Health Status of Individuals With Mild Stroke
Background and Purpose Diminished quality of life and limitations in higher levels of physical functioning are often underestimated in stroke and are not fully captured by measures such as the Barthel Index and the Rankin Outcome Scale. This study used additional measures to assess the health status of 304 persons with mild stroke and to compare these individuals with 184 persons with transient ischemic attack and 654 persons without history of stroke/transient ischemic attack but at elevated risk for stroke (asymptomatic group).
Methods Subjects were recruited from the Academic Medical Center Consortium (inpatients), the Cardiovascular Health Study (population-based sample of community-dwelling persons 65 years and older), and United HealthCare (inpatients and outpatients typically younger than 65 years). Subjects were interviewed by telephone or in person to assess activities of daily living (Barthel Index), depression (Center for Epidemiological Studies Depression Scale), health status (MOS-36), and utility for current health state.
Results Most respondents were independent on all Barthel items. The stroke group was more impaired on the MOS-36 than the asymptomatic group but similar to the group with transient ischemic attack. Health-related quality of life was lowest for persons with stroke. While symptom status and Barthel Index score were the strongest predictors of health status, the Barthel Index showed a consistent ceiling effect when compared with the physical function subscale of the MOS-36.
Conclusions The consequences of even mild stroke affect all dimensions of health except pain. Standardized assessment of persons with stroke must evaluate across the entire continuum of health-related functions.
Three hundred fifty thousand Americans survive a stroke each year, and nearly three million have some degree of stroke-related disability.1 The majority of individuals who survive a stroke have minimal to moderate neurological deficits.2 3 These persons often achieve independence in basic activities of daily living (eg, ambulation, self-care, and continence),4 yet they still report increased dependency in more advanced activities (eg, house management and leisure activities) and diminished quality of life. Depression, which is as strongly related to quality of life as physical dependency,5 has an estimated prevalence of 23% to 63% among survivors of stroke.6 7 8 9
Two of the most prominent outcome measures used in stroke research are the Barthel Index10 and the Rankin Outcome Scale.11 While both measures capture important dimensions of stroke-related disability, each also has definite limitations. The Barthel Index assesses basic activities of daily living yet demonstrates a ceiling effect in individuals with mild or moderate stroke.12 The Rankin scale is a measure of physical ability but does not assess an individual’s psychosocial condition.11 Thus, use of the Barthel and Rankin indices emphasizes basic physical limitations but at the cost of overlooking higher levels of physical functioning and quality of life.
The purpose of this study was to use a broad array of measures to assess the health status of persons with mild stroke deficits. In addition to the Barthel and Rankin scales, measures included a multidimensional health status measure (MOS-36),13 the CESD,14 and a TTO. Comparison groups included persons who had experienced a TIA as well as persons who had no history of TIA/stroke yet were at elevated risk for stroke because of conditions such as atrial fibrillation, hypertension, and/or heart disease.
Subjects and Methods
Subjects were recruited from three sources: the AMCC, UHC, and CHS.
The AMCC identified patients from the 1992 calendar year administrative files of five academic medical centers: Alton Ochsner Medical Foundation (New Orleans, La); Massachusetts General Hospital (Boston); University of California at Los Angeles Medical Center; University of Iowa Hospitals and Clinics (Iowa City); and University of Pennsylvania Health Systems (Philadelphia). AMCC subjects were identified from inpatient records and included approximately equal representation of persons younger than 65 years and aged 65 years and older.
Potential AMCC subjects were identified as having one or more ICD-9 codes indicating high risk for stroke (including codes for TIA, stroke, aortic and mitral valve diseases, arrhythmias, and stenoses of the precerebral vessels). For stroke patients, the selection criteria were targeted toward records with ICD-9 codes of 433, 434, and 436 (ie, ischemic stroke). ICD-9 codes were verified by a chart-based medical record review performed by trained abstractors. To ensure comparable numbers of TIA and stroke patients, we sampled 59% of stroke patients, 53% of TIA patients, and 3% of asymptomatic patients. Potential subjects were sent a letter asking them to return a signed consent form indicating their interest in participating in the study. Nonrespondents were sent a follow-up letter but, by local restriction, were not contacted by telephone. Of 1412 eligible patients, 613 (43%) completed the survey. Interviews took place by telephone with trained interviewers under the direction of the Research Triangle Institute.
UHC identified patients from the 1992 calendar year claims files of five Independent Practice Association model health plans: United HealthCare of Georgia (Atlanta); PHP of Michigan (Lansing); PrimeCare Health Plan (Milwaukee, Wis); United Health Plans of New England (Providence, RI); and United HealthCare of Utah (Salt Lake City). Most patients were younger than 65 years (reflecting the relatively younger population served by this managed care organization).
As with the AMCC sites, potential UHC subjects were identified as having one or more of the ICD-9 codes indicating high risk for stroke, and ICD-9 codes were verified by a chart-based review of the medical record performed by trained abstractors. To ensure sufficient numbers of TIA and stroke patients, we sampled 68% of stroke patients, 62% of TIA patients, and 21% of asymptomatic patients. After selection, patients were sent a letter asking them to return a signed consent form indicating their interest in participating in the study. Nonrespondents were invited to participate by telephone and asked to give oral consent. Of 478 eligible patients, 319 (67%) completed the survey. Interviews took place by telephone with trained interviewers under the direction of the Research Triangle Institute.
The Bowman Gray site of the CHS identified a community-based sample of persons aged 65 years and older.15 CHS patients were eligible if they were currently enrolled in the CHS study; resided in Forsyth County, North Carolina; and had carotid bruit, carotid artery occlusion, carotid artery stenosis exceeding 25%, atrial fibrillation on a 12-lead electrocardiogram, history of TIA, or history of stroke. The Bowman Gray School of Medicine assigned a single interviewer to conduct the survey in the subjects’ homes. Of the 357 persons from the Bowman Gray site of the CHS available for interview, 321 (90% of eligible patients) completed the survey. Cerebrovascular symptom status (ie, stroke, TIA, asymptomatic) for these 321 patients was based on CHS’s baseline examination supplemented by information from prospective follow-up.
The CHS site used in-person interviews instead of the telephone interviews used at the other sites. To ensure that responses from the two interview methods were comparable, telephone interviews were conducted on a 10% sample of CHS respondents. Responses did not differ significantly between telephone and in-person interviews (data not shown).
The survey instrument included questions about demographics, health status, functional status, cerebrovascular symptoms, and utilities.
Functional and health status measures included the Barthel Index,12 the MOS-36,13 and the short form of the CESD.14 The Barthel Index assessed the subjects’ ability to perform basic physical activities of daily living. The eight subscales of the MOS-36 assessed general health, mental health, emotional role, physical role, social function, vitality, bodily pain, and physical function. The CESD assessed depressive symptomatology.
We also solicited utilities (ie, quantitative representations of the relative desirability or undesirability of health states) regarding the individual’s current health state. We measured utilities using the TTO and scale (global) approaches.16 In the TTO, respondents were presented with trade-offs between time in their current health state and time in excellent health. For example, individuals were asked if they would prefer to live 10 years in their current state of health before dying or 9 years in excellent health; in other words, would they trade off 1 year of life in their current state for 9 years of excellent health? The questioning was continued until the subject reached a point of indifference. The TTO was placed on a scale of 0 to 1 by taking the number of years in excellent health at the point of indifference and dividing this number by 10. Respondents were also asked to rate their current quality of life on a scale of 0 to 100, with 0 being death and 100 being a state of excellent health. This global measure of utility was placed on a scale of 0 to 1 by dividing by 100.
Categorical variables were presented as frequencies, and continuous variables were presented as means and standard deviations. Groups were compared with χ2 statistics for categorical variables and ANOVA for continuous variables. Post hoc analysis assessed pairwise differences between (1) asymptomatic patients and stroke patients, (2) stroke patients and TIA patients, and (3) asymptomatic patients and TIA patients. Regression analysis was used to determine whether patient group (stroke, TIA, asymptomatic), other comorbid diseases, and/or Barthel scores were predictive of responses to any of the eight different domains of the MOS-36 and/or the CESD.
Because of the time lag between the medical record review and the actual interview, the study methodology included a classification algorithm that was intended to reconcile discrepancies between the two sources. For example, an individual with no stroke on the medical record would be classified as having had a stroke if he/she reported a stroke and/or stroke-related disabilities during the interview (under the presumption that the stroke had occurred after the medical record review). Other reports from this study have accepted this classification and included all 1253 respondents.17 However, this report excludes 111 patients whose symptom status was based on self-report of stroke but who did not report significant stroke-related disabilities during the interview.
We interviewed 654 asymptomatic individuals, 184 individuals with a history of TIA, and 304 individuals with a prior stroke. Individuals with stroke were interviewed on average 595 (SD, 116) days after stroke; most had had mild strokes.
Based on data from the medical record review, 67% of the respondents with stroke had Rankin scores of 0 or 1; 5% had Rankin scores of 2; 11% had Rankin scores of 3; 15% had Rankin scores of 4; and 2% had Rankin scores of 5.
Table 1⇓ describes selected sociodemographic characteristics of the respondents. Most individuals were white, and slightly more than one half were male. The mean age was approximately 65 years. The three groups were similar in terms of sex, education, and living arrangements. They were notably dissimilar in three areas: (1) The asymptomatic group was older and included a relatively higher percentage of whites; (2) individuals with stroke were less likely to be employed; and (3) the TIA group had relatively less social support.
Table 2⇓ describes the distribution of other comorbid diseases. The stroke and TIA groups were more likely to have diabetes, ischemic heart disease, peripheral vascular disease, chronic obstructive pulmonary disease, renal disease, and rheumatoid arthritis. The asymptomatic and TIA groups had a higher prevalence of cardiac dysrhythmias (which had been part of the selection criteria for the asymptomatic subjects).
Table 3⇓ describes health status. All groups were highly independent in physical activities of daily living: 83% of the asymptomatic individuals, 81% of the individuals with TIA, and 66% of the stroke survivors scored 100 on the Barthel Index. The stroke group was significantly more impaired than the asymptomatic group in every dimension of the MOS-36 except pain. Both individuals with stroke and those with TIA had higher depression ratings on the CESD than did the asymptomatic group. The CESD scores were strongly correlated with MOS-36 mental health score (−0.65) and emotional role function (−0.52). Individuals with TIA were similar to those with mild stroke, except that they were less impaired in physical function and physical roles and reported a higher incidence of pain. Health-related quality of life, as measured by the TTO and global utility indices, was lower for individuals with stroke. The asymptomatic and TIA groups did not differ in health-related quality of life.
Table 4⇓ reports on health status among those respondents who scored 100 on the Barthel Index. Patients in the stroke and TIA groups continued to be significantly more impaired than patients in the asymptomatic group in every dimension of the MOS-36 except pain. There were no significant differences among the groups with respect to utility measures.
Table 5⇓ describes the regression analyses (which incorporated the entire sample of patients). In these analyses, the Barthel Index and a history of stroke were the strongest and most consistent predictors of health status. The regression models accounted for between 6% and 36% of the variance in the health status measures, with the model being most predictive of physical function.
The consequences of mild stroke affect all dimensions of health except pain. Persons with TIA are similarly affected and report health status significantly below that of asymptomatic individuals and not greatly different from that of persons with mild stroke. The major difference between patients with mild stroke and patients with TIA is that those with mild stroke are significantly more impaired in physical functioning and work or leisure activities (physical roles).
Multivariate regression analysis served to confirm the above findings. In addition to symptom status (stroke, TIA, asymptomatic), we found that the Barthel Index was a consistent predictor of health status, including depression. Depression, which has been reported to be highly prevalent in stroke survivors,18 19 may be secondary to loss of abilities rather than a primary manifestation of stroke.
Our study found that the health status of individuals with mild stroke is below the 50th percentile of a norm-based group of men and women aged 65 to 74 years.20 This finding is consistent with other long-term follow-up studies of stroke survivors. Survivors of stroke in the Framingham Study were significantly more disabled than matched control subjects.2 None of the respondents in a 4-year study of stroke survivors reported that their life was back to normal.21
Observed differences among the stroke, TIA, and asymptomatic groups were largest for physical functioning and physical role (limitations in work or leisure activities due to physical health). These findings persisted even when the analysis was limited to patients achieving full independence in basic activities of daily living (Barthel score of 100). In the elderly, multiple impairments, even though mild, may contribute to frailty22 and may put individuals with even the mildest strokes at higher risk for future institutionalization.
Importantly, these findings demonstrate that standardized assessment of individuals with stroke must evaluate across the entire continuum of health-related functions. Measures such as the Barthel Index, which has a ceiling effect and which captures only physical domains of health status, are not adequate for assessing the full impact of stroke-related disability. As Table 6⇓ illustrates, patients who score 100 on the Barthel Index have widely varying scores on the physical function subscale of the MOS-36; for example, fewer than 20% scored the maximum possible value on this subscale. Thus, if the Barthel Index is the only stroke outcome measure used, a decline in many domains of health status will be missed. The Barthel Index will also be ineffective in detecting the psychosocial dimensions of impaired function. Other health status measures (eg, MOS-35,13 the Sickness Impact Profile,23 and Reintegration to Normal Living Index24 ) should be used in addition to the Barthel Index when stroke outcomes are assessed.
Limitations of the study design merit attention. First, contact with individuals was limited for the most part to telephone interviews, a limitation that obviously omitted individuals with language or cognitive dysfunction and could have preferentially selected the fitter of stroke survivors. Second, the assessment of comorbid disease and Barthel Index took place from a medical record review that preceded the date of the interview by a number of months.
Despite these limitations, the study does show that even the fittest stroke survivors tend to have significant impairments in health status. It also shows that stroke and TIA are strong predictors of health status. This finding remains true for stroke even among those survivors who have achieved significant recovery of function. Our results further indicated that the major outcome measure for stroke, the Barthel Index, does not capture important declines in physical or psychosocial function in individuals with mild stroke.
Selected Abbreviations and Acronyms
|AMCC||=||Academic Medical Center Consortium|
|CESD||=||Center for Epidemiological Studies Depression Scale|
|CHS||=||Cardiovascular Health Study|
|ICD-9||=||International Classification of Diseases, 9th Revision|
|TIA||=||transient ischemic attack|
|TTO||=||time trade-off utility|
This study was performed as part of the Stroke Prevention Patient Outcomes Research Team (PORT) and was funded through contract 282-91-0028 from the US Agency for Health Care Policy and Research.
- Received November 1, 1996.
- Revision received December 26, 1996.
- Accepted December 26, 1996.
- Copyright © 1997 by American Heart Association
American Heart Association. Heart and Stroke Facts. Dallas, Tex: American Heart Association; 1991.
Wolf PA, D’Agostino RB, O’Neal MA, Sytkowski P, Kase CS, Belanger A, Kannel WB. Secular trends in stroke incidence and mortality: the Framingham Study. Stroke. 1992;23:1551-1555.
Ahlsio B, Britton M, Murray V, Theorell T. Disablement and quality of life after stroke. Stroke. 1984;15:886-890.
Kelly-Hayes M, Paige C. Assessment and psychologic factors in stroke rehabilitation. Neurology. 1995;S29-S32.
Samsa GP, Matchar DB, Bonito A, Duncan P, Lipscomb J, Enarson C, Witter D, Venus P, Paul J, Weinberger M. Quality of life with stroke: results from a nationally diverse survey of persons at increased risk for stroke. Presented at 1996 Annual National Meeting of the Society for General Internal Medicine; May 3, 1996; Washington, DC.
Parikh R, Lipsey JR, Robinson R, Price TR. Two year long study of post-stroke mood disorders: dynamic changes in correlates of depression at one and two years. Stroke. 1987;18:579-584.
Ware JE. SF-36 Health Survey: Manual and Interpretation Guide. Boston, Mass: The Health Institute, New England Medical Center; 1993.
Becker G. Continuity after a stroke: implications of life-course disruption in old age. Gerontology. 1993;33:148-158.