Predictors of Functional and Gait Outcomes for Persons Poststroke Undergoing Home-based Rehabilitation Faisal Y. Asiri, PhD, PT,*† Gregory F. Marchetti, PhD, PT,‡ Jennifer L. Ellis, PT, DPT, MS, GCS,x Laurie Otis, PT, MBA, MHA,x Patrick J. Sparto, PhD, PT,* Valerie Watzlaf, PhD, RHIA, FAHIMA,jj and Susan L. Whitney, PhD, PT, NCS, ATC, FAPTA*{

Background: The literature on the impact of home-based rehabilitation on functional outcomes for patients after stroke is limited. The purpose of this study was to describe the outcomes of home-based rehabilitation (HBR) on functional and gait performance for patients after stroke and associated factors that contribute to better outcomes after an episode of care. Methods: The nature of the study design was retrospective and the settings used were home care services. The total number of subjects receiving home care services after stroke was 213 (mean age 76.5 6 9 years, 51% female). Treatment records for patients receiving HBR in 2010 were reviewed at the start of care and discharge. The primary outcome measure was a change in a gait speed and activities of daily living (ADL) performance between admission and discharge from home health care services. The composite score to calculate overall functional status (Outcome Information and Assessment Set— version C [OASIS-C]) was used. Mean change in ADL and gait scores and factors predictive of improvement were identified using an analysis of covariance and multivariate linear models. The main outcome measures were change in the OASIS-C composite scores and gait speed. Results: After adjustment for age and ADL score at the start of care, discharge from skilled nursing or long-term facilities, presence of confusion most of the times, cognitive impairment, and memory deficits were negatively associated with an improvement in functional scores (ADL). Living in congregate facilities was also negatively associated with an improvement in gait speed. The best multivariate model included age, baseline ADL composite scores, confusion status, and gait speed at the start of care, which predicted 41% of the variance in ADL score changes over the course of intervention. Conclusions: Gait speed and ADL scores at the start of care had largest influence on functional and gait improvement. Type of discharge facility, confusion status, and living arrangement had effects on HBR outcomes for stroke survivors. Key Words: Stroke rehabilitation—home care services—OASIS-C—gait speed. Ó 2014 by National Stroke Association

From the *Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh, Pittsburgh, PA; †Department of Physical Therapy, College of Applied Medical Sciences, King Khalid University, Abha, Saudi Arabia; ‡Department of Physical Therapy, Rangos School of Health Sciences, Duquesne University, Pittsburgh, PA; xGentiva Health Services Inc, Atlanta, GA; jjDepartment of Health Information Management, School of Health and Rehabilitation Sciences, University of Pittsburgh, Pittsburgh, PA; and {Department of Rehabilitation Sciences, College of Applied Medical Sciences, King Saud University, Riyadh, Saudi Arabia.

Received November 9, 2013; revision received February 18, 2014; accepted February 28, 2014. This study was presented at the Combined Sections Meeting of the American Physical Therapy Association, San Diego, CA 2013. Address correspondence to Faisal Y. Asiri, PhD, PT, Department of Physical Therapy, University of Pittsburgh, 6055 Forbes Tower, Pittsburgh, PA 15260. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.02.025

Journal of Stroke and Cerebrovascular Diseases, Vol. -, No. - (---), 2014: pp 1-9

1

F.Y. ASIRI ET AL.

2

Introduction According to the Centers for Disease Control and Prevention, stroke is the most common condition that leads to long-term disability in the US.1,2 Six months after the onset of stroke, 50% of individuals with stroke continued to demonstrate hemiparesis, 30% were unable to ambulate independently, and 26% had difficulties with activities of daily living (ADL).1,3 The total stroke care costs, both direct and indirect, were estimated to be around $68.9 billion in the US in 2009.1,4 Several factors are associated with stroke recovery. These factors include age,5-10 severity of stroke,4 stroke type,5,6,11, which were stated to be predictive factors for walking recovery within the 30 days of the onset of stroke.5 Cognitive impairment6,12,13 and incontinence9,14 have been negatively associated with functional recovery at discharge from inpatient rehabilitation settings.6 Severity of stroke symptoms, such as moderate to severe hemiparesis, is a predictor for rehospitalization, disability, and increasing mortality rates 5 years after stroke.4,15 The type of discharge facility was shown to have an impact on functional gain for stroke survivors after acute care admissions.16 At 6 months of follow-up, patients with stroke had more functional improvement when they received rehabilitation services from an inpatient rehabilitation facility (IRF) compared with a skilled nursing facility (SNF).16 Home health care (HHC) is also frequently prescribed after stroke. Of the nearly 1.5 million patients who received HHC services delivered by home health agencies in 2007, approximately 3.3% had stroke as a primary diagnosis at admission and 7.1% had stroke listed as any diagnosis.17 HHC services have demonstrated significant reductions in cost when compared with long-term hospital care for chronically ill elderly persons.18 The total cost of inpatient hospital care was 3 times higher than those treated at home.18 There was a small to moderate positive impact of HHC on the number of rehospitalization days, which might be helpful in decreasing the cost of health expenditures with a statistically significant relationship between decreased inpatient hospital care days and HHC.19,20 Decreasing hospital stays is one of the HHC benefits in reducing health care expenditures because inpatient costs represent 70% of the first-year costs of stroke management in the US.21 Home-based rehabilitation (HBR), which involves physical therapists (PTs), occupational therapists, and speech and language pathologists, has an impact on functional recovery and is cost effective.22 In a recent systematic review for stroke rehabilitation at home, the importance of HBR after stroke was emphasized.23 The benefits of rehabilitation at home included cost reductions, enhanced patient satisfaction and functional outcomes,23,24 improved physical health,22,23 and increased independence of ADL.23,25,26 In a clinical trial, Mayo et al22 assigned stroke

survivors into either a home care or usual care group. Home care group consisted of services by an interdisciplinary team (nursing, PT, occupational therapist, and speech and language pathologist), whereas the usual care group consisted of several services (eg, PT outpatient clinics or rehabilitation services in hospital settings) requested by a physician. The home care group had significant improvements compared with the usual care group in the physical health component of the Measuring Outcomes Study Short Form 36, the Older Americans Resource Scale for instrumental ADL, and in the Reintegration to Normal Living scales.22 There were similar improvements between the 2 groups in terms of the Barthel Index and the Timed Up and Go at 1 and 3 months after stroke.22 In addition, the home care intervention program had an impact on delaying mortality rates and functional decline 2 years after intervention.27,28 In the LEAPS (Locomotor Experience Applied Post-stroke) trial, a home-based physical therapy program including balance and strengthening exercise was shown to be equally effective at demonstrating changes in function when compared with a locomotor training program on both treadmill and over ground training.29 The literature on the impact of HBR on physical function outcomes and walking performance for patients after stroke is limited. Consequently, the purpose of this study was as follows: (1) describe the outcomes of HBR on gait speed and functional performance (ADLs) for patients after stroke; and (2) determine the predictors of change in functional performance and gait speed after an episode of care. It was hypothesized that advanced age, living alone, number of comorbidities, cognitive impairment, discharge from IRF settings, and urinary incontinence would be negatively associated with improvements in gait and functional performance after HBR. Better functional performance (ADL) and gait speed at the start of care would be also positively associated with functional improvement at the end of care. Knowledge of which factors affect the outcome in persons who have had a stroke and are undergoing treatment in the home could affect clinical resource allocation, goal setting, the length of the episode of care, and future payment models.

Methods Data Source and Study Design The clinical data were retrospectively collected from 8436 subjects who participated in the Safe Strides program from Gentiva Home Health Services (Atlanta, GA), at various sites in the US in 2010. The Safe Strides program was designed for older adults who are at risk for falling with a focus on fall risk reduction. It includes a falls prevention approach and intervention customized by trained clinicians to improve balance and functional capacity, and to increase the level of functional

PREDICTIVE FACTORS FOR STROKE SURVIVORS

3

Figure 1. Selection process of stroke conditions from the general data set (n 5 8436). Abbreviations: CVA, cerebrovascular accident; HHC, home health care; ICD, International Classification of Disease; OASIS-C, Outcome Information and Assessment Set—version C.

independence of the patient. The inclusion criteria for the Safe Strides program, determined by a trained PT, were a history of falls in the most recent 12 months and/or one or more modifiable fall risk factors. Modifiable fall risk factors addressed through this program have been reported in previous study.28 The data set includes patients’ information from the Outcome Information and Assessment Set—version C (OASIS-C) and gait speed performance at the start of care and discharge. The records were deidentified by an honest broker and were sent to the University of Pittsburgh for collaborative analysis. The University of Pittsburgh’s Institutional Review Board approved the study.

Subjects The full data set of 8436 subjects included a heterogeneous population with at least 17 diagnostic categories, and more than 800 diagnoses based on the International Classification of Disease (ninth version; ICD-9) codes. Therefore, several steps were performed to select the subjects who were included in the study (Fig 1). First, subjects were included in the analysis if they had an ICD-9 code for cerebrovascular accident (430–438) recorded among the diagnoses listed during the inpatient stay within the last 14 days (ie, the OASIS-C M1010 items). This reduced the sample to 447 subjects. Next, the sample was reduced further by selecting those persons who had ICD-9 codes that referred to hemiplegia or hemiparesis (438.20, 438.21, and 438.22) in the primary or other diagnoses at admission to HBR (ie, the OASIS-C M1020 and M1022 items). These steps were used to confirm with reasonable confidence that subjects were treated by the HHC agency because of stroke diagnosis. The total number of subjects included was 213 (mean age 76.5 6 9 years, 51% female). The baseline characteristics of subjects with stroke at the start of care are pre-

sented in Table 1. Fifty-eight percent (n 5 123) of patients were admitted to HHC services as household ambulators (those who walk slower than .4 m/second).30 At admission to HHC, 43% of patients (n 5 92) were discharged from nursing or long-term care (LTC) facilities, 20% (n 5 42) from inpatient rehabilitation facilities, and 37% (n 5 79) from short-stay acute hospitals.

Outcome Measures The main outcome measures used in the study were changes in the performance of ADL and gait speed. To assess changes in the performance of ADLs, the OASISC was used. The OASIS is a comprehensive assessment tool for assessing patient characteristics and measuring patient outcomes in HHC between admission and discharge from the episode of care.31,32 The OASIS-C has 6 domains that can be assessed at admission and discharge from HHC services including sociodemographic, environment, support system, health status, functional status, and behavioral status.32,33 Performance in ADLs was measured by examining the 9 ADL items in the OASIS functional status domain. The lowest score in each ADL item represents independent performance of the activity, whereas the highest score is recorded if an individual is dependent or has difficulties in performing the activity. Because the OASIS-C functional items were not developed for scale scoring,34 the ADL composite score was used to measure overall functional status. The ADL composite score first converts all items into the same scale from 0 to 1 by dividing the unscaled score by the maximum item score. Then all the scaled item scores are summed.35 The ADL composite change score ranges from 29 to 9. The score of 9 indicates optimal performance or functional improvement, 0 indicates no change between admission and discharge, and

F.Y. ASIRI ET AL.

4

Table 1. Baseline characteristics of individuals with stroke (total n 5 213)

Characteristics Gender Age group Under age 65 Age 65-74 Age 75-84 Age . 85 Discharge facility Nursing/LTC facilities Inpatient rehabilitation facility Short-stay acute hospital Anxiety None Less often than daily Daily/constant Confusion status Never In new/complex situations Sometimes or most of the times Cognitive function Alert/oriented Requires prompt Requires assistance Living situation Lives alone Live with other person in home Live in congregate situation Urinary incontinence No Yes Memory deficits No Yes Impaired decision No Yes Number of comorbidities 1 2 3 4 5 Number of episodes Once (60 days) 2 or 3 episodes Speed group Household* Communityy

Subjects (n) 108 Female

Proportion (%) 51%

14 64 96 39

6.6 30.0 45.1 18.3

91 42

42.7 19.7

78

36.6

113 71 29

53.1 33.3 13.6

97 81 35

45.5 38.0 16.4

102 71 40

47.9 33.3 18.8

45 144

21.1 67.6

24

11.3

119 94

55.9 44.1

167 46

78.4 21.6

168 45

78.9 21.1

3 9 26 75 100

1.4 4.2 12.2 35.2 46.9

133 34

80 20

123 90

57.7 42.3

Abbreviation: LTC, long-term care. *Persons who walk less than .4 m/second at the start of care. yPersons who walk at least .4 m/second at the start of care.

29 indicates that the individual’s ADL score got much worse after the episode of care. Gait speed has been used to measure walking performance and overall health.36 Individuals with stroke were asked to take 1-2 strides before and after the timing zone to control acceleration and deceleration. Gait speed was tested over 2.5-6 m based on the home environment. The accuracy of outcome measures data was reviewed by the specialty director and submitted electronically to Gentiva’s Center for Outcome Measures. Gait speed improvements of .16 m/second after stroke have been related to a clinically meaningful change, and patients are more likely to gain improvement in disability level.37 A change in gait speed of .10 m/second in older adults is also considered clinically meaningful in survival rates.38

Independent Variables In the OASIS-C, the independent variables of interest included age, type of inpatient facility from which they were discharged before HHC onset (nursing or LTC facilities, inpatient rehabilitation facilities, and short-stay acute hospitals), current status of anxiety (none, less often than daily, and daily anxious), confusion (never confused, confused in new/complex situations, and confused most of the times), cognitive functioning (oriented, requires prompt, and requires assistance during cognitive tasks), urinary incontinence (presence or absence of urinary incontinence), number of comorbidities, living arrangement (living alone, living with someone, and living in congregate situation), and ADL baseline scores (eg, grooming, dressing upper body, dressing lower body, bathing, toilet transferring, toileting hygiene, transferring, ambulation/ locomotion, and feeding or eating).32 In addition, baseline gait speed was an independent variable. The number of comorbidities was counted based on how many conditions were present at the start of care and included hypertension, diabetes, chronic obstructive pulmonary disease, osteoarthritis, or heart failure.

Statistical Analysis Dependent t-tests were used to examine the difference between admission and discharge scores on both functional (ADL composite) and gait speed scores. Covariates (age, ADL score, and gait speed at the start of care) hypothesized to be predictive of ADL composite change score, and gait speed change were tested with multiple linear regression. Factors hypothesized to be predictive of change in the ADL composite score and gait speed with adjustment for significant covariates were tested using a univariate generalized linear model (GLM), which allows inclusion of categorical predictors and continuous covariates. Univariate predictors with adjustment for covariates were tested independently. Between group post hoc comparisons in mean ADL and gait speed change, based on allocating subjects to the different categories of

PREDICTIVE FACTORS FOR STROKE SURVIVORS

an independent variable, were made using a Bonferroni adjusted type I error rate. All factors were retained for inclusion in a multivariate GLM if found to be significantly associated with univariate analysis at P less than .10. A multivariate GLM with the stepwise procedure was then used to identify the best subset of independent variables for both the ADL composite score and gait speed change between the start of care and discharge from HHC. An independent variable was entered into the model in order based on strength of association, and retained for the final multivariate GLM if significant at the value of P less than .05 level. The relative contribution of each factor to the change in ADL composite score and gait speed was described using the coefficient of determination (R2) from the multivariate GLM. All data analyses were conducted using SPSS version 20 (IBM, Armonk, NY).

Results Descriptive Statistics Two hundred thirteen individuals with stroke received HHC in 2010 and were included in the analysis. The percentage of patients who improved in the ADL composite score at discharge was 95% (n 5 203) compared with 5% (n 5 10) who got worse or who had no change in the ADL score. The mean ADL composite score improved from 3.15 6 1.30 points at the start of HHC to 1.12 6 1.20 points at the end of care (t 5 22.78, P , .001). The mean gait speed improved from .41 6 .26 m/second at admission to .56 6 .29 m/second at discharge (gait speed change .15 6 .23 m/second, t 5 29.71, P , 0.001). The baseline ADL score was a significant covariate predictor of ADL score change from the start of care to discharge (R2 5 .33, P , .001). The covariate most significantly associated with a change in gait speed was baseline gait speed score (R2 5 .076, P , .001). Age was included as covariate for change in ADL and gait speed although not significant (P 5 .081, P 5 .109), respectively, because age was stressed in the literature to be a predictive factor for walking and functional recovery after stroke.

5

Figure 2. Effect of confusion status on adjusted change in ADL scores (mean 6 standard error). Never: patients did not experience confusion within the last 14 days. New/complex situations: patients were confused in new or complex situations only. Sometimes or most of the time: patients were confused on awaking or at night only, during the day and evening, or constantly. Greater change in ADL refers to the improvement in ADL composite score at the end of care. Covariates were adjusted at the following values: age 5 76.58 and baseline ADL 5 3.1459. *P less than .05. Abbreviation: ADL, activities of daily living.

improvement in gait speed (F 5 3.730, P 5 .026) after adjustment for age and baseline gait speed. Figures 2 through 4 show the adjusted mean differences among the levels of independent variables on both ADL and gait speed changes scores. Subjects who were confused sometimes or most of the time made significantly less change in ADL compared with subjects who were confused in new or complex situation (Fig 2). Subjects discharged from short-stay acute hospital showed a significantly greater change in ADL compared with subjects discharged from SNF/LTC facilities (Fig 3). Subjects living in congregate situation had a significantly less change in gait speed (Fig 4) compared with subjects living alone or

Univariate Predictors of Change in ADL and Gait Speed Scores The associations between the independent variables and the change in gait speed and ADL composite scores were analyzed, after first adjusting for the covariates. After adjustment for age and ADL scores at the start of care, the following factors were significantly associated (at P , .10) with ADL change scores: discharge facility (F 5 4.500, P 5 .012), presence of confusion (F 5 5.051, P 5 .007), cognitive function (F 5 4.232, P 5 .016), and memory function (F 5 4.874, P 5 .028). Only living arrangement was significantly associated with an

Figure 3. Effect of discharge facility on adjusted change in ADL scores (mean 6 standard error). Greater change in ADL refers to the improvement in ADL composite score at the end of care. Covariates were adjusted at the following values: age 5 76.53 and baseline ADL 5 3.1675. *P less than 05. Abbreviations: ADL, activities of daily living; LTC, long-term care.

F.Y. ASIRI ET AL.

6

P , .001), confusion status (F 5 5.193, P 5 .007), and gait speed at the start of care (F 5 14.021, P , .001). This multivariate model predicted 41% of the variance in ADL score changes over the course of intervention. The largest influence on the change in the ADL was ADL at the start of care (R2 5 .382), followed by walking speed at the start of care (R2 5 .063), confusion in new/complex situations (R2 5 .047), age (R2 5 .024), and the absence of confusion (R2 5 .024). Age was negatively associated with a change in ADL (R2 5 .024), whereas greater deficits in ADL performance and faster gait speed at the start of care were associated with improvements in ADL composite change scores (Table 2). Predictive factors such as discharge facility, cognitive function, and memory deficits were not associated with the change in ADL when included in the final model because they were intercorrelated with other independent variables in the model. Regarding the change in gait speed, the final multivariate model included living arrangement (alone, with someone, in congregate situation), age, walking speed at the start of care, and ADL scores at the start of care. This model predicted 14% of the variance in gait speed score changes over the course of intervention. The largest influence on gait speed change was walking speed at the start of care (R2 5 .097), which explained 10% of the variance in gait speed change. Slower ambulators at the start of care had larger changes in gait speed at discharge from HHC. Living with someone (R2 5 .027), better ADL scores at the start of care (R2 5 .022), and living alone were associated with improvements in gait speed (R2 5 .019).

Figure 4. Effects of living arrangement on adjusted change in gait speed scores (mean 6 standard error). Congregate situation (eg, assisted living facilities). Covariates were adjusted at the following values: age 5 76.58 and baseline gait speed 5 .411 m/second. *P less than .05.

with other person. There was no significant difference between living alone or with other person, and both did almost the same change in gait speed at the end of care. The number of comorbidities (F 5 .893, P 5 .469; F 5 .924, P 5 .451) and incontinence problems (F 5 1.445, P 5 .231; F 5 .314, P 5 .576) were found not to be significant predictors of ADL and gait speed change, respectively. Living arrangement (F 5 .584, P 5 .559) was also found not to be associated with ADL change at the end of care.

Discussion

Multivariate Model Predictive of Change in ADL and Gait Speed

HBR delivered by home health clinicians had significant positive effects on both ADL function and gait performance. Our results demonstrated that the average ADL composite score improved by two points with 95%

The best multivariate model that predicted the change in ADL composite score included age (F 5 5.099, P 5 .025), baseline ADL composite scores (F 5 127.907,

Table 2. Multivariate linear model (stepwise stepping procedure) for the ADL composite score and gait improvement ADL change

Gait speed change

Factors

B coefficients

P value

Partial R2

B coefficients

P value

Partial R2

Age Gait speed (baseline) ADL composite (baseline) Confusion status* Absence of confusion Confused in new/complex situations Living arrangementy Living alone Living with other person

2.018 .095 .631

.025 ,.001 ,.001

.024 .063 .382

2.002 2.084 2.025

.236 ,.001 .033

.007 .097 .022

.466 .661

.025 .002

.024 .047

— —

ns ns

— —

— —

ns ns

.115 .117

.044 .018

— —

.019 .027

Abbreviation: ns, nonsignificant. P , .05. *The regression coefficients for the confusion status levels were referenced to the highest group (confusion sometimes or most of the times). yThe regression coefficients for the living arrangement levels were referenced to the highest group (living in congregate situation).

PREDICTIVE FACTORS FOR STROKE SURVIVORS

of individuals with stroke improving their ADL composite score, and there was a significant improvement in gait speed with an average change of .15 m/second at discharge. ADL composite score and gait speed at the start of care had the largest influence on functional and gait improvement. Although the predictive factors, such as discharge facility, cognitive function, and memory deficits, were independently associated with the change in ADL, they were not associated when including multiple variables (full model). Living arrangement was a predictive factor for change in gait speed, in that living alone or with other person was positively associated with the change in gait speed. Other hypothesized predictors, such as gender, number of comorbidities, urinary incontinence, did not show a significant association with the change in ADL and gait speed after the end of care. The increase in average gait speed of .15 m/second exceeded the minimally clinically important difference (MCID).39 Perera et al39 determined the MCID (.10 m/second) from a heterogeneous sample of older adults with mobility disability, subjects with stroke, and communitydwelling older adults. A change of .10 m/second is associated with reduced disability and better survival rates among older adults.38,40 The MCID for stroke survivors within 2 months of the onset was .16 m/second,37,40 similar to the mean change of .15 m/second in our study. Our study showed that patients who were ambulating at least .4 m/second (limited community ambulators) demonstrated significantly greater ADL change than those walking at slower speeds of less than .4 m/second (household ambulators). Changes in ADL performance were dependent on several predictors. Forty-one percent of the variance in the ADL change scores was explained by the following combination of predictors: age, ADL scores at the start of care, walking speed at the start of care, and confusion status. The ADL score and walking speed at the start of care alone contribute to the most of the variance of change in ADL composite scores. Patients admitted to HHC from SNFs made less improvement in the mean ADL score when compared with patients admitted directly from acute care facilities (Fig 3). Chan et al reported that patients poststroke discharged from IRFs had better functional outcome scores compared with those from other settings (including home with HHC services) at 6-month follow-up.16 Enhanced functional status for patients poststroke managed in IRF settings may have been related to more intense rehabilitation services.16 Our findings suggest that patients discharged from short-stay acute care hospitals make greater improvements in the mean ADL performance during an episode of HHC. And patients admitted to HHC from SNF did not differ significantly from those admitted from an IRF. A potential explanation for these ADL improvement differences by postacute setting is that patients receiving

7

postacute care in SNFs make the most improvement toward the maximum function in that setting before being admitted to home care services. It is also possible that patients capable of returning home immediately postacute care have less severe poststroke sequelae and/or greater support resources to allow home discharge compared with those referred for management to SNFs. One final point for consideration is that the HHC practice setting offers environmental specificity for the OASIS-C ADL assessment and subsequent functional rehabilitation. Patients receiving HHC services immediately postacute care receive training and are functionally evaluated in a familiar home environment as they resolve poststroke impairments and make progress toward maximal functional potential. The effect of environmental specificity and timing with functional recovery poststroke should be investigated across different postacute rehabilitation settings. Regarding the predictive factors in gait speed change, 14% of the variance in gait speed change scores was explained by the combination of the speed of walking and ADL at the start of care, and living arrangement. Gait speed at the start of care was the highest predictor of change in gait speed at discharge, and faster gait speed was associated with a lesser degree of improvement in gait speed. Others hypothesized that the predictor did not show a significant univariate association with the change in gait speed after the end of care. Predictive factors for an improvement in ADL and gait for stroke survivors in the home care have not previously been described. This study included persons poststroke. Additional studies evaluating the types of interventions provided are needed to generalize the results to the home health population. Additional psychometric work is also needed to validate the use of the ADL composite score. The reliability and validity of the functional items in the OASIS-C have already been investigated but not the ADL composite score. Landis et al41 found that the inter-rater reliability was adequate or better with kappa’s of .60 or higher.42-46 The internal consistency was .88 and higher by using Cronbach’s coefficient alpha for baseline and discharge ADL scores.46 The OASIS functional scores (ADLs items) were compared with the Older American Resources and Services instrument to measure its concurrent validity with an overall correlation of r 5 .71.33 The OASIS functional scores are also highly correlated with the Katz Index of ADL33,46; however, the functional items in the OASIS-C were not developed for scale scoring.34 Also, the ability of the functional items within the OASIS-C to detect functional change over a period of time compared with validated outcome measures has not been examined. Therefore, the ADL composite score was used in our study.35 The clinically important change in ADL composite score (obtained from the OASIS–C) remains to be identified and should be the objective of future studies.

F.Y. ASIRI ET AL.

8

There are some limitations to this study. The nature of the study design was retrospective. Patients were admitted to the rehabilitation program with a high level of impairments at the start of care in terms of number of comorbidities, discharge facility, and their walking speed. Forty-three percent of patients were discharged from nursing or LTC facilities, 94% had three or more comorbidities at the start of care beside their primary condition (stroke), and 58% of patients were walking at a slower speed of less than .4 m/second (household ambulators). In addition, it was impossible to detect the severity of stroke, which limited our ability to describe whether severity of stroke affected outcomes. Stroke severity has been shown in the literature as the main predictor of functional improvement.4 An additional limitation is that the study did not control social situation, for example, number and capacity of family supports, availability of community resources, and financial resources. Finally, we have only subjects in the data set that completed rehabilitation. Future studies must look at factors associated with successful completion of rehabilitation in home versus those who are unable to complete an episode of care.

Conclusions There were significant improvements in ADL function and gait speed in individuals with stroke after receiving HBR. Gait speed and ADL scores at the start of care had the largest influence on functional and gait improvement. The absence of confusion and confusion in new/complex situations were also considered as positive predictors of ADL improvement. In addition, living arrangement, either living alone or with other persons, was associated with a greater change in gait speed. Although fast walkers and impaired patients in terms of ADL scores at the start of care were associated with poor outcomes in the change of gait speed. Knowledge of which factors affect the outcome in persons who have had a stroke and are undergoing treatment in the home could affect intervention planning, clinicians’ judgment, and future payment models. Acknowledgment: We acknowledge the contributions and support of Gentiva Health Services Inc. We also would like to thank Dr Charlotte Weaver for her suggestions on our manuscript.

References 1. Roger VL, Go AS, Lloyd-Jones DM, et al. Heart Disease and Stroke Statistics—2012 Update: A Report From the American Heart Association. Circulation 2012; 125:e2-e220. 2. Centers for Disease Control and Prevention (CDC). Prevalence and most common causes of disability among adults—United States, 2005. MMWR Morb Mortal Wkly Rep 2009;58:421-426.

3. Kelly-Hayes M, Beiser A, Kase CS, et al. The influence of gender and age on disability following ischemic stroke: the Framingham Study. J Stroke Cerebrovasc Dis 2003; 12:119-126. 4. Zorowitz RD, Chen E, Tong KB, et al. Costs and rehabilitation use of stroke survivors: a retrospective study of Medicare beneficiaries. Top Stroke Rehabil 2009;16: 309-320. 5. Craig LE, Wu O, Bernhardt J, et al. Predictors of poststroke mobility: systematic review. Int J Stroke 2011; 6:321-327. 6. Mutai H, Furukawa T, Araki K, et al. Factors associated with functional recovery and home discharge in stroke patients admitted to a convalescent rehabilitation ward. Geriatr Gerontol Int 2012;12:215-222. 7. Inouye M, Kishi K, Ikeda Y, et al. Prediction of functional outcome after stroke rehabilitation. Am J Phys Med Rehabil 2000;79:513-518. 8. Frank M, Conzelmann M, Engelter S. Prediction of discharge destination after neurological rehabilitation in stroke patients. Eur Neurol 2010;63:227-233. 9. Massucci M, Perdon L, Agosti M, et al. Prognostic factors of activity limitation and discharge destination after stroke rehabilitation. Am J Phys Med Rehabil 2006;85: 963-970. 10. Ng YS, Stein J, Salles SS, et al. Clinical characteristics and rehabilitation outcomes of patients with posterior cerebral artery stroke. Arch Phys Med Rehabil 2005;86: 2138-2143. 11. Shen Q, Cordato D, Chan DK, et al. Identifying the determinants of 1-year post-stroke outcomes in elderly patients. Acta Neurol Scand 2006;113:114-120. 12. Wee JY, Hopman WM. Stroke impairment predictors of discharge function, length of stay, and discharge destination in stroke rehabilitation. Am J Phys Med Rehabil 2005; 84:604-612. 13. Denti L, Agosti M, Franceschini M. Outcome predictors of rehabilitation for first stroke in the elderly. Eur J Phys Rehabil Med 2008;44:3-11. 14. Thommessen B, Bautz-Holter E, Laake K. Predictors of outcome of rehabilitation of elderly stroke patients in a geriatric ward. Clin Rehabil 1999;13:123-128. 15. Hankey GJ, Jamrozik K, Broadhurst RJ, et al. Long-term disability after first-ever stroke and related prognostic factors in the Perth Community Stroke Study, 19891990. Stroke 2002;33:1034-1040. 16. Chan L, Sandel ME, Jette AM, et al. Does postacute care site matter? A longitudinal study assessing functional recovery after a stroke. Arch Phys Med Rehabil 2013; 94:622-629. 17. Caffrey C, Sengupta M, Harris-Kojetin L, Valverde R. Home health care and discharged hospice care patients: United States, 2000 and 2007; 2011. http://www.cdc.gov/ nchs/data/nhsr/nhsr038.pdf. Accessed May 30, 2012. 18. Melin AL, Hakansson S, Bygren LO. The costeffectiveness of rehabilitation in the home: a study of Swedish elderly. Am J Public Health 1993;83:356-362. 19. Hughes SL, Ulasevich A, Weaver FM, et al. Impact of home care on hospital days: a meta analysis. Health Serv Res 1997;32:415-432. 20. The National Association for Home Care & Hospice 2010; Basic statistics about home care; 2012. Available at: http://nahc.org/facts/10HC_Stats.pdf. Accessed May 30, 2012. 21. Agrawal N, Johnston SC, Wu YW, et al. Imaging data reveal a higher pediatric stroke incidence than prior US estimates. Stroke 2009;40:3415-3421.

PREDICTIVE FACTORS FOR STROKE SURVIVORS 22. Mayo NE, Wood-Dauphinee S, Cote R, et al. There’s no place like home: an evaluation of early supported discharge for stroke. Stroke 2000;31:1016-1023. 23. Hillier S, Inglis-Jassiem G. Rehabilitation for communitydwelling people with stroke: home or centre based? A systematic review. Int J Stroke 2010;5:178-186. 24. Geddes JM, Chamberlain MA. Home-based rehabilitation for people with stroke: a comparative study of six community services providing co-ordinated, multidisciplinary treatment. Clin Rehabil 2001;15:589-599. 25. Therapy-based rehabilitation services for stroke patients at home. Cochrane Database Syst Rev (Online) 2003; CD002925. 26. Fusco D, Bochicchio GB, Onder G, et al. Predictors of rehabilitation outcome among frail elderly patients living in the community. J Am Med Dir Assoc 2009;10:335-341. 27. Gitlin LN, Hauck WW, Dennis MP, et al. Long-term effect on mortality of a home intervention that reduces functional difficulties in older adults: results from a randomized trial. J Am Geriatr Soc 2009;57:476-481. 28. Whitney SL, Marchetti GF, Ellis JL, Otis L. Improvements in Balance in Older Adults Engaged in a Specialized Home Care Falls Prevention Program. Journal Ger Phys Ther 2012;35:1-10. 29. Duncan PW, Sullivan KJ, Behrman AL, et al. Protocol for the locomotor experience applied post-stroke (LEAPS) trial: a randomized controlled trial. BMC Neurol 2007;7:39. 30. Schmid A, Duncan PW, Studenski S, et al. Improvements in speed-based gait classifications are meaningful. Stroke 2007;38:2096-2100. 31. Marrelli TM, ed. Handbook of home health standards: quality, documentation, and reimbursement. 5th ed. St. Louis, Missouri: Elsevier 2012. 32. OASIS-C. Centers for Medicare & Medicaid Services; 2012. Available at: http://www.cms.gov/Medicare/ Quality-Initiatives-Patient-Assessment-Instruments/OASIS/ index.html. Accessed May 30, 2012. 33. Tullai-McGuinness S, Madigan EA, Fortinsky RH. Validity testing the Outcomes and Assessment Information Set (OASIS). Home Health Care Serv Q 2009;28:45-57. 34. Madigan EA, Gordon N, Fortinsky RH, et al. Predictors of functional capacity changes in a US population of Medi-

9

35.

36.

37.

38.

39.

40.

41. 42.

43.

44.

45.

46.

care home health care (HHC) patients with heart failure (HF). Arch Gerontol Geriatr 2011. Fortinsky RH, Garcia RI, Joseph Sheehan T, et al. Measuring disability in Medicare home care patients: application of Rasch modeling to the outcome and assessment information set. Med Care 2003;41:601-615. Viccaro LJ, Perera S, Studenski SA. Is timed up and go better than gait speed in predicting health, function, and falls in older adults? J Am Geriatr Soc 2011; 59:887-892. Tilson JK, Sullivan KJ, Cen SY, et al. Meaningful gait speed improvement during the first 60 days poststroke: minimal clinically important difference. Phys Ther 2010; 90:196-208. Hardy SE, Perera S, Roumani YF, et al. Improvement in usual gait speed predicts better survival in older adults. J Am Geriatr Soc 2007;55:1727-1734. Perera S, Mody SH, Woodman RC, et al. Meaningful change and responsiveness in common physical performance measures in older adults. J Am Geriatr Soc 2006; 54:743-749. Shumway-Cook A, Woollacott MH, eds. Motor control: translating research into clinical practice. 4th ed. Baltimore, MD: Lippincott Williams & Wilkins 2012. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-174. Madigan EA, Fortinsky RH. Additional psychometric evaluation of the Outcomes and Assessment Information Set (OASIS). Home Health Care Serv Q 2000;18:49-62. Madigan EA, Fortinsky RH. Interrater reliability of the outcomes and assessment information set: results from the field. Gerontologist 2004;44:689-692. Hittle DF, Shaughnessy PW, Crisler KS, et al. A study of reliability and burden of home health assessment using OASIS. Home Health Care Serv Q 2003;22:43-63. Madigan EA, Tullai-McGuinness S, Fortinsky RH. Accuracy in the Outcomes and Assessment Information Set (OASIS): results of a video simulation. Res Nurs Health 2003;26:273-283. Scharpf TP, Madigan EA. Functional status outcome measures in home health care patients with heart failure. Home Health Care Serv Q 2010;29:155-170.

Predictors of functional and gait outcomes for persons poststroke undergoing home-based rehabilitation.

The literature on the impact of home-based rehabilitation on functional outcomes for patients after stroke is limited. The purpose of this study was t...
479KB Sizes 0 Downloads 3 Views