Abstract
Objectives:
Patients with intracerebral hemorrhage (ICH) are more likely to present with severe symptoms than those with ischemic stroke (IS); however, the way in which long-term outcomes differ between groups is less clear. Given that the tissue surrounding ICH is not always irreversibly infarcted, it may have the potential to recover more fully over time. Understanding the differences in expected outcome severity is critical in order to prognosticate and to determine appropriate outcome measures when designing clinical trials.
Materials and Methods:
We used our prospectively collected stroke registry to identify and follow a cohort of 300 patients with ICH and 300 patients with IS, matched by age, sex, lesion size, location, and admission date. Paired t-tests were used to compare modified Rankin Scores (mRS) between groups at hospital discharge, 90-day, and >12-month follow-up time points.
Results:
Not surprisingly, patients with ICH had worse discharge mRS scores compared to individuals with IS (4.20 (SD 0.09) versus 3.42 (SD 0.08)). However, rather than improving, the long-term outcomes remained significantly worse for ICH patients compared to their IS matches (4.02 (SD 0.15) versus 2.89 (SD 0.14) at 90 days, and 4.32 (SD 0.20) versus 3.16 (SD 0.22) at >12 months (p < 0.001 for all analyses)).
Conclusions:
This study longitudinally followed matched cohorts of patients with ICH and IS, confirming that outcomes for those with ICH remain significantly worse over time. Results allow for better long-term prognostication, illustrate the need for further intervention trials to improve outcomes, and inform the development of evidence-based endpoints.
Keywords: stroke, ischemia, hemorrhage, recovery, outcomes
Introduction
Stroke can be broadly broken down into two categories: ischemic, where there is occlusion of a vessel leading to cell death and dysfunction; and hemorrhagic, where there is rupture of a vessel leading not only to neuronal compromise, but an inflammatory cascade that results in edema and compression that further impairs function. Due in part to this inflammatory cascade and the uncontrolled nature of the rupture, the initial presentation of intracerebral hemorrhage (ICH) is often more severe than ischemic stroke (IS). ICH patients present with higher blood pressure, more frequent seizures, and higher rates of intubation.1–4 However, the fact that a hemorrhagic stroke does not necessarily result in neuronal death raises the possibility that in the long-term an individual with an ICH may recover more abilities than a patient after an ischemic lesion of the same size and location.5,6
Though the recovery process can be long and complicated, many stroke patients experience significant improvement in symptoms over time.7 Recent advancements in the acute treatment of ischemic stroke using thrombolytic agents and stentrievers have improved outcomes further, turning large hemispheric lesions into smaller less severe infarcts.8–11 Classically, outcomes have been measured using the modified Rankin score (mRS)12 for acute treatment and rehabilitation trials. The mRS evaluates functionality and independence post-stroke, with a score of 0–2 considered a “good outcome”. However, as the treatment landscape of ischemic stroke has changed, we have found that measures focused not only on the physical manifestations of stroke, but factors such as cognition, mood and quality of life, may be just as important as traditional outcome measures for detecting the effectiveness of treatment.13 The same may be the case for those with hemorrhagic stroke, especially if long-term individuals exhibit better overall recovery once blood within the parenchyma is reabsorbed. Unfortunately, there are not enough longitudinal studies contrasting the severity and functional outcomes of those with ischemic stroke to those with hemorrhage in a similar region for us to draw these comparisons.
The purpose of this study was to delineate the differences in outcomes for patients with IS versus ICH of similar size and location in both the acute and chronic setting using our large, prospectively collected stroke registry. We hypothesized that symptom severity would be worse for patients presenting to the hospital with ICH compared to those with IS and that this pattern would continue through hospital discharge and early recovery stages but shift toward favoring those with ICH with reabsorption of blood products chronically. Results will allow clinicians to counsel patients and their families more effectively and inform our future design of the endpoints of clinical trials to most effectively detect differences in outcome following intervention.
Methods
Study Population
All patients hospitalized at our Comprehensive Stroke Center are scheduled for follow-up in our outpatient stroke clinic one month after discharge and longitudinally followed for at least one year of recovery. Patients are enrolled at hospitalization and their information prospectively entered into a large, IRB-approved and HIPAA-compliant stroke registry that includes: demographics (age, race, sex, occupation, level of education, living situation), IS/ICH characteristics (NIHSS14 on admission, ICH score, lesion volume and location), medical variables (ambulatory status prior to admission, BMI, history of: prior stroke, depression, smoking, hypertension, hyperlipidemia, diabetes, atrial fibrillation, coronary artery disease, coronary artery stenting, and congestive heart failure), and outcomes (length of stay, discharge disposition, mortality, modified Rankin Score (mRS)12 on discharge and 90 days post-discharge). Patients are followed longitudinally, returning at 1-, 6- and 12-months post-stroke. At this time additional data points are collected to evaluate functional abilities and outcomes. For this analysis, we also collected mRS scores >12 months post-stroke through review of the electronic medical record by three individuals (SG, AY, JL) on the study team. These were extrapolated, most commonly using the neurology clinic follow-up note, but if unavailable then through medical or other subspecialty clinic notes or rehabilitation assessments within our health system. Scores were determined after reviewing ambulatory status, level of independence, residual symptoms, and physical abilities. An mRS of 6 was assigned to any individual who had passed away, for each data point following the reported date of death. Values were left blank if there was no specific mention of recovery or insufficient information available to determine an accurate score. There was 5% overlap of record review to ensure adequate inter-rater reliability.
Case Matching
Three hundred cases (ICH) entered into the database between March 2014 and June 2019 were matched to controls (IS) entered between January 2014 and May 2020. This time period was chosen to avoid the potential confounding effects of acute treatment of ischemic stroke in the pre-thrombectomy era and to ensure our findings were most relevant to current clinical practice standards. For each patient with ICH, a patient with IS of the same sex, similar age (within 10 years), lesion size (within 10cc), location (hemisphere, circulation), and admission date (within 5 years) was chosen. Examples of matched cases are shown in Figure 1. The majority of individuals underwent magnetic resonance imaging (MRI) to allow full characterization of their lesion following acute intervention and within the first 3 days of hospitalization, but non-contrast computed tomography (CT) was used for the remaining patients. Lesion volume for ICH was calculated on head CT also obtained within the first 3 days once bleeding was stable using a volumetric analysis package that is part of the Carestream platform.15 Areas of surrounding edema were not included in total lesion volume, but noted separately, along with presence of midline shift or herniation. For ischemic strokes, volume was calculated on diffusion-weighted imaging sequences. Lesion location was scored in terms of hemisphere (right versus left), circulation (supra versus infratentorial), depth (cortical versus deep), and intraventricular involvement of hemorrhage. ICH cases that were bilateral (n=60) and had no similar age/sex match (n= 14 females, 6 males), were discarded; most commonly occurring at the extremes of age. This study was approved by our institutional review board and did not require informed consent given its observational nature.
Figure 1.
Examples of matched IS and ICH patients. Each participant was matched by age, sex, lesion size/location, and admission year.
Statistical Analysis
Data were analyzed using Stata version 14. Paired t-tests and chi-square analyses were used to compare the two groups (for continuous and categorical variables respectively). The pre-determined primary outcome variables included the modified Rankin score at discharge, 90 days, and greater than 12 months. Multivariable regression was used to adjust for potential confounders of outcome.
Results
Comparing Group Characteristics
300 patients with ICH were successfully matched with 300 patients with IS at the time of hospitalization. A complete list of group comparisons can be seen in Table 1. The groups were very similar, with no significant differences in age, sex, or stroke characteristics. Of the 300 matches, approximately half (n=149, 49.7%) were male. There were marginally more black patients in the ICH group with 89 (31.9%) compared to the IS group with only 83 (29.8%). Years of education varied between the groups (13.2 versus 12.2 years respectively (p = 0.006)) along with the calculated average income based on zip code (~$80,000 versus $70,000, p < 0.001) and were therefore considered in subsequent multivariable regression analyses. Both groups also had similar medical histories, including depression, smoking, and hypertension, though there were more patients in the ischemic group who had previously been diagnosed with atrial fibrillation (25.7% compared to 16.0%, p=0.016).
Table 1.
Group Demographics, Medical History, and Stroke Characteristics
| Hemorrhages | Ischemic Strokes | P - Values | |
|---|---|---|---|
| (N=300) | (N=300) | ||
| Age, mean years (SD) | 68.4 (0.9) | 68.7 (0.9) | p = 0.495 |
| Sex, n male (%) | 149 (49.7%) | 149 (49.7%) | p = 1.000 |
| Race, n black (%) | 89 (31.9%) | 83 (29.8%) | P = 0.359 |
| Education, mean years (SD) | 13.2 (0.3) | 12.2 (0.2) | p = 0.006 |
| Income, mean USD (SD) | $82,662 ($2,137) | $71,391 ($1,418) | p < 0.001 |
| Social Support, n yes (%) | 222 (79.3%) | 223 (79.6%) | p = 0.910 |
| p = 0.341 | |||
| 1 - Professional | 20 (9.2%) | 9 (4.1%) | |
| 2 - Intermediate | 19 (8.7%) | 11 (5.1%) | |
| 3 - Skilled | 53 (24.3%) | 44 (20.2%) | |
| 4 - Semiskilled | 52 (23.9%) | 52 (23.9%) | |
| 5 - Unskilled | 44 (20.2%) | 37 (17.0%) | |
| 6 - Unemployed | 7 (3.2%) | 11 (5.1%) | |
| 7 - Retired, unknown occupation | 23 (10.6%) | 54 (24.8%) | |
| Prior Stroke, % | 65 (21.7%) | 78 (26.0%) | p = 0.774 |
| Depression, % | 19 (6.3%) | 33 (11.0%) | p = 0.946 |
| Smoking, % | 57 (19.0%) | 105 (35.0%) | p = 0.211 |
| Hypertension, % | 229 (76.3%) | 237 (79.0%) | p = 0.173 |
| Body Mass Index, mean percentage (SD) | 28.4 (0.4) | 28.8 (0.4) | p = 0.460 |
| Obesity, % | 21 (7.0%) | 34 (11.3%) | p = 0.248 |
| Diabetes, % | 75 (25.0%) | 98 (32.7%) | p = 0.670 |
| Hyperlipidemia, % | 98 (32.7%) | 138 (46.0%) | p = 0.635 |
| Systolic Blood Pressure, mean (SD) | 164.5 (2.1) | 154.2 (1.6) | p < 0.001 |
| Ambulation, % | 199 (92.6%) | 197 (91.6%) | p = 0.793 |
| Atrial Fibrillation, % | 48 (16.0%) | 77 (25.7%) | p = 0.016 |
| Coronary Artery Disease, % | 57 (19.1%) | 48 (16.0%) | p = 0.653 |
| Carotid Artery Stenting, % | 4 (1.3%) | 7 (2.3%) | p = 0.756 |
| Congestive Heart Failure, % | 29 (9.7%) | 35 (11.7%) | p = 0.707 |
| Admission NIHSS, mean points (SD) | 12.54 (0.69) | 8.29 (0.49) | p < 0.001 |
| Stroke Volume, mean cc (SD) | 31.9 (2.5) | 32.5 (2.5) | p = 0.331 |
| Hemisphere, n left (%) | 149 (49.7%) | 148 (49.3%) | p = 1.000 |
| Circulation, n infratentorial (%) | 27 (9.1%) | 29 (9.7%) | p = 1.000 |
| Isolated subcortical location, (%) | 160 (53.7%) | 131 (44.0%) | p = 1.000 |
| Isolated cortical location, (%) | 101 (33.9%) | 99 (33.2%) | p = 1.000 |
| p = 0.077 | |||
| 1 | 6 (3.7%) | 15 (9.3%) | |
| 2 | 51 (31.5%) | 56 (34.6%) | |
| 3 | 48 (29.6%) | 36 (22.2%) | |
| 4 | 28 (17.3%) | 12 (7.4%) | |
| 5 | 13 (8.0%) | 26 (16.1%) | |
| 6 | 6 (3.7%) | 9 (5.6%) | |
| 7 | 9 (5.6%) | 5 (3.1%) | |
| 8 | 1 (0.6%) | 3 (1.9%) |
Comparing Group Outcomes
On admission, the presentation of patients with ICH was more severe than those with IS, with significantly higher stroke severity (NIHSS 12.54 versus 8.29, p<0.001) and systolic blood pressure (164.5mmHg v 154.2mmHg, p<0.001). Similarly, ICH patients required longer hospital stays (10.3 v 6.9 days, p<0.001), exhibited higher discharge mRS scores (4.20 v 3.42, p<0.001), and were less likely to be discharged home from the hospital (20.7% v 35%). Of note, outcomes remained poorer for those with ICH at 90 days post-stroke, with a mean mRS of 4.02 compared to 2.89 for those with IS, and this pattern persisted at greater than 12 months (4.32 v 3.16, p<0.001). Overall mortality was also significantly higher (43.2% v 30.3%, p=0.014). The distributions of long-term trajectories for both patient groups are plotted in Figure 2A. In multivariate regression, differences in mRS score at each time point by group remained consistent when controlling for potential confounders significant in univariate analysis.
Figure 2.
A. A box and whisker plot demonstrating the differences between mRS scores at each time point between groups. B. Individual histograms comparing the distribution of mRS scores at each time point for each group. The normalized distribution of the population is plotted on each graph for reference.
On further comparison of the groups (Figure 2B), patients presenting with ICH were more likely to have an mRS of 5 or 6 on hospital discharge while scores for patients with IS were more normally distributed. At 90 days post-discharge the number of ICH patients with an mRS of 6 continued to increase, though some patients showed improvement, resulting in an overall wider distribution of scores. Most IS patients continued to improve, with a greater proportion falling in the range of “good outcome” (mRS 0–2). Three quarters (74.9%) of the greater than 12-month mRS scores extrapolated were within 90 days of the 1-year post-stroke time point. Over time, both ICH and IS groups migrated toward more extreme scores, resulting in a bimodal distribution; however, a greater proportion of patients with IS displayed “good outcomes” while a greater portion of ICH patients were deceased. Differences in the progression of longitudinal outcomes can be seen in Figure 3. (Figure 3)
Figure 3.
Sankey plot showing the evolution of mRS scores for Ischemic Stroke (A) and ICH (B) patients over time.
Discussion
Our results demonstrate a pattern of recovery for patients with ICH that is distinctive from patients with ischemic stroke and unfortunately does not improve once there is resorption of blood products and a reduction of inflammation. While it is not surprising that their initial presentation was more severe, it was notable that the majority of hemorrhages were initially clustered around mRS scores of 4–6, while the spectrum of scores for patients with IS was much broader. Over time, this difference was magnified further, though both groups ultimately progressed toward a bimodal distribution of either very good or very poor outcome. While mortality rates (mRS 6) for both groups increased, the rate was higher at every time point for those with ICH. Fewer ICH patients progressed to mRS scores of 2 or lower than those with IS, where many demonstrated significant improvement longitudinally.
Results are not inconsistent with prior literature but help clarify longitudinal patterns further and provide significant additional insights. Prior studies have shown that patients with intracerebral hemorrhage present with higher systolic blood pressures and more severe clinical presentations, resulting in a higher mortality rate acutely.1,2 However, following discharge there is less consensus in the literature, and prior to this study these inconsistencies were not well understood. Paolucci et al., demonstrated that ICH patients exhibit functional gains faster during rehabilitation than patients with ischemic stroke.5,6,16 However, while this study included IS and ICH patients of similar “stroke severity”, they were not matched for size or location, and given that to qualify for inpatient rehabilitation programs patients typically have to demonstrate a certain level of function, it is possible that overall the cohort’s stroke severity was less severe. In less severe cases, ICH patients may do better when edema begins to resolve if there is no underlying co-existing infarction.
Other studies including Perna et al., found no differences between the functional scores of IS and ICH patients discharged from rehabilitation. They rationalized that, “while the pathophysiology is different – both ultimately result in ischemic injuries.”17 That is one plausible explanation; however, this study was a general comparison between groups without implementing matching to compare lesion size and location. Similarly, Salvadori et al. determined that both IS and ICH patients appeared similar on hospital discharge, despite any differences they displayed on admission.4 Both of these studies are in contrast to Chiu et al., who found ICH to be a predictor of poorer neurological outcomes (worse mRS scores) at 90 days.18 Heterogeneity of lesion size, location, and severity, along with timing of evaluation and relatively small sample sizes may explain these discrepancies, prompting the need for further clarification in a longitudinal matched study. Our data illustrate that the presentation of ICH remains more severe throughout both the acute and chronic stages of recovery when compared to patients with IS when lesion size and location are matched. Further, our data dispel the notion that if a patient with ICH can survive the initial hospitalization, their recovery course may be better than that of someone with an ischemic infarct of similar size and location. We instead show that the risk of mortality increases throughout the chronic stage of recovery, even after any potential bias for withdrawal of care is removed. This information is important when counseling families and helping to prognosticate post-stroke care.
While the majority of previous studies looked at recovery and prognosis in the acute/subacute setting, focusing on hospital discharge and rehabilitation, we followed patients longitudinally, exploring mRS scores at greater than 12 months post-hospitalization. Our findings suggest that outcomes remain worse at all time points for ICH patients compared to those with ischemic stroke, though both recovery patterns for both groups shared some consistent features. While ICH patients were more likely to have poor outcomes, many exhibited improvement over the first 90 days, similar to patients with IS, broadening the distribution of scores, though to a lesser degree than ischemic lesions.19 Similarly, between 90 days and 12 months, there was much greater variability in recovery for both groups. This resulted in a bimodal distribution of very good and very poor outcomes. However, the favorable peak for ICH patients was much smaller than the unfavorable, demonstrating a skew toward poorer outcomes.20
It is essential to consider potential confounders that may also impact outcomes. In general, the two groups were quite similar. However, the ICH group had a higher average income and education level than patients with IS. This is particularly interesting because one might expect someone with more resources to have a better outcome, especially in the long-term, with greater access to specialty care. Alternatively, higher socioeconomic status may influence decisions regarding withdrawal of care. While this might explain initially higher rates of mRS 6 values, it would be less likely to be a factor later on in recovery and does not explain the continued high rate of mortality in ICH patients. In addition, we must consider that a portion of the patients with IS were treated with thrombolysis and recanalization. While one would expect that the greatest impact on function would be seen by decreasing lesion volume and thus accounted for by matching, it is possible that function is improved by other unidentified mechanisms. Finally, considering differences in life events such as stroke recurrence, trauma, myocardial infarctions (MI), and cancer between stroke subtypes could be important when evaluating long-term prognosis. Unfortunately, these outcomes were not collected as part of our prospective registry and we did not have adequate access to the complete medical record to ensure that all events would be captured. In review of the literature, rates of MI and arrhythmia are reportedly higher in those following ischemic stroke,21 as well as recurrent stroke risk given factors such as atrial fibrillation.22,23 There is no reason to believe that rates of trauma or cancer should vary significantly between groups. This would suggest that those with IS could theoretically have a recurrent event or new health problem, theoretically worsening their prognosis and if anything would have muted our findings that those with ICH experience poorer outcomes at each time point.
Along with directing prognosis, the results of this study are important in considering the design of future clinical trials. First and foremost, it illustrates the significant need for treatment trials focused on the treatment of ICH given the high rates of poor outcome and mortality. In addition, understanding the expected recovery patterns allows us to determine the most appropriate end points to illustrate improvement. Mortality is clearly a reasonable target given the high rates of death in patients with ICH. Further studies should explore the delayed causes of death in patients to improve overall mortality. In addition, essential functions such as ambulation and independence are reasonable based on observed recovery patterns. Conversely, a larger number of patients with IS display mRS scores within the range of 2–4. For this group, more sensitive markers such as cognition and quality of life may be needed to most effectively separate groups into good and poor outcomes. It should also be noted that despite concerns that the mRS may be too coarse as an endpoint to adequately determine differences in outcome, the differences noted between groups for our study (e.g., 4.32 compared to 3.16), represent significantly different levels of function (in this case, the ability to walk unassisted). Independent mobility greatly affects an individual’s ability to perform their activities of daily living, and independence for elderly people is an important contributing factor to their overall happiness and self-esteem.24
Limitations
This study is not without its limitations. The cohorts were recruited from a single center and are composed of a relatively small sample size, though the largest to date. The cohort was recruited mid 2014 or later to allow for major advances in the treatment of both ischemic (mechanical thrombectomy guided by perfusion imaging) and ICH (reversal agents and surgical intervention), though some of the earliest cases likely did not have access. MRI was not obtained for everyone, and lesion volumes can be less precise when calculated using CT. Calculating mRS scores when not explicitly specified can also be challenging, though agreement was good among reviewers (kappa =0.745). Relying on this method also requires that the patient continue to follow within the healthcare system so that records are available within the electronic medical record (EMR). We are a tertiary referral center, so most stroke patients are brought in from surrounding neighborhoods and access their care within our health system. In addition, the state of Maryland has made it possible to access care at other centers across the state through our EMR, minimizing the risk of migration and further loss to follow-up. Nevertheless, by the 12-month time point, ~100 patients with ICH and ~150 with IS did not have adequate information to determine mRS. The characteristics of those with data at all time points was similar with respect to demographics, stroke severity, and co-existing risk factors to those who were lost to follow-up. However, there is still the potential for unidentified bias. Finally, while this study controlled for age, sex, size and location of stroke, and date of admission, we could not control for all variables, for instance baseline level of mobility or income, and it is possible that there were unknown confounders affecting outcome. Multivariable regression was able to account for those identified through univariate analysis.
Despite these limitations, the data clearly show that patients with ICH not only present with a more severe constellation of symptoms than those with IS of similar size and location, longitudinally they are more likely to remain impaired, or even die, than their ischemic counterparts. The results of this study will allow physicians and health care providers to better understand the differences in recovery patterns between intracerebral hemorrhage and ischemic stroke, better prognosticate and assist patients in decision-making, and better design clinical trials that effectively capture the key factors driving outcomes for their specific population.
Table 2.
Group Outcomes
| Hemorrhages | Ischemic Strokes | P - Values | |
|---|---|---|---|
| (N=300) | (N=300) | ||
| Length of Stay, mean points (SD) | 10.3 (0.6) | 6.9 (0.4) | p < 0.001 |
| Discharge mRS, mean score (SD) | 4.20 (0.09) | 3.42 (0.08) | p < 0.001 |
| 90 Day mRS, mean score (SD) | 4.02 (0.15) | 2.89 (0.14) | p < 0.001 |
| 12 Month mRS, mean score (SD) | 4.32 (0.20) | 3.16 (0.22) | p < 0.001 |
| Mortality, % | 127 (43.2%) | 89 (30.3)% | p = 0.014 |
| p = 0.126 | |||
| 1 - Home | 62 (20.7%) | 105 (35.0%) | |
| 2 - Acute Care Facility | 7 (2.3%) | 10 (3.3%) | |
| 3 - Inpatient Rehab | 97 (32.3%) | 126 (42.0%) | |
| 4 - Health Care Facility | 8 (2.7%) | 7 (2.3%) | |
| 5 - Skilled Nursing Facility | 50 (16.7%) | 37 (12.3%) | |
| 6 - Expired | 71 (23.7%) | 12 (4.0%) | |
| 7 - Hospice | 4 (1.3%) | 1 (0.3%) | |
| 8 - Intermediate Care Facility (ICF) | 1 (0.3%) | 0 (0%) | |
| 9 - Long Term Care Hospital (LTCH) | 0 (0%) | 1 (0.3%) | |
| 10 - Other Health Care Facility | 0 (0%) | 1 (0.3%) |
Highlights.
Patients with intracranial hemorrhage present with more severe symptoms than those with ischemic stroke
Outcomes are worse long-term after hemorrhage, compared to ischemia of similar size and location
Etiology is important when advising patients and informing them about recovery
Understanding differences in outcomes will allow for better design of trial endpoints
Acknowledgements
As a member of the Marsh Lab, we would like to thank Annie Yang for her significant role in the data collection for this manuscript.
Funding
Dr. Marsh’s work has been funded in part through the American Heart Association (18IPA34170313) and National Institutes of Health (RF1 AG079324).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Declaration of Interests and Disclosures
We declare no competing interests or additional disclosures.
Contributor Information
Sophia Girgenti, The Johns Hopkins Hospital, Department of Neurology, Baltimore, MD.
Justin Lu, The Johns Hopkins Hospital, Department of Neurology, Baltimore, MD.
Elisabeth Marsh, The Johns Hopkins Hospital, Department of Neurology, Baltimore, MD.
References
- 1.Lattanzi S, Silvestrini M. Blood pressure in acute intra-cerebral hemorrhage. Ann Transl Med. 2016;4(16). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Carlberg B, Asplund K, Hagg E. Comments, Opinions, and Reviews Factors Influencing Admission Blood Pressure Levels in Patients With Acute Stroke. Stroke. 1991;22(4):527–530. [DOI] [PubMed] [Google Scholar]
- 3.Gujjar AR, Deibert E, Manno EM, et al. Mechanical ventilation for ischemic stroke and intracerebral hemorrhage. The American Academy of Neurology. 1998;51(2):447–451. [DOI] [PubMed] [Google Scholar]
- 4.Salvadori E, Papi G, Insalata G, et al. Comparison between ischemic and hemorrhagic strokes in functional outcome at discharge from an intensive rehabilitation hospital. Diagnostics. 2021;11(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kelly PJ, Furie KL, Shafqat S, et al. Functional recovery following rehabilitation after hemorrhagic and ischemic stroke. Arch Phys Med Rehabil. 2003;84(7):968–972. [DOI] [PubMed] [Google Scholar]
- 6.Paolucci S, Antonucci G, Grasso MG, et al. Functional Outcome of Ischemic and Hemorrhagic Stroke Patients after Inpatient Rehabilitation: A Matched Comparison. Stroke. 2003;34(12):2861–2865. [DOI] [PubMed] [Google Scholar]
- 7.Harvey RL, Macko RF, Stein J, et al. Stroke Recovery and Rehabilitation. Demos Medical Publishing; 2008. [Google Scholar]
- 8.Demchuk AM, Goyal M, Menon BK, et al. Endovascular Treatment for Small Core and Anterior Circulation Proximal Occlusion with Emphasis on Minimizing CT to Recanalization Times (ESCAPE) Trial: Methodology. International Journal of Stroke. 2015;10(3):429–438. [DOI] [PubMed] [Google Scholar]
- 9.Berkhemer OA, Fransen PSS, Beumer D, et al. A Randomized Trial of Intraarterial Treatment for Acute Ischemic Stroke. New England Journal of Medicine. 2015;372(1):11–20. [DOI] [PubMed] [Google Scholar]
- 10.Campbell BCV, Mitchell PJ, Kleinig TJ, et al. Endovascular Therapy for Ischemic Stroke with Perfusion-Imaging Selection. New England Journal of Medicine. 2015;372(11):1009–1018. [DOI] [PubMed] [Google Scholar]
- 11.Saver J, Goyal M, Bonafe A, et al. Stent-retriever thrombectomy after intravenous t-PA vs. t-PA alone in stroke . N Engl J Med. 2015;372(24):2285–2295. [DOI] [PubMed] [Google Scholar]
- 12.Rankin J. Cerebral Vascular Accidents in Patients over the Age of 60: II. Prognosis. Scott Med J. 1957;2(5):200–215. [DOI] [PubMed] [Google Scholar]
- 13.Marsh EB, Girgenti S, Llinas EJ, et al. Outcomes in Patients with Minor Stroke: Diagnosis and Management in the Post-thrombectomy Era. Neurotherapeutics. 2023;20(3):732–743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Brott T, Adams HP, Olinger CP, et al. Measurements of Acute Cerebral Infarction: A Clinical Examination Scale. Stroke 1989;20(7):864–870. [DOI] [PubMed] [Google Scholar]
- 15.Carestream Health. Published online 2019.
- 16.Chae J, Zorowitz R, Johnston M. FUNCTIONAL OUTCOME OF HEMORRHAGIC AND NONHEMORRHAGIC STROKE PATIENTS AFTER IN-PATIENT REHABILITATION. Am J Phys Med Rehabil. 1996;75(3):177–186. [DOI] [PubMed] [Google Scholar]
- 17.Perna R, Temple J. Rehabilitation Outcomes: Ischemic versus Hemorrhagic Strokes. Behavioural Neurology. 2015;2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chiu D, Peterson L, Elkind MSV, et al. Comparison of Outcomes after Intracerebral Hemorrhage and Ischemic Stroke. Journal of Stroke and Cerebrovascular Diseases. 2010;19(3):225–229. [DOI] [PubMed] [Google Scholar]
- 19.Marsh EB, Khan S, Llinas RH, et al. Multidomain cognitive dysfunction after minor stroke suggests generalized disruption of cognitive networks. Brain Behav. 2022;12(5). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Girgenti SG, Brunson AO, Marsh EB. Baseline Function and Rehabilitation Are as Important as Stroke Severity as Long-term Predictors of Cognitive Performance Post-stroke. Am J Phys Med Rehabil. 2023;102(2):S43–S50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Cheng C-Y, Hsu C-Y, Wang T-C, Jeng Y-C, Yang W-H. Evaluation of cardiac complications following hemorrhagic stroke using 5-year centers for disease control and prevention (CDC) database. J Clin Med. 2018;7(12):519. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Wong Y-S, Tsai C-F, Ong C-T. Risk factors for stroke recurrence in patients with hemorrhagic stroke. Nature Sci Rep. 2022;12:17151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Flach C, Muruet W, Wolfe CDA, et al. Risk and secondary prevention of stroke recurrence: a population-base cohort study. Stroke 2020;51(8):2435–2444. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Ulfiana E, Saelindra M, Suyatmana SH. Level of Independence Towards Self-Esteem in The Elderly. Atlantis Press. 2017. [Google Scholar]