Association of inpatient rehabilitation with functional outcome in patients with stroke receiving mechanical thrombectomy

Abstract

Introduction

Patients with ischemic stroke that have undergone mechanical thrombectomy (MT) can still have poor functional outcomes. Limited research exists on factors that impact functional outcomes in these patients who are discharged to inpatient rehabilitation. We examined patient characteristics correlated with 90-day outcomes in ischemic stroke patients having undergone MT and undergo acute rehabilitation.

Methods

This is a retrospective study of patients (≥18 years) who were consecutively admitted to inpatient rehabilitation after undergoing MT following an ischemic stroke event from 2015 through 2018. Functional outcome at 90-days were determined by the modified Rankin Score (mRS). Binary logistic regression models were used to measure the association of meaningful Functional Independence Measure (FIM) change with poor-outcome (mRS>2).

Results

Out of 56 stroke patients undergoing MT, 36 had poor outcome (mRS>2) at 90-days. One unit increase in discharge FIM significantly decreased the likelihood of poor outcome by 4% [aOR=0.96; 95% CI=95% CI=0.92, 0.99]. Admission FIM [aOR=0.95; 95% CI=0.90, 1.00] and change in FIM [aOR=0.97; 95% CI=0.93, 1.02] were not significantly associated poor outcome in the adjusted models.

Conclusion

Patients with ischemic stroke who received MT who have worse disability, particularly at discharge, while undergoing inpatient rehabilitation have higher odds of having poor future functional outcomes. This was regardless of age, sex, or race. Future research is needed to understand the mechanisms that can improve functional performance in our patient population at the IRF.

INTRODUCTION

Ischemic stroke can lead to significant functional disabilities.¹ Acute endovascular therapy using mechanical thrombectomy (MT) has been shown to significantly improve functional outcomes in patients with ischemic stroke and large vessel occlusions.² Unfortunately, patients that have undergone MT can still have poor functional outcomes at three months.^3,4 At discharge from acute hospitalization, many patients with persistent deficits require admission to an inpatient rehabilitation facility (IRF).

Patients with stroke who are discharged to IRF can have variable long- term outcomes.^1,5 Prior research has indicated predictors such as interval between stroke onset and hospital admission, more severe hemiparesis upon admission, and lower Functional Independence Measure (FIM) admission scores as potential predictors of poor future outcomes.¹ However, there is limited information on what factors impact future outcomes in patients with stroke after MT and discharged to IRF. The aim of the present study was to determine if disability (admission, discharge, or change) during inpatient acute rehabilitation is correlated with long-term functional outcomes in patients with ischemic stroke who have undergone MT. More specifically, we hypothesized that patients who had lower FIM scores will have higher odds of poor 90-day outcomes.

METHODS

Participants

This is a retrospective study of patients (≥18 years) who were consecutively admitted to an inpatient rehabilitation center after undergoing MT for a primary diagnosis of ischemic stroke from 2015 through 2018. All cases were confirmed on neuroimaging with CT or MRI. For patients with recurrent events, only the incident stroke with MT were included. Patients were excluded if they did not receive MT, were not discharged to the IRF, rehabilitation measures were unavailable, or stroke was due to intracerebral hemorrhage subarachnoid, subdural, or epidural hemorrhage. Institutional Review Board approval was obtained as a part of our Institutional Stroke Registry. Informed consent was waived under the common rule and a waiver of authorization.

Outcome measure

The outcome of interest was functional status following inpatient rehabilitation for stroke. For this purpose, modified Rankin Scale (mRS) measured at 90 days following stroke was used. There are seven levels of mRS in this scale: 0=no symptoms; 1=no significant disability despite symptoms; 2=slight disability; 3=moderate disability; 4=moderately severe disability; 5=severe disability; 6=death. For ease of analysis, mRS was categorized as a dichotomous variable, where a score of ≤2 indicated good functional status, and a score of >2 indicated poor functional status.

Independent variables

Inpatient rehabilitation disability was measured using the Functional Independence Measure (FIM) manual version 3.0, which contains 18 items divided into 6 domains covering activities of daily living, sphincter control, transfers, locomotion, communication, and social cognition. Each item is scored on an ordinal 7-point scale, where a score of 1 indicates total dependency and 7 indicates complete independence, The FIM total sum score ranges from 18 to 126 and can be divided into a motor (13 items) and a cognitive (5 items) sub-scoreⁱⁱ. The measure was evaluated by an experienced physiatrist at admission and discharge from the IRF. The FIM has proven content and construct validity, is responsive to small increments in functional status after stroke, and correlates highly with measures of neurologic impairment after stroke, such as the National Institutes of Health Stroke Scaleⁱⁱⁱ. The FIM instrument was administered by certified and trained team members at admission and discharge. Functional gains were determined by comparing admission and discharge FIM scores. Other variables included age, sex, race, baseline National Institutes of Health Stroke Scale score, and length of hospital stay.

Statistical analysis

The Baseline characteristics of the participants were described as means and standard deviations (SD) for continuous variables and frequencies for categorical variables. The difference in patient characteristics by 90-day mRS were measured using two-sample t-tests and Pearson chi-square tests for continuous and categorical variables, respectively. Three sets of crude and adjusted binary logistic regression models were used to measure the association of each FIM measure of interest (admission FIM, discharge FIM, and change in FIM) with poor outcome (mRS>2). Adjusted models included age, sex, race, National Institutes of Health Stroke Scale (NIHSS) at baseline, and length of stay at IRF as covariates. The covariates for adjustment were selected a priori based on the evidence from previous studies. Multicollinearity between the covariates and the FIM measure in the model were tested using tolerance and variance inflation factors; and no collinearity between the variables was identified. A sensitivity analysis was performed with mRS as an ordinal outcome instead of dichotomous by using logistic regression with proportional odds models to assess whether change in FIM, admission FIM, or discharge FIM were associated with poor mRS outcome (ordinal). The results were considered statistically significant under the α=0.05. All analyses were performed in SAS, version 9.4 (Cary, NC, USA).

RESULTS

A total of 225 patients received MT at our acute hospital, of which 62 patients matched our target population when they were subsequently admitted to IRF (Figure 1). After removal of recurrent events, 56 patients (59.0 ± 15.1 years; 37.5% female; 44.6% African American) with incident ischemic stroke receiving acute inpatient rehabilitation were identified for this analysis. Poor outcome (mRS >2) at 90 days was reported in 36 (55.4%) patients. The demographic and clinical characteristics of the participants are shown in Table 1. The majority of our patients had no significant premorbid disability. We did not observe any statistical differences in demographic characteristics between poor and favorable outcome groups. Length of IRF stay was significantly longer in those with poor outcome (16.7 ± 7.0) compared to those without (12.2 ± 7.0; p=0.0211). The mean admission (Figure 2a; p=0.0016) and discharge (Figure 2b; p=0.0025) FIM scores were significantly lower for those with poor outcome compared to those with favorable outcomes.

Figure 1:

Flow diagram of case ascertainment

This flow diagram describes how the study sample was obtained after removal of recurrent admissions

Table 1.

Participant characteristics

Variables	All	90-day mRS ≤2	90-day mRS >2	p-value
N	56	25	31	-
Age, mean ± SD	59.0 ± 15.1	56.9 ± 17.1	60.7 ± 13.3	0.3533
Female, N(%)	21	8 (32.0%)	13 (41.9%)	0.4452
Race a				0.6156
White	28	11 (44.0%)	17 (54.8%)
Black	25	13 (52.0%)	12 (38.7%)
Other	3	1 (4.0%)	2 (6.5%)
Length of stay, days, mean ± SD	14.7 ± 7.3	12.2 ± 7.0	16.7 ± 7.0	0.0211
Intravenous thrombolytics	18	8 (32.0%)	10 (32.3%)	0.9836
Baseline mRS a, b				0.3662
0	44	22 (91.7)	22 (73.3)
1	4	1 (4.2)	3 (10.0)
2	2	1 (4.2)	1 (3.3)
3	3	0 (0.0)	3 (10.0)
4	1	0 (0.0)	1 (3.3)
Baseline NIHSS	13.8 ± 7.1	12.7 ± 8.2	14.7 ± 6.2	0.3048
FIM Scores
Admission FIM score, mean ± SD	51.4 ± 17.3	59.1 ± 16.4	45.1 ± 15.6	0.0019
Discharge FIM score, mean ± SD	79.3 ± 21.2	88.0 ± 16.4	72.2 ± 22.2	0.0046
Change in FIM score, mean ± SD	27.9 ± 13.4	28.8 ± 11.4	27.1 ± 15.0	0.6327

FIM, functional independence measure; mRS, Modified Rankin scale; NIHSS, National Institutes of Health Stroke Scale; SD, standard deviation

Boldface indicates statistical significance under the α=0.05

^aStatistical significance tested using Fisher’s exact test

^bBaseline mRS values missing for two participants

Figures 2a and 2b.

show the distribution of admission and discharge FIM scores by 90-day modified Rankin Scale (mRS) score categories where poor outcome is mRS>2

The unadjusted logistic regression models indicated that per unit increase in admission FIM led to 5% lower odds (Table 2; 95% CI=0.91, 0.98) and per unit increase in discharge FIM led to 4% lower odds (95% CI=0.93, 0.99) of poor outcome. After adjusting for covariates, discharge FIM remained significantly associated with poor functional outcome, indicating a 4% lower odds of poor outcome per unit increase in discharge FIM (95% CI=0.92, 0.99); however, admission FIM was not significantly correlated with having poor 90-day outcome in the adjusted model. Change in FIM did not significantly impact the odds of poor outcome in the crude or adjusted models. The sensitivity analysis with ordinal mRS outcome indicated a 7% lower odds of worse functional outcome per unit increase in admission FIM (Table 3; 95% CI=0.89, 9.97) and 5% lower odds of worse outcome per unit increase in discharge FIM (95% CI=0.92, 0.98) after covariate adjustment.

Table 2.

Crude and adjusted odds ratios and corresponding 95% confidence intervals for poor mRS outcome (>2, dichotomous) by participant characteristics

	Crude OR (95% CI)	Adjusted OR (95% CI)a
Admission FIM	0.95 (0.91, 0.98)	0.95 (0.90, 1.00)
Discharge FIM	0.96 (0.93, 0.99)	0.96 (0.92, 0.99)
Change in FIM	0.99 (0.95, 1.03)	0.97 (0.93, 1.02)

^aAdjusted for: Age, sex, race, baseline NIHSS score, length of rehabilitation stay

Boldface indicates statistical significance under α = 0.05

Table 3.

Crude and adjusted odds ratios and corresponding 95% confidence intervals for poor mRS outcome (ordinal) by participant characteristics

	Crude OR (95% CI)	Adjusted OR (95% CI)a
Admission FIM	0.95 (0.92, 0.98)	0.93 (0.89, 0.97)
Discharge FIM	0.96 (0.93, 0.98)	0.95 (0.92, 0.98)
Change in FIM	0.99 (0.95, 1.02)	0.97 (0.94, 1.01)

^aAdjusted for: Age, sex, race, baseline NIHSS score, length of rehabilitation stay

Boldface indicates statistical significance under α = 0.05

DISCUSSION

This is the first study to our knowledge to evaluate if disability assessed during inpatient rehabilitation were associated with long-term functional outcomes in patients with ischemic stroke who have undergone MT. We found that patients with a poor 90-day outcome, in both ordinal and dichotomous analysis, had lower FIM scores at admission and discharge. Interestingly, there were no significant differences between change in FIM scores between the poor and favorable 90-day outcome groups. Patients who had lower discharge FIM scores had higher odds of having poor functional outcomes at 90-days in our adjusted model. No other adjusted variable in our model reached significance. Based on our results, functional performance at discharge can reasonably indicate the 90-day outcome in patients who have undergone inpatient rehabilitation with ischemic stroke who have had MT.

There is more interest in identifying clinical, demographic, imaging, and other variables associated with functional outcomes, particularly at discharge.⁶ In clinical practice, most indicators of prognosis and functional recovery are based in clinical impressions, and prognostication remains challenging. The majority of literature has focused on measuring functional outcomes such as upper extremity function^7–9 and walking after stroke.^10,11 Many of these assessments only go so far as predicting outcomes at time of discharge from the acute setting, including from IRF.^12,13 Our data complements existing literature showing that early evaluations, including during rehabilitation, are likely more accurately indicative of future function. Additionally, our results give an understanding for a population that has not specifically been studied, patients with stroke who have undergone MT and are undergoing acute rehabilitation. Because the disability level at discharge was associated with poor functional outcomes at 90-days, future studies should consider individualized rehabilitation plans to determine if longer lengths of stay or other mechanisms to improve disability scores prior to discharge from IRF.

Both our groups with poor and good outcomes at 90 days saw similar improvements in FIM scores during IRF. This is notable that both groups’ change in FIM scores continued to improve during IRF. This would indicate that both groups likely benefitted from their IRF stay, though patients with good outcomes had a shorter length of stay. This could indicate a circular logic as discharge often occurs once a specific level of function is achieved. However, since we are noting post-IRF function, not only discharge function like other studies^12,13, this could also indicate that because their admission FIM scores were also better, a spontaneous recovery process occurring such as proportional recovery.¹⁴ However, this proportional recovery has not always been found to be indicative of future outcomes including with guiding treatment and quality of life.^6,15 Since both groups improved, we did not find that change in FIM scores at IRF was associated with future functional outcomes. Because not all patients with ischemic stroke see such gains in FIM scores during IRF, we propose that the endovascular treatment was the main driving force of this improvement for both groups.

Most studies have found that age and initial stroke severity play an important role in future functional outcomes.^7,8,11 While we also included in our model, we did not see these two variables to be significantly associated with poor 90-day mRS outcome. Interestingly, neither sex nor race, both of which have also been associated with poorer outcomes after ischemic stroke, were significantly associated with functional outcomes in our population.⁶ In studies that have looked at poor outcomes after MT, older patients were more likely to have poor outcomes.^3,4 A key consideration for our results are that the patients we examined were selected for inpatient rehabilitation. Our findings demonstrate that regardless of age, sex, or race, in patients who received MT and undergo inpatient rehabilitation, discharge disability is most significantly correlated to poor outcomes at 90-days.

Study limitations

Study strengths include the validity and the relevance of the measures. Additionally, very few studies in stroke have evaluated these outcomes at and before IRF discharge, a critical time point in patients’ recovery. There are several limitations to this study. This study was retrospectively conducted in a single large academic Stroke Center where MT is commonly performed and IRF in the Southeastern United States. While differences between MT-capable stroke centers, rehabilitation practices, and patient characteristics should not be substantially different, our results might not be generalizable to all settings. Next, we only included patients that underwent inpatient rehabilitation. These patients have usually had an evaluation by a multidisciplinary team (psychiatrists, therapists, and neurologists). Thus, our sample included younger patients and fewer patients with minor or severe impairments. Finally, while we also included patients that received IV-thrombolytics with MT, this study was not designed to evaluate the differences between the population that received MT only compared to intravenous thrombolytics and MT.

CONCLUSION

Patients with ischemic stroke who received MT who have worse disability, particularly at discharge, while undergoing inpatient rehabilitation have higher odds of having poor future functional outcomes. This was regardless of age, sex, or race. With reducing stroke mortality, more and more emphasis in stroke research is being placed upon the need to identify and address the correlates of stroke disability. Vascular interventions such as MT have become an increasingly common clinical approach for stroke management. To that end, our study addresses the need to identify how functional performance during inpatient rehabilitation is indicative of 90-day outcomes in these patients. It sets precedence for future research where studies with a larger sample size can further expand upon the mechanisms that can improve functional performance in our patient population at the IRF.