As the US population ages, more stroke victims will be

Transcription

1 Stroke Location Is Not Associated With Return to Work After First Ischemic Stroke M.A. Wozniak, MD, PhD; S.J. Kittner, MD, MPH; T.R. Price, MD; J.R. Hebel, PhD; M.A. Sloan, MD; J.F. Gardner, SCM Background and Purpose In prior studies, age, race, job category, disability, and cortical functions such as praxis, language, and memory have been associated with vocational outcome, but the influence of stroke location on return to work has never been critically examined. Methods We examined the influence of stroke location on vocational outcome in patients with clinically confirmed acute ischemic stroke from the National Institute of Neurological Disorders and Stroke Stroke Data Bank. Results Of 143 patients working full time at the time of first ischemic stroke, 23 patients were dead and 120 were alive at 1 year. Employment status was known in 109 (mean age, 55 years; 51 [47%] were white, and 82 [75%] were male). Fifty-eight (53%) had returned to work; most (85%) worked full time. Younger age was positively associated with return to work (P 0.05). In an age-adjusted analysis, stroke severity as measured by the Barthel Index 7 to 10 days after stroke was negatively associated with return to work (P 0.001). Higher household income and absence of cortical neurological dysfunction 7 to 10 days after stroke were positively but less strongly associated with return to work (P 0.08). Stroke location, sex, and depression at time of stroke were not associated with vocational outcome. Conclusions Our data suggest that stroke location may be less important than other more easily measured factors in predicting vocational outcome. (Stroke. 1999;30: ) Key Words: cerebral infarction employment stroke, ischemic stroke outcome As the US population ages, more stroke victims will be working at the time of their event. Full Social Security retirement benefits are now payable at age 65 years, but the age for full benefits will increase gradually. People born after 1959 will not be eligible for full benefits until age 67 years. 1 Increasingly, employed patients, their families, employers, and insurers will be questioning physicians about likelihood of employment after ischemic stroke. Return to work after stroke has been little studied, especially in the United States. Simple observations of percentages and demographics of stroke patients who returned to work have been reported from studies primarily focused on other outcomes. 2 6 Although not strictly comparable because of methodological differences, percentages of patients returning to work after stroke range from 21% 7 to 73%. 3 Few studies have attempted to explore the factors influencing return to work after stroke Demographic factors such as age, race, and socioeconomic characteristics have been associated with vocational outcome. These studies have identified specific cortical signs such as apraxia 9 as important predictors of return to work after stroke. However, they did not characterize the anatomic location or vascular territory of the stroke. Subtle abnormalities in language, praxis, and visual fields require a careful neurological examination and may change with time. We hypothesized that a cortical anatomic localization, which underlies these cortical signs, would be a powerful predictor of stroke outcome. To test this hypothesis, we examined data from the National Institute of Neurological Disorders and Stroke (NINDS) Stroke Data Bank, a prospective study of the natural history and prognosis of stroke, which included detailed information on the anatomic location and vascular distribution. Subjects and Methods Subjects The NINDS Stroke Data Bank 11 prospectively collected detailed neurological and functional information on 1200 ischemic stroke patients admitted acutely to 4 hospitals. 11 Patients were examined on admission and again 7 to 10 days after stroke. When higher cortical function was tested, a neurological symptom was assumed absent unless it was specifically reported as present. A cortical deficit was defined as a deficit in at least 1 higher cortical function (neglect, apraxia, homonymous visual field defect, aphasia, or anosognosia). Functional status was measured with the Barthel Index. 12 Patients independent in activities of daily living (Barthel Index 100) were compared with patients with score 100. Depressive symptoms were assessed with the Center for Epidemiologic Studies Depression scale (CES-D), which has been shown to be a valid and reliable Received September 10, 1999; accepted September 10, From the Departments of Neurology (M.A.W., S.J.K., T.R.P., M.A.S., J.F.G.) and Epidemiology (S.J.K., T.R.P., J.R.H., M.A.S.), University of Maryland School of Medicine, Baltimore. Correspondence to Marcella A. Wozniak, MD, PhD, University of Maryland Medical Center, Department of Neurology, 22 South Greene St, Baltimore, MD mwozniak@som.umaryland.edu 1999 American Heart Association, Inc. Stroke is available at

2 Wozniak et al Stroke Location and Work After Ischemic Stroke 2569 Figure 2. Effect of age on percentage returning to work 1 year after stroke. Figure 1. Flow diagram illustrating selection of employed stroke patients from the NINDS Stroke Data Bank. instrument in screening for depression in this stroke population. 13,14 For this analysis, patients were dichotomized into 2 groups on the basis of CES-D score: 1 with a score of 16 (may be depressed) and the other 16 (not likely to be depressed). All information (history, examination, diagnostic testing, neuroimaging) was reviewed by a stroke neurologist. With all these sources of information considered, the specific stroke location was assigned to at least 1 of 16 supratentorial or 6 infratentorial anatomic regions, and the vascular territory of the ischemic stroke was assigned to at least 1 of 50 possible territories. To analyze the potential effect on vocational outcome, specific anatomic and vascular locations were first dichotomized (right supratentorial versus left supratentorial; infratentorial versus supratentorial; anterior circulation versus posterior circulation). Second, strokes were separated into 4 anatomic groups: large cortical ( 1 lobe); small cortical (1 lobe); lacunar; and infratentorial (cerebellar or brain stem). Using the large cortical group as the reference group, we determined the odds of vocational outcome for the other groups. Patients with multiple strokes were excluded from the analysis of stroke location. This analysis was limited to patients working full time at the time of first ischemic stroke (Figure 1). At follow-up, information on employment, residual neurological deficits, and depression was collected. Follow-up contacts were initiated at 3 and 6 months and 1, 2, and 3 years after stroke, although follow-up was not complete at any of these time periods. When the probability of an outcome varies with time as return to work does, 9 the valid analysis of a follow-up study requires either time-to-event methodology 15 or a uniform follow-up time. 15 Since the date of return to work was not available, we defined 2 cohorts with a uniform follow-up time: cohort 1, composed of persons whose return to work status was known at 1 year and cohort 2, composed of persons whose return to work status was unknown at 1 year but known at 2 years. Six patients had no follow-up contact, and 6 had follow-up at 3 or 6 months but not at longer times. Since these patients may or may not have returned to work by 1 or 2 years after stroke, they were excluded from the analysis. Statistical Methods The 2 and t tests were used to test the significance of differences for discrete and continuous factors, respectively. Odds ratios were determined by fitting logistic regression models with return to work as the dependent variable. The probability values derived from the logistic regression analysis were based on the Walds test. 16 Results At the time of first ischemic stroke, 203 patients were employed full time (Figure 1). Mean age was 55.3 years, and there were 136 men (67%). Seventy-seven were white, and 126 were nonwhite (108 black and 18 other nonwhite). Approximately half (51%) of the patients had blue-collar jobs, and half (48%) had white-collar jobs. Three patients had unknown job type. Employed patients had a broad range of education ( high school 40%, high school 30%, high school 30%). Patients employed at the time of first ischemic stroke were significantly younger (55 versus 69 years; P 0.001), more likely to be male (P 0.001), and better educated than patients not working full time. Their racial distribution was similar. At 1 year after cerebral infarction (cohort 1), 23 were known to be dead and 120 were known to be alive. At the 2-year follow-up, 48 patients not contacted at 1 year were interviewed (cohort 2). Patients in cohort 2 were more likely to be nonwhite (P 0.005) and female (P 0.005) than patients in cohort 1. Overall, when cohorts 1 and 2 were combined, data on survival at 1 or 2 years after stroke were available for 191 of 203 patients (94%). In 120 patients known to survive at 1-year follow-up (cohort 1), employment status was known in 109 of 120. Fifty-eight (53%) had returned to work. Of those who had returned to work, most (85%) were working full time. Patients reemployed at follow-up were significantly younger (52 versus 57 years; P 0.05). The probability of returning to work at 1 year decreased as patients passed age 55 years (Figure 2). There was no significant association between return to work and age in patients younger than 54 years, but the number of patients in this age category was small. Since univariate analysis revealed that age was an important factor in return to work, examination of the other factors was performed with the use of an age-adjusted analysis. The age-adjusted associations of demographic variables, aphasia, cortical findings, stroke severity as determined by functional index and motor weakness, depression (CES-D score), stroke location, and job type with return to work are shown in Table 1. Stroke severity as measured by Barthel Index was a strong (P 0.001) negative predictor of return to work. Household income $30 000/y was positively predictive (P 0.02) of return to work. Absence of cortical neurological findings was positively associated with return to work, but this did not reach statistical significance. Race, education, and type of job did not predict return to work at 1 year. Motor weakness,

3 2570 Stroke December 1999 TABLE 1. Univariate Age-Adjusted Analysis of Variables Determined at Acute Stroke and Return to Work at 1 Year Variable No. With Variable/Total Age-Adjusted Odds Ratio P * White race 51/ White collar 51/ NS Income (reference 15K) 15 30K 24/ K 29/ Glasgow Coma Scale score at 7 days (reference not alert 15) Alert 15 92/ NS Motor Strength Score at 7 days 24/ NS (reference full strength) Barthel Index (reference 100) Independent / Aphasia at 7 days 19/ NS Cortical signs at 7 days 34/ Right-sided stroke 40/ NS Supratentorial 86/ NS Anatomic location (reference large cortical) Small cortical 28/ NS Infratentorial 14/ NS Lacune 36/ NS CES-D at 7 days (reference 0 15) 16 15/ NS TACI indicates total anterior circulation infarct; PACI, partial anterior circulation infarct; POCI, posterior circulation infarct; and LACI, lacunar infarct. *NS P 0.2. aphasia, Glasgow Coma Scale score, and depression measured 7 to 10 days after stroke also were not significantly associated with return to work. Return to work did not differ significantly when patients with dichotomized stroke locations were examined (supratentorial versus infratentorial; right supratentorial versus left supratentorial; or anterior circulation versus posterior circulation). Furthermore, a comparison of large cortical to small cortical, lacunar, and infratentorial strokes also revealed no significant association with vocational outcome at 1 year. In the remaining 48 patients known to be alive at 2 years (cohort 2), employment status was known in 47 (Figure 1). A smaller percentage of these patients were working (21 of 48) at 2 years but, similar to cohort 1, most reemployed patients worked full time (18 of 21). Stroke severity as measured by Barthel Index (P ) and any motor weakness (P ) had a strong negative association with return to work (Table 2). Total household income of $30 000/y and normal Glasgow Coma Scale score 7 to 10 days after stroke had a positive association with return to work (P 0.05). Because of the small number of patients, the association of stroke location on employment status was not examined in cohort 2. When age-adjusted odds ratios in the patients from cohort 1 and cohort 2 were compared, congruent associations were seen in Barthel Index score, motor weakness, ethnic class, income, Glasgow Coma Scale score, right-sided stroke, aphasia, and cortical signs (Tables 1 and 2), although not all of these reached statistical significance in either group. Follow-up assessment of some patients included neurological examination and depression scores. In a cross-sectional analysis of patients at 1-year follow-up, persistent motor weakness, aphasia, and cortical dysfunction on examination at 1 year were negatively associated with return to work. Concurrent depression was also more likely in unemployed patients than in those who had returned to work (Table 3). Discussion Factors associated with return to work have been reported in only 3 studies. The North Carolina Comprehensive Stroke Program 6 collected information on stroke patients from 19 counties during Only 15% of patients were working full time at the time of stroke and survived at least 1 year after stroke. At variable follow-up times for 3 to 11 months after stroke, 19% of these survivors had returned to work. Younger, white patients with white-collar jobs and less disabling strokes were more likely to return to work. Side of ischemic infarction was also associated with return to work but differed between whites and blacks. Right hemisphere stroke had a positive association with reemployment in

4 Wozniak et al Stroke Location and Work After Ischemic Stroke 2571 TABLE 2. Univariate Age-Adjusted Analysis of Variables Determined at Acute Stroke and Return to Work at 2 Years Variable No. With Variable/Total Age-Adjusted Odds Ratio P * White race 14/ White collar 24/ NS Income (reference 15K) 15 30K 12/ K 11/ Glasgow Coma Scale score at 7 days (reference not alert 15) Alert 15 37/ Motor Strength Score at 7 days 35/ (reference full strength) Barthel Index (reference 100) Independent / Aphasia at 7 days 6/ NS Cortical signs at 7 days 13/ Right-sided stroke 20/ NS Supratentorial 39/ NS CES-D at 7 days 8/ NS *NS P 0.2. whites and a negative association with reemployment in blacks. Kotila et al 7 described return to work in 58 previously employed stroke patients in a community-based stroke register in Finland in All stroke types (subarachnoid, other hemorrhagic stroke, and ischemic stroke) were included. Baseline measures of strength, memory, and intelligence were obtained, and patients were followed for 12 months. Acute hemiparesis, impaired intelligence, and memory deficits at 3 months were negatively associated with return to work at 1 year. Unlike the North Carolina Study, side of stroke was not associated with return to work. Patients who survived subarachnoid hemorrhage had better outcomes than patients with cerebral infarcts, but this advantage disappeared when adjusted for age. In Japan, Saeki et al 9,10 performed a retrospective cohort analysis of 244 patients admitted with subarachnoid hemorrhage (n 64), hemorrhagic stroke (n 93), or ischemic stroke (n 103). White-collar employment, type of stroke, side of hemiplegia, and absence of hemiparesis or apraxia were associated in univariate analysis with return to work. When data from all stroke types were combined, normal strength, normal praxis, and white-collar employment were associated with improved odds of employment after stroke in a multivariate model. TABLE 3. Univariate Analysis of Variables Determined at 1 Year and Return to Work at 1 Year Variable Age-Adjusted Odds Ratio P Aphasia Cortical dysfunction Motor Strength Score Depression These studies, designed primarily to study other outcomes, have several limitations. First, determination of stroke and stroke subtype varied. Saeki et al 9 used discharge diagnoses to identify acute stroke. The well-documented inaccuracy of these codes may have allowed inclusion of patients with nonacute stroke or patients with nonspecific neurological deficits. 17 In the studies of Howard et al 8 and Kotila et al, 7 stroke diagnosis was confirmed, but the studies were performed before the CT era or in a setting with limited access to CT scan. Stroke type has been shown to influence outcome, including return to work, 7,9,10 suggesting that it is preferable to separate ischemic from hemorrhagic stroke. Second, the Japanese study 9 relied on variable documentation in charts from both the acute and rehabilitation wards of their hospital to assess potential neurological factors. Finally, the 3 prior studies included patients followed for variable periods of time Since patients return to work at different rates over time, 9,10 patients lost to follow-up early might have been working if followed at later times. This could result in misleading results if the factors influencing return to work vary with time and these factors differ systematically between those with a shorter and a longer follow up time. For example, deficits after stroke such as motor weakness will improve over time. If patients without weakness rapidly resumed a normal life and were lost to follow-up and patients with significant weakness were contacted after months of rehabilitation and after return to work, the study might find no effect of weakness on return to work. The present analysis of the NINDS Stroke Data Bank does not suffer from the 3 aforementioned limitations. First, all patients had clinically verified acute ischemic stroke. Almost all patients had a head CT scan to exclude brain hemorrhage. 11 Stroke type and location were determined with the use of standardized criteria. Second, detailed information on

5 2572 Stroke December 1999 absence and presence of cortical deficits was collected prospectively by trained neurologists during the acute stroke period. Finally, results were analyzed in the 2 cohorts that were each followed for a uniform period (cohort 1, 1 year; cohort 2, 2 years). Follow-up information collected 12 to 24 months after stroke was available in 94% of our employed ischemic stroke patients. In addition, a unique strength of our data is the carefully collected and detailed information about stroke location and vascular territory collected prospectively by neurologists with a special interest in stroke. Our findings corroborate prior reports in several respects. Age 8,9 was significantly associated with return to work, and this association was most marked as patients approached traditional retirement age. In patients younger than 55 years, age was not significantly associated with return to work, but power was limited since patient numbers were small. In both groups, functional status was a powerful predictor of return to work. Although white-collar employment and education were not associated, total household income was positively associated, confirming the influence of socioeconomic factors. Surprisingly, level of consciousness, motor strength, and aphasia at the time of stroke were not prospectively significantly associated with return to work. A qualification applicable to each of these examination findings was the time of assessment at 7 days. This assessment time may not be optimal since considerable improvement may occur during the first few weeks. In support of this possibility, other studies have found an association between return to work, impaired memory and intellect, 7 and apraxia 10 when these factors were measured subacutely (3 months or in rehabilitation). A further important factor is that subjects with severe strokes who died or were discharged to nursing homes were excluded from our analyses, thus limiting the impact of these factors, particularly level of consciousness. Analyses of motor function were also constrained by the limited number of persons with weakness at 7 days. Thus, we were not able to examine the possibility that more severe motor disability may be associated with inability to return to work after 1 year. In fact, in cohort 2 (those with 2-year follow-up), in which more people had motor impairment, any motor impairment was strongly predictive of inability to return to work. Analysis from cohorts with 1- and 2-year follow-up did show a nonsignificant trend for aphasia limiting return to work. Again, the small number of aphasic patients, exclusion of most severely affected patients, and time of examination may explain why this association did not reach statistical significance in this analysis. The careful localization of stroke by the Stroke Data Bank allowed testing of the hypothesis that stroke location (supratentorial location versus brain stem; anterior circulation versus posterior circulation; right versus left; large cortical versus small cortical, lacune, and infratentorial) influenced return to work. Although there was a nonsignificant trend for cortical dysfunction to be negatively associated with return to work after stroke, the anatomic location of the stroke alone was not associated with return to work. It seems likely that stroke location affects vocational outcome when it produces persistent cortical deficits. Similarly, depression at the time of acute stroke was not associated with return to work. Depression is negatively associated with return to work after heart attack 18 and interferes with rehabilitation and cognition in stroke patients In the acute stroke period, depression screenings were only obtained in 73 patients, and only a small number of patients were depressed (15 of 74; 20%). The acute stroke period is not the optimal time to assess for depression. Depression occurs during the first year in 30% to 50% of patients, with peak prevalence at 6 months. 22,23 Our data (Table 3) are consistent with the hypothesis that, if measured after acute stroke but before 1 year, aphasia and depression are associated with return to work. As noted earlier, limitations of the present analysis include the fact that neurological examination and anatomic localization of stroke occurred within the first 10 days. Cortical deficits resolving in the first several weeks or months would not be expected to modify long-term outcome. An additional limitation is that, since all cases were drawn from 4 tertiary medical centers in urban areas, they may not be representative of all cases in a defined population or geographic area. However, with regard to our hypothesis regarding anatomic location, this is not likely to be an important limitation. In summary, we find that anatomic location of ischemic stroke is relatively unimportant compared with functional status in predicting return to work. This has enormous implications for predicting and measuring return to work. First, to predict outcome, stroke severity is most usefully quantified in terms of specific functional deficits and not by precise anatomic localization. Second, large amounts of variance are still unexplained. Studies of return to work after myocardial infarct and traumatic injury suggest that factors not related to the severity of physical disease, such as job characteristics, education, personality, and mood disorders, are important in predicting return to work Intrinsic characteristics of the stroke may determine vocational outcome only when the stroke is of devastating severity. Other potentially modifiable factors such as job characteristics, social support, and mood disorders may have great impact in cases with mild or moderate stroke deficits. Acknowledgments This study was supported by National Institutes of Health (NIH) grant K08NS01764 (Dr Wozniak); NIH/NINDS grant P01NS16332 (Drs Price and Kittner); NIH grant RO1DA/NS06625/POINS16332 (Dr Sloan); and Baltimore Veterans Administration Geriatrics Research, Education, and Clinical Center (Dr Kittner). References 1. Social Security Administration. Social Security Retirement Benefits. Baltimore, Md: Social Security Administration; February 1997:5. Publication Kappelle LJ, Adams HP Jr, Heffner ML, Torner JC, Gomez F, Biller J. Prognosis of young adults with ischemic stroke: a long-term follow-up study assessing recurrent vascular events and functional outcome in the Iowa Registry of Stroke in Young Adults. Stroke. 1994;25: Ferro JM, Crespo M. Prognosis after transient ischemic attack and ischemic stroke in young adults. Stroke. 1994;25: Mackay A, Nias BC. Strokes in the young and middle-aged: consequences to the family and to society. J R Coll Physicians Lond. 1979; 13:

6 Wozniak et al Stroke Location and Work After Ischemic Stroke Angeleri F, Angeleri VA, Foschi N, Giaquinto S, Nolfe G. The influence of depression, social activity, and family stress on functional outcome after stroke. Stroke. 1993;24: Heinemann AW, Roth EJ, Cichowski K, Betts HB. Multivariate analysis of improvement and outcome following stroke rehabilitation. Arch Neurol. 1987;44: Kotila M, Waltimo O, Niemi ML, Laaksonen R, Lempinen M. The profile of recovery from stroke and factors influencing outcome. Stroke. 1984; 15: Howard G, Till JS, Toole JF, Matthews C, Truscott BL. Factors influencing return to work following cerebral infarction. JAMA. 1985;253: Saeki S, Ogata H, Okubo T, Takahashi K, Hoshuyama T. Factors influencing return to work after stroke in Japan. Stroke. 1993;24: Saeki S, Ogata H, Okubo T, Takahashi K, Hoshuyama T. Return to work after stroke: a follow-up study. Stroke. 1995;26: Foulkes MA, Wolf PA, Price TR, Mohr JP, Hier DB. The Stroke Data Bank: design, methods, and baseline characteristics. Stroke. 1988;19: Shinar D, Gross CR, Bronstein KS, Licata-Gehr EE, Eden DT, Cabrera AR, Fishman IG, Roth AA, Barwick JA, Kunitz SC. Reliability of the activities of daily living scale and its use in telephone interview. Arch Phys Med Rehabil. 1987;68: Radloff LS. The CES-D scale: a self-report depression scale for researching the general population. J Appl Psychol Meas. 1977;1: Parikh RM, Robinson RG, Lipsey JR, Starkstein SE, Fedoroff P, Price TR. The impact of poststroke depression on recovery in activities of daily living on a 2 year follow-up. Arch Neurol. 1990;47: Rothman KJ. Modern Epidemiology. Boston, Mass: Little, Brown & Co; Rao CR. Linear Statistical Inference and Its Application. 2nd ed. New York, NY: John Wiley & Sons, Inc; Leibson CL, Naessens JM, Brown RD, Whisnant JP. Accuracy of hospital discharge abstracts for identifying stroke. Stroke. 1994;25: Lane RM, Sweeney M, Henry JA. Pharmacotherapy of the depressed patient with cardiovascular and/or cerebrovascular illness. Br J Clin Pract. 1994;48: Morris PLP, Raphael B, Robinson RG. Clinical depression is associated with impaired recovery from stroke. Med J Aust. 1992;157: Morris PLP, Robinson RG, Andrezejewski P, Samuels J, Price TR. Association of depression with ten-year post-stroke mortality. Am J Psychiatry. 1993;150: Robinson RG, Bolla-Wilson K, Kaplan E, Lipsey JR, Price TR. Depression influences intellectual impairment in stroke patients. Br J Psychiatry. 1986;148: Astrom M, Adolfsson R, Asplund K. Major depression in stroke patients: a 3 year longitudinal study. Stroke. 1993;24: Robinson RG, Starr LB, Price TR. A two year longitudinal study of mood disorders following stroke: prevalence and duration at six months follow-up. Br J Psychol. 1984;144: Riegel BJ. Contributors to cardiac invalidism after acute myocardial infarction. Coron Artery Dis. 1993;4: Abbott J, Berry N. Return to work during the year following first myocardial infarction. Br J Clin Pract. 1991;30: Boudrez H, De Backer G, Comhaire B. Return to work after myocardial infarction: results of a longitudinal population-based study. Eur Heart J. 1994;15:32 36.