Saposnik, Gustavo

Abstract Aims We developed new machine learning (ML) models and externally validated existing statistical models [ischaemic stroke predictive risk score (iScore) and totalled health risks in vascular events (THRIVE) scores] for predicting the composite of recurrent stroke or all-cause mortality at 90 days and at 3 years after hospitalization for first acute ischaemic stroke (AIS). Methods and results In adults hospitalized with AIS from January 2005 to November 2016, with follow-up until November 2019, we developed three ML models [random forest (RF), support vector machine (SVM), and extreme gradient boosting (XGBOOST)] and externally validated the iScore and THRIVE scores for predicting the composite outcomes after AIS hospitalization, using data from 721 patients and 90 potential predictor variables. At 90 days and 3 years, 11 and 34% of patients, respectively, reached the composite outcome. For the 90-day prediction, the area under the receiver operating characteristic curve (AUC) was 0.779 for RF, 0.771 for SVM, 0.772 for XGBOOST, 0.720 for iScore, and 0.664 for THRIVE. For 3-year prediction, the AUC was 0.743 for RF, 0.777 for SVM, 0.773 for XGBOOST, 0.710 for iScore, and 0.675 for THRIVE. Conclusion The study provided three ML-based predictive models that achieved good discrimination and clinical usefulness in outcome prediction after AIS and broadened the application of the iScore and THRIVE scoring system for long-term outcome prediction. Our findings warrant comparative analyses of ML and existing statistical method–based risk prediction tools for outcome prediction after AIS in new data sets.

Objectives: Limited information is available on the prevalence of cardiorespiratory arrest (CA) and its predisposing factors in patients with acute ischemic stroke. We analyzed/evaluated clinical characteristics, risk factors, access to care, and survival in stroke patients who developed a cardiac arrest. Methods: This is a multicenter cohort study including consecutive patients with acute ischemic stroke admitted to regional stroke centers participating in the Registry of the Canadian Stroke Network from July 2003 to March 2008. Patients with DNR order on admission were excluded. CA was defined as cardiac or respiratory arrest within 30 days of admission documented by a physician and identified from resuscitation records. The main outcome measures were the 30-day and 1-year survival among patients with CA vs. no CA. We also assessed the association between medical complications during hospitalization and CA. Results: Overall, 9,019 patients with acute ischemic stroke were included. Cardiac arrest was observed in 352 patients (3.9%). In patients with CA, 30-day survival (17.9% vs. 90.7%, P<0.001) and 1-year survival (4.5% vs. 80.1, p<0.001) were significantly lower. Less than 1% of stroke patients with CA were discharged home (vs. 47.3%; p<0.0001). Other outcome measures are summarized in Table 1. On multivariable analyses, older age, more severe stroke, previous history of myocardial infarction, congestive heart failure, atrial fibrillation, and diabetes were associated with CA. Conclusions: Cardiac arrest affects less than 5% of patients admitted with an acute ischemic stroke and is associated with poor outcomes. Less than 1% of patients were independent or discharged home.

Introduction: Pulmonary embolism (PE) is an uncommon medical complication after stroke. Predisposing factors include deep vein thrombosis (DVT) in patients with hemiplegia or an underlying hypercoagulable state. However, little information is known regarding PE in stroke patients. Objective: We evaluated clinical characteristics, predisposing factors, and outcomes in stroke patients who developed PE. Methods: We included patients with an acute ischemic stroke (AIS) admitted to the participating institutions in the Registry of the Canadian Stroke Network between 2003 to 2008. Pulmonary embolism was diagnosed by nuclear imaging within 30 days of the stroke case index. Demographic data and clinical variables were collected. Logistic regression and survival analyses were completed to determine the association of risk factors with the outcomes of interest. Outcome Measures: primary outcome was death or disability at discharge defined as the modified Rankin scale >3. Secondary outcomes include admission to the Intensive Care Unit, disposition, and length of hospital stay, death at 3 months and at 1 year. Results: Among 11,287 patients with AIS, PE was found in 89 (0.78%) of patients. The development of PE was associated with higher risk of death in 30 days (25.8%vs 13.6%;p <0.001) and 1 year, (47.2% vs 24.6%;p <0.001). Disability was also more common in stroke patients with PE (85.4% vs 63.6% without PE; p <0.001). Mean length of stay was longer in stroke patients with PE (36 vs 16 days; p<0.001). Past medical history of cancer or deep vein thrombosis, history of cardiac arrest or deep vein thrombosis during admission were associated with PE. After adjustment, PE was associated with lower survival at 30 days (p value = 0.0012) and 1 year (p value < 0.0001) (Figures 1 & 2 represent survival function). Conclusions: In this large study, PE occurs in approximately 1% of AIS patients. PE was associated with higher disability, longer length of stay and lower short and long-term survival.

Background Since 2015, endovascular therapy has been established as a standard of care for acute stroke. This has caused major challenges regarding the organization of systems of care, which have to meet the increasing demand for thrombectomies. This study aims to evaluate how endovascular therapy decisions made by European and North American physicians under their current local resources differ from those made under assumed ideal conditions. Methods In an international, multidisciplinary survey, physicians involved in acute stroke care were asked to give their treatment decisions to 10 out of 22 randomly assigned stroke case-scenarios. Participants stated (a) their treatment approach under assumed ideal conditions (without any external limitations) and (b) the treatment they would pursue under their current local resources. Resources gaps (ideal minus current endovascular therapy rates) were calculated for different countries/states/provinces and correlated to economic and healthcare key metrics (gross domestic product-per-capita, public or private health insurance coverage, etc.). Results A total of 607 physicians, among them 218 from North America and 136 from 25 European countries, responded to the survey. Resources gaps in the majority of North American states/provinces and European countries were small (<5%). The highest gaps were observed among few European countries, namely Poland (30%) and the United Kingdom (33%). The magnitude of the resources gap did not correlate to national economic or healthcare metrics. Discussion and conclusion In the majority of North American states/provinces and European countries covered in this study, the discrepancy between endovascular therapy decisions under current local resources and assumed ideal conditions seems to be small, even in countries with a limited economic status and healthcare infrastructure.

Background and Purpose— The American Heart Association and the American Stroke Association guidelines for early management of patients with ischemic stroke offer guidance to physicians involved in acute stroke care and clarify endovascular treatment indications. The purpose of this study was to assess concordance of physicians’ endovascular treatment decision-making with current American Heart Association and the American Stroke Association stroke treatment guidelines using a survey-approach and to explore how decision-making in the absence of guideline recommendations is approached. Methods— In an international cross-sectional survey (UNMASK-EVT), physicians were randomly assigned 10 of 22 case scenarios (8 constructed with level 1A and 11 with level 2B evidence for endovascular treatment and 3 scenarios without guideline coverage) and asked to declare their treatment approach (1) under their current local resources and (2) assuming there were no external constraints. The proportion of physicians offering endovascular therapy (EVT) was calculated. Subgroup analysis was performed for different specialties, geographic regions, with regard to physicians’ age, endovascular, and general stroke treatment experience. Results— When facing level 1A evidence, participants decided in favor of EVT in 86.8% under current local resources and in 90.6% under assumed ideal conditions, that is, 9.4% decided against EVT even under assumed ideal conditions. In case scenarios with level 2B evidence, 66.3% decided to proceed with EVT under current local resources and 69.7% under assumed ideal conditions. Conclusions— There is potential for improving thinking around the decision to offer endovascular treatment, since physicians did not offer EVT even under assumed ideal conditions in 9.4% despite facing level 1A evidence. A majority of physicians would offer EVT even for level 2B evidence cases.

Background and purpose Differences in the treatment practice of female and male physicians have been shown in several medical subspecialties. It is currently not known whether this also applies to endovascular stroke treatment. The purpose of this study was to explore whether there are differences in endovascular treatment decisions made by female and male stroke physicians and neurointerventionalists. Methods In an international survey, stroke physicians and neurointerventionalists were randomly assigned 10 case scenarios and asked how they would treat the patient: (A) assuming there were no external constraints and (B) given their local working conditions. Descriptive statistics were used to describe baseline demographics, and the adjusted OR for physician gender as a predictor of endovascular treatment decision was calculated using logistic regression. Results 607 physicians (97 women, 508 men, 2 who did not wish to declare) participated in this survey. Physician gender was neither a significant predictor for endovascular treatment decision under assumed ideal conditions (endovascular therapy was favored by 77.0% of female and 79.3% of male physicians, adjusted OR 1.03, P=0.806) nor under current local resources (endovascular therapy was favored by 69.1% of female and 76.9% of male physicians, adjusted OR 1.03, P=0.814). Conclusion Endovascular therapy decision making between male and female physicians did not differ under assumed ideal conditions or under current local resources.

Background and Purpose— Little is known about the real-life factors that clinicians use in selection of patients that would receive endovascular treatment (EVT) in the real world. We sought to determine patient, practitioner, and health system factors associated with therapeutic decisions around endovascular treatment. Methods— We conducted a multinational cross-sectional web-based study comprising of 607 clinicians and interventionalists from 38 countries who are directly involved in acute stroke care. Participants were randomly allocated to 10 from a pool of 22 acute stroke case scenarios. Each case was classified as either Class I, Class II, or unknown evidence according to the current guidelines. We used logistic regression analysis applying weight of evidence approach. Main outcome measures were multilevel factors associated with EVT, adherence to current EVT guidelines, and practice gaps between current and ideal practice settings. Results— Of the 1330 invited participants, 607 (45.6%) participants completed the study (53.7% neurologists, 28.5% neurointerventional radiologists, 17.8% other clinicians). The weighed evidence approach revealed that National Institutes of Health Stroke Scale (34.9%), level of evidence (30.2%), ASPECTS (Alberta Stroke Program Early CT Score) or ischemic core volume (22.4%), patient’s age (21.6%), and clinicians’ experience in EVT use (19.3%) are the most important factors for EVT decision. Of 2208 responses that met Class I evidence for EVT, 1917 (86.8%) were in favor of EVT. In case scenarios with no available guidelines, 1070 of 1380 (77.5%) responses favored EVT. Comparison between current and ideal practice settings revealed a small practice gap (941 of 6070 responses, 15.5%). Conclusions— In this large multinational survey, stroke severity, guideline-based level of evidence, baseline brain imaging, patients’ age and physicians’ experience were the most relevant factors for EVT decision-making. The high agreement between responses and Class I guideline recommendations and high EVT use even when guidelines were not available reflect the real-world acceptance of EVT as standard of care in patients with disabling acute ischemic stroke.

Background The greatest potential to reduce the burden of stroke is by primary prevention of first-ever stroke, which constitutes three quarters of all stroke. In addition to population-wide prevention strategies (the ‘mass’ approach), the ‘high risk’ approach aims to identify individuals at risk of stroke and to modify their risk factors, and risk, accordingly. Current methods of assessing and modifying stroke risk are difficult to access and implement by the general population, amongst whom most future strokes will arise. To help reduce the burden of stroke on individuals and the population a new app, the Stroke Riskometer™, has been developed. We aim to explore the validity of the app for predicting the risk of stroke compared with current best methods. Methods 752 stroke outcomes from a sample of 9501 individuals across three countries (New Zealand, Russia and the Netherlands) were utilized to investigate the performance of a novel stroke risk prediction tool algorithm (Stroke Riskometer™) compared with two established stroke risk score prediction algorithms (Framingham Stroke Risk Score [FSRS] and QStroke). We calculated the receiver operating characteristics (ROC) curves and area under the ROC curve (AUROC) with 95% confidence intervals, Harrels C-statistic and D-statistics for measure of discrimination, R2 statistics to indicate level of variability accounted for by each prediction algorithm, the Hosmer-Lemeshow statistic for calibration, and the sensitivity and specificity of each algorithm. Results The Stroke Riskometer™ performed well against the FSRS five-year AUROC for both males (FSRS = 75·0% (95% CI 72·3%–77·6%), Stroke Riskometer™ = 74·0(95% CI 71·3%–76·7%) and females [FSRS = 70·3% (95% CI 67·9%–72·8%, Stroke Riskometer™ = 71·5% (95% CI 69·0%–73·9%)], and better than QStroke [males–59·7% (95% CI 57·3%–62·0%) and comparable to females = 71·1% (95% CI 69·0%–73·1%)]. Discriminative ability of all algorithms was low (C-statistic ranging from 0·51–0·56, D-statistic ranging from 0·01–0·12). Hosmer-Lemeshow illustrated that all of the predicted risk scores were not well calibrated with the observed event data ( P < 0·006). Conclusions The Stroke Riskometer™ is comparable in performance for stroke prediction with FSRS and QStroke. All three algorithms performed equally poorly in predicting stroke events. The Stroke Riskometer™ will be continually developed and validated to address the need to improve the current stroke risk scoring systems to more accurately predict stroke, particularly by identifying robust ethnic/race ethnicity group and country specific risk factors.

Background Limited information is available on the frequency of pulmonary embolism ( PE ) in patients with an acute ischemic stroke ( AIS ). We evaluated clinical characteristics, predisposing factors, and outcomes in AIS patients with PE . Methods and Results We included all AIS patients admitted to participating institutions in the R egistry of the C anadian S troke N etwork. Clinically PE was documented by a physician and confirmed by computed tomography pulmonary angiography within 30 days of the stroke case index. The primary outcome was death or disability at discharge. Secondary outcomes included disposition, length of hospital stay, mortality at 3 months and 1 year. Among 11 287 patients with AIS , PE was found in 89 (0.78%) patients. History of cancer, deep vein thrombosis ( DVT )/ PE , and DVT during the hospitalization were associated with PE . PE was associated with higher risk of death at 30 days (25.8% versus 13.6%; P <0.001), at 1 year (47.2% versus 24.6%; P <0.001), and disability at discharge (85.4% versus 63.6%; P <0.001). Mean length of stay was longer in stroke patients with PE (36 versus 16 days; P =0.001). After adjusting for age, sex, and stroke severity, PE remained associated with lower survival at 30 days and 1 year, and death or disability at discharge ( OR 3.02; 95% CI 1.56 to 5.83). Conclusions In this large cohort study, PE occurred in nearly 1% of AIS patients. PE was more common in patients with severe stroke, history of cancer, previous DVT / PE or acute DVT and associated with lower short‐ and long‐term survival, greater disability, and longer length of stay.

The availability of population-based epidemiological data on the incident risk of stroke is very scarce in Argentina and other Latin American countries. In response to the priorities established by the World Health Organization and the United Nations, PREVISTA was envisaged as a population-based program to determine the risk of first-ever and recurrent stroke and transient ischemic attack incidence and mortality in Tandil, Buenos Aires, Argentina. The study will be conducted according to Standardized Tools for Stroke Surveillance (STEPS Stroke) methodology and will enroll all new (incident) and recurrent consecutive cases of stroke and transient ischemic attack in the City of Tandil between May 1st, 2013 and April 30, 2015. The study will include patients with ischemic stroke, non-traumatic primary intracerebral hemorrhage, subarachnoid hemorrhage, and transient ischemic attack. To ensure the inclusion of every cerebrovascular event during an observation period of two years, we will instrument an ‘intensive screening program’, consisting of a comprehensive daily tracking of every potential event of stroke or transient ischemic attack using multiple overlapping sources. Mortality would be determined during follow-up for every enrolled patient. Also, fatal community events would be screened daily through revision of death certificates at funeral homes and local offices of vital statistics. All causes of death will be adjudicated by an ad-hoc committee. The close population of Tandil is representative of a large proportion of Latin-American countries with low- and middle-income economies. The findings and conclusions of PREVISTA may provide data that could support future health policy decision-making in the region.