Swedish Stroke Register Research Articles

Validation of Hospital Performance Measures of Acute Stroke Care Quality. Riksstroke, the Swedish Stroke Register

Background: Registers are increasingly used to monitor stroke care performance. Fair benchmarking requires sufficient data quality. We have validated acute care data in Riksstroke, the Swedish Stroke Register. Methods: Completeness was assessed by comparisons with diagnoses at hospital discharge recorded in the compulsory National Patient Register and content validity by comparisons with (a) key variables identified by European stroke experts, and (b) items recorded in other European stroke care performance registers. Five test cases recorded by 67 hospitals were used to estimate inter-hospital reliability. Results: All 72 Swedish hospitals admitting acute stroke patients participated in Riksstroke. The register was estimated to cover at least 90% of acute stroke patients. It includes 18 of 22 quality indicators identified by international stroke experts and 14 of 15 indicators used by at least 2 stroke performance registers in other European countries. Inter-hospital reliability was high (≥85%) in 77 of 81 Riksstroke items. Conclusions: A nationwide stroke care register can be maintained with sufficient data quality to permit between-hospital performance benchmarking. Our experiences may serve as a model for other stroke registers while evaluating data quality.

Abstract WP356: Eligibility Criteria in Clinical Trials on Intracerebral Hemorrhage Applied to an Unselected Cohort: The Lund Stroke Register

Introduction: Strict patient selection in medical or surgical trials on intracerebral hemorrhage (ICH) is needed to optimize therapeutic benefit but limits trial enrolment as well as overall applicability of results. We studied the applicability of previous, current, and planned large interventional ICH trials by applying each trial’s defined inclusion criteria to an unselected cohort of ICH patients. Methods: Large interventional ICH trials were identified via trial registration databases. To estimate eligibility rates, each trial’s inclusion criteria were applied on an unselected consecutive group of first-ever ICH patients from the prospective hospital-based Lund Stroke Register. Subsequently, 30 day survival status was obtained from the National Census Office and 90 day poor functional outcome (modified Rankin Scale ≥4 or death) from the Swedish Stroke Register or medical files. Results: Among 253 included ICH patients, estimated eligibility rates ranged from 2-38% for the identified 11 large interventional ICH trials (Figure 1). Patients not eligible for any of the trials (N=91, 36%) had: more extensive intraventricular hemorrhage (p<0.001); lower baseline level of consciousness (p<0.001); higher rate of cerebellar ICH and lower rates of lobar ICH (p<0.001). No significant age, sex, or ICH volume differences were observed. The 30 day mortality rates among eligible patients were 0-33% depending on selected trial. The mortality rate for patients not eligible for any trial was 55% vs 19% for patients eligible in ≥1 trial (95% CI: 45-65% vs 13-25%; p<0.001). Non-eligible ICH patients more frequently had poor functional outcome (75% vs 49%; 95% CI: 65-85% vs 41-57%; p<0.001). Conclusions: There is great variation in proportions of unselected ICH patients eligible for inclusion in treatment trials. Even in trials with broad entry criteria only a minority is eligible, which need to be considered when translating ICH-trial results into clinical practice.

Perceived Unmet Rehabilitation Needs 1 Year After Stroke: An Observational Study From the Swedish Stroke Register.

Met care demands are key aspects in poststroke quality of care. This study aimed to identify baseline predictors and 12-month factors that were associated with perceived unmet rehabilitation needs 1 year poststroke. Data on patients who were independent in activities of daily living, hospitalized for acute stroke during 2008 to 2010, and followed up 1 year poststroke through a postal questionnaire were obtained from the Swedish stroke register. Patients reporting fulfilled rehabilitation needs were compared with those with unmet needs (Chi square test). The study included 37 383 patients, 46% female. At 12 months, 8019 (21.5%) patients reported unmet rehabilitation needs. Compared with those with met rehabilitation needs, patients reporting unmet rehabilitation needs were older (75.4 versus 72.4 years; P<0.0001); a higher proportion was activities of daily living-dependent (59% versus 31.9%; P<0.0001) and institutionalized (24.3% versus 11.5%; P<0.0001) at 12 months. Poststroke depression (32.3% versus 24.9%; P<0.0001) and insufficient pain medication were more common in patients with unmet needs (54.5% versus 32.3%; P<0.0001). Baseline predictors of unmet rehabilitation needs at 12 months in an age-adjusted model were severe stroke (odds ratio [OR]=3.04; confidence interval [CI]: 2.39-3.87), prior stroke (OR=1.63; CI: 1.53-1.75), female sex (OR=1.14; CI: 1.07-1.20), diabetes mellitus (OR=1.24; CI: 1.15-1.32), stroke other than ischemic (OR=1.26; CI: 1.20-1.32), and atrial fibrillation (OR=1.19; CI: 1.12-1.27). Unfulfilled rehabilitation needs 1 year poststroke are common and associated with high age, dependency, pain, and depression. Long-term follow-up systems should, therefore, be comprehensive and address multiple domains of poststroke problems, rather than having a single-domain focus.

Intraventricular Extension of Supratentorial Intracerebral Hemorrhage: The Modified Graeb Scale Improves Outcome Prediction in Lund Stroke Register

Background/Aims: The modified Graeb Scale (mGS) is a semi-quantitative method to assess the extension of intraventricular hemorrhage (IVH) in patients with intracerebral hemorrhage (ICH). The mGS has been shown to prognosticate outcome after ICH in cohorts derived from convenience samples. We evaluated the external validity of mGS in supratentorial ICH-patients from an unselected cohort. Methods: ICH-patients were included prospectively and consecutively in Lund Stroke Register. Follow-up survival status was obtained from the National Census Office; functional outcome was obtained from the Swedish Stroke Register or medical records. Using multivariate analyses, we examined if mGS was related to 30-day survival or poor functional outcome (modified Rankin Scale ≥4) at 90 days. Results: Of 198 supratentorial ICH-patients, 86 (43%) had IVH (median mGS 12, range 1-28). In multivariate regression analyses, the mGS independently predicted 30-day mortality (per point; OR 1.16; 95% CI 1.06-1.27; p = 0.002) and poor functional outcome (OR 1.11; 95% CI 1.02-1.20; p = 0.011) after ICH. In receiver-operator characteristic analysis, the addition of mGS tended to be associated with a higher prognostic accuracy for survival (area under curve 0.886 vs. not including mGS 0.812; p = 0.053). Conclusions: The mGS improves outcome prediction after supratentorial ICH beyond other previously established factors in an unselected population.

Validation of Diagnoses of Transient Ischemic Attack in the Swedish Stroke Register (Riksstroke) TIA-Module

Background: In 2010, the Swedish Stroke Register (Riksstroke; RS) established a module for transient ischemic attacks (RS-TIA). We report a diagnostic validation study of patients included in RS-TIA. Methods: During the first year, 7,825 patients were registered at 59 out of 74 Swedish hospitals. A time-based TIA definition was applied. A sample of 180 patients (30 patients each from 6 hospitals), with a similar distribution of age and sex as in RS-TIA, was prepared. Two independent observers assessed medical records for quality of documentation and assigned a diagnosis of likely, possible, unlikely TIA or ischemic stroke, according to prespecified criteria. Results: The 2 observers agreed in 77% of cases that the event was a likely or possible TIA, in 3% that the event was an ischemic stroke, and in 2% that the event was an unlikely TIA. The observers disagreed in 8% of patients on TIA vs. ischemic stroke, and in 11% on a vascular vs. non-vascular cause. Quality of documentation was fair. Conclusions: There was interobserver agreement on diagnosis of TIA in the majority of patients included in RS-TIA. Diagnostic accuracy may be further improved by more systematic documentation of symptoms and signs.

Intracranial Hemorrhage After Ischemic Stroke: Incidence, Time Trends, and Predictors in a Swedish Nationwide Cohort of 196 765 Patients.

Epidemiological data on the risk of intracranial hemorrhage (ICrH) after ischemic stroke are sparse. The aims of this study were to describe incidence, trends over time, and predictors of ICrH within 1 year after ischemic stroke. All patients registered in the Swedish stroke register Riksstroke for 1998 to 2009 were included (n=196 765), and data were combined with the National Patient Register to identify ICrH occurrence. A matched reference population was obtained. Incidence rates and cumulative incidences were calculated. Multivariable regression analyses were used to identify predictors. Analyses were performed separately for the first 30 days and days 31 to 365 after ischemic stroke. The incidence rate was 1.97% per year at risk for the first year (0.13% in the reference population) and 0.85% excluding the first 30 days. Over time, the cumulative incidence increased the first 30 days but decreased over days 31 to 365. Thrombolysis, previous ICrH, atrial fibrillation, and male sex were associated with increased risk of ICrH during the first 30 days. Previous ICrH, increasing age, and male sex were associated with increased risk during days 31 to 365. Statins and antithrombotic treatment did not independently predict ICrH occurrence. The incidence of ICrH within 1 year after ischemic stroke was ≈2% per year at risk, about 15 times higher compared with the reference population. Over the study period, ICrH risk increased within the first 30 days but decreased thereafter. Previous ICrH, thrombolysis, and male sex affected the risk, whereas an increased use of antithrombotic treatments and statins did not.

Socioeconomic Inequalities in the Prescription of Oral Anticoagulants in Stroke Patients With Atrial Fibrillation.

Oral anticoagulants (OACs) are effective against ischemic stroke in patients with atrial fibrillation. Our aim was to investigate differences in the prescribing of OACs after ischemic stroke in patients with atrial fibrillation based on age, sex, country of birth, and socioeconomic status. Patients with first-ever ischemic stroke and atrial fibrillation without OAC treatment were included from the Swedish stroke register from 2009 to 2012. The outcome was OAC prescribed at discharge. Income, education, country of birth, and risk factors were obtained from official registers. Risk factors and health status were controlled for in multivariable logistic regression. Of 12 088 stroke patients, 36.3% were prescribed an OAC. Prescribing was less common with older age and, in patients born in other Nordic countries (odds ratio [OR], 0.82; 95% confidence interval [CI], 0.68-0.98) or countries outside of Europe (OR, 0.65; 95% CI, 0.42-0.99) compared with those born in Sweden. University education (OR, 1.20; 95% CI, 1.05-1.36) and highest income (OR, 1.19; 95% CI, 1.06-1.33) were associated with higher levels of OAC prescribing compared with those with primary school education or lowest income level. Differences by age, income, education, and country of birth were found in the prescribing of OACs after stroke. Differences were not explained by common risk factors. This indicates socioeconomic inequalities in the prescribing of preventive treatment after stroke.

Statin therapy and the risk of intracerebral haemorrhage: a nationwide observational study.

The association between statin therapy and intracerebral haemorrhage is still unclear. The aim was to investigate whether prior use of statin was associated with risk of intracerebral haemorrhage. Between 2006 and 2009, we identified 7696 cases of intracerebral haemorrhage that were first-ever strokes in the Swedish Stroke Register and 14 670 stroke-free controls that were matched on age and gender in the Population Register. Drug therapy at the time of intracerebral haemorrhage was extracted from the Drug Prescription Register. The risk of intracerebral haemorrhage with statins was estimated by conditional logistic regression. In cases and controls, the median age was 73 years and 53% were men. Intracerebral haemorrhage cases had higher prevalence of antithrombotic therapy, hypertension, and diabetes than controls. Statins were used by 1276 (16·6%) of the intracerebral haemorrhage cases and by 2552 (17·4%) of the controls. The crude odds ratios of intracerebral haemorrhage did not differ significantly between patients with and without statins, but after adjustment for antithrombotic therapy, hypertension, and diabetes, patients with statins had a decreased risk of intracerebral haemorrhage (odds ratio = 0·68, 95% confidence interval, 0·63-0·74). The highest proportion (>20%) of antecedent statins was seen in the 70-84 age group, for both cases and controls. In this matched case-controlled study, statin therapy was associated with a decreased risk of incident intracerebral haemorrhage. Future studies on risk of stroke with statin therapy after intracerebral haemorrhage are needed.

Suicide Risks After Stroke

Mood disorders are common after stroke, and risk for suicidal behavior appears to be increased. Now, using data from Swedish national registries, researchers have assessed the occurrence of suicide attempts (hospital admissions for intentional self-harm) and completed suicides among patients first entered in the Swedish Stroke Register between 2001 and 2012. Stroke severity was rated by consciousness level on hospital admission. At 3 months, …

Poststroke suicide attempts and completed suicides: a socioeconomic and nationwide perspective.

We examined attempted and completed suicides after stroke to determine whether they were associated with socioeconomic status, other patient characteristics, or time after stroke. This nationwide cohort study included stroke patients from Riksstroke (the Swedish Stroke Register) from 2001 to 2012. We used personal identification numbers to link the Riksstroke data with other national registers. Suicide attempts were identified by a record of hospital admission for intentional self-harm (ICD-10: X60-X84), and completed suicides were identified in the national Cause of Death Register. We used multiple Cox regression to analyze time from stroke onset to first suicide attempt. We observed 220,336 stroke patients with a total follow-up time of 860,713 person-years. During follow-up, there were 1,217 suicide attempts, of which 260 were fatal. This was approximately double the rate of the general Swedish population. Patients with lower education or income (hazard ratio [HR] 1.37, 95% confidence interval [CI] 1.11-1.68) for primary vs university and patients living alone (HR 1.73, 95% CI 1.52-1.97) had an increased risk of attempted suicide, and patients born outside of Europe had a lower risk compared to patients of European origin. Male sex, young age, severe stroke, and poststroke depression were other factors associated with an increased risk of attempted suicide after stroke. The risk was highest during the first 2 years after stroke. Both clinical and socioeconomic factors increase the risk of poststroke suicide attempts. This suggests a need for psychosocial support and suicide preventive interventions in high-risk groups of stroke patients.

National registry-based case-control study: comorbidity and stroke in young adults.

Stroke is overrepresented in cohorts of young adults with chronic diseases. The prevalence and impact of comorbidity among young stroke patients have not been compared with individuals without stroke. Our aim was to investigate the association between comorbidity and stroke in young adults. A nationwide cohort of patients (aged 15-44 years), registered in the Swedish Stroke Register, (Riksstroke) 2001-2009, was identified. Age- and sex-matched controls were randomly selected from the Population Register of Sweden. Discharge diagnoses were retrieved from the National Patient Register and grouped by chapter in the International Classification of Diseases 10th revision. Associations between ICD-10 chapters and stroke were stratified (age, sex, and stroke type) and analyzed by multivariable logistic regression. In 2599 stroke patients analyzed, the prevalence of vascular risk factors (hypertension 25.3%, dyslipidemia 13.0%, diabetes 9.7%, heart failure 3.2%, and atrial fibrillation 2.8%), all ICD-10 chapters (except pregnancy) and prestroke hospitalizations were more frequent among cases than controls. Independent associations were found between stroke and eight ICD-10 chapters: neoplasms (odds ratios (OR) 1.53, 95% CI 1.15-2.05), blood (OR 1.61, 1.11-2.34), endocrine (OR 2.28, 1.77-2.93), psychiatric (OR 1.50, 1.24-1.81), nervous (OR 1.91, 1.46-2.50), eye (OR 1.67, 1.05-2.64), circulatory (OR 3.05, 2.45-3.80), and symptoms (OR 1.31, 1.13-1.52). The risk of stroke increased by 26% per ICD-10 chapter diagnosed. In addition to vascular risk factors, comorbidity (represented by ICD-10 chapters) was associated with increased risk of stroke in young individuals. The risk of stroke was further increased with the number of diagnosed ICD-10 chapters.

Changes in functional outcome over the first year after stroke: an observational study from the Swedish stroke register.

Large longitudinal studies on stroke outcome are scarce. The aim of this study was to analyze predictors and changes in functional outcome during the first year poststroke. Data on patients who were independent in activities of daily living (ADL) and hospitalized for acute stroke in 2008 to 2010 were obtained from the Swedish Stroke Register. Case fatality was assessed by linkage to the Swedish Population Register. ADL was defined by independence or dependence in dressing, toileting, and indoor mobility and assessed at 3 and 12 months. Predictors of ADL dependency were assessed through multivariate analysis. In total, 64 746 patients were included. Case fatality at 3 months was 13.1% (men 11.6% versus women 14.8%; P<0.0001) and at 12 months 18.2% (men 16.4% versus women 20.3%; P<0,0001). In the 35 064 followed-up survivors, ADL dependency rates at 3 and 12 months were 16.2% (men 15.9% versus women 19.2%; P<0.0001) and 28.3% (men 22.7% versus women 34.9%, P<0.0001), respectively. Factors predicting deterioration to ADL dependency between 3 and 12 months were female sex (relative risk [RR]=1.56; 95% confidence interval [CI], 1.50-1.70), diabetes mellitus (RR=1.50; 95% CI, 1.05-1.60), comatose at admittance (RR=2.34; 95% CI, 1.79-3.05), previous stroke (RR=1.52; 95% CI, 1.43-1.61), hemorrhagic or unspecified stroke (RR=1.14; 95% CI, 1.05-1.25), and atrial fibrillation (RR= 1.11; 95% CI, 1.04-1.17). Transition from ADL independence to dependence was observed in a high proportion of patients between 3 and 12 months, challenging the common belief that functioning after stroke is stable beyond 3 months. Deterioration occurred more commonly in women, among whom 1/6 converted to dependency.

Education level and inequalities in stroke reperfusion therapy: observations in the Swedish stroke register.

Previous studies have revealed inequalities in stroke treatment based on demographics, hospital type, and region. We used the Swedish Stroke Register (Riksstroke) to test whether patient education level is associated with reperfusion (either or both of thrombolysis and thrombectomy) treatment. We included 85 885 patients with ischemic stroke aged 18 to 80 years registered in Riksstroke between 2003 and 2009. Education level was retrieved from Statistics Sweden, and thrombolysis, thrombectomy, patient, and hospital data were obtained from Riksstroke. We used multivariable logistic regression to analyze the association between reperfusion therapy and patient education. A total of 3649 (4.2%) of the patients received reperfusion therapy. University-educated patients were more likely to be treated (5.5%) than patients with secondary (4.6%) or primary education (3.6%; P<0.001). The inequality associated with education was still present after adjustment for patient characteristics; university education odds ratio, 1.14; 95% confidence interval, 1.03 to 1.26 and secondary education odds ratio, 1.08; 95% confidence interval, 1.00 to 1.17 compared with primary education. Higher hospital specialization level was also associated with higher reperfusion levels (P<0.001). In stratified multivariable analyses by hospital type, significant treatment differences by education level existed only among large nonuniversity hospitals (university education odds ratio, 1.20; 95% confidence interval, 1.04-1.40; secondary education odds ratio, 1.14; 95% confidence interval, 1.01-1.29). We demonstrated a social stratification in reperfusion, partly explained by patient characteristics and the local hospital specialization level. Further studies should address treatment delays, stroke knowledge, and means to improve reperfusion implementation in less specialized hospitals.

Warfarin-associated intracerebral hemorrhage after ischemic stroke.

The aim was to investigate the risk of intracerebral hemorrhage (ICH) in patients with ischemic stroke taking warfarin and whether this risk changed over time. Between 2001 and 2008, the Swedish Stroke Register registered 12,790 patients with ischemic stroke discharged on warfarin. The patients was studied in two 4-year periods (inclusion 2001-2004: follow-up until 2005 and inclusion 2005-2008: follow-up until 2009) for which rates of subsequent ICH were calculated. Adjusted hazard ratios, comparing the second period with the first period, were estimated in Cox regression models. Of 6039 patients, 58 patients (1.0%) in the first period and 69 of 6751 patients (1.0%) in the second period had subsequent ICH. Annual rates of ICH ranged from 0.37% in the first period to 0.39% in the second period (adjusted hazard ratio, 1.04; 95% confidence interval, 0.73-1.48). In this nationwide study, the risk of warfarin-associated ICH among ischemic stroke patients was low and did not change during the 2000s.

On shrinkage and model extrapolation in the evaluation of clinical center performance

We consider statistical methods for benchmarking clinical centers based on a dichotomous outcome indicator. Borrowing ideas from the causal inference literature, we aim to reveal how the entire study population would have fared under the current care level of each center. To this end, we evaluate direct standardization based on fixed versus random center effects outcome models that incorporate patient-specific baseline covariates to adjust for differential case-mix. We explore fixed effects (FE) regression with Firth correction and normal mixed effects (ME) regression to maintain convergence in the presence of very small centers. Moreover, we study doubly robust FE regression to avoid outcome model extrapolation. Simulation studies show that shrinkage following standard ME modeling can result in substantial power loss relative to the considered alternatives, especially for small centers. Results are consistent with findings in the analysis of 30-day mortality risk following acute stroke across 90 centers in the Swedish Stroke Register.

COSTS FOR STROKE IN SWEDEN 2009 AND DEVELOPMENTS SINCE 1997

The aim of this study was to estimate direct and indirect excess costs attributable to stroke in Sweden in 2009 and to compare these with similar estimates from 1997. Data on first-ever stoke admissions in the first half of 2009 from the Swedish national stroke register (RS) were used for cost calculations and compared with results from 1997 also using RS data. A societal perspective was taken including the acute and follow-up phase, rehabilitation, stroke re-admissions, drugs, home- and residential care services for activities of daily life (ADL) support, and indirect costs for premature death and productivity losses (2009 prices). Survival was extrapolated to estimate the lifetime present value cost of stroke. The societal lifetime present value cost for stroke in 2009 was €68,800 per patient (ADL support: 59 percent; productivity losses: 21 percent). Women had higher costs than men in all age groups as a result from greater need for ADL support. Patients treated at a stroke unit indicated low incremental cost per life-year gained compared with those who had not. The total lifetime cost increased between 1997 and 2009. Hospitalization costs per patient were stable, while long-term costs for home- and residential care services increased. Changes in patient characteristics, longer expected survival, and possibly in the Swedish stroke care, have led to higher annual and lifetime costs per patient in 2009 compared with 1997. A comprehensive national stroke care performance register like RS may be suitable for health economic assessments.

Trends in Stroke Treatment and Outcome between 1995 and 2010: Observations from Riks-Stroke, the Swedish Stroke Register

Background: Continuous changes in stroke treatment and care, as well as changes in stroke characteristics, may alter stroke outcome over time. The aim of this paper is to describe time trends for treatment and outcome data, and to discuss if any such changes could be attributed to quality changes in stroke care. Methods: Data from Riks-Stroke, the Swedish stroke register, were analyzed for the time period of 1995 through 2010. The total number of patients included was 320,181. The following parameters were included: use of computed tomography (CT), stroke unit care, thrombolysis, medication before and after the stroke, length of stay in hospital, and discharge destination. Three months after stroke, data regarding walking, toileting and dressing ability, as well social situation, were gathered. Survival status after 7, 27 and 90 days was registered. Results: In 1995, 53.9% of stroke patients were treated in stroke units. In 2010 this proportion had increased to 87.5%. Fewer patients were discharged to geriatric or rehabilitation departments in later years (23.6% in 2001 compared with 13.4% in 2010), but more were discharged directly home (44.2 vs. 52.4%) or home with home rehabilitation (0 vs. 10.7%). The need for home help service increased from 18.2% in 1995 to 22.1% in 2010. Regarding prevention, more patients were on warfarin, antihypertensives and statins both before and after the stroke. The functional outcome measures after 3 months did improve from 2001 to 2010. In 2001, 83.8% of patients were walking independently, while 85.6% were independent in 2010. For toileting, independence increased from 81.2 to 84.1%, and for dressing from 78.0 to 80.4%. Case fatality (CF) rates after 3 months increased from 18.7% (2001) to 20.0% (2010). This trend is driven by patients with severe strokes. Conclusions: Stroke outcomes may change over a relatively short time period. In some ways, the quality of care has improved. More stroke patients have CT, more patients are treated in stroke units and more have secondary prevention. Patients with milder strokes may have benefited more from these measures than patients with severe strokes. Increased CF rates for patients with severe stroke may be caused by shorter hospital stays, shorter in-hospital rehabilitation periods and lack of suitable care after discharge from hospital.

Implementing Thrombolytic Guidelines in Stroke Care

We performed a qualitative study to identify facilitators of and barriers to the implementation of national guidelines on thrombolytic therapy for acute ischemic stroke. We interviewed physicians and nurses at nine Swedish hospitals using 16 explorative, semistructured interviews, and selected hospitals based on their implementation rate of new stroke care methods according to data from the Swedish Stroke Register, Riks-Stroke. Through content analysis, we identified facilitators and barriers to implementation, which we classified into three categories: (a) individuals, (b) social interactions and context, and (c) organizational and resource issues. Insights obtained from this study can be used to identify target areas for improving the implementation of thrombolytic therapy and other new methods in stroke care.

Reduced Inequality in Access to Stroke Unit Care over Time: A 15-Year Follow-Up of Socioeconomic Disparities in Sweden

Background: Despite the compelling scientific evidence on the superiority of stroke unit care, far from all acute stroke patients have access to stroke unit care. In congruence with what has been observed when other new methods are introduced in health care, we hypothesized that there has been an inequality in the buildup phase of stroke units but that the gradients between patient groups have decreased as the total capacity of stroke unit care has increased. The purpose of this study was to explore if patients in a national sample who were socioeconomically disadvantaged (low education or low income) had reduced access to stroke unit care and if differences varied over time. Methods: All patients 18-74 years of age registered between 1995 and 2009 in Riks-Stroke, the Swedish stroke register, were included. The Stroke Unit Trialists' definition of a stroke unit has been adopted by Riks-Stroke and hospitals participating in the registry. Basic patient characteristics, stroke risk factors, process and outcome variables are recorded in Riks-Stroke. Socioeconomic data were accessed from Statistics Sweden. Multiple logistic regression analyses were used to calculate odds ratios (ORs) for stroke unit care between prespecified patient subgroups. Results: A total of 319,240 stroke patients were included in Riks-Stroke during the years 1995-2009, and 124,173 were aged between 18 and 74 years; they were included in the final analyses. After adjustment for confounders in a multiple regression model, women were treated in stroke units slightly less often [OR 0.97, 95% confidence interval (CI) 0.95-0.99]. There were no statistically significant associations between stroke unit care and age or between stroke unit care and cohabiting or living alone. The highest level of education predicted access to stroke unit care (secondary vs. primary school: OR 1.04, 95% CI 1.01-1.07; university vs. primary school: OR 1.06, 95% CI 1.02-1.10). Differences according to level of education diminished over time (p = 0.001). Income was not independently associated with stroke unit care, and over time the proportion of patients treated in stroke units increased at a similar rate in all income groups (p = 0.12). Conclusions: Even in a country with modest socioeconomic differences in the general population and public financing of all acute hospital care, socioeconomic inequalities in access to stroke unit care were evident during the early years, but they diminished as the total capacity for stroke unit care increased.

The association between patients’ beliefs about medicines and adherence to drug treatment after stroke: a cross-sectional questionnaire survey

ObjectivesAdherence to preventive drug treatment is a clinical problem and we hypothesised that patients’ beliefs about medicines and stroke are associated with adherence. The objective was to examine associations between...