Learning together for better health using an evidence-based Learning Health System framework: a case study in stroke.

BACKGROUND: In the context of expanding digital health tools, the health system is ready for Learning Health System (LHS) models. These models, with proper governance and stakeholder engagement, enable the integration of digital infrastructure to provide feedback to all relevant parties including clinicians and consumers on performance against best practice standards, as well as fostering innovation and aligning healthcare with patient needs. The LHS literature primarily includes opinion or consensus-based frameworks and lacks validation or evidence of benefit. Our aim was to outline a rigorously codesigned, evidence-based LHS framework and present a national case study of an LHS-aligned national stroke program that has delivered clinical benefit. MAIN TEXT: Current core components of a LHS involve capturing evidence from communities and stakeholders (quadrant 1), integrating evidence from research findings (quadrant 2), leveraging evidence from data and practice (quadrant 3), and generating evidence from implementation (quadrant 4) for iterative system-level improvement. The Australian Stroke program was selected as the case study as it provides an exemplar of how an iterative LHS works in practice at a national level encompassing and integrating evidence from all four LHS quadrants. Using this case study, we demonstrate how to apply evidence-based processes to healthcare improvement and embed real-world research for optimising healthcare improvement. We emphasize the transition from research as an endpoint, to research as an enabler and a solution for impact in healthcare improvement. CONCLUSIONS: The Australian Stroke program has nationally improved stroke care since 2007, showcasing the value of integrated LHS-aligned approaches for tangible impact on outcomes. This LHS case study is a practical example for other health conditions and settings to follow suit.

Cost-Effectiveness of a Government Policy to Incentivise Chronic Disease Management following Stroke: A Modelling Study.

INTRODUCTION: Little is known about the cost-effectiveness of government policies that support primary care physicians to provide comprehensive chronic disease management (CDM). This paper aimed to estimate the potential cost-effectiveness of CDM policies over a lifetime for long-time survivors of stroke. METHODS: A Markov model, using three health states (stable, hospitalised, dead), was developed to simulate the costs and benefits of CDM policies over 30 years (with 1-year cycles). Transition probabilities and costs from a health system perspective were obtained from the linkage of data between the Australian Stroke Clinical Registry (cohort n = 12,368, 42% female, median age 70 years, 45% had CDM claims) and government-held hospital, Medicare, and pharmaceutical claims datasets. Quality-adjusted life years (QALYs) were obtained from a comparable cohort (n = 512, 34% female, median age 69.6 years, 52% had CDM claims) linked with Medicare claims and death data. A 3% discount rate was applied to costs in Australian dollars (AUD, 2016) and QALYs beyond 12 months. Probabilistic sensitivity analyses were used to understand uncertainty. RESULTS: Per-person average total lifetime costs were AUD 142,939 and 8.97 QALYs for those with a claim, and AUD 103,889 and 8.98 QALYs for those without a claim. This indicates that these CDM policies were costlier without improving QALYs. The probability of cost-effectiveness of CDM policies was 26.1%, at a willingness-to-pay threshold of AUD 50,000/QALY. CONCLUSION: CDM policies, designed to encourage comprehensive care, are unlikely to be cost-effective for stroke compared to care without CDM. Further research to understand how to deliver such care cost-effectively is needed.

Risk Factors Associated with Major Adverse Cardiovascular Events after Ischemic Stroke: A Linked Registry Study.

INTRODUCTION: Survivors of stroke are at risk of experiencing subsequent major adverse cardiovascular events (MACE). We aimed to determine the incidence of, and risk factors for, MACE after first-ever ischemic stroke, by age group (18-64 years vs. ≥65 years). METHODS: Observational cohort study using patient-level data from the Australian Stroke Clinical Registry (2009-2013), linked with hospital administrative data. We included adults with first-ever ischemic stroke who had no previous acute cardiovascular admissions and followed these patients for 2 years post-discharge, or until the first post-stroke MACE event. A Fine-Gray sub-distribution hazard model, accounting for the competing risk of non-cardiovascular death, was used to determine factors for incident post-stroke MACE. RESULTS: Among 5,994 patients with a first-ever ischemic stroke (median age 73 years, 45% female), 17% were admitted for MACE within 2 years (129 events per 1,000 person-years). The median time to first post-stroke MACE was 117 days (89 days if aged <65 years vs. 126 days if aged ≥65 years; p = 0.025). Among patients aged 18-64 years, receiving intravenous thrombolysis (sub-distribution hazard ratio [SHR] 0.51 [95% CI, 0.28-0.92]) or being discharged to inpatient rehabilitation (SHR 0.65 [95% CI, 0.46-0.92]) were associated with a reduced incidence of post-stroke MACE. In those aged ≥65 years, being unable to walk on admission (SHR 1.33 [95% CI 1.15-1.54]), and history of smoking (SHR 1.40 [95% CI 1.14-1.71]) or atrial fibrillation (SHR 1.31 [95% CI 1.14-1.51]) were associated with an increased incidence of post-stroke MACE. Acute management in a large hospital (>300 beds) for the initial stroke event was associated with reduced incidence of post-stroke MACE, irrespective of age group. CONCLUSIONS: MACE is common within 2 years of stroke, with most events occurring within the first year. We have identified important factors to consider when designing interventions to prevent MACE after stroke, particularly among those aged <65 years.

Age and sex disparities in cardiovascular risk factor management prior to stroke: Linked registry and general practice data.

INTRODUCTION: There is limited evidence about the management of cardiovascular risk factors within 12 months before stroke/transient ischaemic attack (TIA) in Australian general practices. We evaluated whether age and sex disparities in cardiovascular risk factor management for primary prevention exist in general practice. METHODS: A retrospective cohort study using data from the Australian Stroke Clinical Registry (2014-2018) linked with general practice data from three primary health networks in Victoria, Australia. We included adults who had ≥2 encounters with a general practitioner within 12 months immediately before the first stroke/TIA. Cardiovascular risk factor management within 12 months before stroke/TIA was evaluated in terms of: assessment of risk factors (blood pressure [BP], serum lipids, blood glucose, body weight); prescription of prevention medications (BP, lipid-, glucose-lowering, antithrombotic agents); and attainment of risk factor targets. RESULTS: Of 2,880 patients included (median age 76.5 years, 48.4% women), 80.9% were assessed for BP, 49.9% serum lipids, 46.8% blood glucose, and 39.3% body weight. Compared to patients aged 65-84 years, those aged <65 or ≥85 years were less often assessed for risk factors, with women aged ≥85 years assessed for significantly fewer risk factors than their male counterparts. The most prescribed prevention medications were BP-lowering (64.9%) and lipid-lowering agents (42.0%). There were significant sex differences among those aged <65 years (34.7% women vs. 40.2% men) and ≥85 years (34.0% women vs. 44.3% men) for lipid-lowering agents. Risk factor target attainment was generally poorer in men than women, especially among those aged <65 years. CONCLUSION: Age-sex disparity exists in risk factor management for primary prevention in general practice, and this was more pronounced among younger patients and older women.

In Search of a Gold Standard Tool for Assessing Knowledge of Stroke: A Systematic Review.

BACKGROUND: Knowledge of stroke is essential to empower people to reduce their risk of these events. However, valid tools are required for accurate and reliable measurement of stroke knowledge. We aimed to systematically review contemporary stroke knowledge assessment tools and appraise their content validity, feasibility, and measurement properties. METHODS: The protocol was registered in PROSPERO (CRD42023403566). Electronic databases (MEDLINE, PsycInfo, CINAHL, Embase, Scopus, Web of Science) were searched to identify published articles (1 January 2015-1 March 2023), in which stroke knowledge was assessed using a validated tool. Two reviewers independently screened titles and abstracts prior to undertaking full-text review. COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) methods guided the appraisal of content validity (relevance, comprehensiveness, comprehensibility), feasibility, and measurement properties. RESULTS: After removing duplicates, the titles and abstracts of 718 articles were screened; 323 reviewed in full; with 42 included (N = 23 unique stroke knowledge tools). For content validity, all tools were relevant, two were comprehensive, and seven were comprehensible. Validation metrics were reported for internal consistency (n = 20 tools), construct validity (n = 17 tools), cross-cultural validity (n = 15 tools), responsiveness (n = 9 tools), reliability (n = 7 tools), structural validity (n = 3 tools), and measurement error (n = 1 tool). The Stroke Knowledge Test met all content validity criteria, with validation data for six measurement properties (n = 3 rated "Sufficient"). CONCLUSION: Assessment of stroke knowledge is not standardised and many tools lacked validated content or measurement properties. The Stroke Knowledge Test was the most comprehensive but requires updating and further validation for endorsement as a gold standard.

Access to inpatient mood management services after stroke in Australian acute and rehabilitation hospitals.

OBJECTIVE: Post-stroke mental health impairments are common, but under-assessed and under-treated. We aim to describe trends in the provision of mood management to patients with stroke, and describe factors associated with adoption of national mood management recommendations for stroke within Australian hospitals. DESIGN: Secondary analysis of cross-sectional data from the biennial Stroke Foundation Audit Program. SETTING: Participating acute (2011-2021) and rehabilitation hospitals (2012-2020) in Australia. PARTICIPANTS: In the acute audit, 22,937 stroke cases were included from 133 hospitals. In the rehabilitation audit, 15,891 stroke cases were included from 127 hospitals. MAIN MEASURES: Hospital- and patient-level mood management processes. RESULTS: Among 133 acute hospitals (22,937 stroke episodes), improvements were made between 2011 and 2021 in utilization of mood screening (17% [2011], 33% [2021]; p < 0.001) and access to psychologists during hospital stay (18% [2011], 45% [2021]; p < 0.001). There was no change in access to a psychologist among those with a mood impairment (p = 0.34). Among 127 rehabilitation hospitals (15,891 stroke episodes) improvements were observed for mood screening (35% [2012], 61% [2020]; p < 0.001), and access to a psychologist during hospital stay (38% [2012], 68% [2020]; p < 0.001) and among those with a mood-impairment (30% [2012], 50% [2020]; p < 0.001). Factors associated with receiving mood management processes included: younger age, not requiring an interpreter and longer length of stay. CONCLUSIONS: Adherence to mood management recommendations has improved over 10 years within Australian hospitals. Those aged over 65, requiring an interpreter, or with shorter hospital stays are at risk of missing out on appropriate mood management.

Stroke Learning Health Systems: A Topical Narrative Review With Case Examples.

To our knowledge, the adoption of Learning Health System (LHS) concepts or approaches for improving stroke care, patient outcomes, and value have not previously been summarized. This topical review provides a summary of the published evidence about LHSs applied to stroke, and case examples applied to different aspects of stroke care from high and low-to-middle income countries. Our attempt to systematically identify the relevant literature and obtain real-world examples demonstrated the dissemination gaps, the lack of learning and action for many of the related LHS concepts across the continuum of care but also elucidated the opportunity for continued dialogue on how to study and scale LHS advances. In the field of stroke, we found only a few published examples of LHSs and health systems globally implementing some selected LHS concepts, but the term is not common. A major barrier to identifying relevant LHS examples in stroke may be the lack of an agreed taxonomy or terminology for classification. We acknowledge that health service delivery settings that leverage many of the LHS concepts do so operationally and the lessons learned are not shared in peer-reviewed literature. It is likely that this topical review will further stimulate the stroke community to disseminate related activities and use keywords such as learning health system so that the evidence base can be more readily identified.

Fracture Risk Increases After Stroke or Transient Ischemic Attack and Is Associated With Reduced Quality of Life.

BACKGROUND: Fractures are a serious consequence following stroke, but it is unclear how these events influence health-related quality of life (HRQoL). We aimed to compare annualized rates of fractures before and after stroke or transient ischemic attack (TIA), identify associated factors, and examine the relationship with HRQoL after stroke/TIA. METHODS: Retrospective cohort study using data from the Australian Stroke Clinical Registry (2009-2013) linked with hospital administrative and mortality data. Rates of fractures were assessed in the 1-year period before and after stroke/TIA. Negative binomial regression, with censoring at death, was used to identify factors associated with fractures after stroke/TIA. Respondents provided HRQoL data once between 90 and 180 days after stroke/TIA using the EuroQoL 5-dimensional 3-level instrument. Adjusted logistic regression was used to assess differences in HRQoL at 90 to 180 days by previous fracture. RESULTS: Among 13 594 adult survivors of stroke/TIA (49.7% aged ≥75 years, 45.5% female, 47.9% unable to walk on admission), 618 fractures occurred in the year before stroke/TIA (45 fractures per 1000 person-years) compared with 888 fractures in the year after stroke/TIA (74 fractures per 1000 person-years). This represented a relative increase of 63% (95% CI, 47%-80%). Factors associated with poststroke fractures included being female (incidence rate ratio [IRR], 1.34 [95% CI, 1.05-1.72]), increased age (per 10-year increase, IRR, 1.35 [95% CI, 1.21-1.50]), history of prior fracture(s; IRR, 2.56 [95% CI, 1.77-3.70]), and higher Charlson Comorbidity Scores (per 1-point increase, IRR, 1.18 [95% CI, 1.10-1.27]). Receipt of stroke unit care was associated with fewer poststroke fractures (IRR, 0.67 [95% CI, 0.49-0.93]). HRQoL at 90 to 180 days was worse among patients with prior fracture across the domains of mobility, self-care, usual activities, and pain/discomfort. CONCLUSIONS: Fracture risk increases substantially after stroke/TIA, and a history of these events is associated with poorer HRQoL at 90 to 180 days after stroke/TIA.

Receipt of Mental Health Treatment in People Living With Stroke: Associated Factors and Long-Term Outcomes.

BACKGROUND: Untreated poststroke mood problems may influence long-term outcomes. We aimed to investigate factors associated with receiving mental health treatment following stroke and impacts on long-term outcomes. METHODS: Observational cohort study derived from the Australian Stroke Clinical Registry (AuSCR; Queensland and Victorian registrants: 2012-2016) linked with hospital, primary care billing and pharmaceutical dispensing claims data. Data from registrants who completed the AuSCR 3 to 6 month follow-up survey containing a question on anxiety/depression were analyzed. We assessed exposures at 6 to 18 months and outcomes at 18 to 30 months. Factors associated with receiving treatment were determined using staged multivariable multilevel logistic regression models. Cox proportional hazards regression models were used to assess the impact of treatment on outcomes. RESULTS: Among 7214 eligible individuals, 39% reported anxiety/depression at 3 to 6 months following stroke. Of these, 54% received treatment (88% antidepressant medication). Notable factors associated with any mental health treatment receipt included prestroke psychological support (odds ratio [OR], 1.80 [95% CI, 1.37-2.38]) or medication (OR, 17.58 [95% CI, 15.05-20.55]), self-reported anxiety/depression (OR, 2.55 [95% CI, 2.24-2.90]), younger age (OR, 0.98 [95% CI, 0.97-0.98]), and being female (OR, 1.30 [95% CI, 1.13-1.48]). Those who required interpreter services (OR, 0.49 [95% CI, 0.25-0.95]) used a health benefits card (OR, 0.73 [95% CI, 0.59-0.92]) or had continuity of primary care visits (ie, with a consistent physician; OR, 0.78 [95% CI, 0.62-0.99]) were less likely to access mental health services. Among those who reported anxiety/depression, those who received mental health treatment had an increased risk of presenting to hospital (hazard ratio, 1.06 [95% CI, 1.01-1.11]) but no difference in survival (hazard ratio, 0.86 [95% CI, 0.58-1.27]). CONCLUSIONS: Nearly half of the people living with mood problems following stroke did not receive mental health treatment. We have highlighted subgroups who may benefit from targeted mood screening and factors that may improve treatment access.

Factors Associated With Transition From Community to Permanent Residential Aged Care Following Stroke: A Linked Registry Data Study.

BACKGROUND: Understanding factors that influence the transition to permanent residential aged care following a stroke or transient ischemic attack may inform strategies to support people to live at home longer. We aimed to identify the demographic, clinical, and system factors that may influence the transition from living in the community to permanent residential care in the 6 to 18 months following stroke/transient ischemic attack. METHODS: Linked data cohort analysis of adults from Queensland and Victoria aged ≥65 years and registered in the Australian Stroke Clinical Registry (2012-2016) with a clinical diagnosis of stroke/transient ischemic attack and living in the community in the first 6 months post-hospital discharge. Participant data were linked with primary care, pharmaceutical, aged care, death, and hospital data. Multivariable survival analysis was performed to determine demographic, clinical, and system factors associated with the transition to permanent residential care in the 6 to 18 months following stroke, with death modeled as a competing risk. RESULTS: Of 11 176 included registrants (median age, 77.2 years; 44% female), 520 (5%) transitioned to permanent residential care between 6 and 18 months. Factors most associated with transition included the history of urinary tract infections (subhazard ratio [SHR], 1.41 [95% CI, 1.16-1.71]), dementia (SHR, 1.66 [95% CI, 1.14-2.42]), increasing age (65-74 versus 85+ years; SHR, 1.75 [95% CI, 1.31-2.34]), living in regional Australia (SHR, 31 [95% CI, 1.08-1.60]), and aged care service approvals: respite (SHR, 4.54 [95% CI, 3.51-5.85]) and high-level home support (SHR, 1.80 [95% CI, 1.30-2.48]). Protective factors included being dispensed antihypertensive medications (SHR, 0.68 [95% CI, 0.53-0.87]), seeing a cardiologist (SHR, 0.72 [95% CI, 0.57-0.91]) following stroke, and less severe stroke (SHR, 0.71 [95% CI, 0.58-0.88]). CONCLUSIONS: Our findings provide an improved understanding of factors that influence the transition from community to permanent residential care following stroke and can inform future strategies designed to delay this transition.

The Excess Costs of Hospitalization for Acute Stroke in People With Communication Impairment: A Stroke123 Data Linkage Substudy.

OBJECTIVE: To describe the costs of hospital care for acute stroke for patients with aphasia or dysarthria. DESIGN: Observational study from the Stroke123 project. SETTING: Data from patients admitted with stroke (2009-2013) from 22 hospitals in Queensland participating in the Australian Stroke Clinical Registry (AuSCR) were linked to administrative datasets. PARTICIPANTS: Communication impairments were identified using International Classification of Diseases, 10th Revision, Australian Modification codes. Overall, 1043 of 4195 (25%) patients were identified with aphasia (49% were women; median age 78 years; 83% with ischemic stroke), and 1005 (24%) with dysarthria (42% were women; median age 76 years; 85% with ischemic stroke). INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Linked patient-level, hospital clinical costing related to the stroke, were adjusted to 2013/2014 Australian dollars (AU$, US$ conversion x 0.691) using recommended national price indices and multivariable regression analysis with clustering by hospital performed. RESULTS: Compared with patients without aphasia, the median hospital costs/patient were greater for those with aphasia for medical (aphasia AU$2273 vs AU$1727, P<.001), nursing (aphasia AU$3829 vs AU$2748, P<.001) and allied health services (aphasia AU$1138 vs AU$720, P<.001). Similarly, costs were greater for patients with dysarthria compared with those without dysarthria. Adjusted median total costs were AU$2882 greater for patients with aphasia compared with patients without aphasia (95% confidence interval, AU$1880-3884), and AU$843 greater for patients with dysarthria compared with those without dysarthria (95% confidence interval, AU$-301 to 1987). CONCLUSIONS: People with communication impairment after stroke incur greater hospital costs, in particular for medical, allied health, and nursing resources.

Benefit of linking hospital resource information and patient-level stroke registry data.

Variation in the delivery of evidence-based care affects outcomes for patients with stroke. A range of hospital (organizational), patient, and clinical factors can affect care delivery. Clinical registries are widely used to monitor stroke care and guide quality improvement efforts within hospitals. However, hospital features are rarely collected. We aimed to explore the influence of hospital resources for stroke, in metropolitan and regional/rural hospitals, on the provision of evidence-based patient care and outcomes. The 2017 National Audit organizational survey (Australia) was linked to patient-level data from the Australian Stroke Clinical Registry (2016-2017 admissions). Regression models were used to assess the associations between hospital resources (based on the 2015 Australian National Acute Stroke Services Framework) and patient care (reflective of national guideline recommendations), as well as 90-180-day readmissions and health-related quality of life. Models were adjusted for patient factors, including the severity of stroke. Fifty-two out of 127 hospitals with organizational survey data were merged with 22 832 Australian Stroke Clinical Registry patients with an admission for a first-ever stroke or transient ischaemic attack (median age 75 years, 55% male, and 66% ischaemic). In metropolitan hospitals (n = 42, 20 977 patients, 1701 thrombolyzed, and 2395 readmitted between 90 and 180 days post stroke), a faster median door-to-needle time for thrombolysis was associated with ≥500 annual stroke admissions [-15.9 minutes, 95% confidence interval (CI) -27.2, -4.7], annual thrombolysis >20 patients (-20.2 minutes, 95% CI -32.0, -8.3), and having specialist stroke staff (dedicated medical lead and stroke coordinator; -12.7 minutes, 95% CI -25.0, -0.4). A reduced likelihood of all-cause readmissions between 90 and 180 days was evident in metropolitan hospitals using care pathways for stroke management (odds ratio 0.82, 95% CI 0.67-0.99). In regional/rural hospitals (n = 10, 1855 patients), being discharged with a care plan was also associated with the use of stroke clinical pathways (odds ratio 3.58, 95% CI 1.45-8.82). No specific hospital resources influenced 90-180-day health-related quality of life. Relevant to all international registries, integrating information about hospital resources with clinical registry data provides greater insights into factors that influence evidence-based care.

Temporal and age-specific trends in incidence and 1-year case-fatality of hospitalized ischaemic stroke in Victoria, Australia.

OBJECTIVES: Investigate temporal and age-specific trends in the incidence of ischaemic stroke and case-fatality risk in Victoria, Australia. MATERIALS AND METHODS: Patients hospitalised with first ischaemic stroke between 2012 and 2018 were included. Trends in age-standardised incidence rates of ischaemic stroke were assessed using linear regression models. Cox proportional hazard regression models were used to examine the case-fatality risk. RESULTS: Overall age-standardised incidence of ischaemic stroke was stable from 2012/13 to 2017/18 (87.6 to 84.8 events per 100,000 population; Annual percentage change [APC] -0.32; 95% Confidence interval [CI] -1.13 to 0.50). The incidence declined in females (APC -1.00; 95% CI -1.49 to -0.50), people aged 75-84 years (APC -1.60; 95% CI -2.83 to -0.36) and in metropolitan areas (APC -0.74; 95% CI -1.02 to -0.45). The risk of 1-year case-fatality (HR 0.85; 95% CI 0.78 to 0.93) significantly declined in 2016/17 compared to 2012/13. CONCLUSIONS: Overall ischaemic stroke incidence remained stable while decreasing trends were observed in females, elderly and metropolitan areas. 1-year case-fatality declined from 2012 to 2017.

Measuring Stroke Quality: Methodological Considerations in Selecting, Defining, and Analyzing Quality Measures.

Knowledge about stroke and its management is growing rapidly and stroke systems of care must adapt to deliver evidence-based care. Quality improvement initiatives are essential for translating knowledge from clinical trials and recommendations in guidelines into routine clinical practice. This review focuses on issues central to the measurement of the quality of stroke care, including selection and definition of quality measures, identification of the eligible patient cohorts, optimization of data quality, and considerations for data analysis and interpretation.

Does a History of Cancer Influence the Effectiveness of Statins on Outcomes After Stroke?

BACKGROUND: Evidence is growing on anticancer effects of statins. We investigated whether the effectiveness of treatment with statins after ischemic stroke on mortality is influenced by a history of cancer. METHODS: Analyses of 90-day survivors of ischemic stroke (2012-2016; 45 hospitals) using linked registry and administrative data. Dispense of statins within 90 days postdischarge was determined from pharmaceutical records. Participants were followed from 91 days postdischarge until death or June 30, 2018. History of cancer was determined from hospital data. Propensity score-adjusted Cox proportional hazards regression model was used to determine the association between being dispensed statins and survival. The influence of history of cancer on this association was assessed based on the concepts of (1) statistical interaction and (2) biological interaction using 3 indices: relative excess risk due to interaction>0, attributable proportion due to interaction >0, or synergy index >1. RESULTS: Among 9948 eligible participants (median age=72 years, 42% female), there were 1463 deaths. In adjusted analyses, there was no statistical interaction between being dispensed statins and history of cancer on mortality (P=0.156). However, being dispensed statins had a significant positive biological interaction with having a history of cancer on mortality: relative excess risk due to interaction, 2.80 (95% CI, 1.56-5.05), attributable proportion due to interaction, 0.45 (95% CI, 0.23-0.66), and synergy index, 2.14 (95% CI, 1.32-3.49). CONCLUSIONS: Treatment with statins after ischemic stroke may confer additional survival benefits for people who also have had cancer.

Optimal Measures for Primary Care Physician Encounters after Stroke and Association with Survival: A Data Linkage Study.

BACKGROUND AND PURPOSE: Primary care physicians (PCPs) provide ongoing management after stroke. However, little is known about how best to measure physician encounters with reference to longer term outcomes. We aimed to compare methods for measuring regularity and continuity of PCP encounters, based on survival following stroke using linked healthcare data. METHODS: Data from the Australian Stroke Clinical Registry (2010-2014) were linked with Australian Medicare claims from 2009 to 2016. Physician encounters were ascertained within 18 months of discharge for stroke. We calculated three separate measures of continuity of encounters (consistency of visits with primary physician) and three for regularity of encounters (distribution of service utilization over time). Indices were compared based on 1-year survival using multivariable Cox regression models. The best performing measures of regularity and continuity, based on model fit, were combined into a composite "optimal care" variable. RESULTS: Among 10,728 registrants (43% female, 69% aged ≥65 years), the median number of encounters was 17. The measures most associated with survival (hazard ratio [95% confidence interval], Akaike information criterion [AIC], and Bayesian information criterion [BIC]) were the Continuity of Care Index (COCI, as a measure of continuity; 0.88 [0.76-1.02], p = 0.099, AIC = 13,746, BIC = 13,855) and our persistence measure of regularity (encounter at least every 6 months; 0.80 [0.67-0.95], p = 0.011, AIC = 13,742, BIC = 13,852). Our composite measure, persistent plus COCI ≥80% (24% of registrants; 0.80 [0.68-0.94], p = 0.008, AIC = 13,742, BIC = 13,851), performed marginally better than our persistence measure alone. CONCLUSIONS: Our persistence measure of regularity or composite measure may be useful when measuring physician encounters following stroke.

Real-World Effectiveness of Lipid-Lowering Medications on Outcomes after Stroke: Potential Implications of the New-User Design.

INTRODUCTION: Observational studies are increasingly being used to provide evidence on the real-world effectiveness of medications for preventing vascular diseases, such as stroke. We investigated whether the real-world effectiveness of treatment with lipid-lowering medications after ischemic stroke is affected by prevalent-user bias. METHODS: An observational cohort study of 90-day survivors of ischemic stroke using person-level data from the Australian Stroke Clinical Registry (2012-2016; 45 hospitals) linked to administrative (pharmaceutical, hospital, death) records. The use of, and adherence to (proportion of days covered <80% [poor adherence] vs. ≥80% [good adherence]), lipid-lowering medications within 90 days post-discharge was determined from pharmaceutical records. Users were further classified as prevalent (continuing) or new users, based on dispensing within 90 days prior to stroke. A propensity score-adjusted Cox regression was used to evaluate the effectiveness of lipid-lowering medications on outcomes (all-cause mortality, all-cause and cardiovascular disease readmission) within the subsequent year. Analyses were undertaken using prevalent-user (all users vs. nonusers) and new-user designs (new users vs. nonusers). RESULTS: Of 11,217 eligible patients (median age 72 years, 42% female), 9,294 (83%) used lipid-lowering medications within 90 days post-discharge, including 5,479 new users. In both prevalent-user and new-user designs, nonusers (vs. users) had significantly greater rates of mortality (hazard ratio [HR] 2.35, 95% CI: 1.89-2.92) or all-cause readmissions (HR 1.22, 95% CI: 1.05-1.40) but not cardiovascular disease readmission. In contrast, associations between having poor (vs. good) adherence on outcomes were stronger among new users than all users. Among new users, having poor adherence was associated with greater rates of mortality (HR 1.48, 95% CI: 1.12-1.96), all-cause readmission (HR 1.14, 95% CI: 1.02-1.27), and cardiovascular disease readmission (HR 1.20, 95% CI: 1.01-1.42). CONCLUSIONS: The real-world effectiveness of treatment with lipid-lowering medications after stroke is attenuated when evaluated based on prevalent-user rather than new-user design. These findings may have implications for designing studies on the real-world effectiveness of secondary prevention medications.

Treatment with Multiple Therapeutic Classes of Medication Is Associated with Survival after Stroke.

INTRODUCTION: Treatment with several therapeutic classes of medication is recommended for secondary prevention of stroke. We analyzed the associations between the number of classes of prevention medications supplied within 90 days after discharge for ischemic stroke (IS)/transient ischemic attack (TIA) and survival. METHODS: This is a retrospective cohort study of adults with first-ever IS/TIA (2010-2014) from the Australian Stroke Clinical Registry individually linked with data from national pharmaceutical and Medicare claims. Exposure was the number of classes of recommended medications, i.e., blood pressure-lowering, antithrombotic, or lipid-lowering agents, supplied to patients within 90 days after discharge for IS/TIA. The longitudinal association between the number of classes of medications and survival was evaluated with Cox proportional hazards regression models using the landmark approach. A landmark date of 90 days after hospital discharge was used to separate exposure and outcome periods, and only patients who survived until this date were included. RESULTS: Of 8,429 patients (43% female, median age 74 years, 80% IS), 607 (7%) died in the year following 90 days after discharge. Overall, 56% of patients were supplied all 3 classes of medications, 28% 2 classes of medications, 11% 1 class of medications, and 5% no class of medications. Compared to patients supplied all 3 medication classes, adjusted hazard ratios for all-cause mortality ranged from 1.43 (95% confidence interval [CI]: 1.18-1.72) in those supplied 2 medication classes to 2.04 (95% CI: 1.44-2.88) in those supplied with no medication class. DISCUSSION/CONCLUSION: Treatment with all 3 classes of guideline-recommended medications within 90 days after discharge was associated with better survival. Ongoing efforts are required to ensure optimal pharmacological intervention for secondary prevention of stroke.

Towards better reporting of the proportion of days covered method in cardiovascular medication adherence: A scoping review and new tool TEN-SPIDERS.

Although medication adherence is commonly measured in electronic datasets using the proportion of days covered (PDC), no standardized approach is used to calculate and report this measure. We conducted a scoping review to understand the approaches taken to calculate and report the PDC for cardiovascular medicines to develop improved guidance for researchers using this measure. After prespecifying methods in a registered protocol, we searched Ovid Medline, Embase, Scopus, CINAHL Plus and grey literature (1 July 2012 to 14 December 2020) for articles containing the terms "proportion of days covered" and "cardiovascular medicine", or synonyms and subject headings. Of the 523 articles identified, 316 were reviewed in full and 76 were included (93% observational studies; 47% from the USA; 2 grey literature articles). In 45 articles (59%), the PDC was measured from the first dispensing/claim date. Good adherence was defined as 80% PDC in 61 articles, 56% of which contained a rationale for selecting this threshold. The following parameters, important for deriving the PDC, were often not reported/unclear: switching (53%), early refills (45%), in-hospital supplies (45%), presupply (28%) and survival (7%). Of the 46 articles where dosing information was unavailable, 59% reported how doses were imputed. To improve the transparent and systematic reporting of the PDC, we propose the TEN-SPIDERS tool, covering the following PDC parameters: Threshold, Eligibility criteria, Numerator and denominator, Survival, Presupply, In-hospital supplies, Dosing, Early Refills, and Switching. Use of this tool will standardize reporting of the PDC to facilitate reliable comparisons of medication adherence estimates between studies.

The Allure of Big Data to Improve Stroke Outcomes: Review of Current Literature.

PURPOSE OF REVIEW: To critically appraise literature on recent advances and methods using "big data" to evaluate stroke outcomes and associated factors. RECENT FINDINGS: Recent big data studies provided new evidence on the incidence of stroke outcomes, and important emerging predictors of these outcomes. Main highlights included the identification of COVID-19 infection and exposure to a low-dose particulate matter as emerging predictors of mortality post-stroke. Demographic (age, sex) and geographical (rural vs. urban) disparities in outcomes were also identified. There was a surge in methodological (e.g., machine learning and validation) studies aimed at maximizing the efficiency of big data for improving the prediction of stroke outcomes. However, considerable delays remain between data generation and publication. Big data are driving rapid innovations in research of stroke outcomes, generating novel evidence for bridging practice gaps. Opportunity exists to harness big data to drive real-time improvements in stroke outcomes.