Genetic-Guided Pharmacotherapy for Coronary Artery Disease: A Systematic and Critical Review of Economic Evaluations.

BACKGROUND: Genetic-guided pharmacotherapy (PGx) is not recommended in clinical guidelines for coronary artery disease (CAD). We aimed to examine the extent and quality of evidence from economic evaluations of PGx in CAD and to identify variables influential in changing conclusions on cost-effectiveness. METHODS AND RESULTS: From systematic searches across 6 databases, 2 independent reviewers screened, included, and rated the methodological quality of economic evaluations of PGx testing to guide pharmacotherapy for patients with CAD. Of 35 economic evaluations included, most were model-based cost-utility analyses alone, or alongside cost-effectiveness analyses of PGx testing to stratify patients into antiplatelets (25/35), statins (2/35), pain killers (1/35), or angiotensin-converting enzyme inhibitors (1/35) to predict CAD risk (8/35) or to determine the coumadin doses (1/35). To stratify patients into antiplatelets (96/151 comparisons with complete findings of PGx versus non-PGx), PGx was more effective and more costly than non-PGx clopidogrel (28/43) but less costly than non-PGx prasugrel (10/15) and less costly and less effective than non-PGx ticagrelor (22/25). To predict CAD risk (51/151 comparisons), PGx using genetic risk scores was more effective and less costly than clinical risk score (13/17) but more costly than no risk score (16/19) or no treatment (9/9). The remaining comparisons were too few to observe any trend. Mortality risk was the most common variable (47/294) changing conclusions. CONCLUSIONS: Economic evaluations to date found PGx to stratify patients with CAD into antiplatelets or to predict CAD risk to be cost-effective, but findings varied based on the non-PGx comparators, underscoring the importance of considering local practice in deciding whether to adopt PGx.

Burden of intracerebral haemorrhage in Europe: forecasting incidence and mortality between 2019 and 2050.

Background: Anticipating the burden of intracerebral haemorrhage is crucial for proactive management and building resilience against future health challenges. Prior forecasts are based on population demography and to a lesser extent epidemiological trends. This study aims to utilise selected modifiable risk factors and socio-demographic indicators to forecast the incidence and mortality of intracerebral haemorrhage in Europe between 2019 and 2050. Methods: Three intracerebral haemorrhage risk factors identified in the Global Burden of Diseases, Injuries, and Risk Factors study (GBD 2019)-high systolic blood pressure, high fasting plasma glucose, and high body mass index-were utilised to predict the risk-attributable fractions between 2019 and 2050. Disease burden not attributable to these risk factors was then forecasted using time series models (autoregressive integrated moving average [ARIMA]), incorporating the Socio-demographic Index (SDI) as an external predictor. The optimal parameters of ARIMA models were selected for each age-sex-country group based on the Akaike Information Criterion (AIC). Different health scenarios were constructed by extending the past 85th and 15th percentiles of annualised rates of change in risk factors and SDI across all location-years, stratified by age and sex groups. A decomposition analysis was performed to assess the relative contributions of population size, age composition, and intracerebral haemorrhage risk on the projected changes. Findings: Compared with observed figures in 2019, our analysis predicts an increase in the burden of intracerebral haemorrhage in Europe in 2050, with a marginal rise of 0.6% (95% uncertainty interval [UI], -7.4% to 9.6%) in incident cases and an 8.9% (-2.8% to 23.6%) increase in mortality, reaching 141.2 (120.6-166.5) thousand and 144.2 (122.9-172.2) thousand respectively. These projections may fluctuate depending on trajectories of the risk factors and SDI; worsened trends could result in increases of 16.7% (8.7%-25.3%) in incidence and 31.2% (17.7%-48%) in mortality, while better trajectories may lead to a 10% (16.4%-2.3%) decrease in intracerebral haemorrhage cases with stabilised mortality. Individuals aged ≥80 years are expected to contribute significantly to the burden, comprising 62.7% of the cases in 2050, up from 40% in 2019, and 72.5% of deaths, up from 50.5%. Country-wide variations were noted in the projected changes, with decreases in the standardised rates across all nations but varying crude rates. The largest relative reductions in counts for both incidence and mortality are expected in Latvia, Bulgaria, and Hungary-ranging from -38.2% to -32.4% and -37.3% to -30.2% respectively. In contrast, the greatest increases for both measures were forecasted in Ireland (45.7% and 74.4%), Luxembourg (45% and 70.7%), and Cyprus (44.5% and 74.2%). The modelled increase in the burden of intracerebral haemorrhage could largely be attributed to population ageing. Interpretation: This study provides a comprehensive forecast of intracerebral haemorrhage in Europe until 2050, presenting different trajectories. The potential increase in the number of people experiencing and dying from intracerebral haemorrhage could have profound implications for both caregiving responsibilities and associated costs. However, forecasts were divergent between different scenarios and among EU countries, signalling the pivotal role of public health initiatives in steering the trajectories. Funding: The European Union's Horizon 2020 Research and Innovation Programme under grant agreement No. 754517. The National Institute for Health and Care Research (NIHR) under its Programme Grants for Applied Research (NIHR202339).

Developing and externally validating a machine learning risk prediction model for 30-day mortality after stroke using national stroke registers in the UK and Sweden.

OBJECTIVES: We aimed to develop and externally validate a generalisable risk prediction model for 30-day stroke mortality suitable for supporting quality improvement analytics in stroke care using large nationwide stroke registers in the UK and Sweden. DESIGN: Registry-based cohort study. SETTING: Stroke registries including the Sentinel Stroke National Audit Programme (SSNAP) in England, Wales and Northern Ireland (2013-2019) and the national Swedish stroke register (Riksstroke 2015-2020). PARTICIPANTS AND METHODS: Data from SSNAP were used for developing and temporally validating the model, and data from Riksstroke were used for external validation. Models were developed with the variables available in both registries using logistic regression (LR), LR with elastic net and interaction terms and eXtreme Gradient Boosting (XGBoost). Performances were evaluated with discrimination, calibration and decision curves. OUTCOME MEASURES: The primary outcome was all-cause 30-day in-hospital mortality after stroke. RESULTS: In total, 488 497 patients who had a stroke with 12.4% 30-day in-hospital mortality were used for developing and temporally validating the model in the UK. A total of 128 360 patients who had a stroke with 10.8% 30-day in-hospital mortality and 13.1% all mortality were used for external validation in Sweden. In the SSNAP temporal validation set, the final XGBoost model achieved the highest area under the receiver operating characteristic curve (AUC) (0.852 (95% CI 0.848 to 0.855)) and was well calibrated. The performances on the external validation in Riksstroke were as good and achieved AUC at 0.861 (95% CI 0.858 to 0.865) for in-hospital mortality. For Riksstroke, the models slightly overestimated the risk for in-hospital mortality, while they were better calibrated at the risk for all mortality. CONCLUSION: The risk prediction model was accurate and externally validated using high quality registry data. This is potentially suitable to be deployed as part of quality improvement analytics in stroke care to enable the fair comparison of stroke mortality outcomes across hospitals and health systems across countries.

Real-world effectiveness of steroids in severe COVID-19: a retrospective cohort study.

INTRODUCTION: Randomised controlled trials have shown that steroids reduce the risk of dying in patients with severe Coronavirus disease 2019 (COVID-19), whilst many real-world studies have failed to replicate this result. We aim to investigate real-world effectiveness of steroids in severe COVID-19. METHODS: Clinical, demographic, and viral genome data extracted from electronic patient record (EPR) was analysed from all SARS-CoV-2 RNA positive patients admitted with severe COVID-19, defined by hypoxia at presentation, between March 13th 2020 and May 27th 2021. Steroid treatment was measured by the number of prescription-days with dexamethasone, hydrocortisone, prednisolone or methylprednisolone. The association between steroid > 3 days treatment and disease outcome was explored using multivariable cox proportional hazards models with adjustment for confounders (including age, gender, ethnicity, co-morbidities and SARS-CoV-2 variant). The outcome was in-hospital mortality. RESULTS: 1100 severe COVID-19 cases were identified having crude hospital mortality of 15.3%. 793/1100 (72.1%) individuals were treated with steroids and 513/1100 (46.6%) received steroid ≤ 3 days. From the multivariate model, steroid > 3 days was associated with decreased hazard of in-hospital mortality (HR: 0.47 (95% CI: 0.31-0.72)). CONCLUSION: The protective effect of steroid treatment for severe COVID-19 reported in randomised clinical trials was replicated in this retrospective study of a large real-world cohort.

Risk prediction of 30-day mortality after stroke using machine learning: a nationwide registry-based cohort study.

BACKGROUNDS: We aimed to develop and validate machine learning (ML) models for 30-day stroke mortality for mortality risk stratification and as benchmarking models for quality improvement in stroke care. METHODS: Data from the UK Sentinel Stroke National Audit Program between 2013 to 2019 were used. Models were developed using XGBoost, Logistic Regression (LR), LR with elastic net with/without interaction terms using 80% randomly selected admissions from 2013 to 2018, validated on the 20% remaining admissions, and temporally validated on 2019 admissions. The models were developed with 30 variables. A reference model was developed using LR and 4 variables. Performances of all models was evaluated in terms of discrimination, calibration, reclassification, Brier scores and Decision-curves. RESULTS: In total, 488,497 stroke patients with a 12.3% 30-day mortality rate were included in the analysis. In 2019 temporal validation set, XGBoost model obtained the lowest Brier score (0.069 (95% CI: 0.068-0.071)) and the highest area under the ROC curve (AUC) (0.895 (95% CI: 0.891-0.900)) which outperformed LR reference model by 0.04 AUC (p < 0.001) and LR with elastic net and interaction term model by 0.003 AUC (p < 0.001). All models were perfectly calibrated for low (< 5%) and moderate risk groups (5-15%) and ≈1% underestimation for high-risk groups (> 15%). The XGBoost model reclassified 1648 (8.1%) low-risk cases by the LR reference model as being moderate or high-risk and gained the most net benefit in decision curve analysis. CONCLUSIONS: All models with 30 variables are potentially useful as benchmarking models in stroke-care quality improvement with ML slightly outperforming others.

Incidence and Long-Term Survival of Spontaneous Intracerebral Hemorrhage Over Time: A Systematic Review and Meta-Analysis.

Background and Purpose: Recent epidemiological data indicate that the absolute number of hemorrhagic stroke cases increased by 47% between 1990 and 2010 and continued to cause high rates of death and disability. The last systematic review and meta-analysis of incidence and long-term survival of intracerebral hemorrhage (ICH) were published 11 and 7 years ago, respectively, and lacked comparison between different income groups, therefore, a more up to date analysis is needed. We aim to investigate the ICH incidence and long-term survival data in countries of different income groups. Materials Methods: We systematically searched Ovid Medline for population-based longitudinal studies of first-ever spontaneous ICH published from January 2000 to December 2020. We performed meta-analyses on the incidence and survival rate in countries of 4 different income groups with random-effects models (severe inconsistency). The I 2 was used to measure the heterogeneity. Heterogeneity was further investigated by conducting the meta-regression on the study mid-year. Time trends of the survival rate were assessed by weighted linear regression. Results: We identified 84 eligible papers, including 68 publications reporting incidence and 24 publications on the survival rate. The pooled incidence of ICH per 100,000 per person-years was 26.47 (95% CI: 21.84-32.07) worldwide, 25.9 (95% CI: 22.63-29.63) in high-income countries (HIC), 28.45 (95% CI: 15.90-50.88) in upper-middle-income countries, and 31.73 (95% CI: 18.41-54.7) in lower-middle-income countries. The 1-year pooled survival rate was from 50% (95% CI: 47-54%; n = 4,380) worldwide to 50% (95% CI: 47-54%) in HIC, and 46% (95% CI: 38-55%) in upper-middle income countries. The 5-year pooled survival rate was 41% (95% CI: 35-48%; n = 864) worldwide, 41% (95% CI: 32-50%) in high-income and upper-middle countries. No publications were found reporting the long-term survival in lower-middle-income and low-income countries. No time trends in incidence or survival were found by meta-regression. Conclusion: The pooled ICH incidence was highest in lower-middle-income countries. About half of ICH patients survived 1 year, and about two-fifths survived 5 years. Reliable population-based studies estimating the ICH incidence and long-term survival in low-income and low-middle-income countries are needed to help prevention of ICH. Systematic Review Registration: https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=170140, PROSPERO CRD42020170140.

Genetic-Guided Pharmacotherapy for Atrial Fibrillation: A Systematic and Critical Review of Economic Evaluations.

OBJECTIVES: This study aimed to examine the extent and quality of evidence from economic evaluations (EEs) of genetic-guided pharmacotherapy (PGx) for atrial fibrillation (AF) and to identify variables influential in changing base-case conclusions. METHODS: From systematic searches, we included EEs of existing PGx testing to guide pharmacotherapy for AF, without restrictions on population characteristics or language. Articles excluded were genetic tests used to guide device-based therapy or focused on animals. RESULTS: We found 18 EEs (46 comparisons), all model-based cost-utility analysis with or without cost-effectiveness analysis mostly from health system's perspectives, of PGx testing to determine coumadin/direct-acting anticoagulant (DOAC) dosing (14 of 18), to stratify patients into coumadin/DOACs (3 of 18), or to increase patients' adherence to coumadin (1 of 18) versus non-PGx. Most PGx to determine coumadin dosing found PGx more costly and more effective than standard or clinical coumadin dosing (19 of 24 comparisons) but less costly and less effective than standard DOAC dosing (14 of 14 comparisons). The remaining comparisons were too few to observe any trend. Of 61 variables influential in changing base-case conclusions, effectiveness of PGx testing was the most common (37%), accounted for in the models using time-based or medication-based approaches or relative risk. The cost of PGx testing has decreased and plateaued over time. CONCLUSIONS: EEs to date only partially inform decisions on selecting optimal PGx testing for AF, because most evidence focuses on PGx testing to determine coumadin dosing, but less on other purposes. Future EE may refer to the list of influential variables and the approaches used to account for the effect of PGx testing to inform data collection and study design.

Diabetes As an Independent Risk Factor for Stroke Recurrence in Ischemic Stroke Patients: An Updated Meta-Analysis.

INTRODUCTION: Stroke and its recurrence and diabetes will increase in incidence as the population ages globally. This study explores the relationship between diabetes and stroke recurrence to understand if diabetes is an independent predictor for stroke recurrence in ischemic stroke (IS) patients. METHODS: We conducted a systematic review and meta-analysis of studies on the effect of diabetes on stroke recurrence among patients with IS. We searched population-based studies published before 15th February 2021 in PubMed and EMBASE following PRISMA guidelines. Random-effects estimates of the pooled hazard ratio (HR) and 95% confidence intervals (CIs) of each study were generated. A funnel plot and an Egger test were performed to evaluate publication bias. All statistical analyses were conducted in the R software 4.0.1 and Stata 16.0. RESULTS: The search identified 3,121 citations, of which 27 studies met inclusion criteria. Diabetes was associated with a significant risk of stroke recurrence in all IS patients (pooled HR, 1.50; 95% CI: 1.36-1.65; I2 = 61.0%). Similar results were found in lacunar stroke patients with diabetes (pooled HR, 1.65; 95% CI: 1.41-1.92; I2 = 22.0%). Moreover, we found that the risk of recurrent IS among patients of IS with diabetes was higher than that in those without diabetes (pooled HR, 1.53; 95% CI: 1.30-1.81; I2 = 74.0%). CONCLUSION: Diabetes is an independent risk factor for stroke recurrence among patients with IS.

Impacts of COVID-19 on clinical research in the UK: A multi-method qualitative case study.

BACKGROUND: Clinical research has been central to the global response to COVID-19, and the United Kingdom (UK), with its research system embedded within the National Health Service (NHS), has been singled out globally for the scale and speed of its COVID-19 research response. This paper explores the impacts of COVID-19 on clinical research in an NHS Trust and how the embedded research system was adapted and repurposed to support the COVID-19 response. METHODS AND FINDINGS: Using a multi-method qualitative case study of a research-intensive NHS Trust in London UK, we collected data through a questionnaire (n = 170) and semi-structured interviews (n = 24) with research staff working in four areas: research governance; research leadership; research delivery; and patient and public involvement. We also observed key NHS Trust research prioritisation meetings (40 hours) and PPI activity (4.5 hours) and analysed documents produced by the Trust and national organisation relating to COVID-19 research. Data were analysed for a descriptive account of the Trust's COVID-19 research response and research staff's experiences. Data were then analysed thematically. Our analysis identifies three core themes: centralisation; pace of work; and new (temporary) work practices. By centralising research prioritisation at both national and Trust levels, halting non-COVID-19 research and redeploying research staff, an increased pace in the setup and delivery of COVID-19-related research was possible. National and Trust-level responses also led to widescale changes in working practices by adapting protocols and developing local processes to maintain and deliver research. These were effective practical solutions borne out of necessity and point to how the research system was able to adapt to the requirements of the pandemic. CONCLUSION: The Trust and national COVID-19 response entailed a rapid large-scale reorganisation of research staff, research infrastructures and research priorities. The Trust's local processes that enabled them to enact national policy prioritising COVID-19 research worked well, especially in managing finite resources, and also demonstrate the importance and adaptability of the research workforce. Such findings are useful as we consider how to adapt our healthcare delivery and research practices both at the national and global level for the future. However, as the pandemic continues, research leaders and policymakers must also take into account the short and long term impact of COVID-19 prioritisation on non-COVID-19 health research and the toll of the emergency response on research staff.

Genetic-guided pharmacotherapy for venous thromboembolism: a systematic and critical review of economic evaluations.

Despite the known contributions of genes, genetic-guided pharmacotherapy has not been routinely implemented for venous thromboembolism (VTE). To examine evidence on cost-effectiveness of genetic-guided pharmacotherapy for VTE, we searched six databases, websites of four HTA agencies and citations, with independent double-reviewers in screening, data extraction, and quality rating. The ten eligible studies, all model-based, examined heterogeneous interventions and comparators. Findings varied widely; testing was cost-saving in two base-cases, cost-effective in four, not cost-effective in three, dominated in one. Of 22 model variables that changed decisions about cost-effectiveness, effectiveness/relative effectiveness of the intervention was the most frequent, albeit of poor quality. Studies consistently lacked details on the provision of interventions and comparators as well as on model development and validation. Besides improving the reporting of interventions, comparators, and methodological details, future economic evaluations should examine strategies recommended in guidelines and testing key model variables for decision uncertainty, to advise clinical implementations.

Risk Prediction of Cognitive Decline after Stroke.

BACKGROUND AND PURPOSE: Cognitive decline is one of the major outcomes after stroke. We have developed and evaluated a risk predictive tool of post-stroke cognitive decline and assessed its clinical utility. METHODS: In this population-based cohort, 4,783 patients with first-ever stroke from the South London Stroke Register (1995-2010) were included in developing the model. Cognitive impairment was measured using the Mini Mental State Examination (cut off 24/30) and the Abbreviated Mental Test (cut off 8/10) at 3-months and yearly thereafter. A penalised mixed-effects linear model was developed and temporal-validated in a new cohort consisted of 1,718 stroke register participants recruited from (2011-2018). Prediction errors on discrimination and calibration were assessed. The clinical utility of the model was evaluated using prognostic accuracy measurements and decision curve analysis. RESULTS: The overall predictive model showed good accuracy, with root mean squared error of 0.12 and R2 of 73%. Good prognostic accuracy for predicting severe cognitive decline was observed AUC: (88%, 95% CI [85-90]), (89.6%, 95% CI [86-92]), (87%, 95% CI [85-91]) at 3 months, one and 5 years respectively. Average predicted recovery patterns were analysed by age, stroke subtype, Glasgow-coma scale, and left-stroke and showed variability. DECISION: curve analysis showed an increased clinical benefit, particularly at threshold probabilities of above 15% for predictive risk of cognitive impairment. CONCLUSIONS: The derived prognostic model seems to accurately screen the risk of post-stroke cognitive decline. Such prediction could support the development of more tailored management evaluations and identify groups for further study and future trials.

Stroke Care in the United Kingdom During the COVID-19 Pandemic.

BACKGROUND AND PURPOSE: The coronavirus disease 2019 (COVID-19) pandemic has potentially caused indirect harm to patients with other conditions via reduced access to health care services. We aimed to describe the impact of the initial wave of the pandemic on admissions, care quality, and outcomes in patients with acute stroke in the United Kingdom. METHODS: Registry-based cohort study of patients with acute stroke admitted to hospital in England, Wales, and Northern Ireland between October 1, 2019, and April 30, 2020, and equivalent periods in the 3 prior years. RESULTS: One hundred fourteen hospitals provided data for a study cohort of 184 017 patients. During the lockdown period (March 23 to April 30), there was a 12% reduction (6923 versus 7902) in the number of admissions compared with the same period in the 3 previous years. Admissions fell more for ischemic than hemorrhagic stroke, for older patients, and for patients with less severe strokes. Quality of care was preserved for all measures and in some domains improved during lockdown (direct access to stroke unit care, 1-hour brain imaging, and swallow screening). Although there was no change in the proportion of patients discharged with good outcome (modified Rankin Scale score, ≤2; 48% versus 48%), 7-day inpatient case fatality increased from 6.9% to 9.4% (P<0.001) and was 22.0% in patients with confirmed or suspected COVID-19 (adjusted rate ratio, 1.41 [1.11-1.80]). CONCLUSIONS: Assuming that the true incidence of acute stroke did not change markedly during the pandemic, hospital avoidance may have created a cohort of untreated stroke patients at risk of poorer outcomes or recurrent events. Unanticipated improvements in stroke care quality should be used as an opportunity for quality improvement and to learn about how to develop resilient health care systems.

Trends in prevalence of acute stroke impairments: A population-based cohort study using the South London Stroke Register.

BACKGROUND: Acute stroke impairments often result in poor long-term outcome for stroke survivors. The aim of this study was to estimate the trends over time in the prevalence of these acute stroke impairments. METHODS AND FINDINGS: All first-ever stroke patients recorded in the South London Stroke Register (SLSR) between 2001 and 2018 were included in this cohort study. Multivariable Poisson regression models with robust error variance were used to estimate the adjusted prevalence of 8 acute impairments, across six 3-year time cohorts. Prevalence ratios comparing impairments over time were also calculated, stratified by age, sex, ethnicity, and aetiological classification (Trial of Org 10172 in Acute Stroke Treatment [TOAST]). A total of 4,683 patients had a stroke between 2001 and 2018. Mean age was 68.9 years, 48% were female, and 64% were White. After adjustment for demographic factors, pre-stroke risk factors, and stroke subtype, the prevalence of 3 out of the 8 acute impairments declined during the 18-year period, including limb motor deficit (from 77% [95% CI 74%-81%] to 62% [56%-68%], p < 0.001), dysphagia (37% [33%-41%] to 15% [12%-20%], p < 0.001), and urinary incontinence (43% [39%-47%) to 29% [24%-35%], p < 0.001). Declines in limb impairment over time were 2 times greater in men than women (prevalence ratio 0.73 [95% CI 0.64-0.84] and 0.87 [95% CI 0.77-0.98], respectively). Declines also tended to be greater in younger patients. Stratified by TOAST classification, the prevalence of all impairments was high for large artery atherosclerosis (LAA), cardioembolism (CE), and stroke of undetermined aetiology. Conversely, small vessel occlusions (SVOs) had low levels of all impairments except for limb motor impairment and dysarthria. While we have assessed 8 key acute stroke impairments, this study is limited by a focus on physical impairments, although cognitive impairments are equally important to understand. In addition, this is an inner-city cohort, which has unique characteristics compared to other populations. CONCLUSIONS: In this study, we found that stroke patients in the SLSR had a complexity of acute impairments, of which limb motor deficit, dysphagia, and incontinence have declined between 2001 and 2018. These reductions have not been uniform across all patient groups, with women and the older population, in particular, seeing fewer reductions.

Inverse problem approach to regularized regression models with application to predicting recovery after stroke.

Regression modelling is a powerful statistical tool often used in biomedical and clinical research. It could be formulated as an inverse problem that measures the discrepancy between the target outcome and the data produced by representation of the modelled predictors. This approach could simultaneously perform variable selection and coefficient estimation. We focus particularly on a linear regression issue, Y ∼ N ( X ß , σ I n ) , where ß ∈ R p is the parameter of interest and its components are the regression coefficients. The inverse problem finds an estimate for the parameter ß , which is mapped by the linear operator ( L : ß âŸ¶ X ß ) to the observed outcome data Y = X ß + ε . This problem could be conveyed by finding a solution in the affine subspace L - 1 ( Y ) . However, in the presence of collinearity, high-dimensional data and high conditioning number of the related covariance matrix, the solution may not be unique, so the introduction of prior information to reduce the subset L - 1 ( Y ) and regularize the inverse problem is needed. Informed by Huber's robust statistics framework, we propose an optimal regularizer to the regression problem. We compare results of the proposed method and other penalized regression regularization methods: ridge, lasso, adaptive-lasso and elastic-net under different strong hypothesis such as high conditioning number of the covariance matrix and high error amplitude, on both simulated and real data from the South London Stroke Register. The proposed approach can be extended to mixed regression models. Our inverse problem framework coupled with robust statistics methodology offer new insights in statistical regression and learning. It could open a new research development for model fitting and learning.

Burden of Stroke in Europe: Thirty-Year Projections of Incidence, Prevalence, Deaths, and Disability-Adjusted Life Years.

BACKGROUND AND PURPOSE: Prediction of stroke impact provides essential information for healthcare planning and priority setting. We aim to estimate 30-year projections of stroke epidemiology in the European Union using multiple modeling approaches. METHODS: Data on stroke incidence, prevalence, deaths, and disability-adjusted life years in the European Union between 1990 and 2017 were obtained from the global burden of disease study. Their trends over time were modeled using 3 modeling strategies: linear, Poisson, and exponential regressions-adjusted for the gross domestic product per capita, which reflects the impact of economic development on health status. We used the Akaike information criterion for model selection. The 30-year projections up to 2047 were estimated using the best fitting models, with inputs on population projections from the United Nations and gross domestic product per capita prospects from the World Bank. The technique was applied separately by age-sex-country groups for each stroke measure. RESULTS: In 2017, there were 1.12 million incident strokes in the European Union, 9.53 million stroke survivors, 0.46 million deaths, and 7.06 million disability-adjusted life years lost because of stroke. By 2047, we estimated an additional 40 000 incident strokes (+3%) and 2.58 million prevalent cases (+27%). Conversely, 80 000 fewer deaths (-17%) and 2.31 million fewer disability-adjusted life years lost (-33%) are projected. The largest increase in the age-adjusted incidence and prevalence rates are expected in Lithuania (average annual percentage change, 0.48% and 0.7% respectively), and the greatest reductions in Portugal (-1.57% and -1.3%). Average annual percentage change in mortality rates will range from -2.86% (Estonia) to -0.08% (Lithuania), and disability-adjusted life years' from -2.77% (Estonia) to -0.23% (Romania). CONCLUSIONS: The number of people living with stroke is estimated to increase by 27% between 2017 and 2047 in the European Union, mainly because of population ageing and improved survival rates. Variations are expected to persist between countries showing opportunities for improvements in prevention and case management particularly in Eastern Europe.

Risk and Secondary Prevention of Stroke Recurrence: A Population-Base Cohort Study.

BACKGROUND AND PURPOSE: With recent advances in secondary prevention management, stroke recurrence rates may have changed substantially. We aim to estimate risks and trends of stroke recurrence over the past 2 decades in a population-based cohort of patients with stroke. METHODS: Patients with a first-ever stroke between 1995 and 2018 in South London, United Kingdom (n=6052) were collected and analyzed. Rates of recurrent stroke with 95% CIs were stratified by 5-year period of index stroke and etiologic TOAST (Trial of ORG 10172 in Acute Stroke Treatment) subtype. Cumulative incidences were estimated and multivariate Cox models applied to examine associations of recurrence and recurrence-free survival. RESULTS: The rate of stroke recurrence at 5 years reduced from 18% (95% CI, 15%-21%) in those who had their stroke in 1995 to 1999 to 12% (10%-15%) in 2000 to 2005, and no improvement since. Recurrence-free survival has improved (35%, 1995-1999; 67%, 2010-2015). Risk of recurrence or death is lowest for small-vessel occlusion strokes and other ischemic causes (36% and 27% at 5 years, respectively). For cardioembolic and hemorrhagic index strokes around half of first recurrences are of the same type (54% and 51%, respectively). Over the whole study period a 54% increased risk of recurrence was observed among those who had atrial fibrillation before the index stroke (hazard ratio, 1.54 [1.09-2.17]). CONCLUSIONS: The rate of recurrence reduced until mid-2000s but has not changed over the last decade. The majority of cardioembolic or hemorrhagic strokes that have a recurrence are stroke of the same type indicating that the implementation of effective preventive strategies is still suboptimal in these stroke subtypes.

A systematic review of machine learning models for predicting outcomes of stroke with structured data.

BACKGROUND AND PURPOSE: Machine learning (ML) has attracted much attention with the hope that it could make use of large, routinely collected datasets and deliver accurate personalised prognosis. The aim of this systematic review is to identify and critically appraise the reporting and developing of ML models for predicting outcomes after stroke. METHODS: We searched PubMed and Web of Science from 1990 to March 2019, using previously published search filters for stroke, ML, and prediction models. We focused on structured clinical data, excluding image and text analysis. This review was registered with PROSPERO (CRD42019127154). RESULTS: Eighteen studies were eligible for inclusion. Most studies reported less than half of the terms in the reporting quality checklist. The most frequently predicted stroke outcomes were mortality (7 studies) and functional outcome (5 studies). The most commonly used ML methods were random forests (9 studies), support vector machines (8 studies), decision trees (6 studies), and neural networks (6 studies). The median sample size was 475 (range 70-3184), with a median of 22 predictors (range 4-152) considered. All studies evaluated discrimination with thirteen using area under the ROC curve whilst calibration was assessed in three. Two studies performed external validation. None described the final model sufficiently well to reproduce it. CONCLUSIONS: The use of ML for predicting stroke outcomes is increasing. However, few met basic reporting standards for clinical prediction tools and none made their models available in a way which could be used or evaluated. Major improvements in ML study conduct and reporting are needed before it can meaningfully be considered for practice.

Long-term trends in death and dependence after ischaemic strokes: A retrospective cohort study using the South London Stroke Register (SLSR).

BACKGROUND: There have been reductions in stroke mortality over recent decades, but estimates by aetiological subtypes are limited. This study estimates time trends in mortality and functional dependence by ischaemic stroke (IS) aetiological subtype over a 16-year period. METHODS AND FINDINGS: The study population was 357,308 in 2011; 50.4% were males, 56% were white, and 25% were of black ethnic backgrounds. Population-based case ascertainment of stroke was conducted, and all participants who had their first-ever IS between 2000 and 2015 were identified. Further classification was concluded according to the underlying mechanism into large-artery atherosclerosis (LAA), cardio-embolism (CE), small-vessel occlusion (SVO), other determined aetiologies (OTH), and undetermined aetiologies (UND). Temporal trends in survival rates were examined using proportional-hazards survival modelling, adjusted for demography, prestroke risk factors, case mix variables, and processes of care. We carried out additional regression analyses to explore patterns in case-fatality rates (CFRs) at 30 days and 1 year and to explore whether these trends occurred at the expense of greater functional dependence (Barthel Index [BI] < 15) among survivors. A total of 3,128 patients with first-ever ISs were registered. The median age was 70.7 years; 50.9% were males; and 66.2% were white, 25.5% were black, and 8.3% were of other ethnic groups. Between 2000-2003 and 2012-2015, the adjusted overall mortality decreased by 24% (hazard ratio [HR] per year 0.976; 95% confidence interval [CI] 0.959-0.993). Mortality reductions were equally noted in both sexes and in the white and black populations but were only significant in CE strokes (HR per year 0.972; 95% CI 0.945‒0.998) and in patients aged ≥55 years (HR per year 0.975; 95% CI 0.959‒0.992). CFRs within 30 days and 1 year after an IS declined by 38% (rate ratio [RR] per year 0.962; 95% CI 0.941‒0.984) and 37% (RR per year 0.963; 95% CI 0.949‒0.976), respectively. Recent IS was independently associated with a 23% reduced risk of functional dependence at 3 months after onset (RR per year 0.983; 95% CI 0.968-0.998; p = 0.002 for trend). The study is limited by small number of events in certain subgroups (e.g., LAA), which could have led to insufficient power to detect significant trends. CONCLUSIONS: Both mortality and 3-month functional dependence after IS decreased by an annual average of around 2.4% and 1.7%, respectively, during 2000‒2015. Such reductions were particularly evident in strokes of CE origins and in those aged ≥55 years.