gerontology, 7-430 Morioka-machi, Obu, Aichi 474-8511, Japan. Address correspondence to: H.M.; e-mail: [email protected] Disclosure: nothing to disclose N.K. National Center for Geriatrics and Gerontology, Aichi, Japan; and Department of Physical Therapy, Funabashi Municipal Rehabilitation Hospital, Chiba, Japan Disclosure: nothing to disclose
A.T. Department of Speech Therapy, Funabashi Municipal Rehabilitation Hospital, Chiba, Japan Disclosure: nothing to disclose K.I. Department of Nursing and Care Work, Funabashi Municipal Rehabilitation Hospital, Chiba, Japan Disclosure: nothing to disclose Submitted for publication November 1, 2013; accepted January 17, 2015.
www.pmrjournal.org
Original Research
Use of the Berg Balance Scale to Predict Independent Gait After Stroke: A Study of an Inpatient Population in Japan Hyuma Makizako, PhD, PT, Norihito Kabe, MSc, PT, Asami Takano, ST, Kanako Isobe, CW
Abstract Objective: To examine whether the Berg Balance Scale (BBS) score on admission and after 1 month can be used to predict the Functional Independence Measure (FIM) walking level 6 or 7 (modified independent and independent gait respectively) after 3 months and to determine the optimal BBS cut-off score for predicting independent gait among inpatients with first stroke in a rehabilitation facility. Design: Retrospective cohort study. Setting: Inpatient rehabilitation facility (Kaifukuki). Participants: Two hundred fifty-one consecutive patients with first stroke and nonindependent gait on admission. Outcome Measures: BBS on admission and after 1 month and FIM on admission and after 3 months. Results: Area under the receiver operating characteristic curves for predicting independent gait after 3 months were 0.81 (95% confidence interval [CI], 0.75-0.88) for BBS score on admission; the optimal cut-off score being 13 (63% sensitivity; 90% specificity). Achieving greater cut-off score on admission was significantly associated with subsequent independent gait (odds ratio, 9.7; 95% CI, 4.38-21.36; P < .001). A subanalysis of patients with poor balance on admission (BBS score <13; n ¼ 191) showed area under the receiver operating characteristic curves for predicting independent gait after 3 months of 0.88 (95% CI, 0.80-0.95) for BBS scores 1 month after admission. In this subgroup, BBS scores at 1 month of 27 were significantly associated with subsequent independent gait (odds ratio, 21.6, 95% CI, 6.40-73.20; P < .001). Conclusions: Admission BBS scores predicted FIM walking level 6 or 7, which denotes modified independent or independent gait, after 3 months, the optimal cut-off for BBS scores being 13 among inpatients with first stroke in a rehabilitation facility. Patients with poor balance on admission whose BBS scores had improved to 27 at 1 month after admission were likely to achieve modified independent or independent gait within 3 months after admission.
Introduction One of the most important goals of rehabilitation for patients with stroke is improvement in gait, ideally to the point of achieving independent ambulation without requiring physical assistance or supervision by another person [1]. Predictors of recovery of this ability in patients with stroke reportedly include age [2], activities of daily living scores [2,3], differences in muscle strength between the lower extremities [4], and balance [3]. In particular, initial walking speed is the most commonly used indicator of the eventual ambulatory ability of patients with stroke [5-7]. It is important to ascertain whether eventual ambulatory ability can be improved in those poststroke
inpatients in rehabilitation facilities who need assistance to walk on admission. It is difficult, however, to predict on the basis of admission assessments whether inpatients poststroke with severe limitations in walking on admission will reach functional independence. In the Post-Stroke Rehabilitation Outcomes Project, on admission to a rehabilitation facility, more than 60% of inpatients poststroke who received a 6-hour block of physical therapy reportedly had severe limitations in walking, with scores of 1 or 2 on the walk Functional Independence Measure (FIM) item [8]. Thus, identification of predictors of independent gait would allow refinement of admission assessments of inpatients poststroke with severe limitations in walking.
1934-1482/$ - see front matter ª 2015 by the American Academy of Physical Medicine and Rehabilitation http://dx.doi.org/10.1016/j.pmrj.2015.01.009
H. Makizako et al. / PM R 7 (2015) 392-399
The Berg Balance Scale (BBS) has been identified as the most commonly used balance assessment tool across the continuum of stroke rehabilitation [9]. Balance is one of the key components that determines eventual ambulatory ability after stroke [10], and the BBS score predicts most of the variation in this ability among patients admitted with strokes to rehabilitation facilities [3]. In a prospective study, BBS scores of 20 on admission to an inpatient rehabilitation facility indicated that persons with strokes were highly likely to achieve only household ambulation speeds (<0.4 m/s) by the time of their discharge [3]. These findings suggest that balance performance based on BBS scores can predict future gait speed in inpatients poststroke. In addition to gait speed, walking ability with or without physical assistance or supervision is important for discharge planning. In addition, another prospective study has reported that balance performance predicts the mobility of patients with stroke at discharge [11]; however, in this study the duration of hospitalization varied and the length of observation was not defined. To our knowledge, no studies have yet reported optimal cut-off balance scores for predicting independent gait among poststroke patients who, on admission assessment, cannot walk without physical assistance or supervision from another person. A better understanding of the ability of admission balance performance to predict attainment of independent gait within an established rehabilitation period (eg, 3 months) would provide useful information about inpatients with strokes in rehabilitation facilities. We decided that examining changes in BBS score at 1 month after admission would provide the best predictor of gait independence in these patients because functional outcomes, including balance, can reportedly improve within this time frame in inpatients with stroke in early rehabilitation [12]. Therefore, we were interested in investigating the ability of changes in BBS scores after 1 month to predict such as outcomes in the patients with the lowest initial scores. Because of the medical insurance system in Japan, rehabilitation generally occurs in dedicated convalescent rehabilitation wards that are known as the “Kaifukuki.” In 2000, these new interdisciplinary postacute “Kaifukuki” rehabilitation wards were incorporated into the Japanese medical insurance system, and it was anticipated that these changes in national insurance policies would affect the amount and organization of rehabilitation interventions and improve in patient outcomes [13]. In these wards, convalescent inpatients with stroke undergo rehabilitation programs provided by physical therapists, occupational therapists, speech therapists, and clinical nurses. These patients and their families are given comprehensive monthly rehabilitation plans that include information about achieved and planned goals; rehabilitative approaches to achieving the planned goals; discharge planning; and social
393
resources necessary for discharge home [13]. Therefore, it is important to predict whether inpatients with stroke who need physical assistance or supervision from another person to walk on admission will achieve independent gait and approximately how long this will take. The aims of this study were to determine both whether BBS scores predict a modified independent or independent gait (FIM walking level 6 or 7) after 3 months and the optimal BBS cut-off score for such predictions among inpatients of a Kaifukuki rehabilitation facility. We postulated that balance measured by BBS on admission assessment would predict gait ability after 3 months. In addition, we postulated that, among inpatients with poor balance on admission, changes in BBS scores within 1 month might help to predict improvement in ambulatory ability after 3 months. Because balance performance can improve rapidly and dramatically even when balance is poor on admission, we hypothesized that improvement in 1 month BBS scores would predict independent gait. Methods Patients Data for this retrospective cohort study were obtained from records stored in the Funabashi Rehabilitation Hospital database (patients between April 2009 and September 2012). This study included 251 inpatients who have had a stroke with a mean age of 68.4 years (SD 11.1). The inclusion criteria were (1) aged 40 years or older; (2) first stroke (physicians-determined based on magnetic resonance imaging findings and/or history obtained); (3) diagnosis of cerebral hemorrhage (CH) or cerebral infarction; (4) independent before onset; (5) walk item FIM score 5 on admission; and (6) underwent the inpatient rehabilitation program of the Funabashi Rehabilitation Hospital for more than 3 months. Exclusion criteria were diagnoses of subarachnoid hemorrhage and ability to walk independently on admission assessment (walk item FIM score 6). The study was approved by the Ethics Committees of the Funabashi Rehabilitation Hospital. Study Variables Gait independence had been assessed using the relevant subitem of the FIM motor scale [14]. The FIM motor scale is a 13-item instrument with 4 subscales: self-care (eating, grooming, bathing, dressing upper body, dressing lower body, and toileting), sphincter control (bladder and bowel management), transfers (in and out of bed, toilet and tub/shower), and locomotion (walking and stair climbing). Ratings for each item ranged from 1 (total assistance) to 7 (complete independence). Excellent internal consistency of FIM (Cronbach alpha >0.9) has been confirmed in a previous
394
Berg Balance Scale for Gait After Stroke
study [15]. In this study, all patients exhibited their scores on the subitem concerning walking ranged from 1 (total assistance) to 5 (supervision) at admission. Gait independence at 3 months had been assessed by the FIM walking score 6 (6 ¼ modified independence or 7 ¼ complete independence) [16]. Complete independence (score 7) means that subject can safely walk a minimum of 150 ft (50 m) without assistive devices. Modified independence (score 6) means that the subject can walk a minimum of 150 ft (50 m) without supervision but with the support of a brace (orthosis) or leg prosthesis, specially modified shoes, cane, crutches, or walkerette; takes more than a reasonable time; or there are safety considerations [8]. FIM scores 1-5 are determined by the level of physical assistance required for walking, whereas scores 6 and 7 denote independent walking [17]. Thus, there are 2 gross score classifications: dependent (helper required: scores 1-5) and independent (no helper required: score 6-7) [18]. Patients’ ability to balance on admission and 1 month later have been evaluated previously by physical therapists using the BBS [19]. The BBS is a 14-item test that uses a 4-point ordinal scale to quantify performance on tasks such as standing unsupported, standing with eyes open or closed, and turning 360 . Possible total scores range from 0 to 56, with greater scores denoting better balance. Patients’ cognitive function and aphasia had been assessed by speech therapists. Patients without aphasia were considered to have cognitive impairment if their precognitive scores were 23/30 on the Mini-Mental State Examination [20], whereas for patients with aphasia, lower age-adjusted cutoff values of Raven’s Coloured Progressive Matrices Scores [21] were used. In addition, lower limb paresis had been assessed by physical or occupational therapists using the Brunnstrom motor recovery stage [22]. Rehabilitation All patients in this study had undergone rehabilitation programs administered by physical therapists, occupational therapists, and speech therapists. Rehabilitation programs were based on a comprehensive approach and included physical therapy, occupational therapy, and speech therapy. Patients were provided 6-9 units (1 unit: 20 minutes) per day as necessary. Statistical Analysis Patients were classified as independent (n ¼ 65) or nonindependent (n ¼ 186) based on their ambulatory ability 3 months after admission. Means, SDs, and proportions were calculated. Kolmogorov-Smirnov tests confirmed that all continuous variables followed a normal distribution. Admission characteristics of patients were compared between the 2 groups with a
Student unpaired t-tests for continuous variables and c2 tests for categorical variables to determine whether there were any significant differences between the groups in these characteristics. Receiver operating characteristic (ROC) curves were used to determine the cutoff points for admission BBS scores that best discriminated between patients who did and did not achieve independent gait during the 3month follow-up period. In the subanalysis using ROC curves, BBS scores from a second assessment 1 month after admission of the patients (n ¼ 191) who had had poor balance on admission (below the cutoff BBS score at admission assessment for predicting independent gait) were assessed to determine optimal cutoff points of this second value for predicting independent gait after 3 months. The cutoff value that equally balanced high sensitivity and specificity was defined using the Youden Index [23] and the 3-month outcomes grouped into 2 categories: below and above this cutoff point. Multiple logistic regression analyses were performed to identify the relationships between independent gait and cutoff points for BBS at admission assessment among all patients and 1-month assessments among patients who had had poor balance on admission (below the cutoff BBS score for predicting independent gait on admission assessment). The covariates assessed by multiple logistic regression analyses included age, sex, type of stroke (cerebral hemorrhage or not), days between stroke onset and admission, Brunnstrom motor recovery stage (lower limb stage III or other), cognitive impairment (23/30 on the Mini-Mental State Examination for patients without aphasia or lower ageadjusted cutoff values of Raven’s Coloured Progressive Matrices Scores for patients with aphasia), and aphasia. Relevant clinical characteristics and factors that potentially affect the future functional status of inpatients poststroke were used as covariates [10,24-28]. Predictive equations were generated [29], and odds ratios for cut-off BBS scores were calculated. Statistical analyses were performed using SPSS for Windows, version 19.0 (IBM Corp, Armonk, NY). A P value of < .05 indicated statistical significance. Results Of the 251 patients in this study, 65 (25.8%) regained independent gait by 3 months after admission. Table 1 summarizes relevant patient characteristics and admission measurements according to independent (n ¼ 65) and nonindependent (n ¼ 186) groups as classified by ambulatory ability after 3 months. Patients in the independent group (63.3 11.7 years) were significantly younger than those in the nonindependent group (70.2 10.4 years) (P < .001). At admission assessment, the independent group had significantly better FIM total (independent group ¼ 67.0 17.9 points, nonindependent group ¼ 40.3 18.3 points; P < .001) and
H. Makizako et al. / PM R 7 (2015) 392-399
395
Table 1 Patient characteristics at admission Gait Ability After 3 Months Age, y, SD Gender, female/male Type of stroke, cerebral hemorrhage/cerebral infarction Hemisphere of stroke, left/right/other* Aphasia, yes/no Cognitive impairment, yes/no† Days between stroke onset and admission, SD Brunnstrom stage, n (%) Stage I Stage II Stage III Stage IV Stage V Stage VI Stage III Functional Independent Measure, SD Berg Balance Scale, SD
Total (n ¼ 251)
Independent (n ¼ 65)
Not Independent (n ¼ 186)
P Value
68.4 11.1 103/148 133/118 108/122/21 76/175 174/77 32.0 12.7
63.3 11.7 22/43 35/30 29/31/5 17/48 28/37 30.4 12.7
70.2 10.4 81/105 98/88 79/91/16 59/125 146/40 32.6 12.7
<.001 .171 .872 .945 .400 <.001 .232
16 (6.4) 78 (31.1) 83 (33.1) 24 (9.6) 31 (12.4) 19 (7.6) 157 (62.5) 47.2 21.6 9.0 12.5
0 (0.0) 10 (15.4) 29 (44.6) 9 (13.8) 5 (7.7) 12 (18.5) 55 (84.6) 67.0 17.9 19.3 14.6
16 (8.6) 68 (36.6) 54 (29.0) 15 (8.1) 26 (14.0) 7 (3.8) 102 (54.8) 40.3 18.3 5.4 9.3
<.001 <.001 <.001
Values are mean SD, n (%), or as otherwise indicated. * Other: bilateral strokes and cerebellar/brainstem damage. † Mini-Mental State Examination score 23/30 for patients without aphasia or lower age-adjusted cutoff values of Raven’s Coloured Progressive Matrices scores for patients with aphasia.
BBS scores (independent group ¼ 19.3 14.6 points, nonindependent group ¼ 5.4 9.3 points; P < .001) than the nonindependent group. There were statistically significant between-group differences in the data from admission assessments concerning prevalence of cognitive impairment (independent group ¼ 43.1%, nonindependent group ¼ 78.5%; P < .001) and Brunnstrom motor recovery stage ( Stage III) (independent group ¼ 40.0%, nonindependent group ¼ 25.8%; P ¼ .040). However, there were no statistically significant between-group differences in sex, type of stroke (CH or cerebral infarction), hemisphere of stroke (left, right, or other), aphasia, or days between stroke onset and admission. Table 2 shows cutoff values of BBS for predicting independent gait after 3 months among all patients. Area under the ROC curves for predicting independent gait after 3 months was 0.81 (95% confidence interval [CI], 0.75-0.88) for BBS scores on admission and its optimal cut-off was 13 (63% sensitivity and 90% specificity) (Figure 1A). Of 60 patients with BBS scores 13 on admission, 41 (68.3%) achieved independent gait after 3 months, whereas of 191 patients with BBS scores <13 on admission, only 24 (12.6%) achieved
independent gait after 3 months (Table 3). In the subanalysis of patients (n ¼ 191) with poor balance on admission (BBS score <13), 24 (12.6%) patients regained independent gait by 3 months after admission and area under the ROC curves for predicting independent gait after 3 months was 0.88 (95% CI, 0.80-0.95) for BBS scores 1 month after admission with a cut-off of 27 (75% sensitivity and 88% specificity) (Figure 1B). In this subanalysis, of 38 patients whose BBS score 27 at 1-month assessment, 18 (47.7%) achieved independent gait after 3 months, but of 153 patients whose BBS score <27 at 1-month assessment, only 6 (3.9%) achieved independent gait after 3 months (Table 3). Multiple logistic regression analyses using the factors of age, sex, type of stroke (CH or not), days between stroke onset and admission, Brunnstrom motor recovery stage (lower limb Stage III or other), cognitive impairment, and aphasia as covariates showed that BBS score better than cutoff points of 13 at admission was significantly associated with independent gait (odds ratio, 9.7; 95% CI, 4.38-21.36; P < .001) (Table 3). The final logistic regression equation derived from these data is as follows:
Table 2 The optimal cutoff value for gait independent after 3 months determined from the ROC curve Assessment Time of BBS BBS at admission BBS after 1-month
Analysis Data
Cutoff Value
Sensitivity
Specificity
Area under the ROC Curves (95% CI)
All participants, n ¼ 251 Poor balance function at admission (Admission BBS <13 points), n ¼ 191
13 27
0.63 0.75
0.90 0.88
0.81 (0.75-0.88) 0.88 (0.80-0.95)
ROC ¼ receiver operating characteristic; BBS ¼ Berg Balance Scale; CI ¼ confidence interval.
Berg Balance Scale for Gait After Stroke
396
A
independent gait (odds ratio, 21.6, 95% CI, 6.40-73.20; P < .001) (Table 3). The final logistic regression equation for the subgroup of patients with poor balance on admission is as follows: Logit P (gait independence after 3 months, FIM walking score 6) ¼ 1.981 þ 3.075 (BBS after 1-month 27) 0.044 (age) 0.681 (female) 0.369 (cerebral hemorrhage) 0.031 (days between stroke onset and admission) 0.137 (Brunnstrom stage III) 0.460 (cognitive impairment) 1.161 (aphasia).
1.0
Sensitivity
0.8 0.6
0.4 0.2 0.0
0.0
B
0.2
0.4
0.6
0.8
1.0
1-Specificity 1.0
Sensitivity
0.8 0.6 0.4 0.2 0.0 0.0
0.2
0.4
0.6
1-Specificity
0.8
1.0
Figure 1. Receiver operating characteristic curve analysis of Berg Balance Scale (BBS) score and optimal cut-off at admission of all patients (n ¼ 251) (A) and cut-off BBS score at 1-month assessment of patients with poor admission balance (BBS score <13) for gait independent after 3-month (n ¼ 191) (B).
Logit P (gait independence after 3 months, FIM walking score 6) ¼ 1.992 þ 2.269 (BBS at admission 13) 0.050 (age) 0.398 (female) þ 0.005 (cerebral hemorrhage) 0.011 (days between stroke onset and admission) þ 0.569 (Brunnstrom stage III) 0.904 (cognitive impairment) 0.192 (aphasia). According to the subanalysis of patients with poor balance on admission (n ¼ 191; BBS score <13), BBS score better than the cutoff point of 27 points at the 1 month assessment was significantly associated with
Discussion Our findings on data on inpatients with stroke of a Japanese Kaifukuki rehabilitation ward who needed assistance with or supervision of gait at admission suggest that balance as assessed by BBS scores can predict achievement of complete or modified independence, that is, the ability to walk a minimum of 150 ft (50 m) without supervision within 3 months of admission. ROC curves demonstrated that inpatients with BBS scores of 13 on admission assessment were highly likely to achieve independent gait within 3 months. The sensitivity and specificity of this cutoff point were 63% and 90%, respectively. In addition, of the subgroup of inpatients with below cutoff BBS scores (<13) on admission, those who had improved their BBS score to 27 at the 1 month after admission were highly likely to achieve independent gait by 3 months after admission. The sensitivity and specificity of the BBS score cutoff of 27 at a second assessment (1 month of the subgroup of patients with poor balance on admission [BBS score <13]) were 75% and 88%, respectively. Because FIM scores 1-5 are almost completely driven by the level of physical assistance required for walking, but scores 6 and 7 denote independent walking, we determined gait independence as functionally meaningful when the score for the subitem of the FIM concerning walking was 6 points (modified and complete independence) during the 3-month follow-up. Researchers have used different definitions for gait
Table 3 ORs for the association of cutoff value of BBS with gait independence after 3 months according to logistic regression models
All participants (n ¼ 251) BBS at admission (<13) BBS at admission (13) Poor balance function at admission† (n ¼ 191) BBS after 1 month (<27) BBS after 1 month (27)
No. Patients, n (%)
No. Gait-Independent Patients After 3 Months (%)
Crude OR (95% CI)
P Value
Adjusted OR* (95% CI)
P Value
191 (76.1) 60 (23.9)
24 (12.6) 41 (68.3)
Reference 15.0 (7.52-30.00)
<.001
Reference 9.7 (4.38-21.36)
<.001
153 (80.1) 38 (19.9)
6 (3.9) 18 (47.4)
Reference 22.1 (7.83-62.09)
<.001
Reference 21.6 (6.40-73.20)
<.001
OR ¼ odds ratio; BBS ¼ Berg Balance Scale; CI ¼ confidence interval. * Adjusted for age, gender, type of stroke (cerebral hemorrhage or not), and days between stroke onset and admission, Brunnstrom stage (III or other), cognitive impairment, and aphasia. † Subanalysis of patients (n ¼ 191) with poor balance on admission (BBS score <13).
H. Makizako et al. / PM R 7 (2015) 392-399
improvement or independent gait: some have determined gait improvement by changes in gait speed [3,6] or changes in scores for gait-related measurements, rather than by FIM scores [2,10]. Although more studies are needed for the various recovery phases and over different durations of follow-up our findings may provide useful information regarding inpatients of poststroke rehabilitation facilities, especially those in subacute phases who are without independent gait on admission. Our finding that BBS scores can predict independent gait is consistent with previous studies suggesting that balance is a key function in determining future ambulatory ability of poststroke inpatients [3,11,24]. Improvement in balance and gait of patients with stroke play important roles in increasing their physical activity [30] and activities in the community [31], improving their quality of life [32], and reducing the burden on caregivers [33]. One aspect of ambulatory ability at discharge, gait speed, reportedly correlates with admission gait speed. In addition, admission gait speed predicts discharge destination and length of stay of patients with stroke [7,34]. As expected, it is not possible to predict independent gait on the basis of admission gait speed among inpatients who need assistance with gait on admission. Even patients with the most severely limited admission mobility, however, reportedly spend more than one fourth of their physical therapy time in gait-related activities [8]. Therefore, predicting gait ability among inpatients who have had a stroke and do not have independent gait on admission assessment could provide important information for selecting follow-up strategies. In the present study, BBS scores at 1 month predict independent gait among those patients who needed physical assistance or supervision to walk on admission (BBS scores 13). Thus, our findings provide evidence that ability to balance on admission may predict 3-month ambulatory ability in all inpatients with stroke and that results of a 1-month assessment of balance predict 3-month ambulatory ability in the subgroup with BBS scores 13 on admission. The limitations of this way of determining independence, however, should be recognized: such patients may not be able to access the community independently. Advantages of the present study include establishing the usefulness of admission BBS scores and changes in BBS score 1 month after admission for predicting independent gait by 3 months poststroke. The percentages of patients whose gait improved in our study, however, seem lower than that in previous studies [2,35]. For instance, Kollen et al. reported that, by 6 months poststroke, 62% of patients had regained independent gait [2]. Conversely, in this study, only 26% (n ¼ 65) had regained independent gait by 3 months after admission. There are several possible explanations for these differences. This study included only patients who have
397
had a stroke and needed assistance with or supervision of gait on admission, and we assessed independence of gait at 3 rather than 6 months. Because there are differences between published studies in the means of admission assessment after stroke onset and interventions implemented in rehabilitation facilities, it may not be possible to accurately and directly compare the findings of this study with other published data. For instance, patients in the cited previous studies were assessed within 14 days of stroke onset [2,10], whereas the mean number of days between stroke onset and admission to our unit was 32.0 days in the present study. The shorter duration of follow-up and differences in relevant patient characteristics (eg, motor recovery stage) may in part be responsible for the lower percentage of patients who regained independent gait in our study. Interventions implemented at the Kaifukuki rehabilitation facility take longer than the rehabilitation programs used in traditional or general facilities in Japan. Therefore, there are no relevant data on gait improvement among inpatients with stroke with which to compare the programs of these facilities. Thus, more studies of the various recovery phases (acute, subacute, and chronic) and facilities over different follow-up times (3 months, 6 months, and longer) are needed. Although all inpatients with stroke underwent admission assessment within 1 week of admission to a rehabilitation unit and no differences between outcome groups in days between stroke onset and admission were identified, the interval between stroke onset and admission did vary. In addition, rehabilitation programs were provided individually and were not homogeneous. Therefore, the amount of time spent on gait activities during physical therapy was not standardized. Although our findings suggest that balance as assessed by optimal cutoff BBS scores is a predictor of independent gait within 3 months of admission, we did not assess other factors that potentially affect gait performance, such as muscle strength [36], trunk control [37], homonymous hemianopia [16], and urinary incontinence [38]. Although in the current study we determined the proportion of people who had attained independent gait at 3 months and the optimal admission BBS cut-off score for these patients, we did not ascertain the time to achieve independence. These are important limitations of our study; future studies are needed to clarify this. In this study, FIM was used as an outcome measure of gait independence. In addition, we used Brunnstrom stage to assess patients’ motor recovery. Most published studies concerning stroke rehabilitation have used other assessments such as the Barthel index, timed walk assessment, FugleMeyer assessment, and Ashworth scale [39]. We considered the FIM preferable for assessment of gait independence because it is scored by observation of patients’ activities of daily living associated with walking performance (eg, walking speed)
398
Berg Balance Scale for Gait After Stroke
[3,7]. The assessment of other specific gait variables (eg, gait speed and gait distance), however, would have provided additional objective information concerning gait and likely increased the sensitivity of predicting independent gait among patients in poststroke rehabilitation facilities. Although we recognize these limitations, our findings may provide useful information regarding inpatients of poststroke rehabilitation facilities, especially those in subacute phases and without independent gait on admission to our Kaifukuki wards. Specifically, BBS scores on admission and changes in these during the first month after admission would be useful for predicting gait dependence and refining approaches to rehabilitation of subacute inpatients with strokes. Knowing after assessment at 1 month that a poststroke inpatient with severe limitations of walking on admission is not likely to reach modified independent and independent gait within 3 months from admission assessments would assist clinicians to plan rehabilitation programs within rehabilitation facilities (Kaifukuki) and patients’ environments after discharge. Conclusions Our findings indicate that BBS scores on admission predict independent gait after 3 months, the optimal cut-off score being 13 among inpatients of Kaifukuki rehabilitation ward. Even among patients with poor balance on admission, those who show improvement of BBS scores to 27 by 1 month after admission are significantly more likely to achieve independent gait within 3 months of admission. Future studies with longer follow-up periods (eg, 6 months, on discharge, and after living in the community) and examining a combination of predictors including modifiable factors potentially affecting achievement of independent gait are needed. Acknowledgments We thank the staff and administrators at the Funabashi Municipal Rehabilitation Hospital and Ms. A. Morita (speech therapist) and Dr. M. Ishikawa, Hatsudai Rehabilitation Hospital, for their support for this study. References 1. Bohannon RW, Horton MG, Wikholm JB. Importance of four variables of walking to patients with stroke. Int J Rehabil Res 1991;14: 246-250. 2. Kollen B, Kwakkel G, Lindeman E. Longitudinal robustness of variables predicting independent gait following severe middle cerebral artery stroke: A prospective cohort study. Clin Rehabil 2006; 20:262-268. 3. Bland MD, Sturmoski A, Whitson M, et al. Prediction of discharge walking ability from initial assessment in a stroke inpatient rehabilitation facility population. Arch Phys Med Rehabil 2012; 93:1441-1447.
4. Kluding P, Gajewski B. Lower-extremity strength differences predict activity limitations in people with chronic stroke. Phys Ther 2009;89:73-81. 5. Schmid A, Duncan PW, Studenski S, et al. Improvements in speedbased gait classifications are meaningful. Stroke 2007;38: 2096-2100. 6. Goldie PA, Matyas TA, Kinsella GJ, Galea MP, Evans OM, Bach TM. Prediction of gait velocity in ambulatory stroke patients during rehabilitation. Arch Phys Med Rehabil 1999;80:415-420. 7. Kuys SS, Bew PG, Lynch MR, Morrison G, Brauer SG. Measures of activity limitation on admission to rehabilitation after stroke predict walking speed at discharge: An observational study. Aust J Physiother 2009;55:265-268. 8. Latham NK, Jette DU, Slavin M, Richards LG, Procino A, Smout RJ, et al. Physical therapy during stroke rehabilitation for people with different walking abilities. Arch Phys Med Rehabil 2005; 86:S41-S50. 9. Blum L, Korner-Bitensky N. Usefulness of the Berg Balance Scale in stroke rehabilitation: A systematic review. Phys Ther 2008; 88:559-566. 10. Kollen B, van de Port I, Lindeman E, Twisk J, Kwakkel G. Predicting improvement in gait after stroke: A longitudinal prospective study. Stroke 2005;36:2676-2680. 11. Tsang YL, Mak MK. Sit-and-reach test can predict mobility of patients recovering from acute stroke. Arch Phys Med Rehabil 2004; 85:94-98. 12. Veerbeek JM, van Wegen E, van Peppen R, et al. What is the evidence for physical therapy poststroke? A systematic review and meta-analysis. PLoS One 2014;9:e87987. 13. Miyai I, Sonoda S, Nagai S, et al. Results of new policies for inpatient rehabilitation coverage in Japan. Neurorehabil Neural Repair 2011;25:540-547. 14. Keith RA, Granger CV, Hamilton BB, Sherwin FS. The functional independence measure: A new tool for rehabilitation. Adv Clin Rehabil 1987;1:6-18. 15. Dodds TA, Martin DP, Stolov WC, Deyo RA. A validation of the functional independence measurement and its performance among rehabilitation inpatients. Arch Phys Med Rehabil 1993; 74:531-536. 16. Patel AT, Duncan PW, Lai SM, Studenski S. The relation between impairments and functional outcomes poststroke. Arch Phys Med Rehabil 2000;81:1357-1363. 17. Ditunno JF Jr, Burns AS, Marino RJ. Neurological and functional capacity outcome measures: Essential to spinal cord injury clinical trials. J Rehabil Res Dev 2005;42:35-41. 18. White DK, Wilson JC, Keysor JJ. Measures of adult general functional status: SF-36 Physical Functioning Subscale (PF-10), Health Assessment Questionnaire (HAQ), Modified Health Assessment Questionnaire (MHAQ), Katz Index of Independence in activities of daily living, Functional Independence Measure (FIM), and Osteoarthritis-Function-Computer Adaptive Test (OA-FunctionCAT). Arthritis Care Res (Hoboken) 2011;63:S297-S307. 19. Berg KO, Maki BE, Williams JI, Holliday PJ, Wood-Dauphinee SL. Clinical and laboratory measures of postural balance in an elderly population. Arch Phys Med Rehabil 1992;73:1073-1080. 20. 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. 21. Raven J, Court J, Raven J. Raven’s Colored Progressive Matrices. Oxford: Oxford Psychologists Press; 1995. 22. Brunnstrom S. Movement therapy in hemiplegia: A neurophysiological approach. Hagerstown, MD: Harper & Row; 1970. 23. Schisterman EF, Perkins NJ, Liu A, Bondell H. Optimal cut-point and its corresponding Youden Index to discriminate individuals using pooled blood samples. Epidemiology 2005;16:73-81. 24. Kwakkel G, Wagenaar RC, Kollen BJ, Lankhorst GJ. Predicting disability in strokeda critical review of the literature. Age Ageing 1996;25:479-489.
H. Makizako et al. / PM R 7 (2015) 392-399 25. Reid JM, Gubitz GJ, Dai D, et al. Predicting functional outcome after stroke by modelling baseline clinical and CT variables. Age Ageing 2010;39:360-366. 26. Wandel A, Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Prediction of walking function in stroke patients with initial lower extremity paralysis: The Copenhagen Stroke Study. Arch Phys Med Rehabil 2000;81:736-738. 27. Dong Y, Slavin MJ, Chan BP, et al. Cognitive screening improves the predictive value of stroke severity scores for functional outcome 3-6 months after mild stroke and transient ischaemic attack: An observational study. BMJ Open 2013;3:e003105. 28. Yavuzer G, Kucukdeveci A, Arasil T, Elhan A. Rehabilitation of stroke patients: clinical profile and functional outcome. Am J Phys Med Rehabil 2001;80:250-255. 29. Park SW, Wolf SL, Blanton S, Winstein C, Nichols-Larsen DS. The EXCITE Trial: Predicting a clinically meaningful motor activity log outcome. Neurorehabil Neural Repair 2008;22:486-493. 30. Michael KM, Allen JK, Macko RF. Reduced ambulatory activity after stroke: the role of balance, gait, and cardiovascular fitness. Arch Phys Med Rehabil 2005;86:1552-1556. 31. Alzahrani M, Dean C, Ada L. Relationship between walking performance and types of community-based activities in people with stroke: An observational study. Rev Bras Fisioter 2011; 15:45-51.
399
32. Samsa GP, Matchar DB. How strong is the relationship between functional status and quality of life among persons with stroke? J Rehabil Res Dev 2004;41:279-282. 33. Vincent C, Desrosiers J, Landreville P, Demers L. Burden of caregivers of people with stroke: Evolution and predictors. Cerebrovasc Dis 2009;27:456-464. 34. Rabadi MH, Blau A. Admission ambulation velocity predicts length of stay and discharge disposition following stroke in an acute rehabilitation hospital. Neurorehabil Neural Repair 2005;19:20-26. 35. Wade DT, Wood VA, Heller A, Maggs J, Langton Hewer R. Walking after stroke. Measurement and recovery over the first 3 months. Scand J Rehabil Med 1987;19:25-30. 36. Veerbeek JM, Van Wegen EE, Harmeling-Van der Wel BC, Kwakkel G. Is accurate prediction of gait in nonambulatory stroke patients possible within 72 hours poststroke? The EPOS study. Neurorehabil Neural Repair 2011;25:268-274. 37. Duarte E, Marco E, Muniesa JM, Belmonte R, Aguilar JJ, Escalada F. Early detection of non-ambulatory survivors six months after stroke. NeuroRehabilitation 2010;26:317-323. 38. Barer DH. Continence after stroke: useful predictor or goal of therapy? Age Ageing 1989;18:183-191. 39. Salter KL, Teasell RW, Foley NC, Jutai JW. Outcome assessment in randomized controlled trials of stroke rehabilitation. Am J Phys Med Rehabil 2007;86:1007-1012.
Disclosure H.M. Department of Functional Activation, Center for Gerontology and Social Science, National Center for Geriatrics and Gerontology, 7-430 Morioka-machi, Obu, Aichi 474-8511, Japan. Address correspondence to: H.M.; e-mail: [email protected] Disclosure: nothing to disclose N.K. National Center for Geriatrics and Gerontology, Aichi, Japan; and Department of Physical Therapy, Funabashi Municipal Rehabilitation Hospital, Chiba, Japan Disclosure: nothing to disclose
A.T. Department of Speech Therapy, Funabashi Municipal Rehabilitation Hospital, Chiba, Japan Disclosure: nothing to disclose K.I. Department of Nursing and Care Work, Funabashi Municipal Rehabilitation Hospital, Chiba, Japan Disclosure: nothing to disclose Submitted for publication November 1, 2013; accepted January 17, 2015.