Pragmatic trial of multifaceted intervention (STROKE-CARD care) to reduce cardiovascular risk and improve quality-of-life after ischaemic stroke and transient ischaemic attack –study protocol

In this trial, we test the hypothesis that the pragmatic disease-management program STROKE-CARD designed for ischaemic stroke and TIA patients is capable of (a) preventing recurrent cardiovascular events by optimised guideline-compliant secondary prevention and target level achievement and (b) ameliorating health-related quality of life (QoL) and patient-wellbeing by early detection and consequent treatment of post-stroke complications.

STROKE-CARD is a randomised, controlled, open interventional phase III trial with blinded outcome assessment [ClinicalTrials.gov NCT02156778] conducted in Austria. It initially started as a single-centre study in January 2014 and was extended to a second centre in December 2014. It compares two standards of post-stroke patient care, both complying with the current state-of-the-art. The trial does not involve experimental interventions but compares usual with intensified care as specified below. Primary outcome events are adjudicated based on pre-specified diagnostic criteria blinded to treatment allocation.

The two trial centres are located at the Departments of Neurology at the Innsbruck University Hospital (IUH) and the Hospital St. John of God in Vienna. IUH serves as the comprehensive stroke unit for the entire federal state of Tyrol (catchment area ≈ 1 Mio) and as a primary stroke unit for the city of Innsbruck and 65 surrounding suburban communities with approximately 0.3 Mio inhabitants. Patients from this exclusive catchment area represent an unselected cohort of ischaemic stroke and TIA patients [5]. The Hospital St. John of God Vienna houses one of the four comprehensive stroke units in Vienna. The Viennese stroke network includes additional six primary stroke units and serves the city of Vienna with a catchment area of more than 1.8 Mio inhabitants.

The study population comprises consecutive patients with acute ischaemic stroke or TIA (ABCD2-Score ≥ 3) [17] admitted to the two study centres. Patients are eligible for inclusion irrespective of whether the index event was a first or recurrent event. Ischaemic stroke is ascertained using the American Heart Association criteria based on clinical and imaging features [18].

All patients with ischaemic stroke or TIA admitted to the study centres are screened for potential inclusion in the STROKE-CARD trial during the first days of their hospital stay and in most instances included before discharge after signing an informed consent. Further details on the inclusion and exclusion criteria are provided in Table 1.

We block-randomise patients in a 2:1 ratio to receive the STROKE-CARD program or standard care on the basis of the exact date and time of the qualifying event, with the durations of the blocks ranging from 4 to 8 weeks. The target sample size is 2160 patients overall, 1440 patients in the STROKE-CARD group, and 720 in the control group (Fig. 1). We strive for inclusion of about 500 patients per year. Patients are followed-up for 12 months, but we also obtain consent from the patients to contact them later on and extend follow-up for future research projects. Patients who do not adhere to the extended standard care program, especially do not attend the 3-month assessment, are followed in the same way as adherent patients for primary and secondary outcome events.

The protocol uses validated and field-proven forms, scales and questionnaires. All the study team members are highly experienced and NIHSS- and mRS-certified.

The clinical baseline-assessment (entire study population) and the outpatient appointment (3-month intervention [intervention arm only] and 12-month outcome assessment [entire study population]) will be performed by trained health care professionals of the multidisciplinary study team composed of experienced stroke neurologists, PhD-students, physiotherapists, and stroke nurses. Data are collected prospectively according to pre-specified protocols and questionnaires with electronic data entry and storage and undergo subsequent quality checks and validation (e.g. by hospital records and feedback from the GP). All hospitals in the survey area dispose of electronic records, to which study investigators were granted access via protected data links according to the patients’ informed consent.

The baseline assessment includes detailed documentation of the following components: ischaemic stroke aetiology [TOAST-criteria [26], Causative Classification of Stroke [27]], stroke severity [NIHSS [28] on admission and discharge], clinical syndrome [stroke/TIA symptoms, vessel territories], acute stroke therapy and recanalization procedures, pre-stroke disability and disability on admission [mRS], prior cardiovascular events and comorbidities, cardiovascular risk factors (hypertension, dyslipidaemia, smoking-status [current, former, never] and burden [pack-years, Fagerstroem questionnaire [29]], alcohol consumption [quantity, patterns, type of beverages], diabetes [ADA criteria] [30], atrial fibrillation, physical activity [Baecke Physical Activity Questionnaire] [31]), family history of cardiovascular disease, previous and current medication, cognitive function [Mini-Mental State Examination [32] and Montreal Cognitive Assessment [33]], depression [Beck Depression Inventar, anxiety [Hospital Anxiety and Depression Scale [34]], fatigue [Fatigue Severity Scale [35]], health related quality of life [European Quality of Life-5 Dimensions EQ5D-3 L [36]], anthropometric measures [body mass index, waist–to hip ratio], repeated blood pressure recordings [acute setting and before discharge], ankle-brachial index, atherosclerosis of the carotid arteries [intima-media thickness, plaque burden], [37] pulse wave velocity [Complior® system], and routine laboratory measures [including lipid profile, renal function, liver function, thyroid function, cardiac markers, differential blood count, coagulation diagnostics, and inflammation markers].

The 3- and 12-month visits involve a neurological evaluation with documentation of clinical deficits and functional outcome [NIHSS, mRS, and Barthel Index [38]], routine laboratory examination and assessment of recurrent cardiovascular events, hospital stays and procedures [interventional and surgical], falls and fractures, all incident diseases and morbidities [including bleeding events], current medication, adherence to prescribed drugs, and risk factor control. Cognitive function, depression, anxiety, fatigue, and QoL all are re-evaluated [see baseline assessment]. Carotid ultrasound [see baseline protocol] is obligatorily repeated during the 12-month visit and, if clinically indicated, also after three months.

Blood and urine samples are drawn after an overnight fast and at least 12-h abstinence from smoking [during the hospital stay and 3 and 12 months thereafter], immediately processed, used for routine testing, and stored in a biobank [plasma, serum, whole blood, and urine]. Samples are stored at − 80 °C in compliance with the OECD guidelines and Austrian Bioethics Commission recommendations [39].

Sample size calculations for this trial were based on the primary composite cardiovascular disease endpoint. Assuming a 1-year cumulative risk of 15% in the usual care group, 2045 subjects are required to detect a 5% absolute risk reduction (α = 0.05) with a 90% power.

Expecting an attrition rate of 15% (7.5% dropout rate [withdrawal of consent] and a loss-to-follow-up of 7.5%), we aimed to include a total of 2400 patients. Between January 2014 and June 2017, 1866 patients were recruited, with an attrition rate substantially lower than expected (< 5.0% as opposed to the anticipated 15%). Accordingly, the data monitoring board recommended revision of the recruitment goal from 2400 to 2160 patients. A sample size of 2160 patients provides the trial with a 90% power to detect a difference of 0.03 points on the EQ-5D-3 L overall health utility score (co-primary efficacy endpoint) assuming a standard deviation of 0.2.

The primary efficacy analysis is based on time from hospital discharge to first occurrence of primary composite end-point as defined above. It is an unadjusted survival analysis according to the intention-to-treat principle with use of the log-rank test. Hazard ratios and 95% confidence intervals will be estimated using Cox models with the assumption of proportional hazards checked by means of Schoenfeld residuals. Prespecified sensitivity analyses include adjusted models and a per-protocol analysis. Consistency of intervention effects in four prespecified subgroups (men and women, age groups, type of index event, and patients with or without regular use of My StrokeCard) will be assessed by means of test for interaction. The continuous efficacy endpoint QoL will be analysed with linear regression models.

Owing to the fact that our study attempts to encourage patients to adhere to current evidence-based prevention guidelines and does not include experimental treatment concepts, we do not perform a formal safety analysis. We however analyze whether there are any differences in the occurrence of potential side effects, risks and consequences of intensified secondary prevention in both study arms including (a) major bleeding [40], (b), syncopes and fractures, (c) laboratory abnormalities (ALT, CK, GFR), and (d) muscle-related outcomes. No formal interim analysis will be performed. All reported P values are two-sided.

STROKE-CARD is the first comprehensive pragmatic post-stroke disease management in Austria and, to our knowledge, one of the largest interventions of its kind worldwide. STROKE-CARD is addressing a real-life-cohort of ischaemic stroke- and TIA-patients with the primary objectives of enhancing adherence to prevention guidelines and thereby lowering rates of recurrent vascular events and post-stroke complications, and improving functional outcome and QoL. Whereas disease management programs traditionally rely on expert opinion, our initiative strives for a rigorous scientific evaluation involving outcome and health economy analyses. It pursues the concept that optimal acute stroke care extends to a thorough three-month assessment and individualised counselling. We employ a multidisciplinary but lean and cheap intervention leveraging contemporary e-technology and encouraging patient self-empowerment with the prospect of a widespread implementation in case of success. Access to Internet and social media is now broadly available and STROKE-CARD offers web-based educational patient-operated tools for risk factor to record and manage risk factors combined with linkage of patient data to medical caregivers. The STROKE-CARD concept was presented as an abstract at the European Stroke Organisation Conference 2018 [41].

Previous trials on multimodal secondary prevention strategies in ischaemic stroke or TIA-patients have shown variable improvements in risk factor control or medication adherence but most were not designed and powered to analyse potential effects recurrent cardiovascular disease and stroke events. An overview on trials focusing on multimodal secondary stroke prevention is provided in Table 3. Limitations shared by these trials include the usually limited sample size (443 median, IQR 347.3) [42‐50] and the common focus on TIA or minor stroke patients only [48, 49, 51].

Only one large-scale study has been completed so far [45] and enrolled more than 3800 ischaemic stroke and TIA-patients in 47 hospitals in China. In spite of a better adherence to statin-medications no difference in the hard clinical end points after one year of follow-up could be demonstrated.

Two ongoing studies deserve special consideration: (i) The Comprehensive Post-Acute Stroke Service (COMPASS) study, [52] a cluster-randomised trial involving 40 hospitals in North Carolina, plans enrolment of 6000 ischaemic and haemorrhagic stroke patients and investigates potential effects of the COMPASS care model (telephone follow-up, one clinic visit within two weeks after discharge, and an individualised patient electronic care plan) on the patient-reported 90-day functional outcome (primary endpoint), hospital re-admission rates, and 90-day mortality. (ii) The Intensified Secondary Prevention intending a Reduction of Recurrent Events in TIA and Minor Stroke patients (INSPiRE-TMS) study, [51] from Germany intends to recruit 2082 patients with minor stroke or TIA, provides up to eight appointments in outpatient clinics and focuses on two-year recurrent vascular events or vascular death.

Type and intensity of intervention vary widely among the different trials. Several programs envisage nurse-led home visits or telephone-based assessments for risk factor control [44, 47, 50]. The ICARUSS-strategy [43] in Australia trusts on several pre-arranged GP visits whereas the STANDFIRM-trial (n = 500), [46] a community-based intervention, consists of educational sessions, multiple follow-up assessments, and as telephone interviews conducted by study nurses and general practices. A recent intervention [49] has achieved a substantial and sustained (six months) reduction in global vascular risk (Framingham Risk Score, Cardiovascular Disease Life Expectancy Model) by non-physician provider case management. Digital health interventions (e.g. telemedicine, web-based strategies, email and- mobile phone-reminders) improved risk factor control in vascular diseases other than stroke [16].

Key and in part unique features of STROKE-CARD are its comprehensive focus on both recurrent vascular events (risk factor control and compliance) and post-stroke complications and QoL, its lean and easily applicable intervention, and the inclusion of moderate and severe ischaemic stroke patients (all except those with permanent severe disability).

The primary analysis may be viewed as conservative in two respects: [1] It relies on a short one-year follow-up whereas potential effects are expected to increase over time. Extension of follow-up is a decided goal and fund raising is in progress [2]. Our program is realised within the framework of a highly developed stroke care setting and bears the challenge of a strong comparator group with high-quality management of ischaemic stroke/TIA patients. Although findings of our trial are not easily extrapolated to less developed stroke systems, the potential benefit of the intervention is presumably even higher on the low-quality stroke care background.

Our program is focused on ischaemic stroke but may serve as a model for other diseases sensitive to post-discharge management.

In case of success, the next steps would be a refinement of STROKE-CARD components and identification of subgroups with a health benefit and cost effectiveness in order to end up with a practicable tool for broad clinical implementation in the whole of Tyrol, Austria and beyond.