This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Glass, T. A. Articles by Ware, J. Search for Related Content PubMed PubMed Citation Articles by Glass, T. A. Articles by Ware, J. Related Collections Social Support Neurology Therapeutic Interventions

The Families In Recovery From Stroke Trial (FIRST): Primary Study Results

From the Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland (T.A.G.); Harvard School of Public Health, Boston, Massachusetts (L.F.B., M.M.G., M.E.F., J.W.); and Massachusetts General Hospital, Boston, Massachusetts (E.F.H., K.F.).

Address correspondence and reprint requests to Thomas A. Glass, PhD, Johns Hopkins Bloomberg School of Public Health, Department of Epidemiology, 615 N. Wolfe Street, Baltimore, MD 21205. E-mail: tglass{at}jhsph.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
OBJECTIVE: Social support and family ties are strong predictors of functional recovery after stroke; however, development of successful psychosocial intervention programs has been difficult. This study examined whether a family-systems intervention designed to influence social support and self-efficacy affects functional outcome in older stroke patients.

METHODS: Two hundred ninety-one community-residing survivors of ischemic stroke or nontraumatic cerebral hemorrhage from eight acute-care hospitals and rehabilitation centers were randomized to either psychosocial intervention (PSI) or usual care (UC). PSI involved up to 16 sessions conducted in the home by a mental health worker. Functional recovery (measured by the Barthel Index [BI] at 6 months postrandomization, inability to assess functioning because of illness or death) was the primary end point.

RESULTS: Functional recovery did not differ between UC and PSI in intention-to-treat analyses. In adjusted logistic regression, the odds of being functionally independent at 6 months was 60% higher in the intervention group, but this difference was not statistically significant (p = .31). Subgroup analyses revealed that PSI may be more effective in subjects with better psychologic and cognitive functioning and who required less inpatient rehabilitation.

CONCLUSION: This study does not provide evidence for the efficacy of psychosocial intervention to improve functional recovery in stroke. Although PSI showed greater improvement than UC, the differences were not statistically significant.

Key Words: cerebrovascular disease, • randomized clinical trial, • social factors, • social support, • outcome study.

Abbreviations: BASRS = Boston Aphasia Severity Rating Scale;; BI = Barthel Index;; CES-D = Center for Epidemiologic Studies Depression Scale;; FIRST = Families in Recovery from Stroke Trial;; MMSE = Mini-Mental Status Exam;; NIH = National Institutes of Health;; NIHSSI = National Institutes of Health Stroke Severity Index;; PSI = psychosocial intervention;; REFFI = recovery efficacy;; RSS = received social support;; SIS = Social Isolation Scale;; TOAST = Trial of ORG 10172 in Acute Stroke Treatment;; UC = usual care.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Improvements in survival after stroke, along with population aging, have led to increased prevalence of stroke. These trends have increased the importance of identifying effective interventions to improve functional outcome and reduce disability poststroke. Growing evidence points to the potential role of psychosocial factors, including depression, social support, and social activity, as important determinants of poststroke outcome (1–10). A review of over 70 studies by Kwakkel (11) found that social support was a robust predictor of poststroke functional status. Evidence in favor of the importance of an enriched social environment for optimal neuronal repair after stroke can also be found in animal models (12,13). Efforts to translate these findings into intervention strategies have been limited.

To date, eight randomized trials have been conducted using various psychosocial intervention models designed to foster improved functional outcome after stroke (14–18). Results thus far have not provided compelling evidence for the effectiveness of psychosocial intervention. The Families in Recovery from Stroke Trial (FIRST) was designed to address limitations of earlier trials by attempting to provide sufficient power to detect differences in functional outcome, testing a more intensive intervention, beginning intervention as soon as possible after stroke onset, and targeting aspects of informal support that have been linked to functional recovery. This article reports primary end point results from the FIRST trial.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Study Design
FIRST was a randomized clinical trial cofunded by the National Institute of Neurologic Diseases and Stroke and by the National Institute on Aging designed to test the efficacy of a psychosocial intervention in stroke patients aged 45 or older. Patients were recruited from four acute-care hospitals (Massachusetts General Hospital, Beth Israel Deaconess Medical Center, Brigham and Women’s Hospital, and Mt. Auburn Hospital) and from four rehabilitation care hospitals (Spaulding Rehab Hospital, Youville Rehabilitation Hospital, Braintree Rehabilitation Hospital, and New England Rehabilitation Hospital). Randomization was stratified by acute-care hospital, lesion lateralization (left, right, cerebellar/brainstem, or multifocal), and stroke type (ischemic or hemorrhagic). A permuted blocks algorithm with a block size of four was used. Patients were randomly assigned to receive usual care (UC) or usual care plus the psychosocial intervention (PSI) at a 1:1 ratio. Although participants and interventionists were aware of the patient’s treatment assignment, all end point data and follow-up assessments were performed by study nurses who were blind to treatment assignment.

Patient Recruitment and Eligibility
Patients with ischemic or nontraumatic hemorrhagic stroke as determined by admitting diagnosis and chart review, including imaging reports, were screened for eligibility by study nurses at eight acute-care hospitals and rehabilitation facilities in the Boston area. As a result of differences known to exist in pathophysiology and in prognosis, patients were classified according to stroke type (ischemic or hemorrhagic) and lateralization (left, right, brainstem/cerebellar, or multifocal). In cases in which medical records were ambiguous, study neurologists at each site resolved classification decisions.

Eligibility criteria were drawn from best practices in stroke trials in consultation with study neurologists. Choices were guided by four principals: 1) exclude patients who were unlikely to benefit from the intervention as a result of severe impairments in cognition or language; 2) exclude patients who had no social network with whom interventionists could collaborate, 3) include only patients with a stroke-related deficit sufficient to require the family system to mount a compensatory response; and 4) exclude patients at low likelihood of finishing the trial as a result of grave illness whether or not related to the stroke. The National Institutes of Health Stroke Severity Index (NIHSSI) was developed in an earlier trial (19) and was used to evaluate severity. The Boston Aphasia Severity Rating Scale (BASRS), a short, standardized subscale of the larger Boston Diagnostic Aphasia Exam (20), was used to characterize level of aphasic disability on a scale of 1 to 5. Level of social isolation was measured using the Social Isolation Score (SIS), which assessed availability and proximity of four types of support. This instrument was developed specifically for FIRST by Berkman and Glass who have previous experience in assessing social isolation in population studies (21,22). Patients were excluded if they were 1) globally aphasic or had limited comprehension and expressive aphasia (BASRS = 0 or 1), 2) extremely socially isolated (SIS >9), 3) residing in a nursing home before stroke or discharged to a nursing home, 4) cognitively impaired before stroke by medical record, 5) living outside metropolitan Boston, 6) only mildly impaired (NIH Stroke Severity Index <3), or 7) too severely impaired (NIH Stroke Severity Index >8).

Written informed consent was requested before completion of all baseline assessments to provide a brief run-in period. All enrolled patients provided written informed consent.

Intervention
Because measurable functional recovery from stroke occurs mainly within 6 months, patients were enrolled as rapidly as possible (within 28 days of hospital admission). Patients assigned to the UC arm were given standard educational material on stroke recovery but were not seen by the FIRST interventionists. Patients randomized to PSI began intervention as soon as possible thereafter. A psychologist or social worker trained in family-systems and cognitive behavioral therapy was assigned to the family by the study coordinator (T.A.G.). The intervention consisted of up to 16 meetings conducted over 6 months in the patient’s home (approximately weekly for 12 weeks, followed by triweekly sessions for another 12 weeks). Sessions lasted approximately 1 hour and included, when possible, the entire support system (stroke survivor, primary caregiver, additional family and friends, and professional caregivers).

The intervention protocol was informed by a family-systems perspective in which stroke is treated as a crisis for the entire family system. Four primary domains of psychosocial challenge in families after stroke were identified: 1) the informational, 2) the social, 3) the emotional, and 4) the behavioral domains. Each of these domains was addressed with targeted interventions designed to optimize the functioning of the whole support system. The primary goals of the intervention were to modify four psychosocial mechanisms found to be associated with better stroke recovery. They included: 1) increase self-efficacy through stroke education, 2) optimize social support through social network mobilization, 3) maximize system cohesion and stress reduction, and 4) enhance problem-solving effectiveness through behavioral skills training. Furthermore, based on these domains, 16 content areas of psychosocial adaptation to stroke were identified (9,23) and used as a checklist to ensure that all topics were addressed during the intervention while allowing variation in the sequence and relative emphases in each family system. Additional details of the intervention are described elsewhere (24). A brief eight-step description of the intervention is included as an Appendix. Interventionists were supervised on a weekly basis by a consulting psychologist and the study coordinator to assure quality control and uniformity of intervention.

Measures
The primary end point was the Barthel Index of activities of daily living (BI) measured 6 months postrandomization. The BI is a standard measure of stroke outcome used in many epidemiologic studies and has been found to have generally favorable measurement properties (25–28). For analysis of continuous BI while following intention-to-treat principles, we defined a new variable, functional BI, which had death as the poorest outcome, illness preventing assessment as the second poorest outcome, and all measured BI values were then assigned ranks relative to these outcomes. The BI is also reported and analyzed dichotomously using a cutpoint of 60, a threshold previously shown to predict capacity to live independently (29). We defined a participant to be functionally independent at 6 months if their BI score was greater than or equal to 60. If the participant had died or was too ill to complete the final assessment, they were grouped with patients whose BI was less than 60.

Nurse assessors classified patients according to stroke type, lateralization, and severity (using the NIH standardized Stroke Severity Index). In cases in which medical records were ambiguous, the study neurologist (K.F.) blindly reviewed all available neuroimaging data on 214 (75%) randomized subjects. No significant protocol violations were discovered. However, two patients were misclassified on stroke type and six patients were misclassified as to lesion lateralization.

The NIH Stroke Scale (NIHSS) and Supplemental Motor Score were used to characterize the patient’s deficits in detail. The NIHSS is a standard well-validated clinical assessment tool widely used to characterize the severity of specific neurologic deficits. It has been found to predict functional status 3 months after stroke (30). Nurse assessors received video-based training and certification on the NIHSS (developed in the TOAST trial), which has been shown to improve reliability and validity (31). In addition, medical history and baseline health variables were collected. Prevalent chronic conditions and risk factors were ascertained by review of the patient’s medical record, including presence of diabetes and hypertension, current smoking status and smoking history, previous stroke or myocardial infarction. A chronic disease score was constructed based on a count of the number of the following conditions present: myocardial infarction, vascular disease, pulmonary disease, endocrine disease, renal or liver disease, gastrointestinal disease, cancer, rheumatologic disease, psychiatric condition, or other serious debilitating chronic condition. Physical performance was measured by combining the scores on five timed tests of functional capacity, including writing a sentence, simulated eating, simulated dressing, turning in a circle, and walking 20 feet. Blood pressure was measured at each visit. The Mini-Mental Status Exam (MMSE) assessed cognitive status. A battery of domain-specific neuropsychologic tests was also given but is not reported here. Quality-of-life and self-rated health were each assessed using a five-level, single-item global rating scale.

Psychosocial risk factors hypothesized to be mediators of intervention effect included the Center for Epidemiologic Studies Depression Scale (CES-D), a general measure of depressive symptoms often used in stroke research (32). The CES-D has been found to be a reliable and sensitive scale to assess poststroke depression in epidemiologic research with few false-positives (33,34). Received social support (RSS) was measured using a modified version of Barrera’s Inventory of Socially Supportive Behaviors (35). Recovery efficacy (REFFI) was measured using 10 questions that assessed the patient’s sense of control over his or her recovery. This instrument was created for FIRST based on similarly structured questions developed previously by Seeman and colleagues (36). Each question had a four-point Likert-scale response; the individual items were summed. Other measures were collected but are not reported in this article.

Statistical Analysis
Baseline characteristics were compared between PSI and UC using t tests and chi-squared tests. The Wilcoxon rank test was performed on the functional BI. As a result of the skewed distribution of the ranked outcome (functional BI), the analyses were conducted using a normal score transformation, which allows the use of ordinary least-squares regression. Failure to transform would lead to violation of model assumptions. Final models were constructed by step-up methods, including all covariates unbalanced at baseline and all covariates significant at the .05 level. Finally, treatment group was added and its regression coefficient was evaluated.

Multiple logistic regression was used to compare treatment groups on functional independence at 6 months after adjustment for other factors. Prespecified subgroup analyses were performed to seek patterns of differential treatment effects. The study population was stratified into high and low levels based on clinically meaningful thresholds for MMSE (<24) and CES-D ([me]18) at the median for NIHSSS and number of preexisting conditions and at the lowest tertile (rehab days). Wilcoxon rank tests were performed in each subgroup. Treatment-by-subgroup interactions were also examined using normal score transformation and ordinary least-squares regression. Finally, the hypothesized mediators were compared between PSI and UC, using t tests, to assess whether the intervention changed the targeted mediators.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
A total of 1683 patients with qualifying stroke were identified at eight participating hospitals (see Figure 1). Of these, 486 (29%) met additional eligibility requirements and were approached for participation. A total of 175 patients (36% of eligibles) refused study participation or were unable to provide consent; the remainder (311) provided informed consent. After a brief run-in period, 291 patients completed the baseline assessment and were randomized to UC (n = 145) or PSI (n = 146). Randomization occurred, on average, 22 days after stroke onset (range, 4–34 days).

View larger version (28K): [in this window] [in a new window]   Figure 1. Participant flow in FIRST.

Baseline characteristics by treatment group of those patients who are included in the primary analysis appear in Table 1. The PSI group had a higher proportion of widowed participants than the UC group (p = .03). There were no significant differences between the UC and PSI on any other baseline variable.

The results of primary analysis of intervention efficacy are presented in Table 4. The analysis of functional BI is presented both unadjusted and adjusted for variables imbalanced at baseline (percent widowed) and those covariates that survived stepwise model fitting. Neither analysis showed a statistically significant difference between the UC and PSI groups. Overall, PSI subjects had 20% higher odds of being functionally independent (unadjusted) and 60% higher odds of being independent after adjustment for other factors compared with the UC group. However, this effect did not reach statistical significance in either model.

View larger version (13K): [in this window] [in a new window]   Figure 2. Summary of effects of treatment by preselected subgroups of patients.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Mounting evidence indicates that social support and family functioning may be important in recovery from stroke. A number of randomized trials have been conducted to amplify the availability of formal and informal social support. The FIRST study addressed several limitations of earlier trials; however, these results are consistent with previous trials.

Based on pilot testing, we hypothesized that those with moderately severe strokes who were sufficiently cognitively intact to participate in complex verbal communication would benefit most from the FIRST intervention. However, a higher-than-expected percent of patients who were randomized had sustained minor strokes with no or few functional deficits. Those patients had high BI scores at baseline and throughout follow up, which created a ceiling effect. Anticipating this possibility, we had put in place several strategies to increase enrollment among patients with more substantial disabilities. However, 41% reported the highest possible BI scores at final follow up, which severely limited the variance observed between the trial’s two arms. We believe that our eligibility criteria were appropriate. In practice, however, we often enrolled higher-functioning patients who were likely to improve without intervention. This, coupled with a higher refusal rate among more severely affected patients, produced a significant source of volunteer bias (37–39). Future investigators should consider recruitment and consent procedures that reduce the influence of volunteer bias by recruiting patients who are more like the target population.

A second group of patients may have been too psychologically impaired to benefit from the intervention. Preplanned subgroup analyses showed that those with better cognition and less depression benefited more. When the FIRST study was planned, it was unclear who would benefit the most from such an intervention. Without evidence, our approach was as inclusive as possible. Future interventions should consider carefully the role of cognitive and emotional health in selecting patients who might benefit.

During the course of outcome assessments, an in-home medication audit was performed; we identified all subjects who had been taking antidepressant medication at baseline and follow up. This analysis showed that rates of antidepressant use were higher in the UC group at baseline (27% vs. 26%), and at 6 months (35% vs. 31%). In the UC group, the use of antidepressants increased slightly more over the course of intervention (8% vs. 6%). Importantly, 9% fewer UC participants scored above our cutpoint for significant depressive symptoms compared with the PSI group at both baseline and follow up. Although important, we do not feel that the difference in antidepressant use explains the absence of a treatment effect, although this possibility cannot be ruled out. First, data on antidepressant use are limited by the fact that stroke patients may be put on antidepressants without a true diagnosis of depression, for decreased activation, or even for poststroke pain syndromes. It appears that fewer of the UC patients were depressed. This may be a function of higher antidepressant use or may be explained by a greater willingness among treatment subjects to report affective symptoms to blinded interviewers. Because a goal of the intervention was to normalize feelings of depression in treatment, social desirability bias against reporting depressive symptoms may have been reduced in treatment respondents.

FIRST did not achieve anticipated changes in mediating mechanisms we hypothesized would lead to better recovery (improved social support, increased self-efficacy, and treatment for depression). This may have been the result of an insufficient "dose," especially for some subgroups. Because those who received the highest treatment dose are different from those who received lower dose, the analysis of treatment effectiveness by dose is quite complex and beyond the scope of this article, but will be investigated in a subsequent analysis. We do not yet know how to design interventions that are optimally intensive for changing psychosocial targets. Our results lend support to the view that approximately 12 treatment sessions is sufficient for most patients but may be insufficient for patients with additional medical or psychosocial challenges. We agree with the conclusions of the investigators of the ENRICHD trial (40), which provided 11 individual sessions on average (median), that in patients who are socially isolated or depressed, a longer duration and greater intensity of treatment may be required to alter these behavioral mediators. It is also possible that the mechanisms we targeted may be resistant to change in older populations as a result of longstanding behavioral or physiological adaptations that are difficult to alter. Future studies should identify and test powerful and sustainable interventions known to modify important psychosocial conditions.

Despite the widespread use of the BI in stroke trials, evidence is accumulating that there may be more appropriate assessments of functioning for stroke patients. Duncan (41) argued that in many trials, post hoc analyses revealed evidence of benefit only after alternative outcomes were used (or standard measures like BI were rearranged). Also, among activities of daily living (ADLs) measures, the BI uses what has been called the "human assistance metric," meaning that respondents are asked if they are able to perform tasks alone or with someone’s help. Jette (42) has shown that compared with scales assessing task difficulty, instruments like the BI may seriously underestimate disability. Thus, the BI may be less sensitive to change than other measures. Although the BI was the gold standard when FIRST began, several potentially more sensitive and reliable measures have been developed and could be used in future trials. One particularly promising example is the Late-Life Function and Disability Instrument developed by Jette and colleagues (43,44). This 32-item instrument was designed to address some of the limitations in scales like the BI and has been found to have favorable reproducibility and measurement properties in older, ethnically diverse populations. This instrument has been effectively applied in studies of stroke outcome (45). This instrument could be combined with other measures of stroke outcome that assess social functioning and quality of life (such as the Frenchay Activities Index (46) or the Sickness Impact Profile (47,48). We advise future investigators to assemble a package of specialized instruments such as the three listed here, which could be used together to assess functional recovery across a broad range of domains from self-care to social and occupational functioning. Such an approach is also recommended for stroke trials by Duncan (49)

Observational studies have found social support and aspects of family functioning to be central to recovery in cardiovascular and cerebrovascular disease. There has been a wide-ranging debate about whether the associations reported from prospective observational studies represent causal links or whether they are confounded by an unobserved third factor, or by reverse causation, in which case the disease process itself influences social support. Because the FIRST intervention did not change social support significantly, our study cannot address this question adequately. Most of the observational studies linking social support and health outcomes have done an excellent job at controlling for known confounders and have gone to great lengths to disentangle the temporal sequencing of social exposures leading to adverse health outcomes. Therefore, we do not view our results as invalidating the findings of observational studies. Rather, we suspect that we have not succeeded at changing social interactions sufficiently or at a point in time where such change is likely to influence prognosis. Social interactions and support may influence stroke recovery cumulatively over the life course and may be difficult to change late in life. Social relationships may confer protection that can only be seen over longer periods. Future interventions should have the demonstrated capacity to modify the psychosocial experiences we hypothesize are linked to poor health and focus on earlier phases of disease risk and progression.

	ACKNOWLEDGMENTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

The authors thank the study nurses (Linda Chellali, BS, RN, Kathy Foskett, BS, RN, and Alice Smith, BS, RN) who screened, enrolled, and tested study participants. The authors also acknowledge the team of family interventionists (Sarah Greenberg, EdD, Peter Manso, MSW, David Rintell, EdD, and Carol Roesch, MSW) who provided the intervention faithfully. Important consultative support was provided by Dr. Lawrence Brass (Yale), Dr. J. Philip Kistler (Harvard), Dr. Marilyn Albert (Johns Hopkins), and Dr. Barry Dym.

Received for publication January 13, 2004.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 

1. Evans RL, Bishop DS, Matlock AL, Stranahan S, Halar EM, Noonan WC. Prestroke family interaction as a predictor of stroke outcome. Arch Phys Med Rehabil 1987; 68: 508–12.[Medline]
2. Gottlieb A, Golander H, Bar-Tal Y, et al. The influence of social support and perceived control on handicap and quality of life after stroke. Aging (Milano) 2001; 13: 11–5.[Medline]
3. Maeshima S, Ueyoshi A, Osawa A, Ishida K, Kunimoto K, Shimamoto Y, Matsumoto T, Yoshida M. Mobility and muscle strength contralateral to hemiplegia from stroke: benefit from self-training with family support. Am J Phys Med Rehabil 2003; 82: 456–62.[CrossRef][Medline]
4. Jaracz K, Kozubski W. Quality of life in stroke patients. Acta Neurol Scand 2003; 107: 324–9.[CrossRef][Medline]
5. Friedland J, McColl M. Social support and psychosocial dysfunction after stroke: buffering effects in a community sample. Arch Phys Med Rehabil 1987; 68: 475–80.[Medline]
6. Hendricks RD, Haselkorn JK, Bishop DS, Baldwin D. The family’s role in stroke rehabilitation. A review of the literature. Am J Phys Med Rehabil 1992; 71: 135–9.[Medline]
7. Kinsella GJ, Duffy FD. Psychosocial readjustment in the spouses of aphasic patients. Scand J Rehabil Med 1979; 11: 129–32.[Medline]
8. Palmer S, Glass TA. Family function and stroke recovery: a review. Rehabil Psychol 2003; 48: 255–65.[CrossRef]
9. Glass TA, Matchar DB, Belyea MJ, Feussner JR. Impact of social support on outcome in first stroke. Stroke 1993; 24: 64–70.[Abstract/Free Full Text]
10. Osberg JS, DeJong G, Haley SM, Seward ML, McGinnis GE, Germaine J. Predicting long-term outcome among post-rehabilitation stroke patients. Am J Phys Med Rehabil 1988; 67: 94–103.[Medline]
11. Kwakkel G, Wagenaar RC, Kollen BJ, Lankhorst GJ. Predicting disability in stroke—a critical review of the literature. Age Ageing 1996; 25: 479–89.[Abstract/Free Full Text]
12. Dahlqvist P, Zhao L, Johansson IM, Mattson B, Johansson BB, Seckl JR, Olsson T. Environmental enrichment alters nerve growth factor-induced gene A and glucocorticoid receptor messenger RNA expression after middle cerebral artery occlusion in rats. Neuroscience 1999; 93: 527–35.[CrossRef][Medline]
13. Kempermann G, Gage FH. New nerve cells for the adult brain. Sci Am 1999; 280: 48–53.[Medline]
14. Evans RL, Matlock A-L, Bishop DS, Stranahan S, Pederson C. Family intervention after stroke: does counseling or education help? Stroke 1988; 19: 1243–9.[Abstract/Free Full Text]
15. Friedland JF, McColl M. Social support intervention after stroke: results of a randomized trial. Arch Phys Med Rehabil 1992; 73: 573–81.[Medline]
16. Dennis M, O’Rourke S, Slattery J, Staniforth T, Warlow C. Evaluation of a stroke family care worker: results of a randomised controlled trial. BMJ 1997; 314: 1071–6.[Abstract/Free Full Text]
17. Forster A, Young J. Specialist nurse support for patients with stroke in the community: a randomised controlled trial. BMJ 1996; 312: 1642–6.[Abstract/Free Full Text]
18. Ben-Yishay Y, Diller L. Cognitive remediation in traumatic brain injury: update and issues. Arch Phys Med Rehabil 1993; 74: 204–13.[Medline]
19. Adams HP Jr, Woolson RF, Clarke WR, Davis PH, Bendixen BH, Love BB, Wasek PA, Grimsman KJ. Design of the Trial of Org 10172 in Acute Stroke Treatment (TOAST). Control Clin Trials 1997; 18: 358–77.[CrossRef][Medline]
20. Goodglass H, Kaplan E, Boston Diagnostic Aphasia Examination (BDAE). Philadelphia: Lea and Febiger; 1983.
21. Berkman LF, Glass TA. Social integration, social networks, social support, and health. In: Berkman LF, Kawachi I, eds. Social Epidemiology. New York: Oxford University Press; 2000: 137–73.
22. Glass TA, Mendes de Leon CF, Seeman TE, Berkman LF. Beyond single indicators of social networks: a LISREL analysis of social ties among the elderly. Soc Sci Med 1997; 44: 1503–18.
23. Glass TA, Maddox GL. The quality and quantity of social support: Stroke recovery as psycho-social transition. Soc Sci Med 1992; 34: 1249–61.
24. Glass TA, Dym B, Greenberg S, Rintell D, Roesch C, Berkman LF. Psychosocial intervention in stroke: the Families in Recovery from Stroke Trial (FIRST). Am J Orthopsychiatry 2000; 70: 169–81.[Medline]
25. Collin C, Wade DT, Davies S, Horne V. The Barthel ADL Index: a reliability study. Int Disabil Stud 1988; 10: 61–3.[Medline]
26. Granger CV, Dewis LS, Peters NC, Sherwood CC, Barrett JE. Stroke rehabilitation: analysis of repeated Barthel Index measures. Arch Phys Med Rehabil 1979; 60: 14–7.[Medline]
27. Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index. Md State Med J 1965; 14: 61–5.[Medline]
28. Wyller TB, Sveen U, Bautz-Holter E. The Barthel ADL index one year after stroke: comparison between relatives’ and occupational therapist’s scores. Age Ageing 1995; 24: 398–401.[Abstract/Free Full Text]
29. Patel AT, Duncan PW, Lai SM, Studenski S. The relation between impairments and functional outcomes poststroke. Arch Phys Med Rehabil 2000; 81: 1357–63.[CrossRef][Medline]
30. Adams HP Jr, Davis PH, Leira EC, Chang KC, Bendixen BH, Clarke WR. Baseline NIH Stroke Scale score strongly predicts outcome after stroke: a report of the Trial of Org 10172 in Acute Stroke Treatment (TOAST). Neurology 1999; 53: 126–31.[Abstract/Free Full Text]
31. Lyden P, Brott T, Tilley B, Welch KM, Mascha EJ, Levine S, Haley EC, Grotta J, Marler J. Improved reliability of the NIH Stroke Scale using video training. NINDS TPA Stroke Study Group. Stroke 1994; 25: 2220–6.[Abstract]
32. Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. J Appl Psychol Measure 1977; 1: 385–401.[CrossRef]
33. Parikh RM, Eden DT, Price TR, Robinson RG. The sensitivity and specificity of the Center for Epidemiologic Studies Depression Scale in screening for post-stroke depression. Intl J Psychiatry Med 1988; 18: 169–81.[Medline]
34. Shinar D, Gross CR, Price TR, Banko M, Bolduc PL, Robinson RG. Screening for depression in stroke patients: the reliability and validity of the Center for Epidemiologic Studies Depression Scale. Stroke 1986; 17: 241–5.[Abstract/Free Full Text]
35. Barrera M. A method for the assessment of social support networks in community survey research. Connections 1980; 3: 8–13.
36. McAvay GJ, Seeman TE, Rodin J. A longitudinal study of change in domain-specific self-efficacy among older adults. J Gerontol B Psychol Sci Soc Sci 1996; 51: 243–53.
37. Amori G, Lenox RH. Do volunteer subjects bias clinical trials? J Clin Psychopharmacol 1989; 9: 321–7.[Medline]
38. Edlund MJ, Craig TJ, Richardson MA. Informed consent as a form of volunteer bias. Am J Psychiatry 1985; 142: 624–7.[Abstract/Free Full Text]
39. Gustavsson JP, Asberg M, Schalling D. The healthy control subject in psychiatric research: impulsiveness and volunteer bias. Acta Psychiatr Scand 1997; 96: 325–8.[Medline]
40. Berkman LF, Blumenthal J, Burg M, Carney RM, Catellier D, Cowan MJ, Czajkowski SM, DeBusk R, Hosking J, Jaffe A, Kaufmann PG, Mitchell P, Norman J, Powell LH, Raczynski JM, Schneiderman N. Effects of treating depression and low perceived social support on clinical events after myocardial infarction: the Enhancing Recovery in Coronary Heart Disease Patients (ENRICHD) Randomized Trial. JAMA 2003; 289: 3106–16.[Abstract/Free Full Text]
41. Duncan PW, Jorgensen HS, Wade DT. Outcome measures in acute stroke trials: a systematic review and some recommendations to improve practice. Stroke 2000; 31: 1429–38.[Abstract/Free Full Text]
42. Jette AM. How measurement techniques influence estimates of disability in older populations. Soc Sci Med 1994; 38: 937–42.
43. Jette AM, Haley SM, Coster WJ, Kooyoomjian JT, Levenson S, Heeren T, Ashba J. Late life function and disability instrument: I. Development and evaluation of the disability component. J Gerontol A Biol Sci Med Sci 2002; 57: M209–16.[Abstract/Free Full Text]
44. Haley SM, Jette AM, Coster WJ, Kooyoomjian JT, Levenson S, Heeren T, Ashba J. Late life function and disability instrument: II. Development and evaluation of the function component. J Gerontol A Biol Sci Med Sci 2002; 57: M217–22.[Abstract/Free Full Text]
45. Ouellette MM, LeBrasseur NK, Bean JF, Phillips E, Stein J, Frontera WR, Fielding RA. High-intensity resistance training improves muscle strength, self-reported function, and disability in long-term stroke survivors. Stroke 2004.
46. Wade DT, Legh-Smith J, Langton Hewer R. Social activities after stroke: measurement and natural history using the Frenchay Activities Index. International Rehabilitation Medicine 1985; 7: 176–81.[Medline]
47. de Bruin A, de Witte L, Stevens F, Diederiks J. Sickness Impact Profile: the state of the art of a generic functional status measure. Soc Sci Med 1992; 35: 1003–14.
48. Nydevik I, Hulter-Asberg K. Sickness impact after stroke. Scand J Prim Health Care 1992; 10: 284–9.[Medline]
49. Duncan PW, Lai SM, van Culin V, Huang L, Clausen D, Wallace D. Development of a comprehensive assessment toolbox for stroke. Clin Geriatr Med 1999; 15: 885–915.[Medline]

This article has been cited by other articles:

				A. Ikeda, H. Iso, I. Kawachi, K. Yamagishi, M. Inoue, S. Tsugane, and for the JPHC Study Group Social Support and Stroke and Coronary Heart Disease: The JPHC Study Cohorts II Stroke, March 1, 2008; 39(3): 768 - 775. [Abstract] [Full Text] [PDF]

				J. R. Marler NINDS Clinical Trials in Stroke: Lessons Learned and Future Directions Stroke, December 1, 2007; 38(12): 3302 - 3307. [Abstract] [Full Text] [PDF]

				K.A. Ertel, M.M. Glymour, T.A. Glass, and L.F. Berkman Frailty modifies effectiveness of psychosocial intervention in recovery from stroke Clinical Rehabilitation, June 1, 2007; 21(6): 511 - 522. [Abstract] [PDF]

				J. Redfern, C. McKevitt, and C. D.A. Wolfe Development of Complex Interventions in Stroke Care: A Systematic Review Stroke, September 1, 2006; 37(9): 2410 - 2419. [Abstract] [Full Text] [PDF]

				M. H.L. Lui, F. M. Ross, and D. R. Thompson Supporting Family Caregivers in Stroke Care: A Review of the Evidence for Problem Solving Stroke, November 1, 2005; 36(11): 2514 - 2522. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Glass, T. A. Articles by Ware, J. Search for Related Content PubMed PubMed Citation Articles by Glass, T. A. Articles by Ware, J. Related Collections Social Support Neurology Therapeutic Interventions