Deep learning survival model predicts outcome after intracerebral hemorrhage from initial CT scan.

BACKGROUND: Predicting functional impairment after intracerebral hemorrhage (ICH) provides valuable information for planning of patient care and rehabilitation strategies. Current prognostic tools are limited in making long term predictions and require multiple expert-defined inputs and interpretation that make their clinical implementation challenging. This study aimed to predict long term functional impairment of ICH patients from admission non-contrast CT scans, leveraging deep learning models in a survival analysis framework. METHODS: We used the admission non-contrast CT scans from 882 patients from the Massachusetts General Hospital ICH Study for training, hyperparameter optimization, and model selection, and 146 patients from the Yale New Haven ICH Study for external validation of a deep learning model predicting functional outcome. Disability (modified Rankin scale [mRS] > 2), severe disability (mRS > 4), and dependent living status were assessed via telephone interviews after 6, 12, and 24 months. The prediction methods were evaluated by the c-index and compared with ICH score and FUNC score. RESULTS: Using non-contrast CT, our deep learning model achieved higher prediction accuracy of post-ICH dependent living, disability, and severe disability by 6, 12, and 24 months (c-index 0.742 [95% CI -0.700 to 0.778], 0.712 [95% CI -0.674 to 0.752], 0.779 [95% CI -0.733 to 0.832] respectively) compared with the ICH score (c-index 0.673 [95% CI -0.662 to 0.688], 0.647 [95% CI -0.637 to 0.661] and 0.697 [95% CI -0.675 to 0.717]) and FUNC score (c-index 0.701 [95% CI- 0.698 to 0.723], 0.668 [95% CI -0.657 to 0.680] and 0.727 [95% CI -0.708 to 0.753]). In the external independent Yale-ICH cohort, similar performance metrics were obtained for disability and severe disability (c-index 0.725 [95% CI -0.673 to 0.781] and 0.747 [95% CI -0.676 to 0.807], respectively). Similar AUC of predicting each outcome at 6 months, 1 and 2 years after ICH was achieved compared with ICH score and FUNC score. CONCLUSION: We developed a generalizable deep learning model to predict onset of dependent living and disability after ICH, which could help to guide treatment decisions, advise relatives in the acute setting, optimize rehabilitation strategies, and anticipate long-term care needs.

Time-Dependent Changes in Hematoma Expansion Rate after Supratentorial Intracerebral Hemorrhage and Its Relationship with Neurological Deterioration and Functional Outcome.

BACKGROUND: Hematoma expansion (HE) following an intracerebral hemorrhage (ICH) is a modifiable risk factor and a treatment target. We examined the association of HE with neurological deterioration (ND), functional outcome, and mortality based on the time gap from onset to baseline CT. METHODS: We included 567 consecutive patients with supratentorial ICH and baseline head CT within 24 h of onset. ND was defined as a ≥4-point increase on the NIH stroke scale (NIHSS) or a ≥2-point drop on the Glasgow coma scale. Poor outcome was defined as a modified Rankin score of 4 to 6 at 3-month follow-up. RESULTS: The rate of HE was higher among those scanned within 3 h (124/304, 40.8%) versus 3 to 24 h post-ICH onset (53/263, 20.2%) (p < 0.001). However, HE was an independent predictor of ND (p < 0.001), poor outcome (p = 0.010), and mortality (p = 0.003) among those scanned within 3 h, as well as those scanned 3-24 h post-ICH (p = 0.043, p = 0.037, and p = 0.004, respectively). Also, in a subset of 180/567 (31.7%) patients presenting with mild symptoms (NIHSS ≤ 5), hematoma growth was an independent predictor of ND (p = 0.026), poor outcome (p = 0.037), and mortality (p = 0.027). CONCLUSION: Despite decreasing rates over time after ICH onset, HE remains an independent predictor of ND, functional outcome, and mortality among those presenting >3 h after onset or with mild symptoms.

APOE ε4 and Intracerebral Hemorrhage in Patients With Brain Arteriovenous Malformation.

Importance: Intracerebral hemorrhage (ICH) is a serious complication of brain arteriovenous malformation (AVM). Apolipoprotein E (APOE) ε4 is a well-known genetic risk factor for ICH among persons without AVM, and cerebral amyloid angiopathy is a vasculopathy frequently observed in APOE ε4 carriers that may increase the risk of ICH. Objective: To assess whether APOE ε4 is associated with a higher risk of ICH in patients with a known AVM. Design, Setting, and Participants: This cross-sectional study including 412 participants was conducted in 2 stages (discovery and replication) using individual-level data from the UK Biobank (released March 2012 and last updated October 2023) and the All of Us Research Program (commenced on May 6, 2018, with its latest update provided in October 2023). The occurrence of AVM and ICH was ascertained at the time of enrollment using validated International Classification of Diseases, Ninth Revision and Tenth Revision, codes. Genotypic data on the APOE variants rs429358 and rs7412 were used to ascertain the ε status. Main Outcomes and Measures: For each study, the association between APOE ε4 variants and ICH risk was assessed among patients with a known AVM by using multivariable logistic regression. Results: The discovery phase included 253 UK Biobank participants with known AVM (mean [SD] age, 56.6 [8.0] years, 119 [47.0%] female), of whom 63 (24.9%) sustained an ICH. In the multivariable analysis of 240 participants of European ancestry, APOE ε4 was associated with a higher risk of ICH (odds ratio, 4.58; 95% CI, 2.13-10.34; P < .001). The replication phase included 159 participants with known AVM enrolled in All of Us (mean [SD] age, 57.1 [15.9] years; 106 [66.7%] female), of whom 29 (18.2%) sustained an ICH. In multivariable analysis of 101 participants of European ancestry, APOE ε4 was associated with higher risk of ICH (odds ratio, 4.52; 95% CI, 1.18-19.38; P = .03). Conclusions and Relevance: The results of this cross-sectional study of patients from the UK Biobank and All of Us suggest that information on APOE ε4 status may help identify patients with brain AVM who are at particularly high risk of ICH and that cerebral amyloid angiopathy should be evaluated as a possible mediating mechanism of the observed association.

Suboptimal Sleep Duration Is Associated With Poorer Neuroimaging Brain Health Profiles in Middle-Aged Individuals Without Stroke or Dementia.

BACKGROUND: The American Heart Association's Life's Simple 7, a public health construct capturing key determinants of cardiovascular health, became the Life's Essential 8 after the addition of sleep duration. The authors tested the hypothesis that suboptimal sleep duration is associated with poorer neuroimaging brain health profiles in asymptomatic middle-aged adults. METHODS AND RESULTS: The authors conducted a prospective magnetic resonance neuroimaging study in middle-aged individuals without stroke or dementia enrolled in the UK Biobank. Self-reported sleep duration was categorized as short (<7 hours), optimal (7-<9 hours), or long (≥9 hours). Evaluated neuroimaging markers included the presence of white matter hyperintensities (WMHs), volume of WMH, and fractional anisotropy, with the latter evaluated as the average of 48 white matter tracts. Multivariable logistic and linear regression models were used to test for an association between sleep duration and these neuroimaging markers. The authors evaluated 39 771 middle-aged individuals. Of these, 28 912 (72.7%) had optimal, 8468 (21.3%) had short, and 2391 (6%) had long sleep duration. Compared with optimal sleep, short sleep was associated with higher risk of WMH presence (odds ratio, 1.11 [95% CI, 1.05-1.18]; P<0.001), larger WMH volume (beta=0.06 [95% CI, 0.04-0.08]; P<0.001), and worse fractional anisotropy profiles (beta=-0.04 [95% CI, -0.06 to -0.02]; P=0.001). Compared with optimal sleep, long sleep duration was associated with larger WMH volume (beta=0.04 [95% CI, 0.01-0.08]; P=0.02) and worse fractional anisotropy profiles (beta=-0.06 [95% CI, -0.1 to -0.02]; P=0.002), but not with WMH presence (P=0.6). CONCLUSIONS: Among middle-aged adults without stroke or dementia, suboptimal sleep duration is associated with poorer neuroimaging brain health profiles. Because these neuroimaging markers precede stroke and dementia by several years, these findings are consistent with other findings evaluating early interventions to improve this modifiable risk factor.

Polygenic Susceptibility to Hypertension and Cognitive Performance in Middle-aged Persons Without Stroke or Dementia.

BACKGROUND AND OBJECTIVES: Mounting evidence indicates that hypertension leads to a higher risk of dementia. Hypertension is a highly heritable trait, and a higher polygenic susceptibility to hypertension (PSH) is known to associate with a higher risk of dementia. We tested the hypothesis that a higher PSH leads to worse cognitive performance in middle-aged persons without dementia. Confirming this hypothesis would support follow-up research focused on using hypertension-related genomic information to risk-stratify middle-aged adults before hypertension develops. METHODS: We conducted a nested cross-sectional genetic study within the UK Biobank (UKB). Study participants with a history of dementia or stroke were excluded. We categorized participants as having low (≤20th percentile), intermediate, or high (≥80th percentile) PSH according to results of 2 polygenic risk scores for systolic and diastolic blood pressure (BP) generated with data on 732 genetic risk variants. A general cognitive ability score was calculated as the first component of an analysis that included the results of 5 cognitive tests. Primary analyses focused on Europeans, and secondary analyses included all race/ethnic groups. RESULTS: Of the 502,422 participants enrolled in the UKB, 48,118 (9.6%) completed the cognitive evaluation, including 42,011 (8.4%) of European ancestry. Multivariable regression models using systolic BP-related genetic variants indicated that compared with study participants with a low PSH, those with intermediate and high PSH had reductions of 3.9% (ß -0.039, SE 0.012) and 6.6% (ß -0.066, SE 0.014), respectively, in their general cognitive ability score (p < 0.001). Secondary analyses including all race/ethnic groups and using diastolic BP-related genetic variants yielded similar results (p < 0.05 for all tests). Analyses evaluating each cognitive test separately indicated that reaction time, numeric memory, and fluid intelligence drove the association between PSH and general cognitive ability score (all individual tests, p < 0.05). DISCUSSION: Among nondemented, community-dwelling, middle-aged Britons, a higher PSH is associated with worse cognitive performance. These findings suggest that genetic predisposition to hypertension influences brain health in persons who have not yet developed dementia. Because information on genetic risk variants for elevated BP is available long before the development of hypertension, these results lay the foundation for further research focused on using genomic data for the early identification of high-risk middle-aged adults.

Polygenic Susceptibility to Hypertension and Blood Pressure Control in Stroke Survivors.

BACKGROUND AND OBJECTIVES: Blood pressure (BP) is often not at goal in stroke survivors, leaving individuals vulnerable to additional vascular events. Given that BP is a highly heritable trait, we hypothesize that a higher polygenic susceptibility to hypertension (PSH) leads to worse BP control in stroke survivors. METHODS: We conducted a study within the UK Biobank evaluating persons of European ancestry who survived an ischemic or hemorrhagic stroke. To model the PSH, we created polygenic risk scores (PRSs) for systolic and diastolic BP using 732 genetic variants. We divided the PRSs into quintiles and used linear/logistic regression to test whether higher PSH led to higher observed BP, uncontrolled BP (systolic BP > 140 mm Hg or diastolic BP > 90 mm Hg), and resistant BP (uncontrolled BP despite being on ≥3 antihypertensive drugs). We conducted an independent replication using data from the Vitamin Intervention for Stroke Prevention (VISP) trial. RESULTS: We analyzed 5,940 stroke survivors. When comparing stroke survivors with very low vs very high PSH, the mean systolic BP was 137 (SD 18) vs 143 (SD 20, p < 0.001), the mean diastolic BP was 81 (SD 10) vs 84 (SD 11, p < 0.001), the prevalence of uncontrolled BP was 42.8% vs 57.2% (p < 0.001), and the prevalence of resistant hypertension was 3.9% vs 11% (p < 0.001). Results remained significant using multivariable models (p < 0.001) and were replicated in the VISP study (all tests with p < 0.05). DISCUSSION: A higher PSH is associated with worse BP control in stroke survivors. These findings point to genetic predisposition as an important determinant of poorly controlled BP in this population.

Development and Validation of a Model to Identify Critical Brain Injuries Using Natural Language Processing of Text Computed Tomography Reports.

Importance: Clinical text reports from head computed tomography (CT) represent rich, incompletely utilized information regarding acute brain injuries and neurologic outcomes. CT reports are unstructured; thus, extracting information at scale requires automated natural language processing (NLP). However, designing new NLP algorithms for each individual injury category is an unwieldy proposition. An NLP tool that summarizes all injuries in head CT reports would facilitate exploration of large data sets for clinical significance of neuroradiological findings. Objective: To automatically extract acute brain pathological data and their features from head CT reports. Design, Setting, and Participants: This diagnostic study developed a 2-part named entity recognition (NER) NLP model to extract and summarize data on acute brain injuries from head CT reports. The model, termed BrainNERD, extracts and summarizes detailed brain injury information for research applications. Model development included building and comparing 2 NER models using a custom dictionary of terms, including lesion type, location, size, and age, then designing a rule-based decoder using NER outputs to evaluate for the presence or absence of injury subtypes. BrainNERD was evaluated against independent test data sets of manually classified reports, including 2 external validation sets. The model was trained on head CT reports from 1152 patients generated by neuroradiologists at the Yale Acute Brain Injury Biorepository. External validation was conducted using reports from 2 outside institutions. Analyses were conducted from May 2020 to December 2021. Main Outcomes and Measures: Performance of the BrainNERD model was evaluated using precision, recall, and F1 scores based on manually labeled independent test data sets. Results: A total of 1152 patients (mean [SD] age, 67.6 [16.1] years; 586 [52%] men), were included in the training set. NER training using transformer architecture and bidirectional encoder representations from transformers was significantly faster than spaCy. For all metrics, the 10-fold cross-validation performance was 93% to 99%. The final test performance metrics for the NER test data set were 98.82% (95% CI, 98.37%-98.93%) for precision, 98.81% (95% CI, 98.46%-99.06%) for recall, and 98.81% (95% CI, 98.40%-98.94%) for the F score. The expert review comparison metrics were 99.06% (95% CI, 97.89%-99.13%) for precision, 98.10% (95% CI, 97.93%-98.77%) for recall, and 98.57% (95% CI, 97.78%-99.10%) for the F score. The decoder test set metrics were 96.06% (95% CI, 95.01%-97.16%) for precision, 96.42% (95% CI, 94.50%-97.87%) for recall, and 96.18% (95% CI, 95.151%-97.16%) for the F score. Performance in external institution report validation including 1053 head CR reports was greater than 96%. Conclusions and Relevance: These findings suggest that the BrainNERD model accurately extracted acute brain injury terms and their properties from head CT text reports. This freely available new tool could advance clinical research by integrating information in easily gathered head CT reports to expand knowledge of acute brain injury radiographic phenotypes.

Genetically Determined Low-Density Lipoprotein Cholesterol and Risk of Subarachnoid Hemorrhage.

We evaluated whether genetically elevated low-density lipoprotein cholesterol (LDL-C) levels are associated with lower risk of intracranial aneurysms and subarachnoid hemorrhage (IA/SAH). We conducted a 2-sample Mendelian randomization (MR) study. Our primary analysis used the inverse-variance weighted method. In secondary analyses, we implemented the MR-PRESSO method, restricted our analysis to LDL-C-specific instruments, and performed multivariate MR. A 1-mmol/l increase in genetically instrumented LDL-C levels was associated with a 17% lower risk of IA/SAH (odds ratio = 0.83, 95% confidence interval = 0.73-0.94, p = 0.004). Results remained consistent in secondary and multivariate analyses (all p < 0.05). Our results provide evidence for an inverse causal relationship between LDL-C levels and risk of IA/SAH. ANN NEUROL 2022;91:145-149.

Cardiovascular Health Disparities in Racial and Other Underrepresented Groups: Initial Results From the All of Us Research Program.

Background All of Us is a novel research program that aims to accelerate research in populations traditionally underrepresented in biomedical research. Our objective was to evaluate the burden of cardiovascular disease (CVD) in broadly defined underrepresented groups. Methods and Results We evaluated the latest data release of All of Us. We conducted a cross-sectional analysis combining survey and electronic health record data to estimate the prevalence of CVD upon enrollment in underrepresented groups defined by race, ethnicity, age (>75 years), disability (not able to carry out everyday physical activities), sexual orientation and gender identity lesbian, gay, bisexual, transgender, queer, intersex, and asexual (LGBTQIA+), income (annual household income <$35 000 US dollars) and education (less than a high school degree). We used multivariate logistic regression to estimate the adjusted odds ratio (OR) and product terms to test for interaction. The latest All of Us data release includes 315 297 participants. Of these, 230 577 (73%) had information on CVD and 17 958 had CVD (overall prevalence, 7.8%; 95% CI, 7.7-7.9). Multivariate analyses adjusted by hypertension, hyperlipidemia, type 2 diabetes mellitus, body mass index, and smoking indicated that, compared with White participants, Black participants had a higher adjusted odds of CVD (OR, 1.21; 95% CI, 1.16-1.27). Higher adjusted odds of CVD were also observed in underrepresented groups defined by other factors, including age >75 years (OR, 1.90; 95% CI, 1.81-1.99), disability (OR, 1.60; 95% CI, 1.53-1.68), and income <$35 000 US dollars (OR, 1.22; 95% CI, 1.17-1.27). Sex significantly modified the odds of CVD in several of the evaluated groups. Conclusions Among participants enrolled in All of Us, underrepresented groups defined based on race, ethnicity and other factors have a disproportionately high burden of CVD. The All of Us research program constitutes a powerful platform to accelerate research focused on individuals in underrepresented groups.

Early head CT in post-cardiac arrest patients: A helpful tool or contributor to self-fulfilling prophecy?

OBJECTIVE: Neuroprognostication guidelines suggest that early head computed tomography (HCT) might be useful in the evaluation of cardiac arrest (CA) patients following return of spontaneous circulation. We aimed to determine the impact of early HCT, performed within the first 6 h following CA, on decision-making following resuscitation. METHODS: We identified a cohort of initially unconscious post-CA patients at a tertiary care academic medical center from 2012 to 2017. Variables pertaining to demographics, CA details, post-CA care, including neuroimaging and neurophysiologic testing, were abstracted retrospectively from the electronic medical records. Changes in management resulting from HCT findings were recorded. Blinded board-certified neurointensivists adjudicated HCT findings related to hypoxic-ischemic brain injury (HIBI) burden. The gray-white matter ratio (GWR) was also calculated. RESULTS: Of 302 patients, 182 (60.2%) underwent HCT within six hours of CA (early HCT group). Approximately 1 in 4 early HCTs were abnormal (most commonly HIBI changes; 78.7%, n = 37), which resulted in a change in management in nearly half of cases (46.8%, n = 22). The most common changes in management were de-escalation in care [including transition to do not resuscitate status), withholding targeted temperature management, and withdrawal of life sustaining therapy (WLST)]. In cases with radiographic HIBI, mean [standard deviation] GWR was lower (1.20 [0.10] vs 1.30 [0.09], P < 0.001) and progression to brain death was higher (44.4% vs 2.9%; P < 0.001). The inter-rater reliability (IRR) of early HCT to determine presence of HIBI between radiology and three neurointensivists had a wide range (κ 0.13-0.66). CONCLUSION: Early HCT identified abnormalities in 25% of cases and frequently influenced therapeutic decisions. Neuroimaging interpretation discrepancies between radiology and neurointensivists are common and agreement on severity of HIBI on early HCT is poor (k 0.11).

Perihematomal Edema After Intracerebral Hemorrhage in Patients With Active Malignancy.

Background and Purpose- Patients with active malignancy are at risk for intracerebral hemorrhage (ICH). We aimed to characterize perihematomal edema (PHE) and hematoma volumes after spontaneous nontraumatic ICH in patients with cancer without central nervous system involvement. Methods- Patients with active malignancy who developed ICH were retrospectively identified through automated searches of institutional databases. Control patients were identified with ICH and without active cancer. Demographic and cancer-specific data were obtained by chart review. Hematoma and PHE volumes were determined using semiautomated methodology. Univariate and multivariate linear regression models were created to assess which variables were associated with hematoma and PHE expansion. Results- Patients with cancer (N=80) and controls (N=136) had similar demographics (all P>0.20), although hypertension was more prevalent among controls (P=0.004). Most patients with cancer had received recent chemotherapy (n=45, 56%) and had recurrence of malignancy (n=43, 54%). Patients with cancer were thrombocytopenic (median platelet count 90 000 [interquartile range, 17 500-211 500]), and most had undergone blood product transfusion (n=41, 51%), predominantly platelets (n=38, 48%). Thirty-day mortality was 36% (n=29). Patients with cancer had significantly increased PHE volumes (23.67 versus 8.61 mL; P=1.88×10-9) and PHE-to-ICH volume ratios (2.26 versus 0.99; P=2.20×10-16). In multivariate analyses, variables associated with PHE growth among patients with cancer were ICH volume (ß=1.29 [95% CI, 1.58-1.30] P=1.30×10-5) and platelet transfusion (ß=15.67 [95% CI, 3.61-27.74] P=0.014). Variables associated with 30-day mortality were ICH volume (odds ratio, 1.06 [95% CI, 1.03-1.10] P=6.76×10-5), PHE volume (odds ratio, 1.07 [95% CI, 1.04-1.09] P=7.40×10-6), PHE growth (odds ratio, 1.05 [95% CI, 1.01-1.10] P=0.01), and platelet transfusion (odds ratio, 1.48 [95% CI, 1.22-1.79] P=0.0001). Conclusions- Patients with active cancer who develop ICH have increased PHE volumes. PHE growth was independent of thrombocytopenia but associated with blood product transfusion. Thirty-day mortality was associated with PHE and ICH volumes and blood product transfusion.

Racial/ethnic disparities in the risk of intracerebral hemorrhage recurrence.

OBJECTIVE: To estimate the risk of intracerebral hemorrhage (ICH) recurrence in a large, diverse, US-based population and to identify racial/ethnic and socioeconomic subgroups at higher risk. METHODS: We performed a longitudinal analysis of prospectively collected claims data from all hospitalizations in nonfederal California hospitals between 2005 and 2011. We used validated diagnosis codes to identify nontraumatic ICH and our primary outcome of recurrent ICH. California residents who survived to discharge were included. We used log-rank tests for unadjusted analyses of survival across racial/ethnic groups and multivariable Cox proportional hazards regression to determine factors associated with risk of recurrence after adjusting for potential confounders. RESULTS: We identified 31,355 California residents with first-recorded ICH who survived to discharge, of whom 15,548 (50%) were white, 6,174 (20%) were Hispanic, 4,205 (14%) were Asian, and 2,772 (9%) were black. There were 1,330 recurrences (4.1%) over a median follow-up of 2.9 years (interquartile range 3.8). The 1-year recurrence rate was 3.0% (95% confidence interval [CI] 2.8%-3.2%). In multivariable analysis, black participants (hazard ratio [HR] 1.22; 95% CI 1.01-1.48; p = 0.04) and Asian participants (HR 1.29; 95% CI 1.10-1.50; p = 0.001) had a higher risk of recurrence than white participants. Private insurance was associated with a significant reduction in risk compared to patients with Medicare (HR 0.60; 95% CI 0.50-0.73; p < 0.001), with consistent estimates across racial/ethnic groups. CONCLUSIONS: Black and Asian patients had a higher risk of ICH recurrence than white patients, whereas private insurance was associated with reduced risk compared to those with Medicare. Further research is needed to determine the drivers of these disparities.

Antiplatelet Therapy After Spontaneous Intracerebral Hemorrhage and Functional Outcomes.

Background and Purpose- Observational data suggest that antiplatelet therapy after intracerebral hemorrhage (ICH) alleviates thromboembolic risk without increasing the risk of recurrent ICH. Given the paucity of data on the relationship between antiplatelet therapy after ICH and functional outcomes, we aimed to study this association in a multicenter cohort. Methods- We meta-analyzed data from (1) the Massachusetts General Hospital ICH registry (n=1854), (2) the Virtual International Stroke Trials Archive database (n=762), and (3) the Yale stroke registry (n=185). Our exposure was antiplatelet therapy after ICH, which was modeled as a time-varying covariate. Our primary outcomes were all-cause mortality and a composite of major disability or death (modified Rankin Scale score 4-6). We used Cox proportional regression analyses to estimate the hazard ratio of death or poor functional outcome as a function of antiplatelet therapy and random-effects meta-analysis to pool the estimated HRs across studies. Additional analyses stratified by hematoma location (lobar and deep ICH) were performed. Results- We included a total of 2801 ICH patients, of whom 288 (10.3%) were started on antiplatelet medications after ICH. Median times to antiplatelet therapy ranged from 7 to 39 days. Antiplatelet therapy after ICH was not associated with mortality (hazard ratio, 0.85; 95% CI, 0.66-1.09), or death or major disability (hazard ratio, 0.83; 95% CI, 0.59-1.16) compared with patients not started on antiplatelet therapy. Similar results were obtained in additional analyses stratified by hematoma location. Conclusions- Antiplatelet therapy after ICH appeared safe and was not associated with all-cause mortality or functional outcome, regardless of hematoma location. Randomized clinical trials are needed to determine the effects and harms of antiplatelet therapy after ICH.