- Email: [email protected]
Effects of a Multifactorial Falls Prevention Program for People With Stroke Returning Home After Rehabilitation: A Randomized Controlled Trial
1648
ORIGINAL ARTICLE
Effects of a Multifactorial Falls Prevention Program for People With Stroke Returning Home After Rehabilitation: A Randomized Controlled Trial Frances A. Batchelor, PhD, Keith D. Hill, PhD, Shylie F. Mackintosh, PhD, Catherine M. Said, PhD, Craig H. Whitehead, MD ABSTRACT. Batchelor FA, Hill KD, Mackintosh SF, Said CM, Whitehead CH. Effects of a multifactorial falls prevention program for people with stroke returning home after rehabilitation: a randomized controlled trial. Arch Phys Med Rehabil 2012;93:1648-55. Objectives: To determine whether a multifactorial falls prevention program reduces falls in people with stroke at risk of recurrent falls and whether this program leads to improvements in gait, balance, strength, and fall-related efficacy. Design: A single blind, multicenter, randomized controlled trial with 12-month follow-up. Setting: Participants were recruited after discharge from rehabilitation and followed up in the community. Participants: Participants (N⫽156) were people with stroke at risk of recurrent falls being discharged home from rehabilitation. Interventions: Tailored multifactorial falls prevention program and usual care (n⫽71) or control (usual care, n⫽85). Main Outcome Measures: Primary outcomes were rate of falls and proportion of fallers. Secondary outcomes included injurious falls, falls risk, participation, activity, leg strength, gait speed, balance, and falls efficacy. Results: There was no significant difference in fall rate (intervention: 1.89 falls/person-year, control: 1.76 falls/personyear, incidence rate ratio⫽1.10, P⫽.74) or the proportion of fallers between the groups (risk ratio⫽.83, 95% confidence interval⫽.60 –1.14). There was no significant difference in injurious fall rate (intervention: .74 injurious falls/personyear, control: .49 injurious falls/person-year, incidence rate ratio⫽1.57, P⫽.25), and there were no significant differences between groups on any other secondary outcome. Conclusions: This multifactorial falls prevention program was not effective in reducing falls in people with stroke who
From the National Ageing Research Institute, Parkville, Victoria (Batchelor, Hill); the School of Health Sciences, University of Melbourne, Parkville, Victoria (Batchelor, Said); the School of Physiotherapy, Curtin University, Perth, Western Australia (Hill); School of Health Sciences, University of South Australia, Adelaide, South Australia (Mackintosh); Heidelberg Repatriation Hospital, Austin Health, Heidelberg, Victoria (Said); and Division of Rehabilitation, Aged Care and Allied Health, Repatriation General Hospital, Daw Park, South Australia (Whitehead), Australia. Presented to the Stroke Society of Australasia Annual Scientific Meeting, September 1–3, 2010, Melbourne, Australia, and to the Australian and New Zealand Falls Prevention Conference, November 21–23, 2010, Dunedin, New Zealand. Supported by Australian National Health and Medical Research Council project (grant ID 385002) and training fellowship (ID310612). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated. Correspondence to Frances A. Batchelor, PhD, National Ageing Research Institute, PO Box 2127, The Royal Melbourne Hospital, Parkville, Victoria 3050, Australia, e-mail: [email protected]. Reprints are not available from the author. In-press corrected proof published online on May 19, 2012, at www.archives-pmr.org. 0003-9993/12/9309-01215$36.00/0 http://dx.doi.org/10.1016/j.apmr.2012.03.031
Arch Phys Med Rehabil Vol 93, September 2012
are at risk of falls nor was it more effective than usual care in improving gait, balance, and strength in people with stroke. Further research is required to identify effective interventions for this high-risk group. Key Words: Accidental falls; Exercise; Randomized controlled trial; Rehabilitation; Stroke. © 2012 by the American Congress of Rehabilitation Medicine ALLS ARE COMMON after stroke,1,2 with fall rates in people with stroke higher than in the general older population.3 Studies have reported that between 46% and 73% of people with stroke fall during the first 6 months after discharge from hospital,2,4,5 compared with the 30% of communitydwelling older people who fall in a 1-year period.6 In particular, people with stroke who have impaired balance or have fallen in hospital are at risk of recurrent falls when they return home.4 In community-dwelling people with stroke, the causes of falls are multifactorial. However, the literature on the association between risk factors and falls and the predictors of falls is somewhat inconsistent. Falls risk factors identified in some studies include neglect,3 balance problems,7 motor and sensory impairment,8 and fear of falling,7 whereas other studies have found no difference between fallers and nonfallers on measures of balance and gait.9,10 Despite the high rate of falls in people with stroke, there are few published studies in community-dwelling people with stroke evaluating interventions where falls are the primary outcome. Of the published studies, few have identified effective interventions. A systematic review investigating the effectiveness of falls prevention strategies after stroke in any setting identified only 1 effective falls prevention intervention (vitamin D supplementation for older women with chronic stroke residing in an institutional setting).11 The review highlighted the lack of evidence for any single or multifactorial approaches being effective in reducing falls in the high falls risk group of people with stroke. This is in contrast to the strong evidence that a range of single and multifactorial interventions are effective in preventing falls in community-dwelling older people.12 Hence, there is a need for randomized controlled trials evaluating the impact of interventions on falls as a primary outcome in community-dwelling people with stroke. In particular, the evaluation of interventions incorporating strength and balance training as well as components that address general and stroke-specific falls risk factors is warranted, because people
F
List of Abbreviations CI IRR OEP
confidence interval incidence rate ratio Otago Exercise Programme
FALLS PREVENTION AFTER STROKE, Batchelor
with stroke will vary in their physical, cognitive, and perceptual abilities as well as their social circumstances. Therefore, we aimed to determine whether a tailored multifactorial falls prevention program prevents falls in people with stroke who are at risk of recurrent falls and whether this program leads to improved gait, balance, strength, and fallrelated efficacy. METHODS Participants People with stroke were eligible if they were aged 45 years or more, had been discharged home after rehabilitation, and were at high risk of falls. A person was determined to have high falls risk if he/she either had fallen during hospital admission or had a Step Test13 worse leg score of less than 7, or a Berg Balance Scale14 score of less than 49, because these variables have been shown to predict multiple falls in the first 6 months after discharge from stroke rehabilitation.4 Those discharged to residential care facilities or with homes more than 100 kilometers from study sites were ineligible. We recruited participants from 9 health services across Melbourne and Adelaide, Australia, between October 2006 and November 2010. Rehabilitation physiotherapy staff assessed eligibility, and if participants were agreeable, the research team contacted them within 2 weeks of discharge. The research team undertook all assessment and intervention in participants’ homes. Study Design We conducted a single blind, multicenter, randomized controlled trial with 12-month follow-up. After baseline assessment, participants were allocated into either the control group or the intervention group (1:1 allocation ratio, simple randomization) using a computer-generated random allocation sequence concealed from all researchers in opaque envelopes. Staff independent of the study undertook sequence and concealment. The relevant health authorities/university human research ethics committees approved the study. The trial was registered with the Australian New Zealand Clinical Trials registry (ACTRN12607000398404), and the protocol has been published.15 Outcome Measures The primary outcomes were fall rates and the proportion of fallers in the 12 months after baseline assessment. The secondary falls outcome was the rate of injurious falls. We monitored falls prospectively for 12 months using calendars. A fall was defined as a sudden, unexpected event in which an individual comes to rest on the ground, the floor, or other surface.16 Participants completed information about falls and injuries each month, returning calendar pages using prepaid envelopes. A researcher blind to group allocation phoned participants who did not return diaries within 2 weeks of the due date. Other secondary outcomes included leg strength (Sit-ToStand test17), gait speed over 5 meters,18 balance (Step Test13), activity (Human Activity Profile19), functional independence (FIM20), fear of falling (Falls Efficacy Scale – Swedish Modification21), and falls risk (Falls Risk for Older People – Community setting22). Other measures collected included unilateral neglect (combination of Baking Tray Task/Star Cancellation Test23), visual fields (Visual Field Confrontation Test24), and cognitive status (Abbreviated Mental Test Score25).
1649
The physiotherapists conducting baseline and the follow-up assessment were blind to group allocation. Intervention All intervention participants received usual care from their treating health professionals. Typically, following discharge from rehabilitation, this included referral for ongoing therapy (physiotherapy and occupational therapy) and follow-up by their general medical practitioner. In addition, a physiotherapist provided participants in the intervention group with a multifactorial, individually tailored falls prevention program. This consisted of the following: 1. Individualized home exercise program prescribed by a physiotherapist and based on the Otago Exercise Programme (OEP).26,27 The OEP has been previously shown to reduce falls in older community-dwelling adults.28,29 Exercises were selected from those included in the OEP to address balance and mobility problems identified in the baseline assessment (table 1). Adherence was assessed through exercise diaries completed by participants and discussion with the physiotherapist at each review and following completion of the study. Average adherence over 12 months was categorized as full (ⱖ3 times/wk), partial (1–2 times/wk), or nonadherent (⬍1 time/wk). 2. Falls risk minimization strategies based on general and stroke-specific risk factors identified in the baseline assessment, including those identified from the Falls Risk for Older People – Community setting (see table 1). Guidelines for this component of the intervention are also available in an earlier publication.15 3. Education (written and verbal) for participant and carer about identified falls risk factors and risk minimization. 4. Injury risk minimization strategies for those at high risk of fracture (based on delayed walking after stroke,30 previous diagnosis of osteoporosis). This included hip protector prescription and liaison with the participant’s general practitioner regarding vitamin D/calcium supplementation. Details of the exercise program and falls risk minimization strategies are contained in table 1. We also provided intervention participants with a falls prevention booklet, “A Guide to Preventing Falls.”a Control Participants in the control group received usual care with their treating health professionals, and research staff made no attempt to limit their access to any care. In addition, we provided control participants with the falls prevention booklet as above. Sample Size The literature suggests that between 46% and 73% of people with stroke fall during the first 6 months after discharge from hospital.4,5 Because we were recruiting people with high falls risk only, we estimated that 75% of the usual care group would experience a fall in a 1-year period. On the basis of the intervention reducing falls incidence by one third to 50%, using a power of 80% (P⫽.05, 2-sided), and allowing for 25% dropout, we set the recruitment target at 80 participants in each group. Data Analysis We analyzed the data on the basis of intention to treat. Participants who commenced the study but did not complete the full 12 months were included in the analysis of fall rate Arch Phys Med Rehabil Vol 93, September 2012
1650
FALLS PREVENTION AFTER STROKE, Batchelor Table 1: Intervention Detail Component
Content
1. Home exercise program based on the OEP
30 – 40-min home exercise sessions, 3–5 times weekly Exercises selected by the physiotherapist on the basis of findings of the initial assessment; weights (ankle cuff) prescribed according to ability; level of support for balance exercises (eg, one-hand support) prescribed according to assessment findings Exercises selected from the following: Warm-up exercises: Head/neck movements Back extension Trunk movements Ankle movements Strength and balance exercises: Knee extension/flexion Hip abduction Heel raise/toe raise Knee bends Backward walk Walk and turn Sideways walk Heel-toe stand Heel-toe walk One-leg stand Heel walking/toe walking Heel-toe backward walk Sit to stand Stair walking Walking recommended 3–5 times weekly. Time set according to assessment findings Exercise and walking program monitored and progressed at 4 and 8mo after initial session Combination of direct intervention and referral; with participant’s consent, written summary of falls risk factors and suggested actions provided to medical practitioner; written and verbal summary of strategies provided to participant Examples of possible falls risk factors and management options: Medications (⬎3, taking medications associated with falls): Referral to medical practitioner Visual impairment (eg, visual field deficit): Referral to training program for people with neurological visual loss; referral for vision assessment Hearing impairment: Referral to audiologist Inappropriate footwear: Provide advice about good footwear for falls prevention Assist participant to purchase appropriate footwear Foot problems: Referral to podiatrist Continence issues, including nocturia: Ensure participant has appropriate equipment, eg, commode, urinal Referral to continence clinic/specialist Home safety issues: Referral for occupational therapy home assessment Provide advice on lighting and home safety
2. Falls risk minimization strategies based on falls risk factors identified at baseline from the FROP-Com
Abbreviation: FROP-Com, Falls Risk for Older People – Community setting.
and rate of injurious falls if they had at least 1 month of falls data available. To compare fall rates between groups, we used negative binomial regression.31,32 Total falls were compared taking into consideration the time of exposure. Chi-square analysis was used to compare the proportion of fallers and nonfallers between groups. For this analysis, participants who did not complete the full study and had not reported any falls (in the months for which falls data were available) were excluded, because they could not be correctly classified. Arch Phys Med Rehabil Vol 93, September 2012
Linear regression was used to determine the difference between groups (intervention effect) on the secondary outcomes. Baseline status on each measure and group was included in the model as independent variables. To analyze differences within groups from baseline to follow-up, paired sample t tests were used because the difference in scores between baseline and follow-up followed a normal distribution. To evaluate the difference in the proportion of participants reporting falls prevention and injury minimization strategies at follow-up, we used chi-square analysis.
FALLS PREVENTION AFTER STROKE, Batchelor
The alpha level for all analyses was set at .05. Statistical analyses were undertaken using SPSS 16.0 GradPack.b RESULTS From the 156 participants recruited, 85 were randomized to the control group and 71 to the intervention group (fig 1). On completion of the study, 144 (92%) had at least 1 month of fall data available for analysis. Twelve additional participants (5 control, 7 intervention) did not complete the final assessment. Overall, there was an 85% retention rate. The 2 groups were similar on all assessed characteristics at baseline (table 2). Overall, participants had high falls risk, poor balance and mobility, and low activity levels. Falls Outcomes In the 2 weeks between discharge home and baseline assessment, 28% (44 of 156) of the participants fell. In total, 119 falls were reported in the intervention group (n⫽64) and 140 in the
1651
control group (n⫽80). One intervention participant reported 20 falls, and 4 participants (2 control, 2 intervention) reported 10 or more falls. The rate of falls (falls/person-year) for participants with at least 1 month of falls data (n⫽144) was 1.89 in the intervention group and 1.76 in the control group (table 3). There was no significant difference in fall rate between the groups, with an incidence rate ratio (IRR) for total falls (intervention:control) of 1.10 (95% confidence interval [CI]⫽.63– 1.90). To determine whether the participant with 20 falls skewed the results, the participant’s data were excluded and the analysis was rerun. However, although the fall rate in the intervention group decreased, there was no significant difference between the groups (IRR⫽.92, 95% CI⫽.54 –1.57) (see table 3). Forty-eight percent (29 of 60) of the intervention participants and 58% (46 of 79) of the control participants reported ⱖ1 falls. There was no significant difference in the proportion of fallers between the groups (risk ratio⫽.83, 95% CI⫽.60 –1.14, P⫽.256).
Fig 1. Flow of participants through the study.
Arch Phys Med Rehabil Vol 93, September 2012
1652
FALLS PREVENTION AFTER STROKE, Batchelor
Table 2: Baseline Characteristics of Participants (Nⴝ156) Characteristic
Age (y), mean ⫾ SD Female Stroke type Infarct Hemorrhage Unknown Side of symptoms Right Left Bilateral Other/unknown Time from stroke to baseline (mo), mean ⫾ SD Living arrangements Alone Spouse/carer Family AMTS (/10), median (IQR) FIM (/126), mean ⫾ SD Faller between discharge and baseline Time between discharge and baseline (d), median (IQR) FROP-Com score (/60), mean ⫾ SD FES-S (/10), median (IQR) HAP-AAS (/94), mean ⫾ SD Presence of inattention/neglect Presence of visual field deficit STS, number completed in 30s, median (IQR) Gait speed (m/min), mean ⫾ SD Step Test, no., median (IQR)
Intervention (n⫽71)
Control (n⫽85)
70.8⫾11.4 26 (36.6)
72.2⫾9.9 31 (36.5)
55 (77.5) 13 (18.3) 3 (4.2)
69 (81.2) 15 (17.6) 1 (1.2)
24 (33.8) 43 (60.6) 2 (2.8) 2 (2.8) 3.0⫾1.6
38 (44.7) 44 (51.8) 2 (2.4) 1 (1.2) 3.1⫾1.9
19 (26.8) 34 (47.9) 18 (25.4) 9.0 (2.0) 107.7⫾14.6
18 (21.2) 53 (62.4) 14 (16.5) 9.0 (2.0) 106.9⫾18.3
22 (31.0)
22 (25.9)
12.0 (8.0)
13.0 (11.0)
22.4⫾5.7 7.7 (3.2) 26.1⫾13.4
21.2⫾5.9 7.5 (2.5) 26.0⫾13.3
22 (32.8)
17 (21.5)
26 (36.6)
25 (29.4)
5.3 (7.9)
4.7 (7.4)
32.1⫾16.4 4 (6)
31.9⫾20.0 2 (6)
NOTE. Values are n (%) or as otherwise indicated. Abbreviations: AMTS, Abbreviated Mental Test Score; FES-S, Falls Efficacy Scale – Swedish Modification; FROP-Com, Falls Risk for Older People – Community setting; HAP-AAS, Human Activity Profile –Adjusted Activity Score; IQR, interquartile range; STS, Sit-to-Stand test.
There was no significant difference in the proportion of frequent fallers (ⱖ2 falls) between the groups (risk ratio⫽1.20, 95% CI⫽.74 –1.94, P⫽.454). The number of injurious falls was 46 in the intervention group (n⫽59) and 38 in the control group (n⫽78) (missing data in both groups). The rate of injurious falls was .74 and .49 per
person-year in intervention and control groups, respectively. The estimated IRR was 1.57 (95% CI⫽.73–3.4, P⫽.248), indicating no significant difference between groups. Secondary Outcomes Both groups improved significantly over 12 months on measures of falls risk (Falls Risk for Older People – Community setting), activity (Human Activity Profile19), leg strength (Sitto-Stand test), and gait speed and balance (Step Test) (table 4). Only participants from the intervention group improved significantly on fall-related efficacy (Falls Efficacy Scale – Swedish Modification). There were no significant differences in change scores between groups on any secondary outcome measure (see table 4). Adherence to the Exercise Program Of the 64 intervention participants for whom falls data were available, 16 (25.0%) fully adhered, 36 (56.3%) partially adhered, and 12 (18.7%) did not adhere to the exercise program during their time in the study (table 5). There was a significant difference in the proportion of fallers to nonfallers across adherence categories (2⫽8.288, P⫽.016), with a significantly higher proportion of fallers among those who partially adhered. The rate of falls was 66% lower in those who fully adhered than in those who partially adhered (IRR⫽.34, 95% CI⫽.15– .78, P⫽.01). Those classified as nonadherent also had a lower, but nonsignificant, falls rate than did those who partially adhered (IRR⫽.77, 95% CI⫽.32–1.82) (see table 5). The falls rate was 55% lower in those who fully adhered than in those classified as nonadherent, but again this was not significant (IRR⫽.45, 95% CI⫽.16 –1.28). At follow-up, a significantly higher proportion of intervention participants compared with control participants reported home exercise program participation (64.9% vs 44.6%, P⫽.026). For all other self-reported activities relating to falls prevention, there was no difference between the intervention and control groups. DISCUSSION To the best of our knowledge, this is the first study to report the effects of a multifactorial falls prevention program on falls in people with stroke returning home after rehabilitation. Although the proportion of fallers was less than expected in both groups (48% in the intervention group and 58% in the control group), this study confirmed that people with stroke are at high risk of falls, particularly in the first weeks after discharge from rehabilitation, with almost one third of participants falling between discharge and baseline assessment, a time period of, on average, 2 weeks. The results of the study did not support the hypothesis that a multifactorial falls prevention intervention reduces falls rate or the proportion of fallers in this population. The study did show that people with stroke at risk of recurrent falls significantly improved in the 12 months after discharge on falls risk, activity, strength, gait, and balance.
Table 3: Randomized Controlled Trial Outcome: Fall Rate Rate (falls/person-year) Participant Category
Intervention
Control
IRR InterventionControl (95% CI)*
P
All participants Participants with ⱖ20 falls excluded All participants, injurious falls
1.89 (n⫽64) 1.59 (n⫽63) .74 (n⫽59)
1.76 (n⫽80) 1.76 (n⫽80) .49 (n⫽78)
1.10 (.63–1.90) 0.92 (.54–1.57) 1.57 (.73–3.4)
.738 .752 .248
*From negative binomial regression model.
Arch Phys Med Rehabil Vol 93, September 2012
1653
FALLS PREVENTION AFTER STROKE, Batchelor Table 4: Secondary Outcome Scores Intervention
Characteristic
n
Baseline
Follow-up
FIM, median (IQR)
57
112.0⫾14.0
114.0⫾14.0
FROP-Com†
57
21.7⫾5.4
18.6⫾6.1
FES-S, median (IQR) HAP-AAS
55
7.8⫾3.2
8.1⫾2.1
57
27.2⫾13.0
36.6⫾17.0
57
6.0⫾8.0
7.6⫾5.5
57
32.4⫾15.8
38.1⫾17.0
57
4.0⫾6.5
7.0⫾6.0
STS, number completed in 30s, median (IQR) Gait speed (m/min) Step Test, no., median (IQR)
Control Difference Within Group (baselinefollow-up) (95% CI)
1.6 (⫺2.1 to 5.4) P⫽.379 ⫺3.1 (⫺4.9 to ⫺1.4) P⫽.001 0.5 (.04 to 0.9) P⫽.034 9.4 (4.9 to 13.9) P⬍.001 1.6 (0.5 to 2.6) P⫽.003 5.7 (1.8 to 9.6) P⫽.005 1.8 (1.0 to 2.6) P⬍.001
n
Baseline
Follow-up
75
114.0⫾17.0
115.0⫾20.0
74
20.8⫾6.0
18.4⫾6.7
70
7.9⫾2.4
7.8⫾3.1
75
26.7⫾13.2
34.4⫾20.6
75
5.4⫾7.9
7.8⫾7.3
75
32.3⫾20.4
37.1⫾22.1
75
3.0⫾6.0
6.0⫾9.0
Difference Within Group (baselinefollow-up) (95% CI)
Difference Between Groups (baselinefollow-up)* (95% CI)
0.4 (⫺2.2 to 2.9) 2.5 (⫺1.7 to 6.7) P⫽.756 P⫽.240 ⫺2.4 (⫺3.6 to ⫺1.2) ⫺0.6 (⫺2.5 to 1.3) P⬍.001 P⫽.554 ⫺0.3 (⫺0.7 to 0.2) 0.6 (⫺.03 to 1.1) P⫽.274 P⫽.062 7.7 (4.5 to 10.9) 1.8 (⫺3.6 to 7.1) P⬍.001 P⫽.514 2.4 (1.6 to 3.2) ⫺0.7 (⫺1.9 to 0.6) P⬍.001 P⫽.282 4.9 (2.2 to 7.6) P⫽.001 2.1 (1.2 to 2.9) P⬍.001
0.9 (⫺3.6 to 5.4) P⫽.691 ⫺0.1 (⫺1.3 to 1.1) P⫽.844
NOTE. Values are mean ⫾ SD unless otherwise indicated. Abbreviations: FES-S, Falls Efficacy Scale – Swedish Modification; FROP-Com, Falls Risk for Older People – Community setting; HAP-AAS, Human Activity Profile – Adjusted Activity Score; STS, Sit-to-Stand test. *Calculated from linear regression model with baseline score and group (control or intervention) as independent variables. † Lower score indicates improved performance.
Issues related to sample size, recruitment, heterogeneity of the sample, the intervention, and the potential for similarity in intervention and control conditions may have contributed to the lack of a significant effect. Sample size estimation was based on a reduction in the proportion of fallers from 75% (in the control group) by one third. Our results found that the falls rate in the control group was smaller than anticipated and closer to 60%. In addition, the effect of the intervention was smaller than anticipated and closer to a 10% reduction in fallers. As a consequence, the sample size was not large enough to detect significant differences between groups. To detect a significant effect of the intervention based on reducing the proportion of fallers by one third to 40%, a sample size of 107 per group would be required (power of 80%, ␣⫽.05), larger than the sample size of the current study. However, given that the comparison of fall rate was not close to significant, it is unlikely that increasing the sample size alone would be enough to show significant effects across all comparisons. One possible reason for the smaller than anticipated effect size is that the intervention provided was not sufficiently different to “usual care.” It is also possible that there was contamination between the groups. For example, those in the control group were more likely to have sought out additional therapy or falls prevention information compared with those in the intervention group. This was confirmed by the lack of difference (apart from home exercise program participation) Table 5: Proportion of Fallers and Fall Rates by Adherence to Exercise Program (Intervention Participants), nⴝ64 n (%) Adherence Category
Nonfaller
Faller
IRR* (95% CI)
Nonadherent Partial adherence† Full adherence
9 (75.0) 14 (38.9) 12 (75.0)
3 (25.0) 22 (61.1) 4 (25.0)
0.77 (.32–1.82) 1.00 .34 (.15–.78)
*From negative binomial regression model. † Chi-square⫽8.288, P⫽.016 for the difference in proportions.
between the groups in strategies reported at follow-up. In addition, the provision of information about the project to rehabilitation staff discharging participants may have led to an increase in the number of falls prevention strategies provided to participants in both groups. It is also possible that the OEP was not effective in this population. Although the OEP has been shown to be effective in reducing falls in community-dwelling older adults,28,29 these effects did not necessarily carryover to older people at high risk of falls.33 In addition, although the exercises in the present study were individually tailored, safety considerations (unsupervised exercise at home) may have led physiotherapists to prescribing balance exercises that were not challenging enough to be effective. Adherence to exercise may have been another factor in the lack of effect. Only one quarter of intervention participants fully adhered to the exercise program, while over one half partially adhered. The relationship between falls and adherence to exercise was not linear. The proportion of fallers was significantly lower in those who fully adhered and in those who were nonadherent than in those who partially adhered. This contrasts with the linear relationship between adherence and falls rate reported in a study of community-dwelling older people with poor vision where increased adherence was associated with decreased fall rates.28 A hypothesis for the nonlinearity is that those who did not experience falls saw no need to complete the exercise program and those who fully adhered gained observable benefit. Alternatively, those who did very little exercise may have had low activity levels that decreased exposure to activities that have a higher falls risk. Further exploration into factors that influence adherence and methods of facilitating adherence in this population is required. This study’s findings raise issues that warrant further investigation. There is still no high-level evidence that any intervention prevents falls in community-dwelling people with stroke. Interventions including vitamin D, evaluated in combination with exercise or as part of a multifactorial falls prevention program, are worthy of consideration in the community-dwelling stroke popuArch Phys Med Rehabil Vol 93, September 2012
1654
FALLS PREVENTION AFTER STROKE, Batchelor
lation because this treatment appears to reduce falls in institutionalized women with stroke.34 Evaluation of an intervention that commences prior to discharge and of a more intensive exercise intervention is also warranted because it is possible that the timing and the intensity of the present study’s intervention was a factor in the null finding. Study Limitations Beyond the potential reasons for the lack of significant effect as discussed in the previous section, the study had some limitations. First, because participants were unable to be blinded to the purpose of the study, it is possible that participants were sensitized to prevent falls, leading to all participants being more cautious or less active. This may have contributed to bias toward the null hypothesis. Conversely, there may have been a bias toward the alternative hypothesis because participants were not blind to group allocation, leading to the potential for intervention participants to underreport falls. Second, the heterogeneity of the sample may have been a factor contributing to the observed results. Our inclusion criteria meant that people with all levels of functional ability (as long as they met the high falls risk criteria) could participate in the trial, and this diversity may have masked positive findings. CONCLUSIONS This study did not support the hypothesis that this multifactorial falls prevention program is effective in people with stroke who are at risk of recurrent falls. Despite the lack of overall effect, this study highlights the high risk of falls in this group and provides direction for future studies aiming to prevent falls in people with stroke returning home. Acknowledgments: We thank Caroline Fryer, BAppSc(Physio), GradDipClinEpi for her invaluable assistance with data collection throughout the study. We also thank the physiotherapists who assisted with recruitment. References 1. Mackintosh SF, Hill K, Dodd KJ, Goldie P, Culham E. Falls and injury prevention should be part of every stroke rehabilitation plan. Clin Rehabil 2005;19:441-51. 2. Kerse N, Parag V, Feigin VL, et al, the Auckland Regional Community Stroke Study Group. Falls after stroke: results from the Auckland Regional Community Stroke (ARCOS) study, 2002 to 2003. Stroke 2008;39:1890-3. 3. Mackintosh SFH, Goldie P, Hill K. Falls incidence and factors associated with falling in older, community-dwelling, chronic stroke survivors (⬎1 year after stroke) and matched controls. Aging Clin Exp Res 2005;17:74-81. 4. Mackintosh SF, Hill KD, Dodd KJ, Goldie PA, Culham EG. Balance score and a history of falls in hospital predict recurrent falls in the 6 months following stroke rehabilitation. Arch Phys Med Rehabil 2006;87:1583-9. 5. Forster A, Young J. Incidence and consequences of falls due to stroke: a systematic inquiry. BMJ 1995;311:83-6. 6. Campbell AJ, Borrie MJ, Spears GF, Jackson SL, Brown JS, Fitzgerald JL. Circumstances and consequences of falls experienced by a community population 70 years and over during a prospective study. Age Ageing 1990;19:136-41. 7. Belgen B, Beninato M, Sullivan PE, Narielwalla K. The association of balance capacity and falls self-efficacy with history of falling in community-dwelling people with chronic stroke. Arch Phys Med Rehabil 2006;87:554-61. 8. Yates JS, Lai SM, Duncan PW, Studenski S. Falls in communitydwelling stroke survivors: an accumulated impairments model. J Rehabil Res Dev 2002;39:385-94. Arch Phys Med Rehabil Vol 93, September 2012
9. Harris JE, Eng JJ, Marigold DS, Tokuno CD, Louis CL. Relationship of balance and mobility to fall incidence in people with chronic stroke. Phys Ther 2005;85:150-8. 10. Lamb SE, Ferrucci L, Volapto S, Fried LP, Guralnik JM. Risk factors for falling in home-dwelling older women with stroke: The Women’s Health and Aging Study. Stroke 2003;34:494-500. 11. Batchelor F, Hill K, Mackintosh S, Said C. What works in falls prevention after stroke? Systematic review and meta-analysis. Stroke 2010;41:1715-22. 12. Gillespie L, Robertson M, Gillespie W, et al. Interventions for preventing falls in older people living in the community. Cochrane Database Syst Rev 2009;2:CD007146. 13. Hill K, Bernhardt J, McGann A, Maltese D, Berkovits D. A new test of dynamic standing balance for stroke patients: reliability, validity, and comparison with healthy elderly. Physiother Can 1996;48:257-62. 14. Berg KO, Wood-Dauphinee SL, Williams JI. Measuring balance in the elderly— validation of an instrument. Can J Public Health 1992;83:S7-11. 15. Batchelor FA, Hill KD, Mackintosh SF, Said CM, Whitehead C. The FLASSH study: protocol for a randomised controlled trial evaluating falls prevention after stroke and two sub-studies. BMC Neurol 2009;9:14. 16. Lamb S, Jørstad-Stein E, Hauer K, Becker C, on behalf of the Prevention of Falls Network Europe and Outcomes Consensus Group. Development of a common outcome data set for fall injury prevention trials: the Prevention of Falls Network Europe consensus. J Am Geriatr Soc 2005;53:1618-22. 17. McCarthy EK, Horvat MA, Holtsberg PA, Wisenbaker JM. Repeated chair stands as a measure of lower limb strength in sexagenarian women. J Gerontol A Biol Sci Med Sci 2004;59: 1207-12. 18. Steffen TM, Hacker TA, Mollinger L. Age- and gender-related test performance in community-dwelling elderly people: SixMinute Walk test, Berg Balance Scale, Timed Up & Go test, and gait speeds. Phys Ther 2002;82:128-37. 19. Fix A, Daughton D. Human Activity Profile (professional manual). Odessa: Psychological Assessment Resources, Inc; 1988. 20. Hamilton B, Granger C. Disability outcomes following inpatient rehabilitation for stroke. Phys Ther 1994;74:494-503. 21. Hellström K, Lindmark B. Fear of falling in patients with stroke: a reliability study. Clin Rehabil 1999;13:509-17. 22. Russell MA, Hill KD, Blackberry I, Day LM, Dharmage SC. The reliability and predictive accuracy of the falls risk for older people in the community assessment (FROP-Com) tool. Age Ageing 2008;37:634-9. 23. Bailey MJ, Riddoch MJ, Crome P. Test-retest stability of three tests for unilateral visual neglect in patients with stroke: Star Cancellation, Line Bisection, and the Baking Tray Task. Neuropsychol Rehabil 2004;14:403-19. 24. Elliott DB, North I, Flanagan J. Confrontation visual field tests. Ophthalmic Physiol Opt 1997;17:S17-24. 25. Hodkinson HM. Evaluation of a mental test score for assessment of mental impairment in the elderly. Age Ageing 1972;1:233-8. 26. ACC NZ Web site. Otago Exercise Programme. Available from: http://www.acc.co.nz/preventing-injuries/at-home/older-people/ information-for-older-people/otago-exercise-programme/index. htm. Accessed March 27, 2011. 27. Gardner MM, Buchner DM, Robertson MC, Campbell AJ. Practical implementation of an exercise-based falls prevention programme. Age Ageing 2001;30:77-83. 28. Campbell AJ, Robertson MC, Grow SJA, et al. Randomised controlled trial of prevention of falls in people aged ⱖ75 with severe visual impairment: the VIP trial. BMJ 2005;331:817.
FALLS PREVENTION AFTER STROKE, Batchelor
29. Robertson MC, Campbell AJ, Gardner MM, Devlin N. Preventing injuries in older people by preventing falls: a meta-analysis of individual-level data. J Am Geriatr Soc 2002;50:905-11. 30. Jørgensen L, Jacobsen B, Wilsgaard T, Magnus J. Walking after stroke: does it matter? Changes in bone mineral density within the first 12 months after stroke. A longitudinal study. Osteoporos Int 2000;11:381-7. 31. Robertson MC, Campbell AJ, Herbison P. Statistical analysis of efficacy in falls prevention trials. J Gerontol A Biol Sci Med Sci 2005;60:530-4. 32. Donaldson MG, Sobolev B, Cook WL, Janssen PA, Khan KM. Analysis of recurrent events: a systematic review of randomised controlled trials of interventions to prevent falls. Age Ageing 2009;38:151-5.
1655
33. Elley CR, Robertson MC, Garrett S, et al. Effectiveness of a falls-and-fracture nurse coordinator to reduce falls: a randomized, controlled trial of at-risk older adults. J Am Geriatr Soc 2008;56: 1383-9. 34. Sato Y, Iwamoto J, Kanoko T, Satoh K. Low-dose vitamin D prevents muscular atrophy and reduces falls and hip fractures in women after stroke: a randomized controlled trial. Cerebrovasc Dis 2005;20:187-92. Suppliers a. Peninsula Health, Mt Eliza Centre, Jacksons Rd, Mt Eliza, Victoria 3930, Australia. b. SPSS, Inc, IBM Corporation, 1 New Orchard Rd, Armonk, NY 10504.
Arch Phys Med Rehabil Vol 93, September 2012
ORIGINAL ARTICLE
Effects of a Multifactorial Falls Prevention Program for People With Stroke Returning Home After Rehabilitation: A Randomized Controlled Trial Frances A. Batchelor, PhD, Keith D. Hill, PhD, Shylie F. Mackintosh, PhD, Catherine M. Said, PhD, Craig H. Whitehead, MD ABSTRACT. Batchelor FA, Hill KD, Mackintosh SF, Said CM, Whitehead CH. Effects of a multifactorial falls prevention program for people with stroke returning home after rehabilitation: a randomized controlled trial. Arch Phys Med Rehabil 2012;93:1648-55. Objectives: To determine whether a multifactorial falls prevention program reduces falls in people with stroke at risk of recurrent falls and whether this program leads to improvements in gait, balance, strength, and fall-related efficacy. Design: A single blind, multicenter, randomized controlled trial with 12-month follow-up. Setting: Participants were recruited after discharge from rehabilitation and followed up in the community. Participants: Participants (N⫽156) were people with stroke at risk of recurrent falls being discharged home from rehabilitation. Interventions: Tailored multifactorial falls prevention program and usual care (n⫽71) or control (usual care, n⫽85). Main Outcome Measures: Primary outcomes were rate of falls and proportion of fallers. Secondary outcomes included injurious falls, falls risk, participation, activity, leg strength, gait speed, balance, and falls efficacy. Results: There was no significant difference in fall rate (intervention: 1.89 falls/person-year, control: 1.76 falls/personyear, incidence rate ratio⫽1.10, P⫽.74) or the proportion of fallers between the groups (risk ratio⫽.83, 95% confidence interval⫽.60 –1.14). There was no significant difference in injurious fall rate (intervention: .74 injurious falls/personyear, control: .49 injurious falls/person-year, incidence rate ratio⫽1.57, P⫽.25), and there were no significant differences between groups on any other secondary outcome. Conclusions: This multifactorial falls prevention program was not effective in reducing falls in people with stroke who
From the National Ageing Research Institute, Parkville, Victoria (Batchelor, Hill); the School of Health Sciences, University of Melbourne, Parkville, Victoria (Batchelor, Said); the School of Physiotherapy, Curtin University, Perth, Western Australia (Hill); School of Health Sciences, University of South Australia, Adelaide, South Australia (Mackintosh); Heidelberg Repatriation Hospital, Austin Health, Heidelberg, Victoria (Said); and Division of Rehabilitation, Aged Care and Allied Health, Repatriation General Hospital, Daw Park, South Australia (Whitehead), Australia. Presented to the Stroke Society of Australasia Annual Scientific Meeting, September 1–3, 2010, Melbourne, Australia, and to the Australian and New Zealand Falls Prevention Conference, November 21–23, 2010, Dunedin, New Zealand. Supported by Australian National Health and Medical Research Council project (grant ID 385002) and training fellowship (ID310612). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated. Correspondence to Frances A. Batchelor, PhD, National Ageing Research Institute, PO Box 2127, The Royal Melbourne Hospital, Parkville, Victoria 3050, Australia, e-mail: [email protected]. Reprints are not available from the author. In-press corrected proof published online on May 19, 2012, at www.archives-pmr.org. 0003-9993/12/9309-01215$36.00/0 http://dx.doi.org/10.1016/j.apmr.2012.03.031
Arch Phys Med Rehabil Vol 93, September 2012
are at risk of falls nor was it more effective than usual care in improving gait, balance, and strength in people with stroke. Further research is required to identify effective interventions for this high-risk group. Key Words: Accidental falls; Exercise; Randomized controlled trial; Rehabilitation; Stroke. © 2012 by the American Congress of Rehabilitation Medicine ALLS ARE COMMON after stroke,1,2 with fall rates in people with stroke higher than in the general older population.3 Studies have reported that between 46% and 73% of people with stroke fall during the first 6 months after discharge from hospital,2,4,5 compared with the 30% of communitydwelling older people who fall in a 1-year period.6 In particular, people with stroke who have impaired balance or have fallen in hospital are at risk of recurrent falls when they return home.4 In community-dwelling people with stroke, the causes of falls are multifactorial. However, the literature on the association between risk factors and falls and the predictors of falls is somewhat inconsistent. Falls risk factors identified in some studies include neglect,3 balance problems,7 motor and sensory impairment,8 and fear of falling,7 whereas other studies have found no difference between fallers and nonfallers on measures of balance and gait.9,10 Despite the high rate of falls in people with stroke, there are few published studies in community-dwelling people with stroke evaluating interventions where falls are the primary outcome. Of the published studies, few have identified effective interventions. A systematic review investigating the effectiveness of falls prevention strategies after stroke in any setting identified only 1 effective falls prevention intervention (vitamin D supplementation for older women with chronic stroke residing in an institutional setting).11 The review highlighted the lack of evidence for any single or multifactorial approaches being effective in reducing falls in the high falls risk group of people with stroke. This is in contrast to the strong evidence that a range of single and multifactorial interventions are effective in preventing falls in community-dwelling older people.12 Hence, there is a need for randomized controlled trials evaluating the impact of interventions on falls as a primary outcome in community-dwelling people with stroke. In particular, the evaluation of interventions incorporating strength and balance training as well as components that address general and stroke-specific falls risk factors is warranted, because people
F
List of Abbreviations CI IRR OEP
confidence interval incidence rate ratio Otago Exercise Programme
FALLS PREVENTION AFTER STROKE, Batchelor
with stroke will vary in their physical, cognitive, and perceptual abilities as well as their social circumstances. Therefore, we aimed to determine whether a tailored multifactorial falls prevention program prevents falls in people with stroke who are at risk of recurrent falls and whether this program leads to improved gait, balance, strength, and fallrelated efficacy. METHODS Participants People with stroke were eligible if they were aged 45 years or more, had been discharged home after rehabilitation, and were at high risk of falls. A person was determined to have high falls risk if he/she either had fallen during hospital admission or had a Step Test13 worse leg score of less than 7, or a Berg Balance Scale14 score of less than 49, because these variables have been shown to predict multiple falls in the first 6 months after discharge from stroke rehabilitation.4 Those discharged to residential care facilities or with homes more than 100 kilometers from study sites were ineligible. We recruited participants from 9 health services across Melbourne and Adelaide, Australia, between October 2006 and November 2010. Rehabilitation physiotherapy staff assessed eligibility, and if participants were agreeable, the research team contacted them within 2 weeks of discharge. The research team undertook all assessment and intervention in participants’ homes. Study Design We conducted a single blind, multicenter, randomized controlled trial with 12-month follow-up. After baseline assessment, participants were allocated into either the control group or the intervention group (1:1 allocation ratio, simple randomization) using a computer-generated random allocation sequence concealed from all researchers in opaque envelopes. Staff independent of the study undertook sequence and concealment. The relevant health authorities/university human research ethics committees approved the study. The trial was registered with the Australian New Zealand Clinical Trials registry (ACTRN12607000398404), and the protocol has been published.15 Outcome Measures The primary outcomes were fall rates and the proportion of fallers in the 12 months after baseline assessment. The secondary falls outcome was the rate of injurious falls. We monitored falls prospectively for 12 months using calendars. A fall was defined as a sudden, unexpected event in which an individual comes to rest on the ground, the floor, or other surface.16 Participants completed information about falls and injuries each month, returning calendar pages using prepaid envelopes. A researcher blind to group allocation phoned participants who did not return diaries within 2 weeks of the due date. Other secondary outcomes included leg strength (Sit-ToStand test17), gait speed over 5 meters,18 balance (Step Test13), activity (Human Activity Profile19), functional independence (FIM20), fear of falling (Falls Efficacy Scale – Swedish Modification21), and falls risk (Falls Risk for Older People – Community setting22). Other measures collected included unilateral neglect (combination of Baking Tray Task/Star Cancellation Test23), visual fields (Visual Field Confrontation Test24), and cognitive status (Abbreviated Mental Test Score25).
1649
The physiotherapists conducting baseline and the follow-up assessment were blind to group allocation. Intervention All intervention participants received usual care from their treating health professionals. Typically, following discharge from rehabilitation, this included referral for ongoing therapy (physiotherapy and occupational therapy) and follow-up by their general medical practitioner. In addition, a physiotherapist provided participants in the intervention group with a multifactorial, individually tailored falls prevention program. This consisted of the following: 1. Individualized home exercise program prescribed by a physiotherapist and based on the Otago Exercise Programme (OEP).26,27 The OEP has been previously shown to reduce falls in older community-dwelling adults.28,29 Exercises were selected from those included in the OEP to address balance and mobility problems identified in the baseline assessment (table 1). Adherence was assessed through exercise diaries completed by participants and discussion with the physiotherapist at each review and following completion of the study. Average adherence over 12 months was categorized as full (ⱖ3 times/wk), partial (1–2 times/wk), or nonadherent (⬍1 time/wk). 2. Falls risk minimization strategies based on general and stroke-specific risk factors identified in the baseline assessment, including those identified from the Falls Risk for Older People – Community setting (see table 1). Guidelines for this component of the intervention are also available in an earlier publication.15 3. Education (written and verbal) for participant and carer about identified falls risk factors and risk minimization. 4. Injury risk minimization strategies for those at high risk of fracture (based on delayed walking after stroke,30 previous diagnosis of osteoporosis). This included hip protector prescription and liaison with the participant’s general practitioner regarding vitamin D/calcium supplementation. Details of the exercise program and falls risk minimization strategies are contained in table 1. We also provided intervention participants with a falls prevention booklet, “A Guide to Preventing Falls.”a Control Participants in the control group received usual care with their treating health professionals, and research staff made no attempt to limit their access to any care. In addition, we provided control participants with the falls prevention booklet as above. Sample Size The literature suggests that between 46% and 73% of people with stroke fall during the first 6 months after discharge from hospital.4,5 Because we were recruiting people with high falls risk only, we estimated that 75% of the usual care group would experience a fall in a 1-year period. On the basis of the intervention reducing falls incidence by one third to 50%, using a power of 80% (P⫽.05, 2-sided), and allowing for 25% dropout, we set the recruitment target at 80 participants in each group. Data Analysis We analyzed the data on the basis of intention to treat. Participants who commenced the study but did not complete the full 12 months were included in the analysis of fall rate Arch Phys Med Rehabil Vol 93, September 2012
1650
FALLS PREVENTION AFTER STROKE, Batchelor Table 1: Intervention Detail Component
Content
1. Home exercise program based on the OEP
30 – 40-min home exercise sessions, 3–5 times weekly Exercises selected by the physiotherapist on the basis of findings of the initial assessment; weights (ankle cuff) prescribed according to ability; level of support for balance exercises (eg, one-hand support) prescribed according to assessment findings Exercises selected from the following: Warm-up exercises: Head/neck movements Back extension Trunk movements Ankle movements Strength and balance exercises: Knee extension/flexion Hip abduction Heel raise/toe raise Knee bends Backward walk Walk and turn Sideways walk Heel-toe stand Heel-toe walk One-leg stand Heel walking/toe walking Heel-toe backward walk Sit to stand Stair walking Walking recommended 3–5 times weekly. Time set according to assessment findings Exercise and walking program monitored and progressed at 4 and 8mo after initial session Combination of direct intervention and referral; with participant’s consent, written summary of falls risk factors and suggested actions provided to medical practitioner; written and verbal summary of strategies provided to participant Examples of possible falls risk factors and management options: Medications (⬎3, taking medications associated with falls): Referral to medical practitioner Visual impairment (eg, visual field deficit): Referral to training program for people with neurological visual loss; referral for vision assessment Hearing impairment: Referral to audiologist Inappropriate footwear: Provide advice about good footwear for falls prevention Assist participant to purchase appropriate footwear Foot problems: Referral to podiatrist Continence issues, including nocturia: Ensure participant has appropriate equipment, eg, commode, urinal Referral to continence clinic/specialist Home safety issues: Referral for occupational therapy home assessment Provide advice on lighting and home safety
2. Falls risk minimization strategies based on falls risk factors identified at baseline from the FROP-Com
Abbreviation: FROP-Com, Falls Risk for Older People – Community setting.
and rate of injurious falls if they had at least 1 month of falls data available. To compare fall rates between groups, we used negative binomial regression.31,32 Total falls were compared taking into consideration the time of exposure. Chi-square analysis was used to compare the proportion of fallers and nonfallers between groups. For this analysis, participants who did not complete the full study and had not reported any falls (in the months for which falls data were available) were excluded, because they could not be correctly classified. Arch Phys Med Rehabil Vol 93, September 2012
Linear regression was used to determine the difference between groups (intervention effect) on the secondary outcomes. Baseline status on each measure and group was included in the model as independent variables. To analyze differences within groups from baseline to follow-up, paired sample t tests were used because the difference in scores between baseline and follow-up followed a normal distribution. To evaluate the difference in the proportion of participants reporting falls prevention and injury minimization strategies at follow-up, we used chi-square analysis.
FALLS PREVENTION AFTER STROKE, Batchelor
The alpha level for all analyses was set at .05. Statistical analyses were undertaken using SPSS 16.0 GradPack.b RESULTS From the 156 participants recruited, 85 were randomized to the control group and 71 to the intervention group (fig 1). On completion of the study, 144 (92%) had at least 1 month of fall data available for analysis. Twelve additional participants (5 control, 7 intervention) did not complete the final assessment. Overall, there was an 85% retention rate. The 2 groups were similar on all assessed characteristics at baseline (table 2). Overall, participants had high falls risk, poor balance and mobility, and low activity levels. Falls Outcomes In the 2 weeks between discharge home and baseline assessment, 28% (44 of 156) of the participants fell. In total, 119 falls were reported in the intervention group (n⫽64) and 140 in the
1651
control group (n⫽80). One intervention participant reported 20 falls, and 4 participants (2 control, 2 intervention) reported 10 or more falls. The rate of falls (falls/person-year) for participants with at least 1 month of falls data (n⫽144) was 1.89 in the intervention group and 1.76 in the control group (table 3). There was no significant difference in fall rate between the groups, with an incidence rate ratio (IRR) for total falls (intervention:control) of 1.10 (95% confidence interval [CI]⫽.63– 1.90). To determine whether the participant with 20 falls skewed the results, the participant’s data were excluded and the analysis was rerun. However, although the fall rate in the intervention group decreased, there was no significant difference between the groups (IRR⫽.92, 95% CI⫽.54 –1.57) (see table 3). Forty-eight percent (29 of 60) of the intervention participants and 58% (46 of 79) of the control participants reported ⱖ1 falls. There was no significant difference in the proportion of fallers between the groups (risk ratio⫽.83, 95% CI⫽.60 –1.14, P⫽.256).
Fig 1. Flow of participants through the study.
Arch Phys Med Rehabil Vol 93, September 2012
1652
FALLS PREVENTION AFTER STROKE, Batchelor
Table 2: Baseline Characteristics of Participants (Nⴝ156) Characteristic
Age (y), mean ⫾ SD Female Stroke type Infarct Hemorrhage Unknown Side of symptoms Right Left Bilateral Other/unknown Time from stroke to baseline (mo), mean ⫾ SD Living arrangements Alone Spouse/carer Family AMTS (/10), median (IQR) FIM (/126), mean ⫾ SD Faller between discharge and baseline Time between discharge and baseline (d), median (IQR) FROP-Com score (/60), mean ⫾ SD FES-S (/10), median (IQR) HAP-AAS (/94), mean ⫾ SD Presence of inattention/neglect Presence of visual field deficit STS, number completed in 30s, median (IQR) Gait speed (m/min), mean ⫾ SD Step Test, no., median (IQR)
Intervention (n⫽71)
Control (n⫽85)
70.8⫾11.4 26 (36.6)
72.2⫾9.9 31 (36.5)
55 (77.5) 13 (18.3) 3 (4.2)
69 (81.2) 15 (17.6) 1 (1.2)
24 (33.8) 43 (60.6) 2 (2.8) 2 (2.8) 3.0⫾1.6
38 (44.7) 44 (51.8) 2 (2.4) 1 (1.2) 3.1⫾1.9
19 (26.8) 34 (47.9) 18 (25.4) 9.0 (2.0) 107.7⫾14.6
18 (21.2) 53 (62.4) 14 (16.5) 9.0 (2.0) 106.9⫾18.3
22 (31.0)
22 (25.9)
12.0 (8.0)
13.0 (11.0)
22.4⫾5.7 7.7 (3.2) 26.1⫾13.4
21.2⫾5.9 7.5 (2.5) 26.0⫾13.3
22 (32.8)
17 (21.5)
26 (36.6)
25 (29.4)
5.3 (7.9)
4.7 (7.4)
32.1⫾16.4 4 (6)
31.9⫾20.0 2 (6)
NOTE. Values are n (%) or as otherwise indicated. Abbreviations: AMTS, Abbreviated Mental Test Score; FES-S, Falls Efficacy Scale – Swedish Modification; FROP-Com, Falls Risk for Older People – Community setting; HAP-AAS, Human Activity Profile –Adjusted Activity Score; IQR, interquartile range; STS, Sit-to-Stand test.
There was no significant difference in the proportion of frequent fallers (ⱖ2 falls) between the groups (risk ratio⫽1.20, 95% CI⫽.74 –1.94, P⫽.454). The number of injurious falls was 46 in the intervention group (n⫽59) and 38 in the control group (n⫽78) (missing data in both groups). The rate of injurious falls was .74 and .49 per
person-year in intervention and control groups, respectively. The estimated IRR was 1.57 (95% CI⫽.73–3.4, P⫽.248), indicating no significant difference between groups. Secondary Outcomes Both groups improved significantly over 12 months on measures of falls risk (Falls Risk for Older People – Community setting), activity (Human Activity Profile19), leg strength (Sitto-Stand test), and gait speed and balance (Step Test) (table 4). Only participants from the intervention group improved significantly on fall-related efficacy (Falls Efficacy Scale – Swedish Modification). There were no significant differences in change scores between groups on any secondary outcome measure (see table 4). Adherence to the Exercise Program Of the 64 intervention participants for whom falls data were available, 16 (25.0%) fully adhered, 36 (56.3%) partially adhered, and 12 (18.7%) did not adhere to the exercise program during their time in the study (table 5). There was a significant difference in the proportion of fallers to nonfallers across adherence categories (2⫽8.288, P⫽.016), with a significantly higher proportion of fallers among those who partially adhered. The rate of falls was 66% lower in those who fully adhered than in those who partially adhered (IRR⫽.34, 95% CI⫽.15– .78, P⫽.01). Those classified as nonadherent also had a lower, but nonsignificant, falls rate than did those who partially adhered (IRR⫽.77, 95% CI⫽.32–1.82) (see table 5). The falls rate was 55% lower in those who fully adhered than in those classified as nonadherent, but again this was not significant (IRR⫽.45, 95% CI⫽.16 –1.28). At follow-up, a significantly higher proportion of intervention participants compared with control participants reported home exercise program participation (64.9% vs 44.6%, P⫽.026). For all other self-reported activities relating to falls prevention, there was no difference between the intervention and control groups. DISCUSSION To the best of our knowledge, this is the first study to report the effects of a multifactorial falls prevention program on falls in people with stroke returning home after rehabilitation. Although the proportion of fallers was less than expected in both groups (48% in the intervention group and 58% in the control group), this study confirmed that people with stroke are at high risk of falls, particularly in the first weeks after discharge from rehabilitation, with almost one third of participants falling between discharge and baseline assessment, a time period of, on average, 2 weeks. The results of the study did not support the hypothesis that a multifactorial falls prevention intervention reduces falls rate or the proportion of fallers in this population. The study did show that people with stroke at risk of recurrent falls significantly improved in the 12 months after discharge on falls risk, activity, strength, gait, and balance.
Table 3: Randomized Controlled Trial Outcome: Fall Rate Rate (falls/person-year) Participant Category
Intervention
Control
IRR InterventionControl (95% CI)*
P
All participants Participants with ⱖ20 falls excluded All participants, injurious falls
1.89 (n⫽64) 1.59 (n⫽63) .74 (n⫽59)
1.76 (n⫽80) 1.76 (n⫽80) .49 (n⫽78)
1.10 (.63–1.90) 0.92 (.54–1.57) 1.57 (.73–3.4)
.738 .752 .248
*From negative binomial regression model.
Arch Phys Med Rehabil Vol 93, September 2012
1653
FALLS PREVENTION AFTER STROKE, Batchelor Table 4: Secondary Outcome Scores Intervention
Characteristic
n
Baseline
Follow-up
FIM, median (IQR)
57
112.0⫾14.0
114.0⫾14.0
FROP-Com†
57
21.7⫾5.4
18.6⫾6.1
FES-S, median (IQR) HAP-AAS
55
7.8⫾3.2
8.1⫾2.1
57
27.2⫾13.0
36.6⫾17.0
57
6.0⫾8.0
7.6⫾5.5
57
32.4⫾15.8
38.1⫾17.0
57
4.0⫾6.5
7.0⫾6.0
STS, number completed in 30s, median (IQR) Gait speed (m/min) Step Test, no., median (IQR)
Control Difference Within Group (baselinefollow-up) (95% CI)
1.6 (⫺2.1 to 5.4) P⫽.379 ⫺3.1 (⫺4.9 to ⫺1.4) P⫽.001 0.5 (.04 to 0.9) P⫽.034 9.4 (4.9 to 13.9) P⬍.001 1.6 (0.5 to 2.6) P⫽.003 5.7 (1.8 to 9.6) P⫽.005 1.8 (1.0 to 2.6) P⬍.001
n
Baseline
Follow-up
75
114.0⫾17.0
115.0⫾20.0
74
20.8⫾6.0
18.4⫾6.7
70
7.9⫾2.4
7.8⫾3.1
75
26.7⫾13.2
34.4⫾20.6
75
5.4⫾7.9
7.8⫾7.3
75
32.3⫾20.4
37.1⫾22.1
75
3.0⫾6.0
6.0⫾9.0
Difference Within Group (baselinefollow-up) (95% CI)
Difference Between Groups (baselinefollow-up)* (95% CI)
0.4 (⫺2.2 to 2.9) 2.5 (⫺1.7 to 6.7) P⫽.756 P⫽.240 ⫺2.4 (⫺3.6 to ⫺1.2) ⫺0.6 (⫺2.5 to 1.3) P⬍.001 P⫽.554 ⫺0.3 (⫺0.7 to 0.2) 0.6 (⫺.03 to 1.1) P⫽.274 P⫽.062 7.7 (4.5 to 10.9) 1.8 (⫺3.6 to 7.1) P⬍.001 P⫽.514 2.4 (1.6 to 3.2) ⫺0.7 (⫺1.9 to 0.6) P⬍.001 P⫽.282 4.9 (2.2 to 7.6) P⫽.001 2.1 (1.2 to 2.9) P⬍.001
0.9 (⫺3.6 to 5.4) P⫽.691 ⫺0.1 (⫺1.3 to 1.1) P⫽.844
NOTE. Values are mean ⫾ SD unless otherwise indicated. Abbreviations: FES-S, Falls Efficacy Scale – Swedish Modification; FROP-Com, Falls Risk for Older People – Community setting; HAP-AAS, Human Activity Profile – Adjusted Activity Score; STS, Sit-to-Stand test. *Calculated from linear regression model with baseline score and group (control or intervention) as independent variables. † Lower score indicates improved performance.
Issues related to sample size, recruitment, heterogeneity of the sample, the intervention, and the potential for similarity in intervention and control conditions may have contributed to the lack of a significant effect. Sample size estimation was based on a reduction in the proportion of fallers from 75% (in the control group) by one third. Our results found that the falls rate in the control group was smaller than anticipated and closer to 60%. In addition, the effect of the intervention was smaller than anticipated and closer to a 10% reduction in fallers. As a consequence, the sample size was not large enough to detect significant differences between groups. To detect a significant effect of the intervention based on reducing the proportion of fallers by one third to 40%, a sample size of 107 per group would be required (power of 80%, ␣⫽.05), larger than the sample size of the current study. However, given that the comparison of fall rate was not close to significant, it is unlikely that increasing the sample size alone would be enough to show significant effects across all comparisons. One possible reason for the smaller than anticipated effect size is that the intervention provided was not sufficiently different to “usual care.” It is also possible that there was contamination between the groups. For example, those in the control group were more likely to have sought out additional therapy or falls prevention information compared with those in the intervention group. This was confirmed by the lack of difference (apart from home exercise program participation) Table 5: Proportion of Fallers and Fall Rates by Adherence to Exercise Program (Intervention Participants), nⴝ64 n (%) Adherence Category
Nonfaller
Faller
IRR* (95% CI)
Nonadherent Partial adherence† Full adherence
9 (75.0) 14 (38.9) 12 (75.0)
3 (25.0) 22 (61.1) 4 (25.0)
0.77 (.32–1.82) 1.00 .34 (.15–.78)
*From negative binomial regression model. † Chi-square⫽8.288, P⫽.016 for the difference in proportions.
between the groups in strategies reported at follow-up. In addition, the provision of information about the project to rehabilitation staff discharging participants may have led to an increase in the number of falls prevention strategies provided to participants in both groups. It is also possible that the OEP was not effective in this population. Although the OEP has been shown to be effective in reducing falls in community-dwelling older adults,28,29 these effects did not necessarily carryover to older people at high risk of falls.33 In addition, although the exercises in the present study were individually tailored, safety considerations (unsupervised exercise at home) may have led physiotherapists to prescribing balance exercises that were not challenging enough to be effective. Adherence to exercise may have been another factor in the lack of effect. Only one quarter of intervention participants fully adhered to the exercise program, while over one half partially adhered. The relationship between falls and adherence to exercise was not linear. The proportion of fallers was significantly lower in those who fully adhered and in those who were nonadherent than in those who partially adhered. This contrasts with the linear relationship between adherence and falls rate reported in a study of community-dwelling older people with poor vision where increased adherence was associated with decreased fall rates.28 A hypothesis for the nonlinearity is that those who did not experience falls saw no need to complete the exercise program and those who fully adhered gained observable benefit. Alternatively, those who did very little exercise may have had low activity levels that decreased exposure to activities that have a higher falls risk. Further exploration into factors that influence adherence and methods of facilitating adherence in this population is required. This study’s findings raise issues that warrant further investigation. There is still no high-level evidence that any intervention prevents falls in community-dwelling people with stroke. Interventions including vitamin D, evaluated in combination with exercise or as part of a multifactorial falls prevention program, are worthy of consideration in the community-dwelling stroke popuArch Phys Med Rehabil Vol 93, September 2012
1654
FALLS PREVENTION AFTER STROKE, Batchelor
lation because this treatment appears to reduce falls in institutionalized women with stroke.34 Evaluation of an intervention that commences prior to discharge and of a more intensive exercise intervention is also warranted because it is possible that the timing and the intensity of the present study’s intervention was a factor in the null finding. Study Limitations Beyond the potential reasons for the lack of significant effect as discussed in the previous section, the study had some limitations. First, because participants were unable to be blinded to the purpose of the study, it is possible that participants were sensitized to prevent falls, leading to all participants being more cautious or less active. This may have contributed to bias toward the null hypothesis. Conversely, there may have been a bias toward the alternative hypothesis because participants were not blind to group allocation, leading to the potential for intervention participants to underreport falls. Second, the heterogeneity of the sample may have been a factor contributing to the observed results. Our inclusion criteria meant that people with all levels of functional ability (as long as they met the high falls risk criteria) could participate in the trial, and this diversity may have masked positive findings. CONCLUSIONS This study did not support the hypothesis that this multifactorial falls prevention program is effective in people with stroke who are at risk of recurrent falls. Despite the lack of overall effect, this study highlights the high risk of falls in this group and provides direction for future studies aiming to prevent falls in people with stroke returning home. Acknowledgments: We thank Caroline Fryer, BAppSc(Physio), GradDipClinEpi for her invaluable assistance with data collection throughout the study. We also thank the physiotherapists who assisted with recruitment. References 1. Mackintosh SF, Hill K, Dodd KJ, Goldie P, Culham E. Falls and injury prevention should be part of every stroke rehabilitation plan. Clin Rehabil 2005;19:441-51. 2. Kerse N, Parag V, Feigin VL, et al, the Auckland Regional Community Stroke Study Group. Falls after stroke: results from the Auckland Regional Community Stroke (ARCOS) study, 2002 to 2003. Stroke 2008;39:1890-3. 3. Mackintosh SFH, Goldie P, Hill K. Falls incidence and factors associated with falling in older, community-dwelling, chronic stroke survivors (⬎1 year after stroke) and matched controls. Aging Clin Exp Res 2005;17:74-81. 4. Mackintosh SF, Hill KD, Dodd KJ, Goldie PA, Culham EG. Balance score and a history of falls in hospital predict recurrent falls in the 6 months following stroke rehabilitation. Arch Phys Med Rehabil 2006;87:1583-9. 5. Forster A, Young J. Incidence and consequences of falls due to stroke: a systematic inquiry. BMJ 1995;311:83-6. 6. Campbell AJ, Borrie MJ, Spears GF, Jackson SL, Brown JS, Fitzgerald JL. Circumstances and consequences of falls experienced by a community population 70 years and over during a prospective study. Age Ageing 1990;19:136-41. 7. Belgen B, Beninato M, Sullivan PE, Narielwalla K. The association of balance capacity and falls self-efficacy with history of falling in community-dwelling people with chronic stroke. Arch Phys Med Rehabil 2006;87:554-61. 8. Yates JS, Lai SM, Duncan PW, Studenski S. Falls in communitydwelling stroke survivors: an accumulated impairments model. J Rehabil Res Dev 2002;39:385-94. Arch Phys Med Rehabil Vol 93, September 2012
9. Harris JE, Eng JJ, Marigold DS, Tokuno CD, Louis CL. Relationship of balance and mobility to fall incidence in people with chronic stroke. Phys Ther 2005;85:150-8. 10. Lamb SE, Ferrucci L, Volapto S, Fried LP, Guralnik JM. Risk factors for falling in home-dwelling older women with stroke: The Women’s Health and Aging Study. Stroke 2003;34:494-500. 11. Batchelor F, Hill K, Mackintosh S, Said C. What works in falls prevention after stroke? Systematic review and meta-analysis. Stroke 2010;41:1715-22. 12. Gillespie L, Robertson M, Gillespie W, et al. Interventions for preventing falls in older people living in the community. Cochrane Database Syst Rev 2009;2:CD007146. 13. Hill K, Bernhardt J, McGann A, Maltese D, Berkovits D. A new test of dynamic standing balance for stroke patients: reliability, validity, and comparison with healthy elderly. Physiother Can 1996;48:257-62. 14. Berg KO, Wood-Dauphinee SL, Williams JI. Measuring balance in the elderly— validation of an instrument. Can J Public Health 1992;83:S7-11. 15. Batchelor FA, Hill KD, Mackintosh SF, Said CM, Whitehead C. The FLASSH study: protocol for a randomised controlled trial evaluating falls prevention after stroke and two sub-studies. BMC Neurol 2009;9:14. 16. Lamb S, Jørstad-Stein E, Hauer K, Becker C, on behalf of the Prevention of Falls Network Europe and Outcomes Consensus Group. Development of a common outcome data set for fall injury prevention trials: the Prevention of Falls Network Europe consensus. J Am Geriatr Soc 2005;53:1618-22. 17. McCarthy EK, Horvat MA, Holtsberg PA, Wisenbaker JM. Repeated chair stands as a measure of lower limb strength in sexagenarian women. J Gerontol A Biol Sci Med Sci 2004;59: 1207-12. 18. Steffen TM, Hacker TA, Mollinger L. Age- and gender-related test performance in community-dwelling elderly people: SixMinute Walk test, Berg Balance Scale, Timed Up & Go test, and gait speeds. Phys Ther 2002;82:128-37. 19. Fix A, Daughton D. Human Activity Profile (professional manual). Odessa: Psychological Assessment Resources, Inc; 1988. 20. Hamilton B, Granger C. Disability outcomes following inpatient rehabilitation for stroke. Phys Ther 1994;74:494-503. 21. Hellström K, Lindmark B. Fear of falling in patients with stroke: a reliability study. Clin Rehabil 1999;13:509-17. 22. Russell MA, Hill KD, Blackberry I, Day LM, Dharmage SC. The reliability and predictive accuracy of the falls risk for older people in the community assessment (FROP-Com) tool. Age Ageing 2008;37:634-9. 23. Bailey MJ, Riddoch MJ, Crome P. Test-retest stability of three tests for unilateral visual neglect in patients with stroke: Star Cancellation, Line Bisection, and the Baking Tray Task. Neuropsychol Rehabil 2004;14:403-19. 24. Elliott DB, North I, Flanagan J. Confrontation visual field tests. Ophthalmic Physiol Opt 1997;17:S17-24. 25. Hodkinson HM. Evaluation of a mental test score for assessment of mental impairment in the elderly. Age Ageing 1972;1:233-8. 26. ACC NZ Web site. Otago Exercise Programme. Available from: http://www.acc.co.nz/preventing-injuries/at-home/older-people/ information-for-older-people/otago-exercise-programme/index. htm. Accessed March 27, 2011. 27. Gardner MM, Buchner DM, Robertson MC, Campbell AJ. Practical implementation of an exercise-based falls prevention programme. Age Ageing 2001;30:77-83. 28. Campbell AJ, Robertson MC, Grow SJA, et al. Randomised controlled trial of prevention of falls in people aged ⱖ75 with severe visual impairment: the VIP trial. BMJ 2005;331:817.
FALLS PREVENTION AFTER STROKE, Batchelor
29. Robertson MC, Campbell AJ, Gardner MM, Devlin N. Preventing injuries in older people by preventing falls: a meta-analysis of individual-level data. J Am Geriatr Soc 2002;50:905-11. 30. Jørgensen L, Jacobsen B, Wilsgaard T, Magnus J. Walking after stroke: does it matter? Changes in bone mineral density within the first 12 months after stroke. A longitudinal study. Osteoporos Int 2000;11:381-7. 31. Robertson MC, Campbell AJ, Herbison P. Statistical analysis of efficacy in falls prevention trials. J Gerontol A Biol Sci Med Sci 2005;60:530-4. 32. Donaldson MG, Sobolev B, Cook WL, Janssen PA, Khan KM. Analysis of recurrent events: a systematic review of randomised controlled trials of interventions to prevent falls. Age Ageing 2009;38:151-5.
1655
33. Elley CR, Robertson MC, Garrett S, et al. Effectiveness of a falls-and-fracture nurse coordinator to reduce falls: a randomized, controlled trial of at-risk older adults. J Am Geriatr Soc 2008;56: 1383-9. 34. Sato Y, Iwamoto J, Kanoko T, Satoh K. Low-dose vitamin D prevents muscular atrophy and reduces falls and hip fractures in women after stroke: a randomized controlled trial. Cerebrovasc Dis 2005;20:187-92. Suppliers a. Peninsula Health, Mt Eliza Centre, Jacksons Rd, Mt Eliza, Victoria 3930, Australia. b. SPSS, Inc, IBM Corporation, 1 New Orchard Rd, Armonk, NY 10504.
Arch Phys Med Rehabil Vol 93, September 2012