Stroke disability.

Key Words: Cerebrovascular accident, Disability, Measurement, Outcomes, Stroke.

Incidence and Prevalence of Stroke

Stroke is the third leading cause of death in the United States and a major cause of disability. Each year approximately 500,000 Americans suffer a new or recurrent stroke. Of this number, 350,000 will survive.[1] Although the incidence of stroke is decreasing, its prevalence in the population appears to be increasing because of enhanced stroke survival and a growing elderly population. The prevalence of stroke survivors is currently over 2,000,000.[1]

Within the first 30 days following stroke, the mortality rate is high 30%).[2] It is more severe for hemorrhagic stroke than for ischemic stroke.[3] However, more than 50% of stroke survivors are alive in 5 years.[4] Given this good survival rate and the continuing high incidence of stroke, estimating and understanding disability following stroke becomes a high priority in health care.

Disability Estimates

Exact estimates of disability following stroke are difficult to obtain because patients selected for study may be population based or referral based (acute hospitals or rehabilitation facilities), outcome measures may be inconsistent, and assessments may have been taken at different times during the recovery period. Patient-level variables such as severity and type of stroke, age, and comorbid conditions will also affect stroke disability and vary in different disability studies. The best estimates of disability following stroke are provided by population-based studies. Stroke registries of all patients referred to hospitals are the next best source, whereas studies of patients referred to rehabilitation facilities tend to be the most biased. Patients who are selected for rehabilitation are usually moderately to severely impaired.

The incidence of dependence in activities of daily living (ADL) is highest immediately after a stroke and decreases significantly thereafter, according to various authors.[4-8] A retrospective analysis of data from 292 persons in Rochester, Minn, following their first stroke indicated that 75% were dependent in ADL at onset of stroke. Only 57% of the survivors, however, were dependent at the time they were discharged from the hospital.[4] In a prospective, population-based registry of 976 patients with stroke in the Frenchay Health District in England, Wade and Langton-Hewer[7] found that the incidence of total dependence in ADL decreased from 58% at 1 week poststroke to 9% at 6 months poststroke. In a Japanese study, Kojima et al[8] found similar results, with only 25% of 5-year survivors totally dependent in ADL. Christie[9] reported 6-month disability levels for 296 persons who participated in a population-based study of stroke in Melbourne, Australia. Twenty-five percent of the subjects in that study were independent in ADL, and 17% could drive a car or use public transportation. Of those subjects under age 75 years who did not lose consciousness immediately after the stroke, 54% were independent in ADL and 36% could drive a car or use public transportation.

Several other factors may influence these results. Selective survival of the least seriously disabled individuals may bias these findings. Also, dependencies in ADL may vary by function, making a summated ADL score less representative of limitations in individual activities. For example, the dependence in more complex functions such as bathing is much greater than that in less complex activities such as grooming.[7]

Population-based studies that assess neurological function following stroke support an optimistic view of the patient's prospects for recovery. Bonita and Beaglehole,[10] who assessed the natural history of motor recovery for patients with stroke in Auckland, New Zealand, reported that 88% of the subjects had motor deficits. The proportion of survivors who had persistent motor deficits at 6 months poststroke, however, had declined to 62%, and the majority of these motor deficits were mild. Similarly, in a Finnish population-based study,[6] 73% of all patients with stroke had hemiparesis, but only 37% demonstrated hemiparesis at a 12-month follow-up. An assessment of 148 long-term stroke survivors in the Framingham (Mass) community-based population study[11] revealed that 67 had residual hemiparesis, 4 had bilateral motor weakness, and 77 had no motor deficit.

Although the inability to walk is one of the most common problems in acute stroke, most survivors achieve independence in ambulation. In the population-based Frenchay Health District study, only 27% of patients could functionally ambulate within 1 week of stroke, but at 6 months 85% were independent, although only 25% had regained normal speed of ambulation.[7,12] Gresham et al[11] reported that 78% of the stroke survivors in the Framingham study were ambulatory.

A few studies have examined instrumental ADL and quality of life in stroke survivors. In the Framingham cohort, Gresham et all, compared stroke survivors with age-matched controls and found that 90% of the stroke survivors demonstrated one or more disabilities, compared with 58% of the matched controls. Stroke survivors were more limited in several areas, including household activities, recreational activities, social interaction, and public transportation usage. In a matched control study using the Sickness Impact Profile, Schuling et found that stroke affected household management, leisure activities, and mobility. Christie[9] reported an "imperfect correlation" between residual physical impairment and disability. For example, of the stroke survivors who were employed prior to stroke and who had no residual physical impairments, only 60% returned to work and only 80% continued their prior leisure activities. Of patients who had residual physical impairments, 40% continued to work and 60% engaged in leisure activities.[9]

Other conditions in addition to physical disability undermine the quality of life for persons following stroke. These conditions include depression, dependency on others, and the inability to return to work.[14] Many patients with stroke also have other major comorbid disease, which contributes to their disability. Based on the results of the Framingham disability study of stroke survivors, Jette et al[15] concluded that a history of stroke explained only 12% of the variance in physical disability among men living in the community and only 3% of the variance among women.

The overall prevalence of severe disability among stroke survivors may be overestimated by clinicians whose contact with patients with stroke has been limited to early acute care management, rehabilitation units, or long-term care facilities. In any given patient, the process of disablement may be profound and complex. A broader perspective is needed to understand the nature of stroke-related disablement in the population.

Conceptual Models of Disability

Several conceptual models of disability are available to facilitate the understanding, assessment, measurement, and treatment of stroke-related disabilities. The World Health Organization's (WBO) International Classification of Impairments, Disabilities, and Handicaps (ICIDH)[16] and the Nagi "functional limitation" model[17] are the most frequently presented models of the disablement process. The WHO model classifies disablement in terms of "disease, impairment, disability, and handicap." Nagi's model refers to pathology, impairment, functional limitation, and disability (Fig. 1).

In Nagi's model, pathology or disease refers to the underlying pathologic state that interferes with normal bodily functions or structure.[18] In stroke, the pathology may be due to thrombosis, emboli, or hemorrhage in a particular cerebrovascular distribution.

Impairments are the physiological or psychological consequences, or the signs and symptoms, of the pathology of the disease. Some common impairments after stroke are impaired motor function, sensory deficits, abnormal tone, perceptual deficits, cognitive limitations, aphasia, and depression. Although some impairments are the direct effects of stroke, others may be indirect. For example, shoulder pain may not be a direct effect of a stroke but may instead result from the composite effects of loss of shoulder motor function, loss of range of motion, and altered biomechanics of the shoulder complex.

Functional limitations reflect the functional consequences of the pathology or the abilities lost. Examples of physical functional limitations following stroke are restriction in activities such as transfers and walking.

Disability represents the social and societal consequences of functional limitations. It is defined by a patient's inability to perform ADL and maintain social and family relationships, to continue in a vocation, or to pursue leisure activities.

Assessment of the disablement of stroke should capture not only the patient's ability to perform basic or instrumental ADL, but also the patient's perceptions of his or her emotional, social, and physical functions and the ease with which these activities are performed. This more complex and multidimensional concept of well-being and perception of health is called quality of life. Quality-of-life assessments try to capture how limitations in function affect emotional, social, and physical roles as well as perceptions of health. Interest in these assessments is growing rapidly. Many measures have been developed,[19,20] and one measure has been used specifically with patients with stroke.[21]

The pathways from pathology to disability are not necessarily unidirectional. For example, limitations in functional activities can produce impairments such as increased weakness, restricted range of motion, and deconditioning of the cardiovascular system. Moreover, the disablement process may be modified by many other factors (eg, social support, physical environment, motivation, depression). Figure 2 is a graphical representation of a modified version of the Nagi model. It illustrates the complexity of the physical disablement of patients following a stroke and the relationships among impairment, functional limitation, disability, and quality of life. This model of physical disablement also makes a distinction between functional performance that is observed and functional performance that is self-reported. Under the ideal circumstances often created in rehabilitation settings, the patient may be able to perform a task, but this achievement does not reflect his or her abilities in more variable environments or represent his or her usual performance.

Recovery Patterns

Most patients experience some degree of recovery following a stroke. It is difficult, however, to completely distinguish between recovery from impairments and recovery from disability. The recovery of motor function, sensation, and language are representative of neurological recovery. Recovery of functional skills may be attributable to neurological recovery or behavioral compensation, or both. For example, in behavioral compensation, the unaffected extremities may compensate for the reduced function of the affected extremities or the patient may learn to maximize residual motor control in the affected extremities. Neurological and functional recovery are often parallel,[22-25] but distinctions should be made between neurological recovery and functional recovery.

Although it is often thought that the upper extremity does not recover as well as the lower extremity following stroke, the actual degree of neurological recovery of the upper and lower extremities may be similar (Duncan PW, unpublished observations from the Durham County Stroke Study). The lower extremity, however, can function with less motor control than the upper extremity. Thus, partial motor recovery in the lower extremity may permit many patients with stroke to ambulate independently, although the pattern will not be "normal" in pattern or velocity. Partial recovery of upper-extremity function does not usually translate into functional use.

In general, neurological recovery occurs within the first 1 to 3 months following stroke. Further motor or sensory recovery may continue to occur 6 months to 1 year later; however, these changes are generally limited to individuals with some degree of volitional motor control[10,22,24-27] and may not reach statistical or clinical significance.[4,5,27] A prospective study of recovery patterns in 104 patients with anterior circulation ischemic stroke demonstrated that 86% of the variance in 6-month motor recovery can be predicted in 1 month. The more severely impaired patients continued to experience some measurable recovery of function from 3 to 6 months poststroke, but this recovery was not correlated with clinically meaningful ADL scores (> 60 on the Barthel Index).[25] In the Framingham cohort, recovery of motor function and ADL occurred within 3 months of stroke, then subsequently plateaued. Language and cognitive function improved over longer periods of time.[11]

The courses of recovery previously described do not reflect the natural history of recovery because patients who received rehabilitation were not excluded from the population studied. The profiles of recovery that are reported represent measures of central tendency. Individual recovery patterns may be more variable, and some individuals may show more protracted periods of recovery. In spite of some variations in individuals, the temporal pattern of recovery has now been well documented in several independent cohorts. These studies support the tenet that most neurological recovery occurs early. In addition, the time course of recovery is similar for different levels of stroke severity (Duncan PW, unpublished observations from the Durham County Stroke Study).

The prognosis for neurological and functional recovery following stroke is influenced by a number of factors. In a review of the literature, Dombovy et al28 identified the following factors as predictors of poor functional outcome after stroke: coma at onset, incontinence 2 weeks after stroke, poor cognitive function, severe hemiparesis, no motor return within 1 month, previous stroke, perceptual-spatial disorders, and significant cardiovascular disease. In a similar review of 33 articles, Jongbloed[29] identified older age, history of prior stroke, incontinence, and visual-spatial deficits as prognosticators of poor recovery. Other studies have identified the inability to sit unsupported[30] and an accumulation of motor, sensory, and visual deficits[31] as factors that contribute to more protracted recovery of functional activities as well as poorer functional outcomes. Depression and lack of social support have also been associated with longer and less extensive recovery of function following stroke.[32-34]

Staging Patients

Understanding the patterns of recovery and the predictors of outcome are critical for establishing realistic goals and planning appropriate intervention strategies for each patient. Combined with the results of clinical assessments, this understanding can lead to more accurate and realistic patient staging. Staging of patients is useful as a means of dealing realistically with the patient's and his or her family's expectations of recovery and outcome. Staging is also useful as a way of selecting the most appropriate level of therapeutic intervention. For example, staging could be used to target limited resources for stroke rehabilitation so that intensive rehabilitation would be provided only for those for whom it is most appropriate.

Staging of patients following stroke by degree of motor impairment was introduced by Signe Brunnstrom over 23 years ago35 and has more recently been expanded by Gowland.[36] Staging of patients following stroke by expected outcomes has not been a general practice, but it is currently being attempted in some clinics. Table 1 outlines a classification system I use in clinical management of patients after stroke to stage them by expected outcomes.

The results of standardized assessments of cognitive function, sensation, motor control, perception, mobility status, balance, continence, depression, and comorbid diseases guide my classification of the expected outcomes. Patients classified as being at stage 1 of expected outcomes exhibit two or more of the following factors: moderate to severe cognitive deficits, hemiplegia (Brunnstrom stage 1 or 2), severe sensory deficits, severe perceptual-motor deficits, impaired sitting balance, incontinence of bowel and bladder, major comorbidities, and a Barthel ADL index of < 60. At stage 2, patients may have moderate cognitive deficits, hemiplegia (Brunnstrom stage 2 or 3), mild sensory deficits, impaired standing balance, major comorbidities, bladder incontinence, and a Barthel ADL index of at least 60. At stage 3, patients exhibit mild or no cognitive deficits, hemiplegia (Brunnstrom stage 4 or 5), mild or no sensory deficits, good standing balance, and a limited number of comorbidities; they are continent of bowel and bladder and score >60 on the Barthel Index. Patients at stage 4 will have good cognition, slight hemiparesis Brunnstrom stage 5 or 6), no sensory deficits, good balance, and a limited number of comorbidities and will be continent of bowel and bladder.

A study is in progress at Harmarville Rehabilitation Center, Pittsburgh, Pa, to test the validity and reliability of this outcome staging scheme. Physicians, occupational therapists, and physical therapists are participating in this study. They are also investigating the optimal timing of assessments; that is, they are trying to ascertain how soon after an event the outcomes can be accurately predicted.

Staging, if demonstrated to be reliable and valid in the clinical setting, could guide treatment goals and family education of patients following a stroke. For example, for patients in stages 3 and 4, the therapeutic intervention program would be planned to remediate neurological impairments and to improve physical conditioning, as well as to promote independence in ADL. The program for patients in stage 2 would not focus on remediating impairments but rather on compensatory training in functional tasks and ADL. The program for patients at stage I would emphasize family or caregiver instruction and assistance with basic ADL. The patient's stage of expected outcomes would be considered in conjunction with other influencing factors (time since stroke, social support, medical status, and patient preferences) to help guide selection of the most appropriate site for postacute care. The proposed method of staging and the accuracy of prediction are yet to be validated. Previous work by Komer-Bitensky et al[37] suggests that physical therapists demonstrate a reasonable degree of accuracy in predicting rehabilitation discharge outcomes in patients who have had a stroke. When therapists predicted dependency, they were almost always correct (predictive value of dependence=91.3%-100%). The predictive value of independence was not as great, but was good (70.5%-79.7%).

Recently, Kalra and Crome[38] demonstrated that the Orpington Prognostic Score (Tab. 2) measured at 2 weeks poststroke was very useful in predicting outcomes in patients over 75 years of age. The correlation ([r.sup.2]) between the Orpington score and functional outcome was .89. Patients with an Orpington score of <3.2 were discharged within 3 weeks of stroke, whereas those scoring >5.2 required long-term care.38

The results of the previously mentioned studies suggest that clinicians are able to accurately predict outcomes in most patients. Initially, there may be some hesitation to "stage" patients. This resistance, however, will decrease with recognition that staging can be accurate in predicting recovery and response to specific rehabilitation services. Early staging can serve as a guiding principle but will remain subject to modification if the patient's potential changes.

Measurement

We will never be able to adequately capture the process of disablement following stroke, define the recovery process, or develop predictive models of outcome unless we use well-characterized measures of impairment, functional performance, and disability. The selected measures must be valid, reliable, and sensitive to change. We should also use measures that do not have ceiling effects. One measure that has been used in most assessments of stroke-related disability is the Barthel Index. Although this instrument may be an adequate measure of basic ADL, it may not measure more complex functional loss and disablement, such as a decline in the performance of more advanced ADL or in the patient's self-assessment of quality of life.

Several factors appear to contribute to the current reluctance to utilize standardized assessments in evaluating patients who have had a stroke. Stroke differs widely in its clinical manifestations, and no single assessment instrument can measure the full range of potential impairments, functional limitations, and disability. Adequate evaluation must rely on a battery of instruments. These measures are available, but only a few have demonstrated clinimetric priorities of reliability, validity, and sensitivity to change. Probably the most important barrier, however, has been the reluctance of clinicians to adopt standardized instruments. Recent reports that some formal assessments may be more reliable than clinical impressions may help to overcome this resistance.[39]

Table 3 represents a battery of measures that I recommend for assessment of stroke.[40-50] The list is not comprehensive of all assessments available, but the measures included are practical and have been assessed for reliability and validity in studies of patients who have sustained a stroke. The assessments span the domains of impairments, functional limitations, and disability. Most of these measures can be performed in the acute or postacute care setting or following discharge. Quality-of-life assessment, however, is most relevant in the person's home setting and should preferably be done there. Quality-of-life measures have the specific advantage of capturing more complex functions, which may be compromised poststroke even in the presence of good recovery of basic ADL.

The cognitive and sensorimotor measures of impairment listed in Table 3 may need to be supplemented by assessments of depression, language deficits, and perceptual deficits because these factors can also affect the level of disability. These assessments can be performed by our colleagues in psychology, speech and language pathology, and occupational therapy. For more comprehensive reviews of the measures available for characterizing impairments, functional limitations, and disability following stroke, the reader is referred to Wade.[40]

Modifying the Disablement Process

The goals of physical therapy in stroke rehabilitation should be to maximize function and minimize impairments within the constraints of the patient's pathology, comorbidities, and available resources. The ultimate goal is to reduce the physical contributions to disability.

A primary step in reducing disability is to examine the relationships between impairment and disability. Exact measurements of both are obviously preliminary to this process. A physical therapy evaluation should involve more than just compiling a list of patient impairments or functional deficits. The challenge to the physical therapist is to evaluate findings and to analyze critically the interrelationships among impairments, functional limitations, and disability.

In addition, the variables that moderate these relationships need to be carefully considered.

The movement deficits following stroke are extremely complex and reflect the complexity of normal motor control. One of the tasks for the physical therapist is to observe the patient's functional limitations and to carefully assess the patient to determine which impairments are contributing to the functional limitations. The physical therapist needs to adequately diagnose the cause of the functional limitations and to decide whether the impairments can be remediated. If the impairments cannot be remediated, then the patient should be taught to compensate for the impairments.

A primary question in all rehabilitation programs is: What constitutes a clinically significant change in impairment? Many impairments may be modified to some degree with aggressive programs, but are these changes in impairments clinically meaningful? Clinically meaningful changes in impairments are those associated with changes in function. Therefore, the ultimate goal in the rehabilitation process is to modify function. Any evaluation of the effectiveness of the rehabilitation program should assess functional limitations and level of disability.

We all know patients who have similar levels of impairment and similar functional abilities but quite different levels of disability. The quality of life of a person following a stroke is not determined solely by his or her physical function but may be modified by many other factors (eg, age, general health state, psychosocial, personal motivation, social support, and physical environment). It is beyond the scope of physical therapy to alter or control all of these moderating variables; however, we must consider them in our plans to reduce the disablement of stroke.

Conclusions

In order to better characterize stroke-related disability, we should use population-based studies to assess the extent of impairment, functional limitation, and disability in patients following a stroke. Population-based assessments of disability do not support the typically fatalistic attitude that most stroke survivors are profoundly disabled and need long-term care. Many stroke survivors experience some degree of neurological recovery; they are ambulatory and not totally dependent or in need of long-term care.

Disability following stroke can be adequately characterized only if we use measures across all the domains of disablement (impairment, functional limitations, disability, and quality of life). Impairment-level measures are important to define the factors that are contributing to functional limitations and to guide treatment. Yet, the ultimate stroke outcome measures are not at the impairment level but rather are at the functional and disability level. We need to select our measures of these domains carefully.

Finally, we need to use the measures obtained to reasonably predict outcomes, making allowance in our predictions for the patient's own perceptions and the many variables outside the clinical sphere that may affect the patient's recovery. Using predicted outcomes to stage patients may result in more effective treatment plans and a hopefully enhanced quality of life for the ever-increasing numbers of stroke survivors in the population we serve.

Acknowledgments

I thank Annette Jurgelski, MAT, for editorial assistance and Fikri Yucel for assistance in preparing graphics.

References

[1] 1992 Stroke Facts, Dallas, Tex: American Heart Association; 1992. [2] Garraway WM, Whisnant JP, Drury I. The changing pattern of survival following stroke. Stroke. 1983;14:699-703. [3] Soltero I, Liu K, Cooper R, et al. Trends in mortality from cerebrovascular diseases in the United States, 1960 to 1975. Stroke. 1978;9: 549-558. [4] Dombow ML, Basford JR, Whisnant JP, Bergstralh EJ. Disability and use of rehabilitation services following stroke in Rochester, Minnesota, 1975-1979. Stroke. 1987; 18:830-836. [5] Ahlsio B, Britton M, Murray V, Theorell T. Disablement and quality of life after stroke. Stroke. 1984; 15:886-890. [6] Kotila M, Waltimo O, Niemi M-L, et al. The profile of recovery from stroke and factors influencing outcome. Stroke. 1984;15:1039-1044. [7] Wade DT, Langton-Hewer R. Functional abilities after stroke: measurement, natural history, and prognosis. J Neurol Neurosurg Psychiatry. 1987;50:177-182. [8] Kojima S, Omura T, Wakamatsu W, et al. Prognosis and disability of stroke patients after 5 years in Akita, Japan. Stroke. 1990;21:72-77. [9] Christie D, Aftermath of stroke: an epidemiological study in Melbourne, Australia. J Epidemiol Community Health. 1982;36:123-126. [10] Bonita R, Beaglehole R. Recovery of motor function after stroke. Stroke. 1988;19:1497-1500. [11] Gresham GE, Phillips TF, Wolf PA, et al. Epidemiologic profile of long-term stroke disability: the Framingham Study. Arch Phys Med Rehabil. 1979;60:487-491. [12] Wade DT, Wood VA, Heller A, et al. Walking after stroke: measurement and recovery over the first three months. Scand J Rehabil Med. 1987;19:25-30 [13] Schuling J, Greidanus J, Meyboom-de Jong B. Measuring functional status of stroke patients with the Sickness Impact Profile. Disabil Rehabil. 1993;15:19-23. [14] Viitanen M, Fugl-Meyer KS, Bernspang B, Fugl-Meyer AR. Life satisfaction in long-term survivors after stroke. Scand J Rehabil Med. 1988;20:17-24. [15] Jette AM, Pinsky JL, Branch LG, et al. The Framingham disability study: physical disability among community-dwelling survivors of stroke. J Clin Epidemiol. 1988;41:719-726. [16] International Classification of Impairments, Disabilities, and Handicaps. Geneva, Switzerland: World Health Organization; 1980, [17] Nagi SZ. Disability and Rehabilitation. Columbus, Ohio: Ohio State University Press; 1969. [18] Nagi S. Disability concepts revisited: implications for prevention. In: Pope AM, Tralov AR,. eds, Disability in America: Toward a National Agenda for Prevention. Washington, DC: National Academy Press; 1991. [19] Dehaan R, Aaronson N, Limburt M, et al. Measuring quality of life in stroke. Stroke. 1993;24:320-327. [20] Jette AM. Using health-related quality of life measures in physical therapy outcomes research Phys Ther. 1993;73:528-537. [21] Holbrook M, Skilbeck CE. An activities index for use with stroke patients. Age Ageing. 1983; 12:166-170. [22] Wade DT, Langton-Hewer R, Wood VA, et al. The hemiplegic arm after stroke: measurement and recovery, J Neurol Neurosurg Psychiatry. 1983;46:521-524. [23] Wade DT, Wood VA, Langton-Hewer R. Recovery after stroke: the first three months. J Neurology. 1985;48;7-13. [24] Skilbeck CE, Wade DT, Langton-Hewer R, Wood VA. Recovery after stroke J Neurol Neurosurg Psychiatry. 1983;46:5-8. [25] Duncan PW, Goldstein LB, Matchar D, et al. Measurement of motor recovery after stroke: outcome assessment and sample size requirements. Stroke. 1992;23:1084-1089. [26] Kinsella G, Ford B. Acute recovery patterns in stroke. Med J Aust. 1980;2:662-666. [27] Andrews K, Brocklehurst JC, Richards B, Lavcock PJ. The rate of recovery from stroke and its measurement. International Rehabilitation Medicine. 1981;3:155-161. [28] Dombovy ML, Sandok BA, Basford JR. Rehabilitation for stroke: a review. Stroke. 1986; 17:363-369. [29] Jongbloed L. Prediction of function after stroke: a critical review. Stroke. 1986;17:765-776. [30] Sandin KJ, Smith BS. The measure of balance in sitting in stroke rehabilitation prognosis. Stroke. 1990;21:82-86. [31] Reding MJ, Potes E. Rehabilitation outcome following initial unilateral hemispheric stroke: life table analysis approach. Stroke, 1988;19: 1354-1358. [32] Parikh RM, Robinson RG, Lipsey JR, et al. The impact of post-stroke depression on recovery in activities of daily living over a two-year follow-up. Arch Neurol 1990;47:785-789. [33] Schubert DS, Taylor C, Lee S, et al. Physical consequences of depression in the stroke patient. Gen Hosp Psychiatry 1992; 14:69-76. [34] Glass TA, Matchar DB, Belyea M, Feussner JR. Impact of social support on outcome in first stroke. Stroke. 1993;24:64-70. [35] Brunnstrom S. Movement Therapy in Hemiplegia: A Neurophysiological Approach. New York, NY: Harper & Row; 1970. [36] Gowland CA. Staging motor impairment after stroke. Stroke. 1990;21 (suppl II): II19-II21. [37] Korner-Bitensky N, Mayo N, Cabot R, et al. Motor and functional recovery after stroke: accuracy of physical therapists' predictions. Arch Phys Med Rehabil 1989;70:95-99. [38] Kalra L, Crome P. The role of prognostic scores in targeting stroke rehabilitation in elderly patients. J Am Geriatr Soc. 1993;41:396-400. [39] Tinetti ME, Ginter SF. Identifying mobility dysfunctions in elderly patients. Jama. 1988; 259:1190-1193. [40] Wade DT. Measurement in Neurological Rehabilitation. Oxford, England: Oxford University Press; 1992. [41] Folstein MF, Folstein SE, McHugh PR. "Mini-mental state": a practical method for grading the cognitive state of patients for the clinician. J Pychiatr Res. 1975;12:189-198. [42] Full-Meyer AR, Jaasko L, Leyman I, et al. The post-stroke hemiplegic patient, I: a method for evaluation of physical performance. Scand J Rehabil Med. 1975;7:13-31. [43] Berg K, Wood-Dauphinee S, Williams JI, Gayton D. Measuring balance in the elderly: preliminary development of an instrument. Physiotherapy Canada. 1989;41:304-311. [44] Berg K, Wood-Dauphinee S, Williams JI, Maki B. Measuring balance in the elderly: validation of an instrument. Can J Public Health, 1992;2:S7-S11. [45] Butland RJA, Pang J, Gross ER, et al. Two-, six-, and twelve-minute walking test in respiratory disease. Br Med J. 1982;284:1604-1608, [46] Mahoney FI, Barthel DW, Functional evaluation: the Barthel index Md State Med J 1965; 14:61-65. [47] Hamilton BB, Laughlin JA, Granger CV, Kayton RM. Interrater agreement of the seven-level Functional Independence Measure (FIM). Arch Phys Med Rehabil 1991;72:790, [48] Schuling J, de Haan R, Limburg M, Groenier KH. The Frenchay activities index: assessment of functional status in stroke patients. Stroke. 1993;24:1173-1177. [49] Bergner M, Bobbitt RA, Carter WB, Gilson BS. The Sickness Impact Profile: development and revision of a health status measure. Med Care, 1981;19:275-289. [50] Stewart A, Ware JE, eds. Measuring Functioning and Well-being, The Medical Outcomes Study Approach. Durham, NC: Duke University Press; 1992.

PW Duncan, PhD, PT, is Associate Director for Research, Center for Health Policy Research and Education, Duke University, and Associate Professor, Graduate Program in Physical Therapy, Duke University. Address all correspondence to Dr Duncan at Center for Health Policy Research and Education, Duke University, Erwin Sq, Ste 230, 2200 W Main St, Durham, NC 27705 (USA).