Developing a stroke severity index based on administrative data was feasible using data mining techniques

Racial, Ethnic, and Regional Disparities of Post-Acute Service Utilization After Stroke in the United States

Background and Objectives

Post-acute care is critical for patient functional recovery and successful community transition. This study aimed to understand the current racial, ethnic, and regional disparities in post-acute service utilization after stroke.

Methods

This retrospective cross-sectional study included patients hospitalized for ischemic stroke and intracerebral hemorrhage in 2017-2018 using the National Inpatient Sample. Discharge destinations were classified as follows: (1) facility including inpatient rehabilitation, skilled nursing facility, and facility hospice; (2) home health care (HHC), including home health and home hospice; and (3) home without HHC. Multinomial logistic regression was used to study the odds of discharge to a facility over home and HHC over home without HHC by race, ethnicity, insurance, and census division, adjusting for clinical factors and survey design.

Results

Among the 1,000,980 weighted ischemic stroke admissions, 66.9% were White, 17.6% Black, 9.5% Hispanic, 3.1% Asian American/Pacific Islander, and 0.4% Native American. Relative to private insurance, uninsured patients had the lowest adjusted odds of facility over home discharge (0.44; 95% CI 0.40-0.48) and HHC discharge over home without HHC (0.79; 95% CI 0.71-0.88). Compared with White patients, only Hispanic patients with Medicare/Medicaid insurance or self-pay had lower odds of facility over home discharge (adjusted OR 0.80 and 0.75, respectively; 95% CI 0.76-0.84 and 0.63-0.93). Uninsured Hispanic patients also had lower odds of HHC discharge over home without HHC than White patients (0.74; 95% CI 0.57-0.97). Facility discharge rate was the highest in East North Central (39.2%) and lowest in Pacific (31.2%). HHC discharge rate was the highest in New England (20.2%) and lowest in West North Central (10.3%), which had the highest home without HHC discharge (46.1%). Compared with New England, other census divisions had lower odds of facility over any home discharge with Pacific being the lowest (adjusted OR, 0.66; 95% CI 0.60-0.71) and HHC over home without HHC discharge with West North Central being the lowest (adjusted OR, 0.33; 95% CI 0.29-0.38). Similar patterns were observed in intracerebral hemorrhage.

Discussion

Significant insurance-dependent racial and ethnic disparities and regional variations were evident in post-acute service utilization after stroke. Targeted efforts are needed to improve post-acute service access for uninsured patients especially Hispanic patients and people in certain regions.

Prognosis of Patients With Ischemic Stroke With Prior Anticoagulant Therapy: Direct Oral Anticoagulants Versus Warfarin

BACKGROUND

Direct oral anticoagulants (DOACs) have been the drug of choice for preventing ischemic stroke in patients with atrial fibrillation since 2014. In previous studies, the stroke risk while taking warfarin was 2 per 100 patient-years and 1.5% per year while taking DOACs. We hypothesized that even if ischemic stroke occurred during anticoagulation therapy with DOACs, the prognosis was likely to be better than that with warfarin.

METHODS AND RESULTS

Data from 2002 to 2019, sourced from a nationwide claims database, were used to identify atrial fibrillation patients using International Classification of Diseases codes. Patients who experienced an ischemic stroke during anticoagulation were categorized by the drugs used (warfarin, dabigatran, apixaban, rivaroxaban, and edoxaban). The primary outcome was mortality within 3 months and 1 year after the ischemic stroke. Among the 9578 patients with ischemic stroke during anticoagulation, 3343 received warfarin, and 6235 received DOACs (965 dabigatran, 2320 apixaban, 1702 rivaroxaban, 1248 edoxaban). The DOACs group demonstrated lower risks of 3-month (adjusted hazard ratio [HR], 0.550, [95% CI, 0.473-0.639]; P<0.0001) and 1-year mortality (adjusted HR, 0.596 [95% CI, 0.536-0.663]; P<0.0001) than the warfarin group. Apixaban and edoxaban within the DOAC group exhibited particularly reduced 1-year mortality risk compared with other DOACs (P<0.0001).

CONCLUSIONS

Our study confirmed that DOACs have a better prognosis than warfarin after ischemic stroke. The apixaban and edoxaban groups had a lower risk of death after ischemic stroke than the other DOAC groups.

The association between antithrombotic drug regimen changes and clinical outcomes after stroke in atrial fibrillation

BACKGROUND

The impact of post-stroke antithrombotic regimen in atrial fibrillation (AF) is uncertain.

OBJECTIVE

To describe antithrombotic therapy prescribing patterns following ischemic stroke, and its impact on outcomes.

METHODS

A total of 23,165 AF patients experiencing ischemic stroke were identified. Subsequent post-stroke events included recurrent ischemic stroke, intracranial hemorrhage (ICH), major bleeding, mortality, and composite outcomes.

RESULTS

Among those who were non-anticoagulated before a stroke, 33.5% remained non-anticoagulated and 39.2% were prescribed only antiplatelets (AP) post-stroke. Compared to NOACs post-stroke, there was a significant increase in ischemic stroke and mortality in non-anticoagulated (aHRs 2.09 and 3.92) and antiplatelet users (aHRs 1.32 and 1.28). Post-stroke warfarin was associated with a significantly incresaed risk of major bleeding compared to NOACs (aHR 1.23). Among 769 patients receiving NOACs before stroke and continuing NOAC post-stroke, those switching to a different NOAC were associated with significantly higher risk of ischemic stroke (aHR 2.07) and composite outcomes (aHRs 1.36-1.85) with no difference in ICH, major bleeding or mortality compared to those on the same NOAC post-stroke. Among patients receiving NOACs before stroke, the risks of clinical events were similar between patients on NOACs alone and those on NOAC plus AP post-stroke.

CONCLUSIONS

NOAC alone post-stroke was associated with a better clinical outcome compared to non-anticoagulation, AP or warfarin. Among patients already taking NOACs before stroke, the addition of AP didn't confer additional benefits compared to NOACs alone. A change of NOAC types post-stroke was associated with a two-fold higher risk of ischemic stroke and composite outcomes.

Cilostazol plus Aspirin vs. Clopidogrel plus Aspirin in Acute Minor Stroke or Transient Ischemic Attack

AIM

This study compared the effectiveness, safety, and mortality risks between cilostazol plus aspirin and clopidogrel plus aspirin treatment for patients with acute minor ischemic stroke or transient ischemic attack (TIA).

METHODS

This retrospective cohort study employed a new-user design and utilized data from the nationwide Health and Welfare Database in Taiwan. Patients were included if they were discharged with newly initiated cilostazol plus aspirin or clopidogrel plus aspirin after primary acute minor ischemic stroke or TIA hospitalization between 2009 and 2018. Inverse probability of treatment weighting was applied to balance covariats between study groups. Effectiveness outcomes were the risks of acute ischemic stroke, acute myocardial infarction (AMI), TIA, and composite cardiovascular events; Safety outcomes were the risks of intracranial hemorrhage (ICH), gastrointestinal bleeding, and composite bleeding events; Mortality outcomes were the risks of fatal stroke, cardiovascular mortality, and all-cause mortality.

RESULTS

A total of 3,403 patients were included, of which 578 were treated with cilostazol plus aspirin and 2,825 were treated with clopidogrel plus aspirin. Cilostazol plus aspirin was associated with a higher risk of ICH (HR: 1.82; 95% CI: 1.16-2.84) compared to clopidogrel plus aspirin. No significant differences in the risks of effectiveness or mortality outcomes between the two groups were found.

CONCLUSIONS

The effectiveness and mortality of the two groups were similar for patients with acute minor ischemic stroke or TIA. However, cilostazol plus aspirin was associated with a higher risk of ICH compared to clopidogrel plus aspirin. Patients treated with cilostazol plus aspirin among this population should be monitored carefully to ensure their safety.

Characteristics and risk of stroke in emergency department patients with acute dizziness

Background

Acute dizziness is a common symptom in the emergency department (ED), with strokes accounting for 3 %-5 % of cases. We investigated the risk of stroke in ED patients with acute dizziness and compared stroke characteristics diagnosed during and after the ED visit.

Methods

We identified adult patients with acute dizziness, vertigo, or imbalance using a hospital research-based database. Patients with abdominal or flank pain were used as the comparison group. Patients with dizziness were 1:1 matched to comparison patients. Each patient was traced for up to one year until being hospitalized for a stroke.

Results

Out of the 24,266 eligible patients, 589 (2.4 %) were hospitalized for stroke during the ED visit. For the remaining 23,677 patients, the risk of stroke at 7, 30, 90, and 365 days after ED discharge was 0.40 %, 0.52 %, 0.71 %, and 1.25 % respectively. Patients with dizziness had a higher risk of stroke compared to the comparison group at 7, 30, 90, and 365 days. The risk ratios decreased from 5.69 (95 % confidence interval [CI], 3.34-9.68) to 2.03 (95 % CI, 1.65-2.49). Compared to patients hospitalized for stroke during the ED visit, those hospitalized for stroke after the ED visit had greater stroke severity despite a lower initial triage acuity. Patients with early stroke (≤7 days) after ED discharge were less likely to have hypertension, diabetes, hyperlipidemia, and atrial fibrillation. They mostly experienced posterior circulation stroke. Patients with late stroke (>7 days) were older and less likely to have hypertension and hyperlipidemia but more likely to have a history of prior stroke and ischemic heart disease. Their strokes were mainly located in the anterior circulation territory.

Conclusions

The risk of stroke after ED discharge was higher in patients with dizziness than in the comparison group, with gradually decreasing risk ratios in the following year. Patients hospitalized for stroke during and after the ED visit had different profiles of vascular risk factors and clinical characteristics.

Changing or Retaining Direct Oral Anticoagulant After Ischemic Stroke Despite Direct Oral Anticoagulant Treatment

BACKGROUND

The optimal antithrombotic strategies for patients with atrial fibrillation who experience ischemic stroke (IS) despite direct oral anticoagulant (DOAC) therapy remain inconclusive. This study compared outcomes for patients with DOAC treatment failure who changed or retained their prestroke DOAC.

METHODS AND RESULTS

This retrospective cohort study analyzed data from the National Health Insurance Research Database from 2012 to 2020. Patients with atrial fibrillation who experienced IS during DOAC therapy were assigned to either (1) the DOAC-change group: changing prestroke DOAC or (2) the DOAC-retain group: retaining prestroke DOAC. The primary outcome was a composite of recurrent IS and transient ischemic attack. The secondary outcomes included intracranial hemorrhage, major bleeding, systemic thromboembolism, and all-cause death. Propensity score-based inverse probability of treatment weighting was applied to balance the baseline characteristics between the DOAC-change and DOAC-retain groups. The Cox proportional hazards model compared the risk of outcomes between the 2 groups. In total, 1979 patients were enrolled (609 DOAC-change patients and 1370 DOAC-retain patients). The incidence rates of recurrent IS or transient ischemic attack were 7.20 and 6.56 per 100 person-years in the DOAC-change and DOAC-retain groups, respectively (hazard ratio [HR], 1.07 [95% CI, 0.87-1.30]). A nonsignificantly higher incidence rate of intracranial hemorrhage was observed in the DOAC-change group compared with the DOAC-retain group (0.75 versus 0.53 per 100-person-years; HR, 1.49 [95% CI, 0.78-2.83]). The systemic thromboembolism, major bleeding, and death rates were comparable between the DOAC-change and DOAC-retain groups.

CONCLUSIONS

Changing prestroke DOAC does not reduce the risk of recurrent cerebral ischemia in patients with atrial fibrillation who develop IS during DOAC therapy. However, future studies should continue to observe the potential trends of increased intracranial hemorrhage risk.

Risk of Bleeding Following Non-Vitamin K Antagonist Oral Anticoagulant Use in Patients With Acute Ischemic Stroke Treated With Alteplase

Importance

Current guidelines advise against intravenous alteplase therapy for treatment of acute ischemic stroke in patients previously treated with non-vitamin K antagonist oral anticoagulants (NOACs).

Objective

To evaluate the risk of bleeding and mortality after alteplase treatment for acute ischemic stroke among patients treated with NOACs compared to those not treated with NOACs.

Design, Setting, and Participants

This nationwide, population-based cohort study was conducted in Taiwan using data from Taiwan's National Health Insurance Research Database from January 2011 through November 2020 and included 7483 patients treated with alteplase for acute ischemic stroke. A meta-analysis incorporating the results of the study with those of previous studies was performed, and the review protocol was prospectively registered with PROSPERO.

Exposures

NOAC treatment within 2 days prior to stroke, compared to either no anticoagulant treatment or warfarin treatment.

Main Outcomes and Measures

The primary outcome was intracranial hemorrhage after intravenous alteplase during the index hospitalization (the hospitalization subsequent to alteplase administration). Secondary outcomes were major bleeding events and mortality during the index hospitalization. Propensity score matching was used to control potential confounders. Logistic regression was used to estimate the odds ratio (OR) of outcome events. Meta-analysis was performed using a random-effects model.

Results

Of the 7483 included patients (mean [SD] age, 67.4 [12.7] years; 2908 [38.9%] female individuals and 4575 [61.1%] male individuals), 91 (1.2%), 182 (2.4%), and 7210 (96.4%) received NOACs, warfarin, and no anticoagulants prior to their stroke, respectively. Compared to patients who were not treated with anticoagulants, those treated with NOACs did not have significantly higher risks of intracranial hemorrhage (risk difference [RD], 2.47% [95% CI, -4.23% to 9.17%]; OR, 1.37 [95% CI, 0.62-3.03]), major bleeding (RD, 4.95% [95% CI, -2.56% to 12.45%]; OR, 1.69 [95% CI, 0.83-3.45]), or in-hospital mortality (RD, -4.95% [95% CI, -10.11% to 0.22%]; OR, 0.45 [95% CI, 0.15-1.29]) in the propensity score-matched analyses. Furthermore, the risks of bleeding and mortality were not significantly different between patients treated with NOACs and those treated with warfarin. Similar results were obtained in the meta-analysis.

Conclusions and Relevance

In this cohort study with meta-analysis, compared to no treatment with anticoagulants, treatment with NOACs prior to stroke was not associated with a higher risk of intracranial hemorrhage, major bleeding, or mortality in patients receiving intravenous alteplase for acute ischemic stroke.

Development and validation of a model predicting mild stroke severity on admission using electronic health record data

OBJECTIVE

Initial stroke severity is a potent modifier of stroke outcomes but this information is difficult to obtain from electronic health record (EHR) data. This limits the ability to risk-adjust for evaluations of stroke care and outcomes at a population level. The purpose of this analysis was to develop and validate a predictive model of initial stroke severity using EHR data elements.

METHODS

This observational cohort included individuals admitted to a US Department of Veterans Affairs hospital with an ischemic stroke. We extracted 65 independent predictors from the EHR. The primary analysis modeled mild (NIHSS score 0-3) versus moderate/severe stroke (NIHSS score ≥4) using multiple logistic regression. Model validation included: (1) splitting the cohort into derivation (65%) and validation (35%) samples and (2) evaluating how the predicted stroke severity performed in regard to 30-day mortality risk stratification.

RESULTS

The sample comprised 15,346 individuals with ischemic stroke (n = 10,000 derivation; n = 5,346 validation). The final model included 15 variables and correctly classified 70.4% derivation sample patients and 69.4% validation sample patients. The areas under the curve (AUC) were 0.76 (derivation) and 0.76 (validation). In the validation sample, the model performed similarly to the observed NIHSS in terms of the association with 30-day mortality (AUC: 0.72 observed NIHSS, 0.70 predicted NIHSS).

CONCLUSIONS

EHR data can be used to construct a surrogate measure of initial stroke severity. Further research is needed to better differentiate moderate and severe strokes, enhance stroke severity classification, and how to incorporate these measures in evaluations of stroke care and outcomes.

A double-hit: End-stage renal disease patients suffer worse outcomes in intracerebral hemorrhage

BACKGROUND

Intracerebral hemorrhage (ICH) carries significant morbidity and mortality. Previous single-center retrospective analysis suggests that end-stage renal disease (ESRD) is a risk factor for severe ICH and worse outcomes. This investigation aims to examine the impact of ESRD on ICH severity, complications, and outcomes using a multicenter national database.

METHODS

The International Classification of Disease, Ninth and Tenth Revision Clinical Modification codes were used to query the National Inpatient Sample for patients with ICH and ESRD between 2010 and 2019. Primary endpoints were the functional outcome, length of stay (LOS), and in-hospital mortality. Multivariate variable regression models and a propensity-score matched analysis were established to analyze patient outcomes associated with baseline patient characteristics.

RESULTS

We identified 211,266 patients with ICH, and among them, 7,864 (3.77%) patients had a concurrent diagnosis of ESRD. Patients with ESRD were younger (60.85 vs. 67.64, P < 0.01) and demonstrated increased ICH severity (0.78 vs. 0.77, P < 0.01). ESRD patients experienced higher rates of sepsis (15.9% vs. 6.15%, P < 0.01), acute myocardial infarction (8.05% vs. 3.65%, P < 0.01), and cardiac arrest (5.94% vs. 2.4%, P < 0.01). In addition, ESRD predicted poor discharge disposition (odds ratio [OR]: 2.385, 95% confidence interval [CI]: 2.227-2.555, P < 0.01), longer hospital LOS (OR: 1.629, 95% CI: 1.553-1.709, P < 0.01), and in-hospital mortality (OR: 2.786, 95% CI: 2.647-2.932, P < 0.01).

CONCLUSIONS

This study utilizes a multicenter database to analyze the effect of ESRD on ICH outcomes. ESRD is a significant predictor of poor functional outcomes, in-hospital mortality, and prolonged stay in the ICH population.

Weekend effect on 30-day mortality for ischemic and hemorrhagic stroke analyzed using severity index and staffing level

BACKGROUND AND PURPOSE

Previous studies on the weekend effect-a phenomenon where stroke outcomes differ depending on whether the stroke occurred on a weekend-mostly targeted ischemic stroke and showed inconsistent results. Thus, we investigated the weekend effect on 30-day mortality in patients with ischemic or hemorrhagic stroke considering the confounding effect of stroke severity and staffing level.

METHODS

We retrospectively analyzed data of patients hospitalized for ischemic or hemorrhagic stroke between January 1, 2015, and December 31, 2018, which were extracted from the claims database of the National Health Insurance System and the Medical Resource Report by the Health Insurance Review & Assessment Service. The primary outcome measure was 30-day all-cause mortality.

RESULTS

In total, 278,632 patients were included, among whom 84,240 and 194,392 had a hemorrhagic and ischemic stroke, respectively, with 25.8% and 25.1% of patients, respectively, being hospitalized during the weekend. Patients admitted on weekends had significantly higher 30-day mortality rates (hemorrhagic stroke 16.84%>15.55%, p<0.0001; ischemic stroke 5.06%>4.92%, p<0.0001). However, in the multi-level logistic regression analysis adjusted for case-mix, pre-hospital, and hospital level factors, the weekend effect remained consistent in patients with hemorrhagic stroke (odds ratio [OR] 1.05, 95% confidence interval [CI] 1.00-1.10), while the association was no longer evident in patients with ischemic stroke (OR 1.01, 95% CI 0.96-1.06).

CONCLUSIONS

Weekend admission for hemorrhagic stroke was significantly associated with a higher mortality rate after adjusting for confounding factors. Further studies are required to understand factors contributing to mortality during weekend admission.

Depression Risk in Post-Stroke Aphasia Patients: A Nationwide Population-Based Cohort Study

INTRODUCTION

Depression and aphasia impair the quality of life after a stroke. Studies linking depression risk to post-stroke aphasia (PSA) lacked confirmation using a large database.

METHODS

Using Taiwan's National Health Insurance claims data, we identified ≥18-year-old patients hospitalized for stroke from 2005 to 2009, and those diagnosed with aphasia during hospitalization or within 3 months after discharge were selected to form the aphasic group. We estimated depression incidence by December 31, 2018, and used the Cox proportional hazards model to estimate aphasia group to non-aphasia group hazard ratios (HRs).

RESULTS

With a median follow-up period of 7.91 and 8.62 years for aphasia (n = 26,754) and non-aphasia groups (n = 139,102), respectively, the incidence of depression was higher in the aphasia group than in the non-aphasia group (9.02 vs. 8.13 per 1,000 person-years), with an adjusted HR (95% confidence intervals [CI]) of 1.21 (1.15-1.29) for depression. The adjusted HRs (95% CI) of depression were homogenous for females, 1.26 (1.15-1.37); for males, 1.18 (1.09-1.27); for hemorrhagic stroke, 1.22 (1.09-1.37); and for ischemic stroke, 1.21 (1.13-1.30). Results in analyzing 25,939 propensity score-matched pairs demonstrated an equivalent effect.

CONCLUSION

Patients with PSA are at an increased risk of developing depression, regardless of sex or stroke type.

Incidence, prescription patterns and risk factors of antipsychotic initiation in elderly stroke survivors

OBJECTIVES

Epidemiological data regarding antipsychotic initiation in elderly patients with stroke are limited. We aimed to investigate the incidence, prescription patterns and determinants of antipsychotic initiation in elderly patients with stroke.

METHODS

We conducted a retrospective cohort study to identify patients aged above 65 years who had been admitted for stroke from the National Health Insurance Database (NHID). The index date was defined as the discharge date. The incidence and prescription pattern of antipsychotics were estimated using the NHID. To evaluate the determinants of antipsychotic initiation, the cohort identified from the NHID was linked to the Multicenter Stroke Registry (MSR). Demographics, comorbidities and concomitant medications were obtained from the NHID. Information including smoking status, body mass index, stroke severity and disability was retrieved by linking to the MSR. The outcome was antipsychotic initiation after the index date. Hazard ratios for antipsychotic initiation were estimated using the multivariable Cox model.

RESULTS

In terms of prognosis, the first 2 months after a stroke was the highest-risk period for antipsychotic use. A high burden of coexisting diseases carried an increased risk of antipsychotic use; in particular, chronic kidney disease (CKD) had the highest adjusted hazard ratio (aHR = 1.73; 95% CI 1.29-2.31) as compared with other risk factors. Furthermore, stroke severity and disability were significant risk factors for antipsychotic initiation.

CONCLUSIONS

Our study indicated that elderly stroke patients with chronic medical conditions, particularly CKD, and a higher stroke severity and disability were at greater risk of psychiatric disorders during the first 2 months after a stroke.

CLINICAL TRIAL REGISTRATION

NA.

Can markers of disease severity improve the predictive power of claims-based multimorbidity indices?

BACKGROUND

Claims-based measures of multimorbidity, which evaluate the presence of a defined list of diseases, are limited in their ability to predict future outcomes. We evaluated whether claims-based markers of disease severity could improve assessments of multimorbid burden.

METHODS

We developed 7 dichotomous markers of disease severity which could be applied to a range of diseases using claims data. These markers were based on the number of disease-associated outpatient visits, emergency department visits, and hospitalizations made by an individual over a defined interval; whether an individual with a given disease had outpatient visits to a specialist who typically treats that disease; and ICD-9 codes which connote more versus less advanced or symptomatic manifestations of a disease. Using Medicare claims linked with Health and Retirement Study data, we tested whether including these markers improved ability to predict ADL decline, IADL decline, hospitalization, and death compared to equivalent models which only included the presence or absence of diseases.

RESULTS

Of 5012 subjects, median age was 76 years and 58% were female. For a majority of diseases tested individually, adding each of the 7 severity markers yielded minimal increase in c-statistic (≤0.002) for outcomes of ADL decline and mortality compared to models considering only the presence versus absence of disease. Gains in predictive power were more substantial for a small number of individual diseases. Inclusion of the most promising marker in multi-disease multimorbidity indices yielded minimal gains in c-statistics (<0.001-0.007) for predicting ADL decline, IADL decline, hospitalization, and death compared to indices without these markers.

CONCLUSIONS

Claims-based markers of disease severity did not contribute meaningfully to the ability of multimorbidity indices to predict ADL decline, mortality, and other important outcomes.

A Machine Learning Approach to Support Urgent Stroke Triage Using Administrative Data and Social Determinants of Health at Hospital Presentation: Retrospective Study

BACKGROUND

The key to effective stroke management is timely diagnosis and triage. Machine learning (ML) methods developed to assist in detecting stroke have focused on interpreting detailed clinical data such as clinical notes and diagnostic imaging results. However, such information may not be readily available when patients are initially triaged, particularly in rural and underserved communities.

OBJECTIVE

This study aimed to develop an ML stroke prediction algorithm based on data widely available at the time of patients' hospital presentations and assess the added value of social determinants of health (SDoH) in stroke prediction.

METHODS

We conducted a retrospective study of the emergency department and hospitalization records from 2012 to 2014 from all the acute care hospitals in the state of Florida, merged with the SDoH data from the American Community Survey. A case-control design was adopted to construct stroke and stroke mimic cohorts. We compared the algorithm performance and feature importance measures of the ML models (ie, gradient boosting machine and random forest) with those of the logistic regression model based on 3 sets of predictors. To provide insights into the prediction and ultimately assist care providers in decision-making, we used TreeSHAP for tree-based ML models to explain the stroke prediction.

RESULTS

Our analysis included 143,203 hospital visits of unique patients, and it was confirmed based on the principal diagnosis at discharge that 73% (n=104,662) of these patients had a stroke. The approach proposed in this study has high sensitivity and is particularly effective at reducing the misdiagnosis of dangerous stroke chameleons (false-negative rate <4%). ML classifiers consistently outperformed the benchmark logistic regression in all 3 input combinations. We found significant consistency across the models in the features that explain their performance. The most important features are age, the number of chronic conditions on admission, and primary payer (eg, Medicare or private insurance). Although both the individual- and community-level SDoH features helped improve the predictive performance of the models, the inclusion of the individual-level SDoH features led to a much larger improvement (area under the receiver operating characteristic curve increased from 0.694 to 0.823) than the inclusion of the community-level SDoH features (area under the receiver operating characteristic curve increased from 0.823 to 0.829).

CONCLUSIONS

Using data widely available at the time of patients' hospital presentations, we developed a stroke prediction model with high sensitivity and reasonable specificity. The prediction algorithm uses variables that are routinely collected by providers and payers and might be useful in underresourced hospitals with limited availability of sensitive diagnostic tools or incomplete data-gathering capabilities.

Effectiveness and Safety of Clopidogrel vs Aspirin in Elderly Patients With Ischemic Stroke

OBJECTIVE

To evaluate the risks of recurrent stroke and major bleeding events with clopidogrel and aspirin use among patients aged 80 years or older.

PATIENTS AND METHODS

This retrospective cohort study was conducted using the Full Population Data of the Health and Welfare Database in Taiwan. Patients aged 80 years or older who received monotherapy with clopidogrel or aspirin following hospitalization for primary acute ischemic stroke between January 1, 2009, and December 31, 2018, were included. Inverse probability of treatment weighting was used to balance measured covariates between clopidogrel and aspirin users. Measured outcomes included recurrent acute ischemic stroke, acute myocardial infarction, composite cardiovascular events (recurrent stroke or acute myocardial infarction), intracranial hemorrhage, major gastrointestinal tract bleeding, and composite major bleeding events (intracranial hemorrhage or major gastrointestinal tract bleeding).

RESULTS

A total of 15,045 patients were included in the study, 1979 of whom used clopidogrel and 13,066 who used aspirin following hospitalization for primary acute ischemic stroke. Clopidogrel use was associated with significantly lower risk of recurrent acute ischemic stroke (hazard ratio [HR], 0.89; 95% CI, 0.83 to 0.96; P=.002), composite cardiovascular events (HR, 0.88; 95% CI, 0.82 to 0.95; P<.001), intracranial hemorrhage (HR, 0.71; 95% CI, 0.56 to 0.90; P=.005), and composite major bleeding events (HR, 0.89; 95% CI, 0.80 to 0.99; P=.04) compared with aspirin use.

CONCLUSION

In patients aged 80 years or older with primary acute ischemic stroke, clopidogrel users had lower risks of recurrent stroke and the composite cardiovascular events compared with aspirin users. Clopidogrel users also had lower risks of intracranial hemorrhage and the composite major bleeding events compared with aspirin users.

The weekend effect in stroke mortality: evidence from Austrian acute care hospitals

Many studies provide evidence for the so-called weekend effect by demonstrating that patients admitted to hospital during weekends show less favourable outcomes such as increased mortality, compared with similar patients admitted during weekdays. The underlying causes for this phenomenon are still discussed controversially. We analysed factors influencing weekend effects in inpatient care for acute stroke in Austria. The study analysed secondary datasets from all 130 public acute care hospitals in Austria between 2010 and 2014 (Austrian DRG Data). The study cohort included 86,399 patient cases admitted with acute ischaemic stroke. By applying multivariate regression analysis, we tested whether patient, treatment or hospital characteristics drove in-hospital mortality on weekends and national holidays. We found that the risk to die after an admission at weekend was significantly higher compared to weekdays, while the number of admissions following stroke was significantly lower. Adjustment for patient, treatment and hospital characteristics substantially reduced the weekend effect in mortality but did not eliminate it. We conclude that the observed weekend effect could be explained either by lower quality of health care or higher severity of stroke admissions at the weekend. In depth analyses supported the hypothesis of higher stroke severity in weekend patients as seen in other studies. While DRG data is useful to analyse stroke treatment and outcomes, adjustment for case mix and severity is essential.

Risk Factors for Postischemic Stroke Epilepsy in Young Adults: A Nationwide Population-Based Study in Taiwan

Background

The incidence of ischemic stroke has been increasing in the young population over the past 20 years. Poststroke epilepsy (PSE) is a common complication after stroke. However, few population-based studies with sufficient follow-up have investigated factors associated with PSE, especially factors related to comorbidities and unhealthy lifestyles in the modern young population. Accordingly, this study aimed to determine the long-term incidence and these risk factors for PSE young adults.

Methods

This cohort study was conducted using data from the Taiwan National Health Insurance Research Database (NHIRD) from 2002 to 2018. All patients aged between 19 and 44 years and diagnosed with ischemic stroke from 2002 to 2015 were retrospectively enrolled with a follow-up of at least 3 years. Multivariable Cox regression models were performed to identify predictors of PSE, including patients' demographics, baseline conditions, stroke severity, etiologies, comorbidities, and unhealthy behaviors.

Results

Among 6,512 ischemic stroke patients, 402 cases (6.2%) developed PSE who were with a mean follow-up period of 8.3 years (SD = 4.3 years). During the overall follow-up, stroke severity and manifestations were associated with PSE, including National Institutes of Health Stroke Scale (NIHSS) score ≥10 (aHR, 1.98; 95% CI, 1.50–2.61), seizure at first stroke admission [adjusted hazard ratio (aHR), 57.39; 95% confidence interval (CI), 43.02–76.55], length of hospital stay ≥14 days (aHR, 1.60; 95% CI, 1.26–2.02), recurrent stroke (aHR, 2.32; 95% CI, 1.85–2.90), aphasia (aHR, 1.77; 95% CI, 1.20–2.60), and malignancy (aHR, 2.05; 95% CI, 1.30–3.24). Furthermore, stroke patients with drug abuse were 2.90 times more likely to develop PSE than those without (aHR, 2.90; 95% CI, 1.53–5.50). By contrast, statin use (aHR, 0.62; 95% CI, 0.48–0.80) was associated with a lower risk of PSE. The risk factors at 1-year and 5-year PSE were similar to that of an overall follow-up.

Conclusions

Stroke severity, aphasia, malignancy, and drug abuse were associated increased risk of PSE and statin use may protect against PSE in young adults. Reducing the severity of stroke, statin use and controlling unhealthy behaviors might be able to decrease the development of PSE. Since PSE is associated with poor outcomes, early identification or intervention of PSE based on the risk factors might reduce the harmful effects of PSE.

External validation of the Passive Surveillance Stroke Severity Indicator

BACKGROUND

The Passive Surveillance Stroke Severity (PaSSV) Indicator was derived to estimate stroke severity from variables in administrative datasets but has not been externally validated.

METHODS

We used linked administrative datasets to identify patients with first hospitalization for acute stroke between 2007-2018 in Alberta, Canada. We used the PaSSV indicator to estimate stroke severity. We used Cox proportional hazard models and evaluated the change in hazard ratios and model discrimination for 30-day and 1-year case fatality with and without PaSSV. Similar comparisons were made for 90-day home time thresholds using logistic regression. We also linked with a clinical registry to obtain National Institutes of Health Stroke Scale (NIHSS) and compared estimates from models without stroke severity, with PaSSV, and with NIHSS.

RESULTS

There were 28,672 patients with acute stroke in the full sample. In comparison to no stroke severity, addition of PaSSV to the 30-day case fatality models resulted in improvement in model discrimination (C-statistic 0.72 [95%CI 0.71-0.73] to 0.80 [0.79-0.80]). After adjustment for PaSSV, admission to a comprehensive stroke center was associated with lower 30-day case fatality (adjusted hazard ratio changed from 1.03 [0.96-1.10] to 0.72 [0.67-0.77]). In the registry sample (N = 1328), model discrimination for 30-day case fatality improved with the inclusion of stroke severity. Results were similar for 1-year case fatality and home time outcomes.

CONCLUSION

Addition of PaSSV improved model discrimination for case fatality and home time outcomes. The validity of PASSV in two Canadian provinces suggests that it is a useful tool for baseline risk adjustment in acute stroke.

Detailed risks and characteristics of postepilepsy stroke in non-traumatic adult-onset epilepsy

BACKGROUND

Patients with epilepsy have an increased risk of stroke. However, the detailed risk and characteristics of postepilepsy stroke have not been investigated.

METHODS

This study utilized the National Health Insurance Research Database in Taiwan. We classified adult patients with newly diagnosed epilepsy from 2003 to 2016 as the epilepsy cohort. Patients in the nonepilepsy cohort were selected with propensity score matching at a case-control ratio of 1:5. The incidence, hazard ratio (HR), period-specific HR, recurrent HR in the Wei-Lin-Weissfeld model, stroke severity index, complications, and mortality of all stroke, ischemic stroke (IS) and hemorrhagic stroke events in the two cohorts were analyzed.

RESULTS

We enrolled 23,810 patients in the epilepsy cohort and 119,050 persons in the nonepilepsy cohort. The period-specific HRs of all stroke, IS and hemorrhagic stroke peaked immediately after epilepsy diagnosis and trended downward [Adjusted HRs of all stroke: 4.88 (3.88-6.14), 4.47 (3.50-5.70), 3.17 (2.62-3.84), 2.81 (2.27-3.48), 2.81 (2.36-3.34) and 2.33 (2.07-2.62) in 0-0.5, 0.5-1, 1-2, 2-3, 3-5 and ≥5 years after epilepsy diagnosis, respectively]. The recurrent stroke HRs in the epilepsy cohort were >1 from the first [3.06 (2.71-3.34)] to the fourth events [6.33 (1.08-37.03)]. IS events in the epilepsy cohort were associated with a younger onset age, a higher IS severity index, a higher rate of urinary tract infection, a lower in-hospital mortality, while 90-day stroke mortality was similar between the 2 cohorts.

CONCLUSION

Since the increased risk of stroke in epilepsy cohort peaked immediately after epilepsy diagnosis, early implementation of prevention strategies is considered.

High-Intensity Post-Stroke Rehabilitation Is Associated with Lower Risk of Pressure Ulcer Development in Patients with Stroke: Real-World Evidence from a Nationwide, Population-Based Cohort Study

Background and Objectives: Multiple factors are associated with pressure ulcer (PU) development, including limited mobility following stroke. We performed a nationwide cohort study to investigate the impact of rehabilitation intensity on the incidence of post-stroke PU. Materials and Methods: Data of patients diagnosed with stroke between 2000 and 2012 were collected from the 2000 Longitudinal Health Insurance Database (Taiwan). Based on the number of rehabilitation sessions attended within 90 days of discharge, the rehabilitation intensity was classified as low, medium, or high. After adjusting for sociodemographic factors and comorbidities, the Cox proportional hazards model evaluated the risk of PU development during the 12-year follow-up period. Kaplan−Meier curves were used to estimate the cumulative incidence of PUs. Results: Our study included 18,971 patients who had their first episode of stroke. Of these, 9829 (51.8%) underwent rehabilitation therapy after discharge. Female patients and patients with a National Institutes of Health Stroke Scale (NIHSS) score >13 points, who commenced high-intensity post-stroke rehabilitation after discharge had a significantly lower risk of PU development than those who underwent low-intensity post-stroke rehabilitation after discharge. Cumulative survival analysis showed a significantly lower cumulative incidence of PU during the 12-year follow-up period in the high-intensity rehabilitation group. Conclusion: Compared with low-intensity post-stroke rehabilitation, high-intensity post-stroke rehabilitation after discharge from hospital is associated with a lower risk of post-stroke PU development, especially in female stroke patients and patients with a NIHSS score >13 points. High-intensity rehabilitation is also associated with a significantly lower cumulative incidence of PU events during the 12-year follow-up period.

Development of Deep Learning Models for Predicting In-Hospital Mortality Using an Administrative Claims Database: Retrospective Cohort Study

BACKGROUND

Administrative claims databases have been used widely in studies because they have large sample sizes and are easily available. However, studies using administrative databases lack information on disease severity, so a risk adjustment method needs to be developed.

OBJECTIVE

We aimed to develop and validate deep learning-based prediction models for in-hospital mortality of acute care patients.

METHODS

The main model was developed using only administrative claims data (age, sex, diagnoses, and procedures on the day of admission). We also constructed disease-specific models for acute myocardial infarction, heart failure, stroke, and pneumonia using common severity indices for these diseases. Using the Japanese Diagnosis Procedure Combination data from July 2010 to March 2017, we identified 46,665,933 inpatients and divided them into derivation and validation cohorts in a ratio of 95:5. The main model was developed using a 9-layer deep neural network with 4 hidden dense layers that had 1000 nodes and were fully connected to adjacent layers. We evaluated model discrimination ability by an area under the receiver operating characteristic curve (AUC) and calibration ability by calibration plot.

RESULTS

Among the eligible patients, 2,005,035 (4.3%) died. Discrimination and calibration of the models were satisfactory. The AUC of the main model in the validation cohort was 0.954 (95% CI 0.954-0.955). The main model had higher discrimination ability than the disease-specific models.

CONCLUSIONS

Our deep learning-based model using diagnoses and procedures produced valid predictions of in-hospital mortality.

Long-Term Rehabilitation Utilization Pattern Among Stroke Patients Under the National Health Insurance Program

OBJECTIVE

The aim of this study was to understand the frequency of patients receiving rehabilitation services at various periods after stroke and the possible medical barriers to receiving rehabilitation.

DESIGN

A retrospective cohort study was conducted using a nationally representative sample in Taiwan. A total of 14,600 stroke patients between 2005 and 2011 were included. Utilization of physical therapy or occupational therapy at different periods after stroke onset was the outcome variable. Individual and geographic characteristics were investigated to determine their effect on patients' probability of receiving rehabilitation.

RESULTS

More severe stroke or more comorbid diseases increased the odds of receiving physical therapy and occupational therapy; older age was associated with decreased odds. Notably, sex and stroke type influenced the odds of rehabilitation only in the early period. Copayment exemption lowered the odds of rehabilitation in the first 6 mos but increased the odds in later periods. Rural and suburban patients had significantly lower odds of receiving physical therapy and occupational therapy, as did patients living in areas with fewer rehabilitation therapists.

CONCLUSIONS

Besides personal factors, geographic factors such as urban-rural gaps and number of therapists were significantly associated with the utilization of post-stroke rehabilitation care. Furthermore, the influence of certain factors, such as sex, stroke type, and copayment exemption type, changed over time.

Utilization of screening and treatment for osteoporosis among stroke survivors

BACKGROUND

Stroke survivors are prone to osteoporosis and fractures. However, bone mineral density (BMD) testing and osteoporosis treatment were underutilized in patients with recent stroke. We aimed to examine whether stroke has an impact on the utilization of BMD testing and osteoporosis treatment as well as the determinants of their utilization in stroke patients using nationwide population-based data in Taiwan.

METHODS

We identified patients aged 55 years and older who were hospitalized for hemorrhagic or ischemic stroke as the stroke cohort, and age- and sex-matched patients hospitalized for reasons other than stroke, fracture, or fall as the non-stroke cohort. We used the Fine-Gray sub-distribution hazard competing risk regression model to determine the predictors for BMD testing and osteoporosis treatment.

RESULTS

A total of 32997 stroke patients and 32997 age- and sex-matched controls comprised the stroke and non-stroke cohorts, respectively. BMD testing and osteoporosis treatment were performed in 1.0% and 5.2% of the stroke patients, respectively, within one year after hospitalization while these measures were performed in 0.8% and 4.7% of the controls. Stroke patients were more likely to receive BMD testing (adjusted hazard ratio [HR] 1.33; 95% confidence interval [CI] 1.11-1.58) and osteoporosis treatment (adjusted HR 1.19; 95% CI 1.11-1.29). Female sex, osteoporosis, prior BMD testing, and low-trauma fractures after stroke increased the likelihood of using BMD testing and osteoporosis treatment whereas greater stroke severity reduced the likelihood of receiving both measures.

CONCLUSIONS

Both BMD testing and osteoporosis treatment were underutilized among stroke survivors even though they had a higher chance of receiving both measures than non-stroke patients.

Stroke Disease Detection and Prediction Using Robust Learning Approaches

Stroke is a medical disorder in which the blood arteries in the brain are ruptured, causing damage to the brain. When the supply of blood and other nutrients to the brain is interrupted, symptoms might develop. According to the World Health Organization (WHO), stroke is the greatest cause of death and disability globally. Early recognition of the various warning signs of a stroke can help reduce the severity of the stroke. Different machine learning (ML) models have been developed to predict the likelihood of a stroke occurring in the brain. This research uses a range of physiological parameters and machine learning algorithms, such as Logistic Regression (LR), Decision Tree (DT) Classification, Random Forest (RF) Classification, and Voting Classifier, to train four different models for reliable prediction. Random Forest was the best performing algorithm for this task with an accuracy of approximately 96 percent. The dataset used in the development of the method was the open-access Stroke Prediction dataset. The accuracy percentage of the models used in this investigation is significantly higher than that of previous studies, indicating that the models used in this investigation are more reliable. Numerous model comparisons have established their robustness, and the scheme can be deduced from the study analysis.

Oral Anticoagulation Timing in Patients with Acute Ischaemic Stroke and Atrial Fibrillation

BACKGROUND AND PURPOSE

Oral anticoagulants (OACs) prevent stroke recurrence and vascular embolism in patients with acute ischaemic stroke (AIS) and atrial fibrillation (AF). Current guidance recommends a "1-3-6-12 day"' rule to resume OACs after AIS, based mainly on empirical consensus. This study investigated the suitability of guideline-recommended timing for OAC initiation.

METHODS

To overcome immortal time bias, we emulated a sequence of randomized placebo-controlled trials and constructed 90 propensity score-matched cohorts of 12,307 patients with AF and AIS from 2012 to 2016. We compared the risk of composite effectiveness and safety outcome in the early vs no OAC use group and in the delayed vs no OAC use. Indirect comparison between early and delayed use was conducted using a network meta-analysis.

RESULTS

Across the groups of AIS severity, the risks of composite outcome or effectiveness outcome were lower in the OAC use group than the no use group and the risks were similar between the early and delayed use groups. In patients with severe AIS, those receiving early OACs use had an increased risk of safety outcome, with HR of 2.10 (CI: 1.13-3.92) compared with those without OAC use, and HR of 1·44 (CI: 0·99-2·09) compared with those receiving delayed use.

CONCLUSIONS

In AF patients with severe AIS, early OAC use before the guideline-recommended days appeared to increase the risk of bleeding events, although the OAC initiation time seemed not to affect the risk of serious vascular events. The optimal severity-specific timing for OAC initiation after AIS requires further evaluation.

Mechanical thrombectomy reduces the gap in treatment outcomes of ischemic stroke between hospital levels of care: analysis of a Korean nationwide data

Background

Mechanical thrombectomy (MT) of ischemic stroke was demonstrated to be effective in clinical trials and was reported to have favorable outcomes in real clinical settings since 2015. We aimed to determine the national trends of MT and compare the outcomes between the different levels of treating hospital.

Methods

We obtained data from the nationwide database from 2008 to 2017. Patients with ischemic stroke who received MT were identified using the International Classification of Disease Codes. Good outcome was defined as discharge to home, and a poor outcome was defined as cerebral hemorrhage, physical disability, or death. The study period was divided into three (off-label MT, transitional, MT period). Hospital groups where MT was performed were divided into tertiary and non-tertiary hospitals.

Results

In MT period, 47.0% of the MT procedures were performed in non-tertiary hospitals compared with 36.1% in off-label MT period. Comparison of the 3-month mortality between patients who were treated in tertiary vs. non-tertiary hospitals revealed significant lower mortality in tertiary hospital through all period. The incidence of cerebral hemorrhage and physical disability did not differ between hospital groups. However, the percentage of patients discharged home was 41.4% for tertiary hospitals and 42.4% for non-tertiary hospitals, which was not statistically different in MT period (P=0.4671).

Conclusions

Analysis of the nationwide data confirmed that the extent of increase in MT was higher in non-tertiary hospitals than tertiary hospitals. In addition, no significant difference was revealed in the number of favorable clinical outcome between the hospital groups during MT period.

Functional and Cognitive Impairments Increased Risks of Outcomes of Healthcare Utilization in Patients With Stroke Receiving Home and Community-Based Care in Taiwan

Aim: Stroke is a leading cause of disability; however, little is known about the outcomes of the utilization of long-term care (LTC) recipients in Taiwan. This study aimed to quantify the burdens of disease of stroke survivors receiving LTC by evaluating the outcomes of their utilization including mortality, readmissions, and re-emergency within 1 year after diagnoses of strokes. Methods: By interlinkages among the national mortality registry, LTC dataset (LTC-CM), and the National Health Insurance Research Dataset (NHIRD), the outcomes and the factors associated with receiving LTC up to 1 year were explored. Patients were aged 50 years and over with an inpatient claim of the first diagnosis of stroke of intracerebral hemorrhage (ICH) and ischemic stroke during 2011-2016. Outcomes of the healthcare utilization include rehospitalization and re-emergency. Results: There were 15,662 patients with stroke who utilized the LTC services in the dataset among the stroke population in NHIRD. Stroke survivors receiving LTC showed no difference in clinical characteristics and their expected years of life loss (EYLL = 7.4 years) among those encountered in NHIRD. The LTC recipients showed high possibilities to be rehospitalized and resent to emergency service within 1 year after diagnosis. Apart from the comorbidity and stroke severity, both the physical and mental functional disabilities and caregiving resources predicted the outcomes of the utilization. Conclusions: For stroke survivors, both severe functional impairments and cognitive impairments were found as important factors for healthcare utilizations. These results regarding reserving functional abilities deserve our consideration in making the decision on the ongoing LTC policy reform in the aged society of Taiwan.

Comparative safety of antipsychotic medications in elderly stroke survivors: A nationwide claim data and stroke registry linkage cohort study

BACKGROUND

Antipsychotics remain the first choice of treatment for post-stroke psychosis, despite an increased risk of mortality reported in elderly patients. We aimed to compare the mortality risk among antipsychotics in elderly patients with stroke using the stroke registry for external adjustment.

METHODS

We conducted a retrospective cohort study to identify patients aged above 65 years who were admitted for stroke in the National Health Insurance Database (NHID) from 2002 to 2014. The first date of antipsychotic use after the stroke hospitalization was defined as the index date. Covariates including diseases, medications and external information on smoking, BMI, stroke severity and disability, that were unavailable in the NHID were obtained from the linked Multicenter Stroke Registry (MSR) and used for propensity score calibration (PSC). The main outcome was one-year all-cause mortality.

RESULTS

Stroke patients in the NHID prescribed with haloperidol, quetiapine and risperidone numbered 22,235, 28,702 and 8 663, respectively. In the PSC-adjusted analyses, haloperidol [adjusted hazard ratio (aHR) = 1.22; 95% CI 1.18-1.27] and risperidone (aHR = 1.31; 95% CI 1.24-1.38) users had a higher mortality risk than quetiapine users. When the dosage was higher than 0.5 defined daily dose (DDD), haloperidol and risperidone users had a significant mortality risk as compared with those taking a lower dose.

CONCLUSIONS

In post-stroke elderly patients, quetiapine would pose less mortality risk than risperidone and haloperidol at doses higher than 0.5 DDD. When haloperidol or risperidone is indicated, starting with a lower dose is suggested to avoid excess risk.

Associations between stroke type, stroke severity, and pre-stroke osteoporosis with the risk of post-stroke fracture: A nationwide population-based study

Background Recognizing the post-stroke fracture risk factors is crucial for targeted intervention and primary fracture prevention. We aimed to investigate whether stroke types, stroke severity, and pre-stroke osteoporosis are associated with post-stroke fracture. Methods In a nationwide cohort, we identified previously fracture-free patients who suffered from first-ever stroke, either acute ischemic stroke (AIS) or intracerebral hemorrhage (ICH), between 2003 and 2015. Information regarding stroke severity, osteoporosis, comorbidity, and medication information was collected. The outcomes analyzed included hip fracture, spine fracture, and other fractures. Cumulative incidence functions (CIFs) were used to estimate the cumulative incidence of fractures over time after accounting for competing risk of death. Multivariable Fine and Gray models were used to determine the adjusted hazard ratio (HR) and 95% confidence interval (CI). Results Of the 41,895 patients with stroke, the 5-year CIFs of any incident fracture, hip fracture, spine fracture, and other fractures were 8.03%, 3.42%, 1.87%, and 3.05%, respectively. The fracture risk did not differ between patients with AIS and ICH. While osteoporosis increased the risk of post-stroke fracture (adjusted HR [95% CI],1.42 [1.22-1.66]), stroke severity was inversely associated with post-stroke fracture (moderate, 0.88 [0.81-0.96] and severe, 0.39 [0.34-0.44], compared with mild stroke severity). Conclusions Stroke survivors had an over 8% fracture risk at 5 years after stroke. Mild stroke severity and osteoporosis were significantly associated with post-stroke fracture risk, whereas stroke type was not. Our results call for effective measures for bone health screening and fracture prevention in patients with stroke.

Healthcare costs before and after stroke in patients with non-valvular atrial fibrillation who initiated treatment with rivaroxaban or warfarin

AIMS

Rivaroxaban reduces stroke compared with warfarin in patients with non-valvular atrial fibrillation (NVAF). This study compared healthcare costs before and after stroke in NVAF patients treated with rivaroxaban or warfarin.

MATERIALS AND METHODS

Using de-identified IBM MarketScan Commercial and Medicare databases, this retrospective cohort study (from 2011 to 2019) included patients with NVAF who initiated rivaroxaban or warfarin within 30 days after initial NVAF diagnosis. Patients who developed stroke were identified, and stroke severity was determined by the National Institutes of Health Stroke Scale (NIHSS) score, imputed by a random forest method. Total all-cause per-patient per-year (PPPY) costs of care were determined for patients: (1) who developed stroke during the pre- and post-stroke periods and (2) who remained stroke-free during the follow-up period. Treatment groups were balanced using inverse probability of treatment weighting.

RESULTS

A total of 13,599 patients initiated rivaroxaban and 39,861 initiated warfarin, of which 272 (2.0%) and 1,303 (3.3%), respectively, developed stroke during a mean follow-up of 28 months. Among patients who developed stroke, PPPY costs increased from the pre-stroke to post-stroke period, with greater increases in the warfarin cohort relative to the rivaroxaban cohort. Overall, the costs increased by 1.78-fold for rivaroxaban vs 3.07-fold for warfarin; for less severe strokes (NIHSS < 5), costs increased 0.88-fold and 1.05-fold, respectively. Cost increases for more severe strokes (NIHSS ≥ 5) among rivaroxaban patients were half those for warfarin patients (3.19-fold vs 6.37-fold, respectively). Among patients without stroke, costs were similar during the follow-up period between the two treatment groups.

CONCLUSIONS

Total all-cause costs of care increased in the post-stroke period, and particularly in the patients treated with warfarin relative to those treated with rivaroxaban. The lower rate of stroke in the rivaroxaban cohort suggests that greater pre- to post-stroke cost increases result from more strokes occurring in the warfarin cohort.

Comparison of hospital length of stay of acute ischemic stroke patients with non-valvular atrial fibrillation started on rivaroxaban or warfarin treatment during hospitalization

OBJECTIVE

To compare the hospital length of stay (LOS) between rivaroxaban and warfarin in hospitalized acute stroke patients with non-valvular atrial fibrillation (NVAF) in Japan.

METHODS

This was a retrospective, observational study using a Japanese hospital claims database. Data of NVAF patients who were started on oral anticoagulant (OAC) treatment during hospitalization were extracted and LOS-OAC (period from the initiation of index OAC therapy to the end of hospitalization or censoring date) and medical costs were compared between rivaroxaban and warfarin treatments. To compare LOS-OAC, a time-to-event analysis was performed using the Kaplan-Meier method. The analysis period was from April 2012 to December 2015.

RESULTS

This study included 773 rivaroxaban users and 1077 warfarin users. After the propensity score matching, 546 patients for each treatment constituted the matched cohorts. Although the rivaroxaban users had a similar LOS-OAC to warfarin users (median, 18 vs. 19 days, p = .657) in the matched cohorts, 3 days shorter LOS-OAC was observed in the rivaroxaban users (median, 17 vs. 20 days, p = .043) after IPTW adjustment. Subgroup analysis by the severity of stroke after IPTW adjustment demonstrated that rivaroxaban users had a shorter LOS-OAC than warfarin users among patients with mild (median, 10 vs. 14 days) and moderate stroke severity (22 vs. 27 days), but not among those with severe stroke severity (26 vs. 25 days).

LIMITATIONS

It is not possible to say that the only confounder was stroke severity and therefore other possible known and unknown confounders could not be ruled out.

CONCLUSIONS

The rivaroxaban users had a 3-day shorter LOS-OAC after IPTW-adjustment. Using rivaroxaban was associated with 4-5 days shorter LOS-OAC than using warfarin in patients with mild or moderate stroke, though treatment selection did not have a large impact in patients with severe stroke.

Predicting clinically significant motor function improvement after contemporary task-oriented interventions using machine learning approaches

BACKGROUND

Accurate prediction of motor recovery after stroke is critical for treatment decisions and planning. Machine learning has been proposed to be a promising technique for outcome prediction because of its high accuracy and ability to process large volumes of data. It has been used to predict acute stroke recovery; however, whether machine learning would be effective for predicting rehabilitation outcomes in chronic stroke patients for common contemporary task-oriented interventions remains largely unexplored. This study aimed to determine the accuracy and performance of machine learning to predict clinically significant motor function improvements after contemporary task-oriented intervention in chronic stroke patients and identify important predictors for building machine learning prediction models.

METHODS

This study was a secondary analysis of data using two common machine learning approaches, which were the k-nearest neighbor (KNN) and artificial neural network (ANN). Chronic stroke patients (N = 239) that received 30 h of task-oriented training including the constraint-induced movement therapy, bilateral arm training, robot-assisted therapy and mirror therapy were included. The Fugl-Meyer assessment scale (FMA) was the main outcome. Potential predictors include age, gender, side of lesion, time since stroke, baseline functional status, motor function and quality of life. We divided the data set into a training set and a test set and used the cross-validation procedure to construct machine learning models based on the training set. After the models were built, we used the test data set to evaluate the accuracy and prediction performance of the models.

RESULTS

Three important predictors were identified, which were time since stroke, baseline functional independence measure (FIM) and baseline FMA scores. Models for predicting motor function improvements were accurate. The prediction accuracy of the KNN model was 85.42% and area under the receiver operating characteristic curve (AUC-ROC) was 0.89. The prediction accuracy of the ANN model was 81.25% and the AUC-ROC was 0.77.

CONCLUSIONS

Incorporating machine learning into clinical outcome prediction using three key predictors including time since stroke, baseline functional and motor ability may help clinicians/therapists to identify patients that are most likely to benefit from contemporary task-oriented interventions. The KNN and ANN models may be potentially useful for predicting clinically significant motor recovery in chronic stroke.

Lifetime risks and health impacts of hemorrhagic and ischemic stroke in South Korea

This study is aimed toward estimating the lifetime risks, life expectancy, expected years of life lost (EYLL), and lifetime costs related to different subtypes of stroke in South Korea. We included 13,994 patients diagnosed with stroke (ICD-10, I60-I63) in the National Health Insurance Service-National Sample Cohort of Korea between 2006 and 2015. Lifetime risks were calculated using the cumulative incidence rate for patients aged 18–84. Lifetime survival data were obtained through the Kaplan–Meier method and extrapolated with a rolling-over extrapolation algorithm. The lifetime costs were estimated by multiplying the average monthly expenditures with the survival probabilities and adding the values over lifetime. The lifetime risks of stroke in Korea have been decreasing consistently over the last decade with the exception of subarachnoid hemorrhage in females, which appears to have slightly increased. The EYLL is higher in hemorrhagic stroke than in ischemic stroke (6–9.7 vs. 4.7). Expected lifetime costs reimbursed by the NHIS would amount to about $71,406 accompanied with $14,921 copayment from the patients for hemorrhagic stroke, and $50,551 and $11,666, respectively, for ischemic stroke. Further studies are warranted to combine survival with quality of life and functional disability to obtain a more detailed outcome assessment of the potential impact of the prevention of stroke.

Predictors for the use of traditional Chinese medicine among inpatients with first-time stroke: a population-based study

BACKGROUND

Stroke is one of the major causes of death and disability. The treatments that are provided to patients during hospitalization after an acute stroke are very important in stabilizing their medical condition and enabling the recovery of their motor functions. However, limited information is available regarding the use of traditional Chinese medicine (TCM) during hospitalization for first-time stroke patients. The researchers aimed to investigate the factors affecting TCM use and to provide clinicians with comprehensive information on TCM use among first-time stroke inpatients in Taiwan.

METHODS

The researchers collected and analyzed data, including patient characteristics, TCM use, and TCM prescription patterns, from the National Health Insurance Research Database in Taiwan for first-time stroke inpatients between 2006 and 2012.

RESULTS

Among the 89,162 first-time stroke patients, 7455 were TCM users, and 81,707 were TCM nonusers. The predictors for TCM use were as follows: age, 45-64 or < 45 years; men; living in a level 2, 4, or 7 urbanized area; insured amount ≥ 576 USD per month; ischemic stroke; hospitalized for first-time stroke for 8-14 days, 15-28 days, or ≥ 29 days; stroke severity index score 0-9 or 10-19; Charlson-Deyo comorbidity index score 0 or 1-2; hospitalization in a regional or community hospital; receiving rehabilitation; and previous experience with outpatient TCM use. An increase in the number of TCM users was observed from 2006 to 2012. Furthermore, 68.8-79.7% of TCM users used acupuncture only, while 17.8-26.1% used both acupuncture and Chinese herbal medicine.

CONCLUSIONS

An increasing number of first-time stroke patients have been choosing TCM as a complementary treatment during hospitalization. Moreover, TCM use is associated with demographic, clinical, and socioeconomic characteristics. These findings may help clinicians comprehensively understand the trend and the important factors affecting TCM utilization among patients who are hospitalized due to first-time stroke.

Holiday Season and Weekend Effects on Stroke Mortality: A Nationwide Cohort Study Controlling for Stroke Severity

Background The effect of holiday season admission for stroke on mortality has not been investigated. Thus, we aimed to evaluate whether "holiday season" and "weekend" effects exist on mortality risk for stroke admission. Methods and Results A nationwide cohort study was conducted using Taiwan's National Health Insurance Research Database. We identified all patients admitted for stroke between 2011 and 2015 in Taiwan, and categorized them according to the admission date: holiday season (at least 4 days off) (n=3908), weekend (n=13 774), and weekday (n=49 045). We analyzed in-hospital, 7-day, and 30-day mortality using multivariable logistic regression, adjusting for stroke severity and other confounders. Compared with weekday admissions, holiday season admission for stroke was significantly associated with a 20%, 33%, and 21% increase in in-hospital, 7-day, and 30-day mortality, respectively. Compared with weekend admissions, holiday season admissions were associated with a 24%, 30%, and 22% increased risk of in-hospital, 7-day, and 30-day mortality, respectively. However, mortality did not differ significantly between weekend and weekday admissions. Subanalyses after stratification for age, sex, and stroke type also revealed similar trends. Conclusions We report for the first time a "holiday season effect" on stroke mortality. Patients admitted during holiday seasons had higher mortality risks than those admitted on weekends and weekdays. This holiday season effect persisted even after adjusting for stroke severity and other important confounders. These findings highlight the need for healthcare delivery systems with a consistent quality of round-the-clock care for patients admitted for stroke.

Home-Time as a Surrogate Measure for Functional Outcome After Stroke: A Validation Study

Purpose

Home-time has been found to correlate well with modified Rankin Scale (mRS) scores in patients with stroke. This study aimed to determine its correlations in patients with different types of stroke at various time points after stroke in a non-Western population.

Methods

This study used linked data from multi-center stroke registry databases and a nationwide claims database of health insurance. Functional outcomes as measured with the modified Rankin Scale were obtained from the registry databases and home-time was derived from the claims database. Spearman correlation coefficients were used to assess the correlation between home-time and mRS scores. The area under the receiver operating characteristic curve (AUC) was used to evaluate the performance of home-time in predicting good functional outcome.

Results

This study included 7959 patients hospitalized for stroke or transient ischemic attack (TIA), for whom mRS scores were available in 6809, 6694, and 4330 patients at 90, 180, and 365 days, respectively. Home-time was highly correlated with mRS scores at the three time-points in patients with ischemic (Spearman's rho -0.69 to -0.83) or hemorrhagic (Spearman's rho -0.86 to -0.88) stroke, but the correlation was only weak to moderate in those with TIA (Spearman's rho -0.32 to -0.58). Home-time predicted good functional outcome with excellent discrimination in patients with ischemic (AUCs >0.8) or hemorrhagic (AUCs >0.9) stroke but less so in those with TIA (AUCs >0.7).

Conclusion

Home-time was highly correlated with mRS scores and showed excellent discrimination in predicting good functional outcome in patients with ischemic or hemorrhagic stroke. Home-time could serve as a valid surrogate measure for functional outcome after stroke.

Efficacy and safety of clopidogrel and aspirin do not differ in patients with stable ischemic stroke

BACKGROUND

The current study compared the efficacy and safety of clopidogrel vs aspirin in the secondary prevention of ischemic stroke (IS).

METHODS

We included patients from the Taiwan National Health Insurance Research Database who were aged between 20 and 80 years, had their first ever IS, had no diagnosis of atrial fibrillation, and had not used an oral anticoagulant before the index IS between 2002 and 2010. We excluded patients who died or were admitted to a hospital due to acute myocardial infarction, recurrent IS, or major bleeding within 3 months of IS. Patients were then classified into clopidogrel as aspirin users. Propensity score matching was adopted to select clopidogrel and aspirin groups with similar baseline characteristics (n = 8457 vs 16,914, mean follow-up period of 2.1 years and 1.9 years, respectively). Conditional Cox proportional hazard regression was used to compare risks of all-cause death, cardiovascular death, recurrent stroke, acute myocardial infarction, and major bleeding in clopidogrel users and aspirin users.

RESULTS

The risks of all-cause death, cardiovascular death, recurrent stroke, and acute myocardial infarction did not differ between clopidogrel and aspirin users. Subgroup analyses revealed that the results were consistent regardless of age, disease severity, or comorbidity.

CONCLUSION

According to real-world data, the efficacy and safety of clopidogrel and aspirin for secondary prevention of stable IS did not differ.

Deriving a Passive Surveillance Stroke Severity Indicator From Routinely Collected Administrative Data: The PaSSV Indicator

BACKGROUND

Adjusting for stroke severity is crucial for stroke outcomes research. However, this information is not available in administrative healthcare data. We aimed to derive an indicator of baseline stroke severity using these data.

METHODS AND RESULTS

We identified patients with stroke enrolled in a population-based registry in Ontario, Canada, and used the Canadian Neurological Scale (CNS), documented in the registry, as a measure of stroke severity. We derived an estimated CNS from a linear regression model in which we regressed the observed CNS on predictor variables: age, sex, arrival by ambulance, interhospital transfer, mechanical ventilation, and an emergency department triage score. The effect of stroke severity on the estimated hazard ratios for 30-day mortality was determined in 3 Cox-proportional hazards models with (1) no CNS, (2) observed CNS, and (3) estimated CNS, all adjusted for age, sex, Charlson index, and stroke type. We assessed model discrimination using C statistics. To assess for construct validity, we repeated these analyses in a subset of patients with documented National Institute of Health Stroke Scale and in a cohort of patients with stroke external to the registry. We derived the estimated stroke severity in 41 481 patients (48.7% female, median age of 75 years [interquartile range, 64- 83]). The magnitude of the association between stroke severity and mortality was similar for the observed and estimated CNS. The discriminative ability of the Cox-proportional hazards models to predict mortality was highest when the observed CNS was included (C statistic, 0.82 [95% CI, 0.81-0.82]), moderate with estimated CNS (0.76 [0.75-0.76]), and lowest without CNS (0.69 [0.69-0.70]. Our findings were replicated with the National Institute of Health Stroke Scale and in the external cohort.

CONCLUSIONS

We derived an estimated measure of stroke severity using administrative data. This can be applied for risk adjustment in population-based stroke outcomes research and in assessments of health system performance.

Risk Factors Associated with Outcomes of Recombinant Tissue Plasminogen Activator Therapy in Patients with Acute Ischemic Stroke

Ischemic stroke is the most common type of stroke, and early interventional treatment is associated with favorable outcomes. In the guidelines, thrombolytic therapy using recombinant tissue-type plasminogen activator (rt-PA) is recommended for eligible patients with acute ischemic stroke. However, the risk of hemorrhagic complications limits the use of rt-PA, and the risk factors for poor treatment outcomes need to be identified. To identify the risk factors associated with in-hospital poor outcomes in patients treated with rt-PA, we analyzed the electronic medical records of patients who were diagnosed with acute ischemic stroke and treated for rt-PA at Chang Gung Memorial Hospitals from 2006 to 2016. In-hospital death, intensive care unit (ICU) stay, or prolonged hospitalization were defined as unfavorable treatment outcomes. Medical history variables and laboratory test results were considered variables of interest to determine risk factors. Among 643 eligible patients, 537 (83.5%) and 106 (16.5%) patients had favorable and poor outcomes, respectively. In the multivariable analysis, risk factors associated with poor outcomes were female gender, higher stroke severity index (SSI), higher serum glucose levels, lower mean corpuscular hemoglobin concentration (MCHC), lower platelet counts, and anemia. The risk factors found in this research could help us study the treatment strategy for ischemic stroke.

Assessing stroke severity using electronic health record data: a machine learning approach

Background

Stroke severity is an important predictor of patient outcomes and is commonly measured with the National Institutes of Health Stroke Scale (NIHSS) scores. Because these scores are often recorded as free text in physician reports, structured real-world evidence databases seldom include the severity. The aim of this study was to use machine learning models to impute NIHSS scores for all patients with newly diagnosed stroke from multi-institution electronic health record (EHR) data.

Methods

NIHSS scores available in the Optum© de-identified Integrated Claims-Clinical dataset were extracted from physician notes by applying natural language processing (NLP) methods. The cohort analyzed in the study consists of the 7149 patients with an inpatient or emergency room diagnosis of ischemic stroke, hemorrhagic stroke, or transient ischemic attack and a corresponding NLP-extracted NIHSS score. A subset of these patients (n = 1033, 14%) were held out for independent validation of model performance and the remaining patients (n = 6116, 86%) were used for training the model. Several machine learning models were evaluated, and parameters optimized using cross-validation on the training set. The model with optimal performance, a random forest model, was ultimately evaluated on the holdout set.

Results

Leveraging machine learning we identified the main factors in electronic health record data for assessing stroke severity, including death within the same month as stroke occurrence, length of hospital stay following stroke occurrence, aphagia/dysphagia diagnosis, hemiplegia diagnosis, and whether a patient was discharged to home or self-care. Comparing the imputed NIHSS scores to the NLP-extracted NIHSS scores on the holdout data set yielded an R² (coefficient of determination) of 0.57, an R (Pearson correlation coefficient) of 0.76, and a root-mean-squared error of 4.5.

Conclusions

Machine learning models built on EHR data can be used to determine proxies for stroke severity. This enables severity to be incorporated in studies of stroke patient outcomes using administrative and EHR databases.

Risks of Stroke and Mortality in Atrial Fibrillation Patients Treated With Rivaroxaban and Warfarin

Background and Purpose- Oral anticoagulation therapy is standard of care for patients with nonvalvular atrial fibrillation to prevent stroke. This study compared rivaroxaban and warfarin for stroke and all-cause mortality risk reduction in a real-world setting. Methods- This retrospective cohort study (2011-2017) included de-identified patients from the Optum Clinformatics Database who started treatment with rivaroxaban or warfarin within 30 days following initial diagnosis of nonvalvular atrial fibrillation. Before nonvalvular atrial fibrillation diagnosis, patients had 6 months of continuous health plan enrollment and CHA₂DS₂-VASc score ≥2. Stroke severity was determined by the National Institutes of Health Stroke Scale, imputed based on machine learning algorithms. Stroke and all-cause mortality risks were compared by treatment using Cox proportional hazard regression, with inverse probability of treatment weighting to balance cohorts for baseline risk factors. Stratified analysis by treatment duration was also performed. Results- During a mean follow-up of 27 months, 175 (1.33/100 patient-years [PY]) rivaroxaban-treated and 536 (1.66/100 PY) warfarin-treated patients developed stroke. The inverse probability of treatment weighting model showed that rivaroxaban reduced stroke risk by 19% (hazard ratio [HR], 0.81 [95% CI, 0.73-0.91]). Analysis by stroke severity revealed risk reductions by rivaroxaban of 48% for severe stroke (National Institutes of Health Stroke Scale score, 16-42; HR, 0.52 [95% CI, 0.33-0.82]) and 19% for minor stroke (National Institutes of Health Stroke Scale score, 1 to <5; HR, 0.81 [95% CI, 0.68-0.96]), but no difference for moderate stroke (National Institutes of Health Stroke Scale score, 5 to <16; HR, 0.93 [95% CI, 0.78-1.10]). A total of 41 (0.31/100 PY) rivaroxaban-treated and 147 (0.44/100 PY) warfarin-treated patients died poststroke, 12 (0.09/100 PY) and 67 (0.20/100 PY) of whom died within 30 days, representing mortality risk reductions by rivaroxaban of 24% (HR, 0.76 [95% CI, 0.61-0.95]) poststroke and 59% (HR, 0.41 [95% CI, 0.28-0.60]) within 30 days. Conclusions- After the initial diagnosis of atrial fibrillation, patients treated with rivaroxaban versus warfarin had significant risk reduction for stroke, especially severe stroke, and all-cause mortality after a stroke. Findings from this observational study may help inform anticoagulant choice for stroke prevention in patients with nonvalvular atrial fibrillation.

Economic evaluation of transferring first-stroke survivors to rehabilitation wards: A 10-year longitudinal, population-based study

Background: Transferring stroke survivors to the rehabilitation ward for rehabilitation reduces long-term mortality; however, the long-term economic impact remains unknown.Objective: We aimed to assess the 10-year economic outcome of transferring first-stroke survivors to the rehabilitation ward.Methods: In this population-based, retrospective study, we examined the incremental costs per life year gained (ICLYG) for stroke survivors who were transferred to the rehabilitation ward (TR) as compared to that for those who underwent rehabilitation without being transferred to the rehabilitation ward (R) and those who did not undergo rehabilitation (NR). The differences in the daily medical expenditures among the three groups during the 10-year post-stroke period were examined.Results: After balancing characteristics of the three groups, the data of 14,544 first-stroke survivors between 1999 and 2003 were collected. The medical expenditure of index hospitalization was the lowest and the survival period was the longest in the TR group. The ICLYG of TR vs. NR (reference) was -388.5 (95% CI -396.2, -380.8) USD/year and that of TR vs. R (reference) was -121.5 (95% CI -130.4, -112.6) USD/year. The daily medical expenditure of the post-stroke survival period was significantly lower in the TR group (median 11.0, IQR 5.7-22.5 USD) than in the R (median 14.2, IQR 6.4-41.4 USD) and NR (median 19.5, IQR 6.4-88.2 USD) groups.Conclusions: The 10-year post-stroke follow-up showed that transferring patients to the rehabilitation ward is more cost effective than rehabilitation without transfer to the rehabilitation ward and no rehabilitation.

Association between Clinical and Laboratory Markers and 5-year Mortality among Patients with Stroke

Factors influencing long-term stroke mortality have not been comprehensively investigated. This study aimed to identify the baseline clinical, laboratory, demographic/socioeconomic, and hospital factors influencing 5-year mortality in patients with first stroke. Total 3,956 patients with first-stroke hospitalization from 2004 to 2008 were connected to the longitudinal National Health Insurance Research Database. Post-admission baseline data that significantly increased 5-year mortality were red cell distribution width (RDW) >0.145 (adjusted hazard ratio [aHR] = 1.71), hemoglobin <120 g/L (aHR = 1.25), blood sugar <3.89 mmol/L (70 mg/dL)(aHR = 2.57), serum creatinine >112.27 μmol/L (aHR = 1.76), serum sodium <134 mmol/L (aHR = 1.73), body mass index (BMI) < 18.5 kg/m² (aHR = 1.33), Glasgow Coma Scale <15 (aHR = 1.43), Stroke Severity Index ≥20 (aHR = 3.92), Charlson–Deyo Comorbidity Index ≥3 (aHR = 4.21), no rehabilitation (aHR = 1.86), and age ≥65 years (aHR = 2.25). Hemoglobin, RDW, blood sugar, serum creatinine and sodium, BMI, consciousness, stroke severity, comorbidity, rehabilitation, and age were associated with 5-year mortality in patients with first stroke.

Taiwan's National Health Insurance Research Database: past and future

Taiwan's National Health Insurance Research Database (NHIRD) exemplifies a population-level data source for generating real-world evidence to support clinical decisions and health care policy-making. Like with all claims databases, there have been some validity concerns of studies using the NHIRD, such as the accuracy of diagnosis codes and issues around unmeasured confounders. Endeavors to validate diagnosed codes or to develop methodologic approaches to address unmeasured confounders have largely increased the reliability of NHIRD studies. Recently, Taiwan's Ministry of Health and Welfare (MOHW) established a Health and Welfare Data Center (HWDC), a data repository site that centralizes the NHIRD and about 70 other health-related databases for data management and analyses. To strengthen the protection of data privacy, investigators are required to conduct on-site analysis at an HWDC through remote connection to MOHW servers. Although the tight regulation of this on-site analysis has led to inconvenience for analysts and has increased time and costs required for research, the HWDC has created opportunities for enriched dimensions of study by linking across the NHIRD and other databases. In the near future, researchers will have greater opportunity to distill knowledge from the NHIRD linked to hospital-based electronic medical records databases containing unstructured patient-level information by using artificial intelligence techniques, including machine learning and natural language processes. We believe that NHIRD with multiple data sources could represent a powerful research engine with enriched dimensions and could serve as a guiding light for real-world evidence-based medicine in Taiwan.

Association between continuity of care and long-term mortality in Taiwanese first-ever stroke survivors: An 8-year cohort study

Background

Continuity of care is considered to be an important principle of stroke care; however, few analyses of empirically related outcomes have been reported.

Objective

This study examined the correlation between the continuity of care for outpatients after a stroke event and the survival of stroke patients over the year following hospital discharge.

Research design

Data from the Taiwan National Health Insurance Database were used in this study. We defined stroke as the ICD-9-CM codes 430 to 437, and all patients were followed up regarding their survival for at least one year. The modified modified continuity index (MMCI) was used as the indicator of continuity of care. Cox proportional hazard models with robust sandwich variance estimates were employed to analyze the correlation between continuity of care and stroke-related death.

Results

A total of 9,252 stroke patients were included in the analysis. Those patients who had a high and a completed COC had a higher percentage of survival (97.25% and 95.39%) compared to the other two groups. After controlling for other variables, compared with the low-level continuity of care group, the moderate-level, high-level and completed continuity of care groups still showed a significantly lower risk of death HR (95% CI) were: 0.63 (0.49–0.80), 0.56 (0.40–0.79) and 0.50 (0.39–0.63), respectively.

Conclusion

Continuity of care may increase the survival among stroke patients and therefore plays an important role in management of stroke after survival.

Taiwan's National Health Insurance Research Database: past and future

Taiwan’s National Health Insurance Research Database (NHIRD) exemplifies a population-level data source for generating real-world evidence to support clinical decisions and health care policy-making. Like with all claims databases, there have been some validity concerns of studies using the NHIRD, such as the accuracy of diagnosis codes and issues around unmeasured confounders. Endeavors to validate diagnosed codes or to develop methodologic approaches to address unmeasured confounders have largely increased the reliability of NHIRD studies. Recently, Taiwan’s Ministry of Health and Welfare (MOHW) established a Health and Welfare Data Center (HWDC), a data repository site that centralizes the NHIRD and about 70 other health-related databases for data management and analyses. To strengthen the protection of data privacy, investigators are required to conduct on-site analysis at an HWDC through remote connection to MOHW servers. Although the tight regulation of this on-site analysis has led to inconvenience for analysts and has increased time and costs required for research, the HWDC has created opportunities for enriched dimensions of study by linking across the NHIRD and other databases. In the near future, researchers will have greater opportunity to distill knowledge from the NHIRD linked to hospital-based electronic medical records databases containing unstructured patient-level information by using artificial intelligence techniques, including machine learning and natural language processes. We believe that NHIRD with multiple data sources could represent a powerful research engine with enriched dimensions and could serve as a guiding light for real-world evidence-based medicine in Taiwan.

Variation in Hospital-Based Rehabilitation Services Among Patients With Ischemic Stroke in the United States

BACKGROUND

Little is known about variation in use of rehabilitation services provided in acute care hospitals for people who have had a stroke.

OBJECTIVE

The objective was to examine patient and hospital sources of variation in acute care rehabilitation services provided for stroke.

DESIGN

This was a retrospective, cohort design.

METHODS

The sample consisted of Medicare fee-for-service beneficiaries with ischemic stroke admitted to acute care hospitals in 2010. Medicare claims data were linked to the Provider of Services file to gather information on hospital characteristics and the American Community Survey for sociodemographic data. Chi-square tests compared patient and hospital characteristics stratified by any rehabilitation use. We used multilevel, multivariable random effect models to identify patient and hospital characteristics associated with the likelihood of receiving any rehabilitation and with the amount of therapy received in minutes.

RESULTS

Among 104,295 patients, 85.2% received rehabilitation (61.5% both physical therapy and occupational therapy; 22.0% physical therapy only; and 1.7% occupational therapy only). Patients received 123 therapy minutes on average (median [SD] = 90.0 [99.2] minutes) during an average length of stay of 4.8 [3.5] days. In multivariable analyses, male sex, dual enrollment in Medicare and Medicaid, prior hospitalization, ICU stay, and feeding tube were associated with lower odds of receiving any rehabilitation services. These same variables were generally associated with fewer minutes of therapy. Patients treated by tissue plasminogen activator, in limited-teaching and nonteaching hospitals, and in hospitals with inpatient rehabilitation units, were more likely to receive more therapy minutes.

LIMITATION

The findings are limited to patients with ischemic stroke.

CONCLUSION

Only 61% of patients with ischemic stroke received both physical therapy and occupational therapy services in the acute setting. We identified considerable variation in the use of rehabilitation services in the acute care setting following a stroke.

Big data analytics for preventive medicine

Medical data is one of the most rewarding and yet most complicated data to analyze. How can healthcare providers use modern data analytics tools and technologies to analyze and create value from complex data? Data analytics, with its promise to efficiently discover valuable pattern by analyzing large amount of unstructured, heterogeneous, non-standard and incomplete healthcare data. It does not only forecast but also helps in decision making and is increasingly noticed as breakthrough in ongoing advancement with the goal is to improve the quality of patient care and reduces the healthcare cost. The aim of this study is to provide a comprehensive and structured overview of extensive research on the advancement of data analytics methods for disease prevention. This review first introduces disease prevention and its challenges followed by traditional prevention methodologies. We summarize state-of-the-art data analytics algorithms used for classification of disease, clustering (unusually high incidence of a particular disease), anomalies detection (detection of disease) and association as well as their respective advantages, drawbacks and guidelines for selection of specific model followed by discussion on recent development and successful application of disease prevention methods. The article concludes with open research challenges and recommendations.