Clinical Characteristics and Rehabilitation Outcomes of Patients With Posterior Cerebral Artery Stroke

2138

ORIGINAL ARTICLE

Clinical Characteristics and Rehabilitation Outcomes of Patients With Posterior Cerebral Artery Stroke Yee Sien Ng, MD, MRCP, Joel Stein, MD, Sara S. Salles, DO, Randie M. Black-Schaffer, MD, MA ABSTRACT. Ng YS, Stein J, Salles SS, Black-Schaffer RM. Clinical characteristics and rehabilitation outcomes of patients with posterior cerebral artery stroke. Arch Phys Med Rehabil 2005;86:2138-43. Objectives: To describe the demographics, clinical profile, and functional outcomes in posterior cerebral artery (PCA) stroke and to identify factors associated with functional change during rehabilitation and discharge disposition. Design: Retrospective study of patients with PCA stroke admitted to a rehabilitation hospital over an 8-year period. Setting: Free-standing urban rehabilitation hospital in the United States. Participants: Eighty-nine consecutive patients with PCA stroke (48 men, 41 women; mean age, 71.5y) met inclusion criteria. Intervention: Inpatient multidisciplinary comprehensive rehabilitation program. Main Outcome Measures: Demographic, clinical, and discharge disposition information were collected. Functional status was measured using the FIM instrument, recorded at admission and discharge. The main outcome measures were the discharge total FIM score, the change in total FIM score (⌬FIM), and the discharge disposition. Multiple and logistic regression analyses were performed to identify factors associated with the main outcome measures. Results: The most common impairments were motor paresis (65%), followed by visual field defects (54%) and confusion or agitation (43%). The mean discharge total FIM score ⫾ standard deviation was 88.3⫾28.2. The mean ⌬FIM was 23.3⫾16.4. Fifty-five (62%) patients were discharged home. On multiple regression analysis, higher admission total FIM score, longer length of stay (LOS), and a rehabilitation stay free of interruptions were associated with higher discharge total FIM score and greater ⌬FIM. Absence of diabetes mellitus and younger age were also associated with higher discharge total FIM scores, and male sex had greater ⌬FIM. On logistic regression analysis, younger patients, higher discharge FIM scores, presence of a caregiver, and the nonnecessity for 24hour support were associated with a discharge to home. Conclusions: Motor, visual, and cognitive impairments are common in PCA stroke, and good functional gains are achievable after comprehensive rehabilitation. Higher admission FIM

From the Department of Physical Medicine and Rehabilitation, Spaulding Rehabilitation Hospital, Harvard Medical School, Boston, MA (Ng, Stein, Salles, BlackSchaffer); the Department of Rehabilitation Medicine, Singapore General Hospital, Republic of Singapore (Ng); and the Department of Physical Medicine and Rehabilitation, University of Kentucky; Lexington, KY (Salles). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Correspondence to Yee Sien Ng, MD, MRCP, Dept of Rehabilitation Medicine, Singapore General Hospital, Block 6, Level 9, Outram Rd, Singapore 169608, Republic of Singapore, e-mail: [email protected]. Reprints are not available from the author. 0003-9993/05/8611-10092$30.00/0 doi:10.1016/j.apmr.2005.07.289

Arch Phys Med Rehabil Vol 86, November 2005

scores, longer LOS, and younger and male patients were associated with better functional outcomes. Most patients were discharged home, particularly those with caregivers and those for whom 24-hour support was not required. Further research should aim at the development of functional outcome measures of greater breadth and sensitivity to visual and cognitive deficits and should compare PCA stroke outcomes with outcomes of strokes in other vascular territories. Key Words: Posterior cerebral artery; Rehabilitation; Stroke. © 2005 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation TROKE IN THE TERRITORY OF the posterior cerebral artery (PCA) accounts for 5% to 10% of all strokes, with S actual rates depending partly on the inclusion criteria used to

define the brain territory supplied by the PCA.1,2 Recent studies have focused on elucidating the etiology of PCA stroke. There is currently a paradigm shift from the earlier concept that PCA stroke is due to hemodynamic compromise with hypoperfusion to the current belief that it is more often embolic in origin. The etiology remains elusive in about one quarter of cases.1,3 A wide variety of clinical signs due to PCA stroke have been described.1,2,4 Sensory, motor, balance, and visual field deficits are common. Unusual and often confusing visual cognitive impairments occur. These include optic apraxia (inability to direct gaze to an object of interest), optic ataxia (inability to bring the hand to an object under visual guidance), simultagnosia (ability to see part of a scene in detail but not the whole), achromotopsia (color blindness), palinopsia (the illusionary visual spread of moving objects), and prosopagnosia (inability to recognize faces).1-6 Aphasia, alexia with or without agraphia, and acalculia can occur.2,7,8 Neuropsychiatric manifestations, specifically hallucinations, confusion, and agitated delirium, are common and may be the only presenting signs.9,10 These variable clinical presentations can lead to diagnostic difficulty, especially in the elderly.10 The middle cerebral artery stroke syndrome of unilateral hemiparesis, hemisensory loss, hemianopia, hemispatial neglect, and aphasia can be mimicked by and confused with PCA stroke.11 Neuroanatomic-clinical correlations suggest that the variety of presenting symptoms and signs in PCA stroke are due to the multiple distinct structures this artery supplies. These include the thalamic nuclei, limbic system, posterior limb of the internal capsule, cerebral peduncles, and mesencephalon, as well as portions of the parietal, temporal, and occipital cortex.6,12 The recognition and management of the wide range of impairments caused by PCA stroke can be challenging to professionals at all levels of the continuum of care. There is little research on the clinical course and functional outcome of rehabilitation in PCA stroke. The lack of data reports most likely has occurred because PCA stroke is relatively uncommon; the largest reported series have included just 79 to 127 patients.2,3,13 In addition, it is widely assumed that

2139

REHABILITATION OUTCOMES OF PCA STROKES, Ng Table 1: Cerebrovascular Risk Factors in Patients With PCA Stroke Risk Factor

n

%

Hypertension Coronary artery disease Diabetes mellitus Atrial fibrillation Smoking Migraine Others* No risk factors 1 risk factor 2 risk factors 3 risk factors 4 risk factors

61 31 27 21 18 2 7 9 18 39 18 5

68.5 34.8 30.3 23.6 20.2 2.2 7.9 10.1 20.2 43.8 20.2 5.6

*Others include 3 patients on stroke-associated pharmacologic agents and 1 case each of arteriovenous malformation, hypercoagulable state, collagen vascular disease, and infection-related vasculitis.

few of these patients experience functional disability.2 In this study, we describe the demographic data, clinical profile, and functional outcomes in a series of patients with PCA stroke admitted to the Spaulding Rehabilitation Hospital, Boston, MA. In addition, we aim to identify factors associated with functional change during rehabilitation and with discharge disposition in this cohort. METHODS Stroke was defined based on clinical features consistent with stroke and supported by computed tomography (CT) or magnetic resonance imaging (MRI) findings. Our rehabilitation hospital has a dedicated stroke rehabilitation unit, which admits stroke patients across a wide range of severities. Admission criteria to the stroke rehabilitation unit include (1) recent hemorrhagic or ischemic stroke, (2) resultant significant impairment and disability that may benefit from inpatient rehabilitation, and (3) sufficient medical stability allowing for continued care in a stroke rehabilitation setting. The hospital’s nurse liaisons apply these admission criteria and coordinate transfer of stroke rehabilitation candidates from referring acute hospitals. All patients with stroke admitted to the stroke rehabilitation unit were included in the hospital’s stroke registry. We reviewed the stroke registry for the 8-year period from January 1, 1996, to December 31, 2003. There were 1552 stroke patients during this period, of whom 89 (5.7%) had ischemic or hemorrhagic PCA stroke. There were 2 deaths in the PCA cohort, and clinical and functional data for these patients until time of

Table 2: Clinical Signs in Patients With PCA Strokes Clinical Sign

n, %

Hemiparesis Visual field defect Confusion or agitation Aphasia Dysarthria Visual neglect Motor ataxia Sensory loss Visual agnosia

58, 65.2 48, 53.9 38, 42.7 22, 24.7 18, 20.2 17, 19.1 13, 14.6 9, 10.1 2, 2.2

Table 3: Admission and Discharge FIM Scores of All PCA Strokes (Nⴝ89) FIM Score

Admission

Discharge

P*

Total FIM score 65.0⫾25.3 (18–114) 88.3⫾28.2 (22–126) ⬍.001 FIM motor score 43.3⫾18.5 (13–82) 63.4⫾21.9 (13–91) ⬍.001 FIM cognitive score 21.7⫾8.9 (5–35) 24.9⫾8.1 (7–35) ⬍.001 NOTE. Values are mean ⫾ SD (range). *Paired t test.

death were included in the analyses. All patients with PCA stroke had CT imaging and 56 patients (63%) had MRI studies. Lesion location was determined by review of CT and MRI reports. We further categorized patients into superficial PCA (PCA-S), superficial and deep PCA (PCA-SD), and PCA-plus (PCA-P) strokes.1-3,5 Patients with PCA-S strokes had lesions in 1 or more cortical territories supplied by the PCA (ie, occipital, parietooccipital, medial-inferior temporal lobes). Patients with PCA-SD strokes had, in addition, involvement of deep noncortical territories supplied by the PCA (eg, thalamus, mesencephalon). Finally, patients with PCA-P strokes had lesions in territories not supplied by the PCA (eg, cerebellum, frontal lobe) in addition to PCA-S or PCA-SD involvement. The rationale for this classification in prior studies was to examine if the etiologies in these subcategories were different.1-3,5 We classified our cohort similarly to evaluate whether functional outcomes in these subcategories are different. We then examined patients’ medical records to verify and record further clinical characteristics, lesion characteristics, sociodemographic data, and functional outcomes. Clinical characteristics recorded included hemiparesis, visual field deficit, confusion or agitation, aphasia, dysarthria, visual neglect, motor ataxia, sensory loss, and visual agnosia.1-6 Cardiovascular risk factors documented included hypertension, diabetes mellitus, smoking, coronary artery disease, and atrial fibrillation. Lesion characteristics recorded were lesion location, laterality, and stroke type (hemorrhagic vs ischemic). Sociodemographic variables documented were age, sex, race and ethnicity, presence of a fulltime caregiver at home, and need for 24-hour supervision on discharge. Discharge destinations were to home, to a skilled nursing facility (SNF), or back to an acute unit. Home discharge included discharge to a patient’s own or a family member’s home. Death during rehabilitation was recorded separately as a discharge other than home. Length of stay (LOS) in rehabilitation was recorded, and an interrupted rehabilitation stay, whereby patients required transfer back to an acute unit for emergent medical or surgical conditions, was also noted. The functional outcome measures used were the discharge FIM instrument, total FIM score change (⌬FIM), FIM efficiency, and home discharge. The FIM instrument is a widely used general disability scale, consisting of 13 motor and 5 cognitive items, with established content validity, construct validity, sensitivity, and interrater reliability for the measurement of functional ability.14,15 The ⌬FIM is the difference between the discharge and admission total FIM scores. The FIM efficiency is the ⌬FIM per day of stay in rehabilitation and measures rate of functional improvement. Spaulding Rehabilitation Hospital is a FIM-credentialed facility. All physicians, nurses, or therapists performing FIM assessments are trained and credentialed in FIM scoring. FIM assessments and scoring were performed for all patients at rehabilitation admission and discharge. Arch Phys Med Rehabil Vol 86, November 2005

2140

REHABILITATION OUTCOMES OF PCA STROKES, Ng Table 4: Comparison Among PCA Stroke Subcategories Group Variable

PCA-S (n⫽48)

PCA-SD (n⫽22)

PCA-P (n⫽19)

P*

Age Admission total FIM score Admission FIM motor score Admission FIM cognitive score Discharge total FIM score Discharge FIM motor score Discharge FIM cognitive score ⌬FIM FIM efficiency

73.7⫾10.9 58.8⫾24.7 39.3⫾17.9 19.5⫾8.7 80.6⫾29.6 57.6⫾23.3 23.0⫾8.1 21.8⫾15.6 1.057⫾1.002

68.7⫾16.5 72.7⫾25.8 48.4⫾19.0 24.4⫾8.9 95.6⫾24.9 68.6⫾18.5 27.1⫾8.3 22.9⫾12.9 1.155⫾0.725

69.3⫾12.7 67.2⫾26.3 44.5⫾19.4 22.7⫾9.4 93.4⫾29.2 67.7⫾21.8 25.7⫾8.5 25.9⫾15.8 1.117⫾0.851

.23 .09 .15 .08 .07 .08 .13 .61 .91

NOTE. Values are mean ⫾ SD. *Analysis of variance.

All patients received a comprehensive rehabilitation program, including medical and nursing management and an average of 3 hours a day of physical, occupational, and speech and language therapies. Weekly multidisciplinary staff meetings were conducted to assess progress, review functional goals and plan further therapies. Family case conferences were conducted to educate family members and plan patients’ aftercare needs. This study was approved by the hospital’s institutional review board. Statistical Analysis Data were collected in Access 2000a for Windows and analyzed using SPSS, version 10.0,b for Windows. Means and standard deviations (SDs) or counts and percentages were used for descriptive statistics. The Student t test and analysis of variance were performed to assess differences in group means. The chi-square test was used to evaluate differences in group proportions. In the regression models, the regression coefficients and standard error of the mean are reported. We performed multiple linear regression analysis to identify independent clinical variables associated with discharge FIM score. Independent variables were chosen based on clinical judgment and prior literature review. These variables were age, sex, year of admission, LOS, cerebrovascular risk factors, stroke laterality, stroke type (infarct or hemorrhage), visual impairment, admission total FIM score, and an interrupted rehabilitation program. These variables were then entered simultaneously into the linear regression model. Only significant associations, defined as P values of less than .05, were reported in the final model. Multiple linear regression analysis with the same independent variables listed above was also performed to identify significant associations with ⌬FIM. In both these linear models, assessments for violation of assumptions were made, including analyses of normality of the residuals and linearity of the continuous variables. The calculation of P values and confidence intervals in linear regression assumes that the residuals are normally distributed, and this was tested using the Shapiro-Wilk statistic.16 In addition, variables were mathematically transformed when linearity was violated. This occurred for admission total FIM score as an independent variable in the regression on discharge total FIM scores, where a quadratic element was added to address this violation. The adjusted R2 was calculated for these models to assess whether they were good predictors of the dependent variables. In addition, logistic regression analysis was performed to identify variables associated with home discharge. Patients Arch Phys Med Rehabil Vol 86, November 2005

who were placed in an SNF, transferred back to an acute unit, or died were coded as a discharge other than home. The independent variables entered into this model included age, sex, presence of a caregiver, need for 24-hour supervision, and total discharge FIM scores. The Hosmer-Lemeshow test was used to assess for goodness of fit of the logistic regression. RESULTS Forty-eight (54%) of the PCA stroke patients were men. The mean age ⫾ SD was 71.5⫾13.1 years. Seventy-four (83%) were white, 6 (7%) were African American, and the rest were of other races. In this entire cohort of PCA strokes, 48 (54%) had PCA-S, 22 (25%) had PCA-SD, and 19 (21%) had PCA-P strokes. The thalamus (23 patients [26%]) was the most frequently associated lesion outside the PCA-S stroke territory, followed by the cerebellum (14 patients [16%]). Twenty-seven (32%) patients had right-sided cerebral involvement, in 28 (43%) patients it was left sided, and the rest had bilateral PCA strokes. Most strokes (81 patients [91%]) were ischemic. Frequencies of cerebrovascular risk factors are described in table 1. The most common risk factor was hypertension (69%). Most patients (90%) had at least 1 risk factor, and more than two thirds (70%) had 2 or more. Table 2 lists the commonly encountered neurologic examination findings. The most common impairments in our series were motor paresis (65%), visual field defects (54%), and confusion or agitation (43%). The mean acute hospital and rehabilitation hospital LOSs were 11.4⫾8.8 and 29.2⫾20.9 days, respectively. The mean admission total FIM score was 65.0⫾25.3, and the mean discharge FIM score was 88.3⫾28.2. This gain in FIM score was highly significant (P⬍.001). Highly significant gains in the motor and cognitive FIM subscores were also achieved (table 3). Fiftythree patients (60%) had discharge total FIM scores of 90 or more (indicating modified independence or better), and 64 patients (72%) had discharge FIM scores of 72 or more (indicating minimal assistance or better). The mean ⌬FIM was 23.3⫾16.4, and the mean LOS efficiency was 1.1⫾0.9. Fifty-five (62%) patients were discharged home, 27 (30%) patients were discharged to SNFs, 5 (6%) patients required a transfer back to an acute care facility, and 2 (2%) patients died during their rehabilitation stay. There were no significant differences in the admission and discharge FIM scores between patients with right- and leftsided PCA lesions. Nine (10%) patients who had interrupted rehabilitation stays had substantially lower admission total FIM scores compared with the rest of the cohort (means of 37.7⫾25.8 vs 67.2⫾24.5, respectively; P⫽.001).

2141

REHABILITATION OUTCOMES OF PCA STROKES, Ng Table 5: Multiple Linear Regression Analysis Model on Total Discharge FIM Score Variable

B*

SE

t†

P

Constant Admission total FIM score Age Diabetes mellitus Rehabilitation LOS Interrupted rehabilitation stay

44.8 0.99 ⫺0.33 ⫺7.27 0.22 ⫺17.73

12.26 0.08 0.11 3.01 0.09 4.59

3.66 12.47 ⫺2.96 ⫺2.42 2.34 ⫺3.86

.000 ⬍.001 .004 .018 .022 ⬍.001

NOTE. Shapiro-Wilk statistic for normality of residuals equals .978 (P⫽.13), adjusted R2 equals .81. Abbreviation: SE, standard error. *Unstandardized regression coefficient. † Student t test.

The admission and discharge cognitive, but not motor, FIM scores of the 38 (32%) patients who were confused or agitated were significantly lower than those of the patients who were not confused (P⫽.005 and P⬍.001, respectively). Otherwise, there were no differences in admission and discharge FIM scores between patients who had hemiparesis, visual field defect, aphasia, or neglect compared with those who did not. Comparing the PCA-S, PCA-SD, and PCA-P stroke subgroups, there were no significant differences in the total admission and discharge FIM scores or their corresponding motor and cognitive components (table 4). There were also no significant differences in the ⌬FIM and the FIM efficiency among these 3 groups. In the multiple linear regression analysis on total discharge FIM score, lower admission total FIM score, older age, shorter rehabilitation stay, diabetes mellitus, and interrupted rehabilitation stay were associated with lower discharge FIM score (table 5). These variables accounted for 81% of the variance in the discharge FIM scores. The regression model on ⌬FIM estimated about a third of the variance in this variable (table 6). Female sex, shorter rehabilitation stay, and interrupted rehabilitation stay were associated with lower ⌬FIM score. The admission FIM score was also significantly associated with the ⌬FIM score in a quadratic relation. Increasing admission FIM score was associated with increasing FIM score gain up to admission FIM score of 61. For higher values, the magnitude of FIM score gain incrementally decreased. This relation is schematically depicted in figure 1. In the logistic regression model on home discharge (table 7), the presence of a caregiver was strongly associated with a successful home discharge (odds ratio [OR]⫽16.59; 95% confidence interval [CI], 3.25– 84.60), as was no need for 24-hour

Table 6: Multiple Linear Regression Analysis Model on ⌬FIM Score Variable

Constant Sex LOS in rehabilitation Total admission FIM score Total admission FIM score (squared) Interrupted stay

B*

SE

t†

Fig 1. An increasing admission FIM score is associated with increasing ⌬FIM up to an admission total FIM score of 61; thereafter, ⌬FIM incrementally decreases.*Plot of x versus .855 x–.007 x2, where x is the admission total FIM score (see table 6).

support (OR⫽12.52; 95% CI, 2.42– 64.89). Younger age and higher total discharge FIM score were also significantly associated with a home disposition. DISCUSSION Many of the findings from our study were similar to those reported in the acute PCA stroke literature. The low number (5.7%) of PCA strokes in our database is consistent with the reported low incidence of stroke in this vascular territory.1,2 About half of our PCA stroke patients have PCA-S strokes, and this percentage falls within the range of 33% to 59% reported.2,3 Our study is further consistent with previous reports that the thalamus is involved in about a quarter of PCA stroke patients and is the most frequent associated location of PCA stroke outside the PCA-S territory.3,5 The common stroke risk factors of hypertension, smoking, cardiac disease, and diabetes mellitus reported in larger epidemiologic studies are similarly represented in our PCA stroke cohort.17,18 Just over half the patients in our study were men. In other studies there were more male patients with PCA stroke as well, rising to almost the two-third majority reported by Cals2 and Yamamoto3 and colleagues. The often-described association of female sex with migraine and posterior circulation stroke suggests that women are more prone to PCA strokes. However, this was not the case in this or other series.19,20 Sensory losses were uncommon in our cohort, as in other PCA series.2,3 Confusion or agitation occurred in more than one third of our

P

3.585 7.402 0.279 0.855

14.829 2.633 0.089 0.273

0.242 2.811 3.148 3.129

.810 .006 .002 .002

⫺0.007 ⫺12.683

0.002 4.452

⫺3.307 ⫺2.849

.001 .006

NOTE. Shapiro-Wilk statistic for normality of residuals equals .985 (P⫽.374); adjusted R2 equals .349. *Unstandardized regression coefficient. † Student t test.

Table 7: Logistic Regression Analysis Model on Discharge Disposition Variable

B

SE

Age ⫺0.094 .042 Total discharge FIM score 0.037 .013 Presence of caregiver 2.808 .832 24-h support not required 2.527 .839

P

.024

OR

95% CI

0.910 0.839–0.988

.005 1.037 .001 16.578

1.01–1.07 3.25–84.60

.003 12.521

2.42–64.89

NOTE. Hosmer-Lemeshow test (P⫽.734).

Arch Phys Med Rehabil Vol 86, November 2005

2142

REHABILITATION OUTCOMES OF PCA STROKES, Ng

patients. Frequent occurrence of these symptoms has been described in the acute hospital literature on PCA territory stroke as well.13 Motor impairments were more common than visual field defects in our patients compared with the acute hospital series, where the frequencies are reversed.2,3,13 The likely reason for this discrepancy is selection bias, whereby patients with motor impairments are more likely to require inpatient rehabilitation than those with visual field defects. More than two thirds of patients needed minimal to no assistance by the time of discharge from rehabilitation. These results confirm that patients with PCA stroke can make meaningful functional recovery.2,13 We found no significant differences in functional outcomes among PCA-S, PCA-SD, and PCA-P subcategories in our sample (see table 4). This finding may be influenced by small sample size in the subcategories or may reflect a mix of deficits in each subgroup that together have a similar level of impact on function. The first regression model on discharge total FIM score explained a substantial 81% of the variance in this variable using 5 independent variables. These were the admission total FIM score, age, diabetes, LOS, and interrupted rehabilitation stay. The important association of admission functional score with discharge functional score has been found in numerous prior studies of stroke outcome.21,22 Age was a significant predictor of discharge total FIM score in this model, with a regression coefficient of –.33 (see table 5). This means that a 3-year increase in age is associated with approximately a 1-point decrease in discharge FIM score. This is similar to findings in other stroke outcome studies.23-25 An association between diabetes mellitus and poorer functional outcomes after stroke has also been previously described. It may be due to diabetes’ propensity to accelerate microangiopathic and macroangiopathic disease, thereby compromising circulation-dependent recovery processes.26,27 Longer LOS was associated with higher discharge FIM scores. This association may have several interpretations. The most straightforward interpretation is that the longer they are exposed to an active rehabilitation program, the more patients will improve because of the effect of the program or natural recovery processes. Another explanation is that only patients who are continuing to improve are allowed to stay longer in active rehabilitation. Experienced clinicians will know of individual patients for whom each of these explanations was operative. An interrupted rehabilitation stay for a medical complication has been reported to be strongly associated with greater neurologic deficit.17 Consistent with this, we found that our patients with PCA stroke who had interrupted rehabilitation stays had lower FIM scores on admission and discharge. The lower scores may reflect greater neurologic deficit or medical comorbidity in these patients, with resulting poorer functional ability. The regression model on ⌬FIM explained 35% of the variance in that outcome. In this model male sex, admission FIM score, longer LOS, and lack of interruption in rehabilitation stay were associated with greater improvement in total FIM score. The association of male sex with greater ⌬FIM was intriguing. There were no significant differences in age (P⫽.230) or admission FIM scores (P⫽.205) between the sexes in this sample to explain the finding. The same association has been reported previously and may be due to factors such as lower frequency of prior hospitalization, lesser frailty, differing therapeutic interventions, or more social support among male patients.28,29 The admission total FIM score was significantly associated with the ⌬FIM in this model, but in a quadratic manner (see fig 1). Arch Phys Med Rehabil Vol 86, November 2005

This is due to a ceiling effect in the rehabilitation process, whereby patients are usually independent enough for home discharge when the FIM score has reached about 100, and so those admitted with FIM scores closer to 100 will not stay in inpatient rehabilitation long enough to make the same numeric gains in FIM points as those with lower admission FIM scores.25 It is widely believed that the availability of social support is an important determinant of home discharge.30 Prior studies have also shown higher functional score on admission and discharge to be predictive of a home disposition rather than an SNF. Age, sex, lesion site, and type of stroke (hemorrhagic vs ischemic), on the other hand, have shown uncertain predictive value.25,30,31 Our finding that caregiver support is an important predictor of home discharge in patients requiring inpatient rehabilitation after PCA stroke concurs with previous work on stroke rehabilitation.30,31 Similarly, lower discharge FIM score—reflecting more disability—and the need for 24-hour support were negative predictors of home discharge in our study, as in others.30,31 Increasing age was also a negative predictor, and this is in agreement with recent stroke outcome studies on aging specifically.25 Limitations in this study include a small sample size and potential type II error. In addition, across the 8-year period of the study, important parameters of inpatient rehabilitation throughout the United States have changed.32 These include shorter LOSs and better FIM efficiencies for patients with stroke in recent years. We did not find an association between the year of admission and functional outcome measures in this study. This could be a reflection of our small sample size. In addition to the selection bias inherent in including only patients who were admitted to an inpatient rehabilitation unit, there is the potential further bias of including patients from just 1 rehabilitation unit, with its particular referral patterns and admission criteria. A further limitation of this study is that the sample included patients with strokes affecting territories additional to the PCA (patients with PCA-P), which could have influenced our results. Finally, although the FIM has a proven record as a general measure of ability in mobility and selfcare,14,15 more specific measures of ability in motor, visual, and cognitive domains pertinent to PCA stroke may have provided more sharply focused results. CONCLUSIONS Although strokes in the territory of the PCA are uncommon, they lead to a high frequency of multiple impairments in cognition, vision, language, and motor function. The patients in our series made significant motor and cognitive functional gains in inpatient rehabilitation. A lower admission FIM score, shorter LOS, increasing age, female sex, and an interrupted rehabilitation stay were associated with lower discharge FIM scores and ⌬FIM. A home disposition was more likely when a caregiver was present, when 24-hour supervision was not required, in young patients, and in patients with better discharge FIM scores. Further research of interest would include study of comparative functional outcomes of stroke in different vascular territories and development of functional outcomes measures of greater breadth and sensitivity to the visuocognitive impairments found in patients with PCA territory stroke. Acknowledgment: We thank Richard Goldstein, PhD, for his assistance with the statistical analysis in this study. References 1. Brandt T, Steinke W, Thie A, Pessin MS, Caplan LR. Posterior cerebral artery territory infarcts: clinical features, infarct topogra-

REHABILITATION OUTCOMES OF PCA STROKES, Ng

2.

3.

4.

5.

6.

7.

8.

9. 10.

11. 12.

13. 14.

15.

16. 17.

18.

phy, causes and outcomes. Multicenter results and a review of the literature. Cerebrovasc Dis 2000;10:170-82. Cals N, Devuyst G, Afsar N, Karapanayiotides T, Bogousslavsky J. Pure superficial posterior cerebral artery territory infarction in The Lausanne Stroke Registry. J Neurol 2002;249:855-61. Yamamoto Y, Georgiadis AL, Chang HM, Caplan LR. Posterior cerebral artery territory infarcts in the New England Medical Center Posterior Circulation Registry. Arch Neurol 1999;56:824-32. Georgiadis AL, Yamamoto Y, Kwan ES, Pessin MS, Caplan LR. Anatomy of sensory findings in patients with posterior cerebral artery territory infarction. Arch Neurol 1999;56:835-8. Steinke W, Mangold J, Schwartz A, Hennerici M. Mechanisms of infarction in the superficial posterior cerebral artery territory. J Neurol 1997;244:571-8. Kumral E, Bayulkem G, Akyol A, Yunten N, Sirin H, Sagduyu A. Mesencephalic and associated posterior circulation infarcts. Stroke 2002;33:2224-31. Servan J, Verstichel P, Catala M, Yakovleff A, Rancurel G. Aphasia and infarction of the posterior cerebral artery territory. J Neurol 1995;242:87-92. De Renzi E, Zambolin A, Crisi G. The pattern of neuropsychological impairment associated with left posterior cerebral artery infarcts. Brain 1987;110:1099-116. Vatsavayi V, Malhotra S, Franco K. Agitated delirium with posterior cerebral artery infarction. J Emerg Med 2003;24:263-6. Nicolai A, Lazzarino LG. Acute confusional states secondary to infarctions in the territory of the posterior cerebral artery in elderly patients. Ital J Neurol Sci 1994;15:91-6. North K, Kan A, de Silva M, Ouvrier R. Hemiplegia due to posterior cerebral artery occlusion. Stroke 1993;24:1757-60. Chambers BR, Brooder RJ, Donnan GA. Proximal posterior cerebral artery occlusion simulating middle cerebral artery occlusion. Neurology 1991;41:385-90. North K, Kan A, de Silva M, Ouvrier R. Hemiplegia due to posterior cerebral artery occlusion. Stroke 1993;24:1757-60. Granger CV, Hamilton BB, Linacre JM, Heinemann AW, Wright BD. Performance profiles of the functional independence measure. Am J Phys Med Rehabil 1993;72:84-9. Granger CV, Hamilton BB, Keith RA, Zielezny M, Sherwin FS. Advances in functional assessment for medical rehabilitation. Top Geriatr Rehabil 1986;1:59-74. Shapiro SS, Wilk MB. An analysis of variance test for normality (complete samples). Biometrika 1965;52:591-611. Roth EJ, Lovell L, Harvey RL, Heinemann AW, Semik P, Diaz S. Incidence of and risk factors for medical complications during stroke rehabilitation. Stroke 2001;32:523-9. Sacco RL, Benjamin EJ, Broderick JP, et al. American Heart Association Prevention Conference. IV. Prevention and Rehabilitation of Stroke. Risk factors. Stroke 1997;28:1507-17.

2143

19. Moen M, Levine SR, Newman DS, Dull-Baird A, Brown GG, Welch KM. Bilateral posterior cerebral artery strokes in a young migraine sufferer. Stroke 1988;19:525-8. 20. Tzourio C, Kittner SJ, Bousser MG, Alperovitch A. Migraine and stroke in young women. Cephalalgia 2000;20:190-9. 21. Ween JE, Mernoff ST, Alexander MP. Recovery rates after stroke and their impact on outcome prediction. Neurorehabil Neural Repair 2000;14:229-35. 22. Bohannon RW, Lee N, Maljanian R. Postadmission function best predicts acute hospital outcomes after stroke. Am J Phys Med Rehabil 2002;81:726-30. 23. Kugler C, Altenhoner T, Lochner P, Ferbert A; Hessian Stroke Data Bank Study Group ASH. Does age influence early recovery from ischemic stroke? A study from the Hessian Stroke Data Bank. J Neurol 2003;250:676-81. 24. Nakayama H, Jorgensen HS, Raaschou HO, Olsen TS. The influence of age on stroke outcome. The Copenhagen Stroke Study. Stroke 1994;25:808-13. 25. Black-Schaffer RM, Winston C. Age and functional outcome after stroke. Top Stroke Rehabil 2004;11:23-32. 26. Weimar C, Ziegler A, Konig IR, Diener HC. Predicting functional outcome and survival after acute ischemic stroke. J Neurol 2002; 249:888-95. 27. Desmond DW, Moroney JT, Sano M, Stern Y. Recovery of cognitive function after stroke. Stroke 1996;27:1798-803. 28. Wyller TB, Sodring KM, Sveen U, Ljunggren AE, Bautz-Holter E. Are there gender differences in functional outcome after stroke? Clin Rehabil 1997;11:171-9. 29. Di Carlo A, Lamassa M, Baldereschi M, et al. Sex differences in the clinical presentation, resource use, and 3-month outcome of acute stroke in Europe: data from a multicenter multinational hospital-based registry. European BIOMED Study of Stroke Care Group. Stroke 2003;34:1114-9. 30. Agarwal V, McRae MP, Bhardwaj A, Teasell RW. A model to aid in the prediction of discharge location for stroke rehabilitation patients. Arch Phys Med Rehabil 2003;84:1703-9. 31. Brosseau L, Potvin L, Philippe P, Boulanger YL. Post-stroke inpatient rehabilitation. II. Predicting discharge disposition. Am J Phys Med Rehabil 1996;75:431-6. 32. Ottenbacher KJ, Smith PM, Illig SB, Linn RT, Ostir GV, Granger CV. Trends in length of stay, living setting, functional outcome, and mortality following medical rehabilitation. JAMA 2004;292: 1687-95. Suppliers a. Microsoft Inc, One Microsoft Way, Redmond, WA 98052. b. SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.

Arch Phys Med Rehabil Vol 86, November 2005