Interventions for improving modifiable risk factor control in the secondary prevention of stroke

  • Review
  • Intervention

Authors


Abstract

Background

People with stroke or transient ischaemic attack (TIA) are at increased risk of future stroke and other cardiovascular events. Evidence-based strategies for secondary stroke prevention have been established. However, the implementation of prevention strategies could be improved.

Objectives

To assess the effects of stroke service interventions for implementing secondary stroke prevention strategies on modifiable risk factor control, including patient adherence to prescribed medications, and the occurrence of secondary cardiovascular events.

Search methods

We searched the Cochrane Stroke Group Trials Register (April 2013), the Cochrane Effective Practice and Organisation of Care Group Trials Register (April 2013), CENTRAL (The Cochrane Library 2013, issue 3), MEDLINE (1950 to April 2013), EMBASE (1981 to April 2013) and 10 additional databases. We located further studies by searching reference lists of articles and contacting authors of included studies.

Selection criteria

We included randomised controlled trials (RCTs) that evaluated the effects of organisational or educational and behavioural interventions (compared with usual care) on modifiable risk factor control for secondary stroke prevention.

Data collection and analysis

Two review authors selected studies for inclusion and independently extracted data. One review author assessed the risk of bias for the included studies. We sought missing data from trialists.

Main results

This review included 26 studies involving 8021 participants. Overall the studies were of reasonable quality, but one study was considered at high risk of bias. Fifteen studies evaluated predominantly organisational interventions and 11 studies evaluated educational and behavioural interventions for patients. Results were pooled where appropriate, although some clinical and methodological heterogeneity was present. The estimated effects of organisational interventions were compatible with improvements and no differences in the modifiable risk factors mean systolic blood pressure (mean difference (MD) -2.57 mmHg; 95% confidence interval (CI) -5.46 to 0.31), mean diastolic blood pressure (MD -0.90 mmHg; 95% CI -2.49 to 0.68), blood pressure target achievement (OR 1.24; 95% CI 0.94 to 1.64) and mean body mass index (MD -0.68 kg/m2; 95% CI -1.46 to 0.11). There were no significant effects of organisational interventions on lipid profile, HbA1c, medication adherence or recurrent cardiovascular events. Educational and behavioural interventions were not generally associated with clear differences in any of the review outcomes, with only two exceptions.

Authors' conclusions

Pooled results indicated that educational interventions were not associated with clear differences in any of the review outcomes. The estimated effects of organisational interventions were compatible with improvements and no differences in several modifiable risk factors. We identified a large number of ongoing studies, suggesting that research in this area is increasing. The use of standardised outcome measures would facilitate the synthesis of future research findings.

Plain language summary

Healthcare interventions for reducing the risk of future stroke in people with previous stroke or transient ischaemic attack (TIA)

Question

We wanted to assess the effects of interventions for implementing secondary stroke prevention strategies to reduce the risk of further stroke or other cardiovascular events in people who have already had a stroke or transient ischaemic attack (TIA).

Background

Stroke and TIA are diseases that are caused by interruptions in the blood supply to the brain. People who experience a stroke or TIA are at risk of future stroke and other cardiovascular events. Several medications and lifestyle changes can be used to lower stroke risk by improving the control of modifiable risk factors. In this review, modifiable risk factors for stroke included systolic and diastolic blood pressure, blood lipids, atrial fibrillation, diabetes management, body mass index (BMI) and the use of preventive medications as prescribed. Research evidence suggests that modifiable risk factors are often not managed effectively following a stroke or TIA. Therefore, it is important to identify healthcare interventions that can facilitate stroke prevention by improving modifiable risk factor control and patient adherence to prescribed medications. The categories of interventions considered in this systematic review were educational and behavioural interventions for patients; educational and behavioural interventions for stroke service providers; organisational interventions.

Study characteristics

We identified 26 studies, up to April 2013, for inclusion in the review. These studies included a total of 8021 participants with cerebrovascular disease and were undertaken in the USA, Canada, Europe, Asia and Australia. The mean or median age of participants ranged from 60 to 73 years. Six studies included participants with a diagnosis of ischaemic stroke, whereas three studies included participants with either ischaemic or haemorrhagic stroke, or did not specify stroke subtype. The majority of studies were set in primary care or community settings. Eleven studies involved educational or behavioural interventions for participants and 15 studies involved predominantly organisational interventions. The majority of interventions had a duration of between three and 12 months.

Key results

Analysis of the effects of changes to the organisation of healthcare services was compatible with meaningful improvements in systolic blood pressure, diastolic blood pressure, blood pressure target achievement and BMI; although the imprecision of these estimates meant that absence of improvements could not be ruled out. The effects of these interventions on changes in blood lipids, diabetes management, use of preventive medications as prescribed, or the occurrence of stroke and other cardiovascular events were imprecise and consistent with benefit and harm. Changes to healthcare services that addressed only patient education or behaviour, without any changes to the organisation of patient care, were generally not associated with clear evidence of changes in modifiable risk factors for stroke. We identified a large number of ongoing studies, suggesting that research in this area is increasing.

Quality of the evidence

We judged all but one of the included studies to be of reasonable quality.

Laički sažetak

Zdravstvene intervencije za smanjenje rizika novih moždanih udara u ljudi s prethodnim moždanim udarom ili tranzitornom ishemijskom atakom (TIA-om)

Istraživačko pitanje

Autori Cochrane sustavnog pregleda procijenili su učinke intervencija za usvajanje strategija za sekundarnu prevenciju moždanog udara kako bi se smanjio rizik novih moždanih udara ili drugih srčano-žilnih (kardiovaskularnih bolesti) u ljudi koji su ranije imali moždani udar ili tranzitornu ishemijsku ataku (TIA-u)

Dosadašnje spoznaje

Moždani udar i TIA su bolesti koje uzrokuje prekid dotoka krvi u mozak. Osobe koje su imale moždani udar ili TIA-u imaju veći rizik za nove moždane udare i druge kardiovaskularne bolesti. Nekolicina lijekova i promjene stila življenja mogu se koristiti u smanjivanju rizika od moždanog udara poboljšanjem kontrole rizičnih čimbenika. U ovom Cochrane sustavnom pregledu, uključeni čimbenici rizika za moždani udar bili su sistolički i dijastolički krvni tlak, masnoće u krvi, fibrilacija atrija, kontrola šećerne bolesti, indeks tjelesne mase (ITM) i upotreba preventivnih lijekova prema uputama liječnika. Dokazi iz istraživanja pokazuju često neučinkovitu kontrolu čimbenika rizika nakon moždanog udara ili TIA-e. Stoga je važno utvrditi koje zdravstvene intervencije mogu olakšati prevenciju moždanog udara poboljšanjem kontrole čimbenika rizika i pridržavanjem uzimanja propisanih lijekova od strane pacijenta. Vrste intervencija koje su uzete u obzir u ovom sustavnom Cochrane pregledu su edukacijske i bihevioralne intervencije za pacijente; edukacijske i bihevioralne intervencije za zdravstvene radnike; organizacijske intervencije.

Značajke istraživanja

Pronađeno je 26 istraživanja objavljenih do travnja 2013. koja su uključena u ovaj sustavni pregled. Ta istraživanja uključila su ukupno 8.021 ispitanika s bolestima krvnih žila u mozgu (cerebrovaskularnim bolestima), a rađena su u SAD-u, Kanadi, Europi, Aziji i Australiji. Prosječna dob sudionika bila je u rasponu od 60 do 73 godine. Šest istraživanja uključili su sudionike s dijagnozom ishemijskog moždanog udara, dok su tri istraživanja uključila sudionike bilo s ishemijskim ili hemoragijskim moždanim udarom, ili nije navedena vrsta moždanog udara. Većina istraživanja su rađena u primarnoj zdravstvenoj zaštiti ili u zajednici. 11 istraživanja imala su edukacijske ili bihevioralne intervencije za sudionike a 15 istraživanja imala su uglavnom organizacijske intervencije. Većina intervencija trajala je od tri do 12 mjeseci.

Ključni rezultati

Analiza učinaka promjena u organizaciji zdravstvenih usluga bila je u skladu sa značajnim poboljšanjima sistoličkog krvnog tlaka, dijastoličkog krvnog tlaka, ciljnog krvnog tlaka i ITM-a; iako se zbog nepreciznosti tih procjena ne može isključiti izostanak poboljšanja. Učinci tih intervencija na promjene u koncentraciji lipida u krvi, kontrolu šećerne bolesti, korištenje preventivnih lijekova kako je propisano, odnosno pojava moždanog udara i drugih kardiovaskularnih bolesti bili su neprecizni i dosljedni kad je u pitanju procjena njihovih korisnih i štetnih učinaka. Promjene zdravstvenih usluga koje su uključile samo edukaciju pacijenata ili njihovo ponašanje, bez ikakvih promjena u organizaciji skrbi za bolesnike, uglavnom nisu bile povezane s jasnim dokazima promjena čimbenika rizika za moždani udar. Pronađen je velik broj istraživanja koja su u tijeku, što ukazuje na povećanje broja istraživanja u ovom području.

Kvaliteta dokaza

Procijenjeno je kako su sva uključena istraživanja osim jednog bila razumne kvalitete.

Bilješke prijevoda

Hrvatski Cochrane
Preveo: Mato Lakić
Ovaj sažetak preveden je u okviru volonterskog projekta prevođenja Cochrane sažetaka. Uključite se u projekt i pomozite nam u prevođenju brojnih preostalih Cochrane sažetaka koji su još uvijek dostupni samo na engleskom jeziku. Kontakt: cochrane_croatia@mefst.hr

Background

Description of the condition

Stroke is defined as a rapidly developing neurological deficit, of presumed vascular origin, lasting for over 24 hours or leading to death (WHO 1978). Transient ischaemic attack (TIA) is an expression traditionally used to describe comparable neurological deficits lasting for less than 24 hours (Albers 2002; Easton 2009). More recently a new definition of TIA has been proposed, omitting the arbitrary 24 hour time frame and identifying TIA as a "transient episode of neurological dysfunction caused by focal brain, spinal cord, or retinal ischaemia, without acute infarction" (Easton 2009).

The World Health Organization has reported that cerebrovascular disease (stroke) is the second leading cause of mortality and disease burden among adults aged 60 years and over (WHO 2003). Stroke survivors are 15 times more likely to have a stroke in the year following the primary event compared with the general population (Burn 1994). Furthermore, TIA patients constitute a high risk group for subsequent stroke with an early risk of 9.9%, 13.4% and 17.3% at two, 30 and 90 days, respectively (Wu 2007). Long-term cohort studies have demonstrated that the risk of cardiovascular events remains high for at least 10 years after stroke or TIA (Touze 2005; Van Wijk 2005). Secondary prevention strategies aim to prevent recurrent events by improving modifiable risk factor control. National stroke guidelines identify clinical conditions (hypertension, hyperlipidaemia, atrial fibrillation, diabetes and obesity) and lifestyle factors (smoking, physical inactivity, unhealthy diet and excess alcohol consumption) as significant modifiable risk factors that should be targeted for secondary prevention (ESO 2008; Furie 2011; Lindsay 2010; National Stroke Foundation 2010; RCP 2012; SIGN 2008).

Description of the intervention

For the purposes of this review, stroke services are considered to include all services responsible for providing acute and follow-up care to stroke and TIA patients. Stroke services exist as part of diverse healthcare systems, with specific treatment goals varying according to national clinical guidelines. Acute stroke services include organised inpatient (stroke unit) care and specialist TIA clinics (RCP 2012; Stroke Unit Trialists' Collaboration 2013). Recommendations for secondary prevention can be initiated as part of a co-ordinated treatment programme during acute hospitalisation (Ovbiagele 2004). However, primary care services are well placed to monitor patient risk factors, encourage lifestyle change and review secondary prevention medications on an ongoing basis (RCP 2012). Primary care aims to be characterised by person-centredness, comprehensiveness, continuity of care and community participation (Starfield 2002; WHO 2008). Social care services and voluntary sector organisations can also work in partnership with primary care to deliver healthy living support (NAO 2005). Stroke service interventions are termed complex interventions since they often contain several interacting components and may require complex behaviours, organisational change or the assessment of numerous outcome measures (Craig 2008; Redfern 2008).

How the intervention might work

Stroke services addressing secondary prevention aim to improve patient compliance with medication regimens and lifestyle advice. Several classes of medication reduce stroke incidence by modifying cardiovascular risk. For example, long-term antiplatelet medication in those with a history of stroke or TIA is associated with a significant 25% reduction in secondary vascular events (Antithrombotic Trialists' Collaboration 2002). Similarly, antihypertensive and statin medications are associated with improvements in secondary prevention (Manktelow 2009; Rashid 2003). Meta-analyses report that moderate to high physical activity (Lee 2003), moderate alcohol consumption (Reynolds 2003), reduction of salt intake (He 2004) and specific dietary changes (He 2004a; He 2006) can also facilitate stroke prevention and cardiovascular risk reduction. An international case-control study identified five modifiable risk factors accounting for 83% of the population attributable risk (PAR) for stroke (O'Donnell 2010). Targeting multiple risk factors may have additive benefits for secondary prevention; a modelling study predicted that an 80% cumulative risk reduction in recurrent vascular events could be achieved by combining dietary modification, exercise, aspirin, a statin, and an antihypertensive agent (Hackam 2007).

Why it is important to do this review

Most stroke patients have at least one cardiovascular risk factor and hypertension, hyperlipidaemia, diabetes, smoking and obesity are often inadequately managed during follow-up (Alvarez-Sabin 2009; Li 2008; Mouradian 2002). Although the effectiveness of secondary prevention medications is well-established, non-treatment rates for antithrombotic, antihypertensive and statin therapies remain high after stroke (Raine 2009) and TIA (Lager 2012). Only 31% of stroke patients and 35% of TIA patients receive combination treatment with all three medication classes (Ramsay 2007). Adherence to secondary prevention medications falls progressively as time since the primary stroke elapses (Glader 2010). As strategies for stroke prevention are not optimally implemented, substantial benefits stand to be gained from improving the use of evidence-based interventions (Goldstein 2008).

Several studies have revealed inequalities in the provision of stroke care with older patients being less likely to receive or adhere to secondary prevention medication (De Schryver 2005; Raine 2009; Ramsay 2007). Similarly, stroke patients with more severe disability (Barthel scores of 14 or less) are less likely to receive appropriate secondary prevention than those with mild disability (Barthel score 15 to 20) (Rudd 2004). Ethnic groups are also reported to differ with respect to patterns in behavioural risk factors for stroke (Dundas 2001). These subgroups of patients may require targeted interventions in order to improve risk factor control.

Service interventions used for other conditions, particularly secondary prevention of ischaemic heart disease, may be relevant to the secondary prevention of stroke (Buckley 2010). However, more direct evidence is needed to guide improvements in follow-up care after stroke or TIA. For example, stroke commonly results in cognitive impairments or physical disabilities that are likely to influence both intervention design and outcomes. To date, there are no systematic reviews that have considered the impact of stroke service interventions on cardiovascular risk factor control or adherence to secondary prevention medications. An assessment of the quality and outcomes of previous studies in this field will inform the development of new interventions.

Objectives

The objective of this review was to assess the effects of stroke service interventions for implementing secondary stroke prevention strategies on modifiable risk factor control, including poor adherence to prescribed medications, and the occurrence of secondary cardiovascular events.

Poor adherence can compromise secondary prevention by reducing the clinical benefits of long-term therapies (WHO 2003a). In the context of this review, 'modifiable risk factors' therefore refer to:

  • clinical conditions (hypertension, hyperlipidaemia, atrial fibrillation, diabetes and obesity);

  • patient non-adherence to secondary prevention medications.

Methods

Criteria for considering studies for this review

Types of studies

We included published or unpublished randomised controlled trials (RCTs) with a minimum follow-up of three months after the start of the intervention. Parallel group trials, cluster-randomised trials and cross-over trials were eligible for inclusion in the review.

Types of participants

We included in this review adults (aged 18 years and over) with a confirmed diagnosis of ischaemic stroke, haemorrhagic stroke or TIA.

Types of interventions

For the purposes of this review, we defined stroke service interventions as alternative models of care that are implemented in order to improve patient outcomes following stroke or TIA. We included stroke service interventions that were intended to improve modifiable risk factor control. We focused on interventions that aimed to improve modifiable risk factor control through increased adherence to existing recommendations for secondary stroke prevention (for example recommendations in international stroke guidelines).

We considered the following intervention categories (pre-specified in the review protocol).

  • Educational and behavioural interventions for patients.

  • Educational and behavioural interventions for stroke service providers.

  • Organisational interventions (subdivided into the following categories developed by Wensing 2006):

    • revision of professional roles, e.g. involvement of non-physician staff in prevention clinics;

    • collaboration between multidisciplinary teams, e.g. interventions promoting effective liaison between primary and secondary care teams;

    • integrated care services, e.g. disease and case management programs where patient care follows protocols for screening, education and treatment or monitoring;

    • knowledge management systems, e.g. computerised decision support on medication prescribing, shared medical records;

    • quality management, e.g. guideline and protocol development;

    • financial incentives, e.g. the UK Quality and Outcomes Framework (NHS 2009).

We excluded interventions that were intended to improve physical rehabilitation or knowledge of stroke in general, surgical interventions, and interventions testing new pharmacological therapies. We have also excluded exercise training programs for stroke or TIA patients as these are the subject of other Cochrane reviews (MacKay-Lyons 2010; Saunders 2009).

Types of outcome measures

Primary outcomes
  • Quantitative between-group differences (or target achievement) for blood pressure, lipid profile (total cholesterol, high density lipoprotein (HDL), low density lipoprotein (LDL), triglycerides), glycaemic control in diabetes mellitus (HbA1c), body mass index (BMI) or validated cardiovascular risk score.

  • Any indicator of patient adherence to secondary prevention medications, e.g. self-reported medication adherence or medication persistence, medication possession, individual patient data on prescriptions, pharmacy claims, electronic monitoring, drug tracers in blood or urine.

Secondary outcomes
  • Secondary cardiovascular events: stroke, myocardial infarction, or vascular death.

Search methods for identification of studies

See the 'Specialised register' section in the Cochrane Stroke Group module. We searched for trials in all languages and arranged for translation of relevant papers published in languages other than English.

Electronic searches

We searched the following electronic databases to identify relevant trials:

  • Cochrane Stroke Group Trials Register (April 2013);

  • Cochrane Effective Practice and Organisation of Care Group Trials Register (April 2013);

  • Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library (2013, Issue 3) (Appendix 1);

  • MEDLINE in Ovid (1950 to April 2013) (Appendix 2);

  • EMBASE in Ovid (1981 to April 2013) (Appendix 3);

  • CINAHL in EBSCO (1981 to April 2013) (Appendix 4);

  • AMED in Ovid (1985 to April 2013) (Appendix 5);

  • British Nursing Index (BNI) in Ovid (1985 to April 2013) (Appendix 6);

  • Web of Science Conference Proceedings Citation Index - Science (1970 to April 2013) (Appendix 7);

  • Index to UK theses (April 2013) (http://www.theses.com/);

  • BiblioMap (health promotion research) (April 2013) (http://eppi.ioe.ac.uk/webdatabases/Intro.aspx?ID=7).

In addition, we searched the following databases of ongoing trials and grants registers:

We developed the search strategy with the help of Brenda Thomas (Cochrane Stroke Group Trials Search Co-ordinator) and Dr Brian Buckley of the National University of Ireland (Buckley 2010). We adapted the MEDLINE search strategy (Appendix 2) to search other databases.

Searching other resources

We used the Science Citation Index Cited Reference Search to search for studies citing included trials. We also checked the reference lists of included trials, relevant systematic reviews and relevant meta-analyses. We contacted authors and trialists involved in included trials in order to facilitate identification of ongoing trials and unpublished studies.

Data collection and analysis

Selection of studies

Two review authors (KL and a second review author) independently assessed the titles, abstracts and keywords of all records retrieved from the electronic searches and excluded obviously irrelevant studies. We obtained the full texts of the remaining studies and two authors independently selected studies for inclusion based on the following criteria. The study:

  1. was an RCT;

  2. restricted participants to TIA or stroke patients, or reported outcomes separately for TIA or stroke patient subgroups;

  3. evaluated a stroke service intervention;

  4. stated or clearly implied that the intention of an intervention was to improve modifiable risk factor control;

  5. assessed one or more of the defined outcome measures;

  6. did not include physical rehabilitation programs, new pharmacological therapies, surgical procedures, exercise training programmes, or educational programmes intended to improve knowledge of stroke in general.

We resolved any disagreements regarding study eligibility by discussion between all review authors.

Data extraction and management

Two review authors independently extracted outcome data for each eligible trial using a pre-specified data extraction form. One review author extracted data for all eligible studies (KL) and a second review author (AKS, VH) independently repeated data extraction for each study. We resolved disagreements by discussion to reach consensus, with review authors referring back to the original article.

We recorded the following information for each study.

  1. General information: published or unpublished, title, authors, journal or source, publication date, country of origin, publication language.

  2. Study methods: unit of randomisation (and method), allocation concealment (and method), blinding (outcome assessors), validation of questionnaires.

  3. Participants: sampling (random or convenience), place of recruitment, total sample size, numbers randomised, inclusion criteria, exclusion criteria, demographic characteristics (age, gender, ethnicity, socioeconomic or sociodemographic status), disability (modified Rankin score, Barthel score), co-morbidities, similarity between groups at baseline, dropout and withdrawal rates.

  4. Intervention details: components, length, frequency, location, mode of delivery, personnel responsible for delivery, timing post-stroke, details of control protocol.

  5. Outcomes: pre-specified outcomes defined above, follow-up intervals from start of intervention, units of measurement, missing data.

  6. Results: results for pre-specified outcomes, number of participants assessed, method of analysis (intention-to-treat analysis, per protocol analysis).

  7. Intervention category: pre-specified in the review protocol.

Assessment of risk of bias in included studies

One review author (KL) assessed the quality of all eligible trials. The review author was not blinded to study details (for example author, journal, results) when assessing the methods. We assessed the quality of each RCT according to the Cochrane Collaboration's tool for assessing risk of bias. We excluded the blinding of participants and healthcare providers from the assessment since these criteria were unlikely to be met given the nature of the interventions under consideration. We assessed the risk of bias across six domains (sequence generation, allocation concealment, blinding of outcome assessment, incomplete outcome data, selective outcome reporting and other sources of bias) and summarised the results narratively. We contacted study authors to retrieve missing information. If study authors did not provide the requested information, we recorded the relevant items on the risk of bias assessment as 'unclear'. We have summarised the risk of bias according to the following criteria (Higgins 2011).

  1. Low risk of bias: low risk of bias for all domains.

  2. Unclear risk of bias: unclear risk of bias for one or more domains.

  3. High risk of bias: high risk of bias for one or more domains.

Measures of treatment effect

A mixture of continuous outcomes and dichotomous outcomes were reported by studies included in this review. Where possible, we have reported data in terms of mean difference (MD) and 95% confidence interval (CI) for continuous data. For dichotomous data, we have reported risk ratios (RR) or odds ratios (OR) and 95% CIs. If individual studies reported continuous and dichotomous data for the same outcome, we have included both variables in the review. We used RevMan 5.2 to carry out statistical analyses (RevMan 2012).

Unit of analysis issues

We analysed cluster-RCTs by reporting effect estimates from analyses that accounted for the cluster design. Where necessary, we calculated effective sample sizes for cluster-RCTs and combined these with parallel RCTs in meta-analyses (Higgins 2011a). Where studies included repeated measurements for participants at several time points, we have reported the outcomes recorded at the end of the study per protocol.

Dealing with missing data

We contacted authors of included studies in order to request missing data. We planned to perform sensitivity analyses, where appropriate, to explore the effects of including or excluding studies with incomplete data. We imputed missing summary data (for example standard deviations) based on recommendations in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). We used sensitivity analyses, where appropriate, to investigate the effect of entering assumed values.

Assessment of heterogeneity

We identified heterogeneity from forest plots using the Chi2 test and a significance level of alpha = 0.1. We also quantified heterogeneity using the I2 statistic, where I2 values of 50% or more indicate a substantial level of heterogeneity (Higgins 2002; Higgins 2003). Where appropriate, we assessed possible sources of heterogeneity using sensitivity analyses.

Assessment of reporting biases

We used funnel plots to assess publication bias. Visual inspection suggested no asymmetry.

Data synthesis

Studies identified by the review were heterogeneous in terms of interventions, settings, patient characteristics and outcome measurements. We used consensus methods to decide whether meta-analysis of study results was appropriate. Where there were sufficient comparable data, we combined results for each outcome to give an overall estimate of treatment effect. We conducted meta-analyses separately for each intervention category, in order to reduce clinical heterogeneity among the studies that were combined to produce pooled estimates. We pre-specified intervention categories in the review protocol. Where meta-analysis was not possible or appropriate, we have presented a qualitative synthesis of intervention effects.

Subgroup analysis and investigation of heterogeneity

We planned to analyse outcomes according to the following subgroups:

  • patient age (under 65 years, 65 years and over);

  • condition (ischaemic stroke, haemorrhagic stroke or TIA);

  • stroke severity (e.g. according to the National Institute of Health Stroke Scale (NIHSS)) or disability (e.g. according to the Barthel score or modified Rankin Score (mRS));

  • specific risk factor management strategy (e.g. blood pressure lowering interventions).

However, subgroup analyses were not possible since relevant data were not available from included studies.

Sensitivity analysis

We planned to carry out sensitivity analyses if the results of meta-analysis were significant. Where appropriate, we used the following sensitivity analyses to consider whether identified factors were associated with different effect sizes:

  1. repeating analyses excluding unpublished studies;

  2. repeating analyses excluding studies at high or unclear risk of bias;

  3. repeating analyses excluding very large studies to investigate the extent to which they dominated the results;

  4. repeating analyses using different measures of effect size (risk difference, odds ratio etc) and different statistical models (fixed-effect and random-effects models);

  5. repeating analyses to investigate whether the conclusions were affected by assumptions made when dealing with missing data.

Results

Description of studies

See: Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; Characteristics of ongoing studies

Results of the search

We carried out searches in April 2013 and identified a total of 7757 records after the removal of duplicates (see Figure 1). Title and abstract screening identified 82 studies (211 records) that were potentially eligible for this review.

Figure 1.

Study flow diagram.

Eight potentially eligible studies reported collective outcome data for participants with a broad range of cardiovascular diseases (Amariles 2012; Brotons 2011; Evans 2010; Goessens 2006; Ma 2009; Palanco 2011; Strandberg 2006; Vernooij 2012). We contacted study authors to request outcome data separately for stroke and TIA patients. Two study authors provided unpublished outcome data for stroke and TIA patients and we included these studies in the review (Brotons 2011; Evans 2010). Two study authors reported that separate outcome data for stroke and TIA patients were unavailable (Amariles 2012; Vernooij 2012). The remaining study authors did not respond to requests for additional data and we subsequently excluded four studies from the review (Goessens 2006; Ma 2009; Palanco 2011; Strandberg 2006).

A further 40 studies of potential relevance to this review were not associated with any manuscripts containing outcome data. We therefore attempted to obtain information about the status of these studies by emailing the main study contacts. Five authors supplied unpublished data and we included these studies in the review (Eames 2013; Flemming 2013; Lowrie 2010; O'Carroll 2011; Slark 2013). There were 14 completed trials for which further study information was unavailable (see Characteristics of studies awaiting classification). Correspondence with study contacts confirmed the status of 15 ongoing trials and a further six studies have been classified as ongoing due to their status on clinical trial registers (see Characteristics of ongoing studies).

We excluded a total of 21 studies from this review (see Characteristics of excluded studies for a list of excluded studies and the reasons for exclusion).

Included studies

Twenty-six RCTs met the inclusion criteria for this review, of which 23 used a parallel group design (Adie 2010; Allen 2002; Allen 2009; Boter 2004; Boysen 2009; Chanruengvanich 2006; Chiu 2008; Eames 2013; Ellis 2005; Evans 2010; Flemming 2013; Hornnes 2011; Joubert 2009; Kerry 2013; Kim 2013; Lowe 2007; Maasland 2007; MacKenzie 2013; Markle-Reid 2011; O'Carroll 2011; Slark 2013; Wang 2005; Welin 2010) and three used a cluster design (Brotons 2011; Johnston 2010; Lowrie 2010). Detailed information on each can be found in the Characteristics of included studies table.

Participants

The trials included a total of 8021 participants with cerebrovascular disease. The mean or median age of participants ranged from 60 to 73 years. Six studies included participants with a diagnosis of ischaemic stroke (Allen 2009; Boysen 2009; Chiu 2008; Johnston 2010; Kim 2013; Slark 2013) whereas three studies included participants with either ischaemic or haemorrhagic stroke (Lowe 2007; Welin 2010) or did not specify stroke subtype (Wang 2005). Ten trials included a broader range of participants with a diagnosis of either stroke or TIA (Allen 2002; Boter 2004; Eames 2013; Ellis 2005; Flemming 2013; Hornnes 2011; Joubert 2009; MacKenzie 2013; Markle-Reid 2011; O'Carroll 2011), with the proportion of TIA patients ranging from 1% (Eames 2013) to 46% (Flemming 2013). Four studies focused only on individuals with minor stroke or TIA (Adie 2010; Chanruengvanich 2006; Kerry 2013; Maasland 2007). Other studies included participants with a history of cardiovascular disease or elevated cardiovascular risk factors, and provided separate unpublished data for stroke and TIA patients (Brotons 2011; Evans 2010; Lowrie 2010).

Location

Four of the included trials were conducted in the USA (Allen 2002; Allen 2009; Flemming 2013; Johnston 2010), three in Canada (Evans 2010; MacKenzie 2013; Markle-Reid 2011), 12 in Europe (Adie 2010; Boter 2004; Brotons 2011; Ellis 2005; Hornnes 2011; Kerry 2013; Lowe 2007; Lowrie 2010; Maasland 2007; O'Carroll 2011; Slark 2013; Welin 2010), two in Australia (Eames 2013; Joubert 2009) and four in Asia (Chanruengvanich 2006; Chiu 2008; Kim 2013; Wang 2005). One study was a multi-centre trial conducted in five centres in China and Europe (Boysen 2009).

Setting

The majority of studies were set in primary care or community settings (Adie 2010; Allen 2002; Allen 2009; Boter 2004; Boysen 2009; Brotons 2011; Chanruengvanich 2006; Evans 2010; Hornnes 2011; Kerry 2013; Kim 2013; Lowrie 2010; MacKenzie 2013; Markle-Reid 2011; O'Carroll 2011; Wang 2005). Four studies were set in outpatient clinics (Chiu 2008; Ellis 2005; Flemming 2013; Welin 2010) and one was incorporated into a TIA service that provided screening and diagnostic work-up in a single day (Maasland 2007). Another two interventions were performed during hospitalisation for acute stroke (Johnston 2010; Slark 2013). Three further studies were initiated in the hospital setting (Eames 2013; Joubert 2009; Lowe 2007) with two subsequently continuing the intervention in the community (Eames 2013; Joubert 2009).

Interventions

See the Characteristics of included studies table for details of interventions (components, length, frequency).

Intervention categories

To facilitate analysis and interpretation of study results, we have described interventions according to categories pre-specified in the review protocol (educational and behavioural interventions for patients; educational and behavioural interventions for healthcare providers; organisational interventions as defined according to the taxonomy developed by Wensing 2006). We have summarised the categorisation of interventions in Table 1. All but three studies included educational or behavioural interventions for patients. Fifty per cent of the studies included integrated care services where patient care was delivered according to protocols for screening, education and treatment or monitoring. Thirty five per cent of studies included educational or behavioural interventions for healthcare providers, which usually involved the provision of guidelines or specification of individual patient targets. Less common intervention elements included revision of professional roles (changes in the tasks carried out by pharmacists), collaboration between multidisciplinary teams, knowledge management systems and quality management. No studies included financial interventions.

Table 1. Table of intervention categories
StudyEducational/behavioural interventions for patientsEducational/behavioural interventions for service providersOrganisational interventions

 

Predominant intervention category

Revision of professional rolesCollaboration between multidisciplinary teamsIntegrated care servicesKnowledge management systemsQuality managementFinancial incentives
Allen 2002XX XX   Organisational
Allen 2009XX XX   Organisational
Boter 2004X   X   Organisational
Brotons 2011XX  X   Organisational
Ellis 2005X   X   Organisational
Evans 2010X X X   Organisational
Flemming 2013XXXXX   Organisational
Hornnes 2011X   X   Organisational
Johnston 2010 X    X Organisational
Joubert 2009XX XXX  Organisational
Kerry 2013X    X  Organisational
Lowrie 2010 X    X Organisational
Markle-Reid 2011 X XXX  Organisational
Wang 2005X   X   Organisational
Welin 2010X   X   Organisational
Adie 2010X       Educational/behavioural intervention for patients
Boysen 2009X       Educational/behavioural intervention for patients
Chanruengvanich 2006X       Educational/behavioural intervention for patients
Chiu 2008X       Educational/behavioural intervention for patients
Eames 2013X       Educational/behavioural intervention for patients
Kim 2013X       Educational/behavioural intervention for patients
Lowe 2007X       Educational/behavioural intervention for patients
Maasland 2007X       Educational/behavioural intervention for patients
MacKenzie 2013XX      Educational/behavioural intervention for patients
O'Carroll 2011X       Educational/behavioural intervention for patients
Slark 2013X       Educational/behavioural intervention for patients

The majority of interventions were multifaceted and contained components that were associated with more than one category. However, in order to summarise evidence effectively we categorised interventions according to their predominant components. We decided final category assignments through discussion between review authors to reach consensus. We identified two broad categories of interventions: educational or behavioural interventions for patients and organisational interventions. Predominant intervention categories are highlighted in Table 1. Several interventions included organisational elements with varying amounts of education (directed towards patients or healthcare professionals). We have categorised these as predominantly organisational interventions since organisational elements were considered to have facilitated or permitted the delivery of education (for example patient education is often a component of integrated care services (Wensing 2006)). Conversely, we classified interventions as predominantly educational or behavioural interventions if they were implemented without changes to the organisation of patient care. A summary of the interventions in each category is provided below.

Educational or behavioural interventions for patients

Eleven studies involved educational and behavioural interventions for patients (Adie 2010; Boysen 2009; Chanruengvanich 2006; Chiu 2008; Eames 2013; Kim 2013; Lowe 2007; Maasland 2007; MacKenzie 2013; O'Carroll 2011; Slark 2013). None of these interventions incorporated organisational elements. The content of nine interventions was largely focused on modifiable risk factors for stroke (Adie 2010; Boysen 2009; Chanruengvanich 2006; Chiu 2008; Kim 2013; Maasland 2007; MacKenzie 2013; O'Carroll 2011; Slark 2013). Two interventions delivered education about secondary stroke prevention as part of broader stroke education programmes (Eames 2013; Lowe 2007).

Organisational interventions

Fifteen studies involved predominantly organisational interventions. Five interventions addressed secondary stroke prevention as part of a wider set of study aims encompassing post-stroke rehabilitation (interventions with a broad focus) (Allen 2002; Allen 2009; Boter 2004; Markle-Reid 2011; Welin 2010). Although these organisational interventions generally provided some patient education about secondary stroke prevention, this appeared to be delivered on only one occasion (Allen 2002; Allen 2009) or on an opportunistic basis (Boter 2004; Welin 2010). Conversely, secondary prevention was the main aim of the remaining 10 organisational interventions (interventions specifically targeting secondary prevention). Eight of these interventions included an element of patient education or behavioural counselling directed towards secondary stroke prevention (Brotons 2011; Ellis 2005; Evans 2010; Flemming 2013; Hornnes 2011; Joubert 2009; Kerry 2013; Wang 2005). The two remaining interventions did not specify the inclusion of patient education elements but instead directed secondary prevention education towards healthcare professionals (Johnston 2010; Lowrie 2010).

Timing

Fifteen studies recruited participants immediately following diagnosis of acute stoke or TIA. These studies initiated interventions prior to hospital discharge (Eames 2013; Johnston 2010; Joubert 2009; Lowe 2007; Maasland 2007; Slark 2013), within one week post-discharge (Allen 2002; Allen 2009; Boter 2004; Wang 2005), within one month post-discharge (Adie 2010; Hornnes 2011) or within three months post-discharge (Boysen 2009; Chanruengvanich 2006; Ellis 2005; Flemming 2013; O'Carroll 2011; Welin 2010). Six studies recruited participants from primary care, outpatient or community settings. Four of these studies initiated interventions within nine months (Kerry 2013), 12 months (Brotons 2011; Kim 2013) or 18 months (Markle-Reid 2011) of stroke or TIA diagnosis; one study initiated the intervention when participants had been attending an outpatient clinic for at least 12 months (Chiu 2008) and two studies did not specify intervention timing (Evans 2010; Lowrie 2010).

Four studies involved interventions that were delivered on a single occasion (Lowe 2007; Maasland 2007; Slark 2013) or on two occasions (O'Carroll 2011). The remaining studies implemented interventions over a time frame ranging from three to 36 months. The majority of interventions had a duration of between three and 12 months.

Outcomes

Details of outcomes are provided in the Characteristics of included studies table.

Excluded studies

Reasons for excluding studies are provided in the Characteristics of excluded studies table.

Risk of bias in included studies

We assessed the risk of bias according to the Cochrane Collaboration's tool for assessing risk of bias. For each study we extracted information about methods of randomisation and allocation concealment, blinding of outcome assessors, incomplete outcome data, selective outcome reporting and any other potential sources of bias. We have presented a detailed assessment of the risk of bias for individual studies in the Characteristics of included studies table. Summary assessments are shown in Figure 2.

Figure 2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Allocation

Inclusion criteria for this review required studies to be randomised. All but three studies reported adequate generation of allocation sequence. Two studies were reported as RCTs but did not provide details of randomisation methods (Chanruengvanich 2006; Chiu 2008). One study reported that participants were "randomly divided into intervention group (146 cases) and control group (52 cases)" (Wang 2005). Although the use of randomised methods can be inferred from this statement, the large imbalances in group size were not explained and this study has therefore been included but considered at high risk of bias.

Criteria for adequate allocation concealment were met by all but five studies. Three trials failing to report randomisation methods also provided no information about allocation concealment (Chanruengvanich 2006; Chiu 2008; Wang 2005). Another two studies with adequate sequence generation contained no information about allocation concealment (Allen 2002; Kim 2013).

Blinding

Ten studies reported blinding of outcome assessors for all outcomes (Allen 2009; Boter 2004; Boysen 2009; Chanruengvanich 2006; Eames 2013; Ellis 2005; Hornnes 2011; Kerry 2013; Markle-Reid 2011). A further three studies reported blinding during assessment of selected outcomes (Allen 2002; Johnston 2010; Welin 2010). There were thus 16 studies for which at least some data were collected by unblinded outcome assessors (Adie 2010; Allen 2002; Brotons 2011; Chiu 2008; Evans 2010; Flemming 2013; Johnston 2010; Joubert 2009; Kim 2013; Lowrie 2010; Maasland 2007; MacKenzie 2013; O'Carroll 2011; Slark 2013; Wang 2005; Welin 2010). Following consideration of these 16 studies, the review authors judged that non-blinding of outcome assessors was unlikely to affect the measurement of objective outcomes such as physiological data (for example blood pressure), information extracted from medical records or information measured through validated questionnaires (Adie 2010; Brotons 2011; Chiu 2008; Evans 2010; Flemming 2013; Kim 2013; Lowrie 2010; MacKenzie 2013; O'Carroll 2011; Slark 2013; Wang 2005; Welin 2010). However, it was unclear whether non-blinding could have affected outcomes obtained from participants via self-report (for example adherence to medication and self-reported cardiovascular events) (Flemming 2013; Joubert 2009; Kim 2013; Maasland 2007; MacKenzie 2013; Slark 2013).

Incomplete outcome data

The proportion of study participants completing follow-up ranged from 70% (Brotons 2011) to 100% (Adie 2010; MacKenzie 2013). Two studies did not report the proportion of participants completing follow-up (Chiu 2008; Wang 2005) and in one study information was only available for those participants with baseline and follow-up data (Lowrie 2010). Two studies reported no missing outcome data (Adie 2010; MacKenzie 2013). Nineteen studies reported the reasons for missing outcome data and the review authors judged that reasons were unlikely to be related to the study outcomes (Boter 2004; Boysen 2009; Brotons 2011; Chanruengvanich 2006; Eames 2013; Ellis 2005; Evans 2010; Flemming 2013; Hornnes 2011; Johnston 2010; Joubert 2009; Kerry 2013; Kim 2013; Lowe 2007; Maasland 2007; Markle-Reid 2011; O'Carroll 2011; Slark 2013; Welin 2010). The five remaining studies did not provide enough information about missing outcome data to permit judgement (Allen 2002; Allen 2009; Chiu 2008; Lowrie 2010; Wang 2005).

Selective reporting

Protocols were available for 18 of the 26 studies, and 17 appeared to be free of selective outcome reporting (Adie 2010; Allen 2009; Boter 2004; Boysen 2009; Brotons 2011; Chanruengvanich 2006; Eames 2013; Evans 2010; Flemming 2013; Hornnes 2011; Kerry 2013; Lowrie 2010; Maasland 2007; MacKenzie 2013; O'Carroll 2011; Slark 2013; Welin 2010). One study reported the primary outcomes in the pre-specified way, although some secondary outcomes were not reported (Johnston 2010). Protocols could not be obtained for the remaining eight studies (Allen 2002; Chiu 2008; Ellis 2005; Joubert 2009; Kim 2013; Lowe 2007; Markle-Reid 2011; Wang 2005).

Other potential sources of bias

We identified no other potential sources of bias.

Effects of interventions

Blood pressure

Twenty studies reported data on blood pressure, of which eight evaluated educational or behavioural interventions for participants (Adie 2010; Chanruengvanich 2006; Chiu 2008; Lowe 2007; Maasland 2007; MacKenzie 2013; O'Carroll 2011; Slark 2013) and 12 evaluated organisational interventions (Allen 2002; Allen 2009; Brotons 2011; Ellis 2005; Evans 2010; Flemming 2013; Hornnes 2011; Johnston 2010; Joubert 2009; Kerry 2013; Wang 2005; Welin 2010).

Educational and behavioural interventions for patients

Pooled data from eight studies (Adie 2010; Chanruengvanich 2006; Chiu 2008; Lowe 2007; Maasland 2007; MacKenzie 2013; O'Carroll 2011; Slark 2013) indicated that educational and behavioural interventions for participants were not associated with significant changes in mean systolic blood pressure (Analysis 1.1) or mean diastolic blood pressure (Analysis 1.2). However, the analyses included one large study that was independently associated with reductions in systolic and diastolic blood pressure (Chiu 2008) and consequently the pooled results were associated with a substantial level of statistical heterogeneity (Analysis 1.1; Analysis 1.2). Chiu 2008 only reported outcome data for a subgroup of participants with hypertension, so that baseline blood pressure levels were higher and therefore easier to improve upon. When this study was removed from the analyses, pooled data from the remaining studies did not indicate any intervention effects and statistical heterogeneity was reduced. The three studies that reported data on the achievement of blood pressure targets (< 140/90 mmHg or < 130/80 mmHg) indicated that educational and behavioural interventions for patients were not associated with a significant change in the proportion of participants who attained adequate blood pressure control (Adie 2010; Chiu 2008; MacKenzie 2013) (Analysis 1.3).

Organisational interventions

Pooled data from nine studies (Allen 2002; Brotons 2011; Ellis 2005; Evans 2010; Flemming 2013; Hornnes 2011; Joubert 2009; Kerry 2013; Welin 2010) indicated that organisational interventions were associated with a trend towards mean systolic blood pressure reduction, although this did not reach statistical significance (MD -2.57; 95% CI -5.46 to 0.31) (Analysis 2.1). Pooled data from seven studies (Brotons 2011; Ellis 2005; Evans 2010; Hornnes 2011; Joubert 2009; Kerry 2013; Welin 2010) indicated that organisational interventions were also associated with a non-significant trend towards mean diastolic blood pressure reduction (MD -0.90; 95% CI -2.49 to 0.68) (Analysis 2.2). The three studies (Ellis 2005; Flemming 2013; Joubert 2009) that were associated with the greatest reductions in systolic blood pressure (values ranging from -6.00 mmHg to -12.09 mmHg) combined integrated care with comprehensive patient education (involving promotion and tracking of adherence to medications and healthy lifestyle behaviours for secondary stroke prevention). These studies focused specifically on secondary stroke prevention and involved regular patient appointments (with a nurse or general practitioner (GP)) and review of multiple stroke risk factors (by a nurse case manager). Nurse case managers informed participants (Ellis 2005) or their GPs (Flemming 2013; Joubert 2009) if risk factors deviated from recommended targets (although nurses themselves did not influence medication prescribing). Consideration of other studies included in the meta-analysis of systolic blood pressure data showed that most evaluated interventions were not focused specifically on secondary stroke prevention due to wider study aims (Allen 2002; Welin 2010) or the inclusion of participants with a range of other cardiovascular diseases (Brotons 2011; Evans 2010). Two other studies that did focus specifically on secondary stroke prevention had a more narrow objective since they focused largely on blood pressure control rather than multiple risk factor reduction (Hornnes 2011; Kerry 2013).

Seven studies evaluating organisational interventions reported data on blood pressure control (Allen 2009; Brotons 2011; Flemming 2013; Hornnes 2011; Johnston 2010; Joubert 2009; Wang 2005). Blood pressure targets varied by study and according to participant co-morbidities, with the majority of studies specifying a blood pressure target of ≤ 140/90 mmHg or ≤ 130/80 mmHg for participants with diabetes. Pooled data indicated that organisational interventions were associated with a borderline significant increase in the proportion of patients who attained blood pressure targets (OR 1.24; 95% CI 0.94 to 1.64) (Analysis 2.3). However, one study included in the meta-analysis independently demonstrated that a larger proportion of intervention group participants attained a target systolic blood pressure of < 140 mmHg at follow-up when compared with control group participants (OR 2.19; 95% CI 1.16 to 4.15) (Joubert 2009).

Cholesterol

Total cholesterol

Thirteen studies reported cholesterol data, of which six included educational and behavioural interventions for patients (Adie 2010; Chanruengvanich 2006; Chiu 2008; Kim 2013; Maasland 2007; Slark 2013) and seven included predominantly organisational interventions (Allen 2009; Brotons 2011; Ellis 2005; Evans 2010; Joubert 2009; Lowrie 2010; Wang 2005).

Educational and behavioural interventions for patients

Pooled data from six studies indicated that educational and behavioural interventions for patients were not associated with changes in mean total cholesterol levels (Adie 2010; Chanruengvanich 2006; Chiu 2008; Kim 2013; Maasland 2007; Slark 2013) (Analysis 1.4). Only one study reported achievement of total cholesterol targets (total cholesterol ≤ 4 mmol/L) and found no significant difference between the intervention and control groups (Adie 2010) (Analysis 1.5).

Organisational interventions

Organisational interventions were not associated with changes in mean total cholesterol levels (Brotons 2011; Ellis 2005; Evans 2010; Joubert 2009; Lowrie 2010) (Analysis 2.4). Pooled data from four studies (Allen 2009; Joubert 2009; Lowrie 2010; Wang 2005) indicated that organisational interventions were not associated with changes in the achievement of total cholesterol targets, although the substantial level of statistical heterogeneity observed in this analysis (I2 = 85%) meant that results should be interpreted with caution. It should be noted in this meta-analysis that the outlying study with the largest effect sizewas considered at high risk of bias due to concerns about the adequacy of the randomisation procedures (Wang 2005). Furthermore, the authors of this trial did not specify risk factor targets, stating instead that the results of blood fat tests were either classified as qualified or disqualified. When we removed this study from the meta-analysis, we saw no changes in the achievement of total cholesterol targets (varying from < 4.0 to < 5.0 mmol/L) when the data from the remaining three studies were pooled, and statistical heterogeneity was reduced (I2 = 15%).

Low density lipoprotein (LDL)

Five studies reported LDL data, of which two evaluated educational and behavioural interventions for patients (Chiu 2008; Maasland 2007) and three evaluated organisational interventions (Brotons 2011; Evans 2010; Flemming 2013).

Educational and behavioural interventions for patients

Pooled data from two studies (Chiu 2008; Maasland 2007) indicated that educational and behavioural interventions for patients were not associated with changes in mean LDL levels, although a substantial level of statistical heterogeneity was observed in this analysis (I2 = 61%) (Analysis 1.6). One of the two studies included in the analysis was individually associated with improvements in LDL levels (MD -0.39 mmol/L; 95% CI -0.73 to -0.05). However, data were only presented for a subgroup of study participants with hypercholesterolaemia (i.e. those with the greatest potential for improvement) (Chiu 2008). Maasland 2007 reported significant reductions in LDL during the course of the study for both the intervention and control groups, with no significant differences between the groups. Only Chiu 2008 presented data on the achievement of LDL targets (LDL < 2.6 mmol/L or, if LDL not available, TC < 4.1 mmol/L) and no significant improvements were reported (Chiu 2008).

Organisational interventions

There were no significant effects of organisational interventions on mean LDL levels (Brotons 2011; Evans 2010; Flemming 2013) (Analysis 2.6). Flemming 2013 reported data on the achievement of LDL targets (fasting LDL <  2.6 mmol/L; optional goal < 1.8 mmol/L) and no significant differences were observed between the intervention and control groups (Analysis 2.7).

High density lipoprotein (HDL)

Four studies reported data on HDL, of which one evaluated an educational or behavioural intervention for patients (Chanruengvanich 2006) and three evaluated organisational interventions (Brotons 2011; Evans 2010; Flemming 2013).

Educational and behavioural interventions for patients

One study reported mean HDL levels and no significant intervention effect was observed (Chanruengvanich 2006) (Analysis 1.7). No studies reported data on HDL target achievement.

Organisational interventions

No significant intervention effects on mean HDL levels were observed when data from three studies were pooled (Brotons 2011; Evans 2010; Flemming 2013) (Analysis 2.8). Flemming 2013 reported data on HDL target achievement (fasting HDL > 1.0 mmol/L in men; > 1.3 mmol/L in women) and no significant differences were observed between the intervention and control groups.

Triglycerides

Six studies reported data on triglycerides. Three studies involved educational and behavioural interventions for patients (Chiu 2008; Kim 2013; Maasland 2007) and three involved organisational interventions (Brotons 2011; Evans 2010; Flemming 2013).

Educational and behavioural interventions for patients

There were no effects of patient educational and behavioural interventions on mean triglyceride levels (Chiu 2008; Kim 2013; Maasland 2007) (Analysis 1.8). No studies reported data on triglyceride target achievement.

Organisational interventions

There were no effects of organisational interventions on mean triglyceride levels (Brotons 2011; Evans 2010; Flemming 2013) (Analysis 2.10). Flemming 2013 reported data on the achievement of triglyceride targets (fasting triglycerides < 1.7 mmol/L) and no significant differences were observed between the intervention and control groups (Analysis 2.11).

HbA1c

Six studies reported data on HbA1c outcomes. These outcomes were not restricted to individuals with diabetes. One study evaluated a patient educational or behavioural intervention (Chiu 2008) while five studies evaluated organisational interventions (Allen 2009; Ellis 2005; Evans 2010; Flemming 2013; Wang 2005).

Educational and behavioural interventions for patients

No studies reported mean HbA1c levels. One study reported an outcome relating to HbA1c target achievement (HbA1c < 7% or fasting blood glucose < 7.0 mmol/L or random postprandial blood glucose < 11.1 mmol/L) and no significant differences between the intervention and control groups were observed (Chiu 2008) (Analysis 1.9).

Organisational interventions

Pooled data from three studies (Ellis 2005; Evans 2010; Flemming 2013) indicated that there were no effects of organisational interventions on mean HbA1c levels (Analysis 2.12). Data from three studies (Allen 2009; Flemming 2013; Wang 2005) could be pooled for achievement of HbA1c targets and no significant intervention effect was observed, although a considerable level of statistical heterogeneity was present (I2 = 96%) (Analysis 2.13). The outlying study in this analysis was considered to be at high risk of bias due to concerns about the adequacy of randomisation procedures (Wang 2005). When the data reported by Wang 2005 were removed from the meta-analysis, no significant differences in the achievement of HbA1c targets (varying from ≤ 6.5% to ≤ 7.0%) were observed between the intervention and control groups and statistical heterogeneity was reduced (I2 = 38%).

Body mass index (BMI)

Five studies reported BMI results, of which one evaluated a patient educational or behavioural intervention (Maasland 2007) and four evaluated organisational interventions (Brotons 2011; Flemming 2013; Joubert 2009; Wang 2005).

Educational and behavioural interventions for patients

One study reported data on mean BMI and no significant intervention effects were observed (Maasland 2007) (Analysis 1.10). No data were available on BMI target achievement.

Organisational interventions

Pooled data from three studies (Brotons 2011; Flemming 2013; Joubert 2009) indicated a trend towards a reduction in mean BMI levels (MD -0.68 kg/m2; 95% CI -1.46 to 0.11) (Analysis 2.14) although this did not reach statistical significance. Two studies measured the achievement of BMI targets (Flemming 2013; Wang 2005) (Analysis 2.15). In one study, the intervention was associated with improvements in BMI target achievement that bordered on statistical significance (OR 1.73; 95% CI 0.93 to 3.25). However, the study was considered at high risk of bias and the BMI target was not specified (Wang 2005). In the second study no significant differences in the achievement of the specified BMI target (< 25 kg/m2) were observed between the intervention and control groups (Flemming 2013).

Cardiovascular risk score

Organisational interventions

One study involving an organisational intervention reported data on the Framingham cardiovascular risk scores (Flemming 2013). The Framingham point score can be translated to provide an estimate of an individuals's 10-year risk of developing cardiovascular disease (Anderson 1991; Wilson 1998). Flemming 2013 reported that the intervention group demonstrated a significantly greater reduction in Framingham cardiovascular risk score when compared with the control group (MD -6.50; 95% CI -10.22 to -2.78) (Analysis 2.16), although it was not possible from the available study data to discern the magnitude of cardiovascular risk reduction.

Adherence to secondary prevention medications

Thirteen studies measured adherence to secondary prevention medications, of which six involved educational and behavioural interventions for patients (Eames 2013; Kim 2013; Maasland 2007; MacKenzie 2013; O'Carroll 2011; Slark 2013) and seven involved organisational interventions (Allen 2009; Boter 2004; Ellis 2005; Flemming 2013; Hornnes 2011; Johnston 2010; Joubert 2009).

Educational and behavioural interventions for patients

Six studies reported the effects of patient education on adherence to secondary prevention medications (Eames 2013; Kim 2013; Maasland 2007; MacKenzie 2013; O'Carroll 2011; Slark 2013). These data could not be pooled due to methodological heterogeneity (differences in outcome measurements). It should be noted that only Eames 2013 reported adequate blinding of outcome assessors. It was judged that non-blinding of outcome assessors may have influenced the data collected by four studies that assessed participants' self-reported medication adherence during face-to-face or telephone interviews with outcome assessors (Kim 2013; Maasland 2007; MacKenzie 2013; Slark 2013). However, it was judged that non-blinding of outcome assessors was unlikely to affect the adherence outcome data collected by O'Carroll 2011 since this was obtained using: (1) a previously validated questionnaire that was administered to patients, and (2) electronic pill containers. Similarly, it was judged that non-blinding of outcome assessors was unlikely to affect adherence outcome data obtained via a pharmacist review of prescription renewal patterns (MacKenzie 2013). Please see the Characteristics of included studies table for full evaluations of the risk of bias for the included studies.

The majority of studies measuring medication adherence outcomes found no significant differences between the intervention and control groups on any indicator of adherence. The studies by Eames 2013, Kim 2013, Maasland 2007 and Slark 2013 found no significant differences between the intervention and control groups in participants' self-reported adherence to secondary prevention medications. MacKenzie 2013 evaluated adherence to antihypertensive medication through participants' self-reported missed medication doses and a pharmacist-led review of participants' prescription renewal patterns: no significant differences in the number of missed pills or prescription renewals were observed between the intervention and control groups.

Only one study (O'Carroll 2011) reported significant differences in medication adherence between the participants in the intervention and control groups. O'Carroll 2011 conducted a repeated measures analysis of self-reported adherence to antihypertensive medication over a time frame of three months, assessed using the Medication Adherence Report Scale (Horne 2006), and reported a "significantly greater improvement in the intervention group" with regards to total medication adherence (P = 0.027), although the clinical implications of this effect could not be discerned from the available study data. Conversely, this study found no significant differences between groups in terms of non-intentional non-adherence and intentional non-adherence. O'Carroll 2011 also evaluated antihypertensive medication adherence by obtaining data from electronic pill containers to determine the "percentage of doses taken", "percentage of days on which the correct dose was taken" and "percentage of doses taken on schedule". The trialists reported that "the intervention group had higher adherence on all measures than the control group, although this was only significant for percentage doses taken on schedule (P = 0.048)”. More specifically, it was reported that the intervention group took 9.79% (SD 16.59) more doses on schedule when compared with the control group (O'Carroll 2011).

Organisational interventions

Four studies reported data on the proportion of participants compliant with warfarin therapy (Johnston 2010; Joubert 2009), anticoagulants (Allen 2009) or antithrombotic medication (Flemming 2013). Two studies measured compliance with antihypertensive medication (Hornnes 2011; Johnston 2010) and two measured compliance with statin medication (Flemming 2013; Johnston 2010). Two further studies reported the proportion of participants using secondary prevention medications as prescribed (Boter 2004; Ellis 2005). Medication compliance was either measured through participant self-report (Allen 2009; Boter 2004; Ellis 2005; Flemming 2013; Hornnes 2011; Joubert 2009) or through an analysis of filled prescription data and International Normalised ratio (INR) blood test records (Johnston 2010). Four of the five studies reported blinding of outcome assessors when collecting data on medication compliance (Boter 2004; Ellis 2005; Hornnes 2011; Johnston 2010; Joubert 2009), whereas Joubert et al did not provide any information regarding this (Joubert 2009). Data were not pooled since there was substantial heterogeneity in the methods that were used to obtain outcome data. No individual study reported significant differences in medication compliance between the intervention and control groups.

Secondary cardiovascular events

Eleven studies reported data on secondary cardiovascular events. Of these, eight evaluated organisational interventions (Allen 2002; Boysen 2009; Brotons 2011; Ellis 2005; Hornnes 2011; Markle-Reid 2011; Wang 2005; Welin 2010) and two evaluated educational and behavioural interventions for patients (MacKenzie 2013; Slark 2013).

Secondary stroke
Educational and behavioural interventions for patients

Two studies reported data on the proportion of participants who experienced at least one recurrent stroke (MacKenzie 2013; Slark 2013) (Analysis 1.11). Blinding of outcome assessors was not reported in either study. MacKenzie 2013 observed no significant difference in the number of recurrent strokes (assessed via a clinical record review) between the intervention and control groups. Slark 2013 observed no recurrent strokes in either the intervention group nor the control group for the duration of the study follow-up.

Organisational interventions

Four studies recorded the proportion of participants who experienced at least one recurrent stroke (Allen 2002; Kerry 2013; Wang 2005; Welin 2010). In three of these studies, data on the incidence of recurrent stroke were obtained by blinded outcome assessors via clinical record review (Allen 2002; Welin 2010) or administration of patient questionnaires (Kerry 2013). One study did not specify the method used to determine recurrent stroke events and no blinding of outcome assessors was reported (Wang 2005). Pooled data from all four studies suggested that organisational interventions were not associated with changes in the proportion of participants who experienced at least one recurrent stroke (Analysis 2.21). However, the analysis was associated with substantial statistical heterogeneity (I2 = 77%) due to an outlying study that was considered at high risk of bias (Wang 2005). When the study by Wang et al was removed from the analysis no intervention effect was observed among the three remaining studies.

Three studies provided data on the number of secondary strokes (Boysen 2009; Hornnes 2011; Markle-Reid 2011) that occurred during follow-up (measured at end of study per protocol). Data on secondary stroke events were obtained by blinded outcome assessors following a review of clinical records (Boysen 2009; Hornnes 2011) or face-to-face interviews with study participants (Markle-Reid 2011). Data were pooled as rate ratios (Deeks 2011) and no significant intervention effect was observed (Analysis 2.18). In one further study, an additional follow-up was conducted by blinded outcome assessors after a mean duration of 3.6 years (Ellis 2005). The study found that the number of self-reported secondary strokes were similar between groups (OR 0.72; 95% CI 0.12 to 4.32) (Analysis 2.19) but a significantly higher number of TIAs were reported in the intervention group (OR 2.49; 95% CI 1.18 to 5.22) (Analysis 2.20). The trialists state that "one patient in the intervention group reported 10 possible transient ischaemic attacks in the interval between studies" and that "no objective confirmation of these reported symptoms was possible" (Ellis 2005).

Secondary cardiovascular events
Organisational interventions

One study reported data on the proportion of participants who experienced at least one secondary cardiovascular event during follow-up (Brotons 2011). The data were collected by non-blinded outcome assessors following a review of clinical records and interviews with study participants. No significant intervention effect was observed (Analysis 2.21).

Two studies reported data on the number of secondary cardiovascular events that occurred before the end of the study per protocol (Flemming 2013; Hornnes 2011). Secondary events were documented by blinded (Hornnes 2011) or non-blinded (Flemming 2013) outcome assessors following clinical record review. Flemming 2013 observed no secondary cardiovascular events in the intervention group nor the control group. Similarly, Hornnes 2011 observed no significant differences between groups in the number of cardiovascular events (Analysis 2.22). In an additional follow-up interview with study participants, conducted by blinded outcome assessors after a mean duration of 3.6 years, Ellis 2005 observed a significantly higher number of self-reported cardiovascular events in the intervention group (OR 2.08; 95% CI 1.06 to 4.06) (Analysis 2.22). However, this finding was likely to reflect the increased number of TIAs observed (discussed in the previous section).

Myocardial infarction and ischaemic heart disease
Organisational interventions

Two studies reported the number of myocardial infarctions that occurred during follow-up (Boysen 2009; Hornnes 2011) and no significant intervention effect was seen (Analysis 2.23). Ellis 2005 observed no significant differences in the number of ischaemic heart disease events after a mean follow-up duration of 3.6 years (Analysis 2.23). Data were collected by blinded outcome assessors in all three studies following clinical record review (Boysen 2009; Hornnes 2011) or interviews with study participants (Ellis 2005).

Vascular death
Organisational interventions

Two studies reported data on vascular death that were obtained by blinded (Boysen 2009) and non-blinded (Brotons 2011) outcome assessors following clinical record review. Boysen 2009 observed no significant differences in the number of vascular deaths occurring in the intervention and control groups (Analysis 2.24). Similarly, Brotons 2011 observed no significant differences between intervention and control groups in the proportion of patients experiencing vascular death (Analysis 2.25).

Discussion

Summary of main results

This review produced mixed findings regarding the effectiveness of stroke service interventions for the secondary prevention of stroke. We performed meta-analyses (where appropriate) for the outcomes of blood pressure, lipid profile, HbA1c, BMI and recurrent cardiovascular events. We carried out a qualitative analysis for medication adherence outcomes.

Educational and behavioural interventions for patients were not generally associated with clear differences in any of the review outcomes, with only two exceptions. The pharmacist education program evaluated by Chiu 2008 was associated with significant improvements in mean systolic blood pressure, mean diastolic blood pressure and mean LDL levels. However, this study only presented data for a subgroup of participants with hypertension or hypercholesterolaemia who therefore had the greatest potential for improvement. It may be that educational interventions are more effective for participants with uncontrolled risk factors, and these participants could be targeted in future studies. The studies currently included in this review do not contain sufficient data to evaluate this. Additionally, O'Carroll 2011 was the only study included in the review to demonstrate significant differences between the intervention and control groups in adherence to secondary prevention medications. Although this study demonstrated improvements in two indicators of participant adherence to antihypertensive therapy, further research is required to determine the clinical impact of these findings.

The estimated effects of organisational interventions were compatible with improvements and no differences in mean systolic blood pressure (MD -2.57 mmHg; 95% CI -5.46 to 0.31), mean diastolic blood pressure (MD -0.90 mmHg; 95% CI -2.49 to 0.68) and blood pressure target achievement (OR 1.24; 95% CI 0.94 to 1.64). Blood pressure reductions of this magnitude may be associated with improved clinical outcomes since the Post-stroke Antihypertensive Treatment Study (PATS) provided evidence that blood pressure lowering of 5/2 mmHg with indapamide treatment reduced the incidence of secondary stroke by 29% (PATS Collaborating Group 1995). Additionally, a meta-analysis of individual patient data from 61 prospective studies concluded that a 2 mmHg reduction in usual systolic blood pressure could reduce stroke mortality by approximately 10% (Lewington 2002). In the meta-analysis of systolic blood pressure data presented in this review, the largest blood pressure reductions were associated with three interventions that contained common elements of integrated care and comprehensive patient education (involving promotion and tracking of behaviours for secondary stroke prevention). Another finding from this review is that the effects of organisational interventions (including common elements of integrated care and patient education) on BMI were compatible with a small clinical benefit (MD -0.68 kg/m2; 95% CI -1.46 to 0.11), although the imprecision of this estimate means that an absence of improvement could not be ruled out.

Overall completeness and applicability of evidence

A limitation of the included studies was the lack of consistently used outcome measures (for example some studies measured mean blood pressure whereas some measured target achievement with a variety of acceptable ranges). Combining all results in meta-analyses was therefore problematic. A second limitation was related to variations in study follow-up duration. This review pooled data collected at the end of the study per protocol. However, follow-up duration varied from three to 36 months. The results should therefore be interpreted with some caution since shorter studies may not allow enough time for the interventions to produce an impact on modifiable risk factors.

Intervention intensity is a poorly defined concept (Francis 2011; Greaves 2011), although it may incorporate factors such as "total number of contacts, total contact time, duration of the intervention and the number of behaviour change techniques used" (Greaves 2011). Differences in intervention intensity are considered to represent a potential source of heterogeneity in the context of complex interventions (O'Connor 2011). Correspondingly, the stroke service interventions included in this review were found to differ considerably in terms of aims (for example degree of focus on secondary stroke prevention), duration, components and mode of delivery. Pre-determined strategies for categorising intervention intensity may therefore facilitate the synthesis of future research findings (Baker 2011). However, systematic reviews of complex interventions have to date failed to demonstrate consistent associations between intervention intensity and effectiveness (Baker 2011; Beswick 2008; Greaves 2011). It has been asserted that methods for adequately describing and understanding complex interventions need to be "further developed and tested with the expectation that they will complement existing systematic review methodology" (Shepperd 2009).

Quality of the evidence

We analysed data from 26 trials involving 8021 participants with stroke or TIA. All 26 trials were conducted over the previous 10 years. One study did not report randomisation methods and had unequal group sizes, which raised questions about the validity of the findings (Wang 2005).

Potential biases in the review process

We have attempted to identify all RCTs of potential relevance to the review. In addition to a comprehensive search strategy, we attempted to contact the authors of all included trials in order to identify further published, unpublished and ongoing studies. Visual inspection of funnel plots did not raise any concerns regarding publication bias. We have included all eligible RCTs regardless of publication language (we arranged for the full translation of one study not published in English).

Agreements and disagreements with other studies or reviews

Buckley 2010 conducted a systematic review of the effects of service organisation interventions for the secondary prevention of ischaemic heart disease. Only interventions delivered in primary care were included. The review found that interventions involving certain elements (regular planned patient appointments, patient education and monitoring of medication and risk factors) may be associated with improved control of total cholesterol and blood pressure levels. However, the authors recommend that the results should be interpreted with caution due to significant clinical and statistical heterogeneity. In contrast to Buckley 2010, this systematic review included interventions that were not delivered in primary care, and therefore different types of interventions were included (for example implementation of discharge orders). However, the conclusions of this review are in accordance with Buckley 2010 since organisational interventions including elements of integrated care and patient education were associated with the greatest improvements in blood pressure control.

The positive effects of integrated care services in this review are also supported by the findings of another review of organisational interventions. Wensing 2006 reported that "integrated care services are particularly promising" when considering strategies to improve patient care. This is attributed to the typical multifaceted nature of these interventions. The authors suggested that the incorporation of numerous intervention components may "address a wide range of potential barriers for change". They also stated that "further work should focus on analysing the contributions of the specific components in integrated care services, to identify which particularly contribute to their effectiveness" (Wensing 2006).

Authors' conclusions

Implications for practice

This review has highlighted potential benefits of organisational interventions on mean systolic blood pressure, mean diastolic blood pressure, blood pressure target achievement and mean BMI (analysis of the effects of organisational interventions on these outcomes were imprecise and compatible with improvements and no differences). However, the conclusions of this review do not provide any clear evidence that organisational interventions can improve other modifiable risk factors (lipid profile, HbA1c, medication adherence) or reduce the incidence of recurrent cardiovascular events. Results also suggest that interventions including patient education alone are unlikely to lead to improvements in modifiable risk factor control or the prevention of recurrent cardiovascular events.

Implications for research

Future research should focus on the development of more effective interventions to translate secondary prevention recommendations into practice. The findings from this review suggest that educational and behavioural interventions for patients delivered in the absence of organisational change may not be the most effective means of achieving this. Instead, future research should evaluate the effects of organisational interventions (including common elements of integrated care and patient education). We have identified 21 ongoing studies and 14 studies that are awaiting assessment, so a future review update may lead to more robust conclusions.

Acknowledgements

We would like to thank all of the authors who kindly provided us with additional data about their studies. We are grateful to Brenda Thomas, Janette Camosso-Stefinovic and Dr Brian Buckley for their contributions to the search strategies. We are also grateful for the important contributions made by Phi-Anh Tonnu and Chokanan Thaitirarot to study selection and data collection.

Data and analyses

Download statistical data

Comparison 1. Educational or behavioural interventions for patients versus usual care
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Mean systolic blood pressure8621Mean Difference (IV, Random, 95% CI)-2.06 [-7.40, 3.28]
2 Mean diastolic blood pressure8621Mean Difference (IV, Random, 95% CI)0.35 [-2.38, 3.08]
3 Blood pressure target achievement3266Odds Ratio (M-H, Random, 95% CI)1.34 [0.70, 2.59]
4 Mean total cholesterol6405Mean Difference (IV, Random, 95% CI)0.17 [-0.31, 0.65]
5 Total cholesterol target achievement156Odds Ratio (M-H, Random, 95% CI)0.56 [0.19, 1.68]
6 Mean low density lipoprotein2139Mean Difference (IV, Random, 95% CI)-0.15 [-0.76, 0.47]
7 Mean high density lipoprotein162Mean Difference (IV, Random, 95% CI)0.0 [-0.15, 0.15]
8 Mean triglycerides3182Mean Difference (IV, Random, 95% CI)-0.08 [-0.38, 0.23]
9 HbA1C target achievement167Odds Ratio (M-H, Random, 95% CI)0.65 [0.25, 1.75]
10 Mean BMI157Mean Difference (IV, Random, 95% CI)-0.24 [-0.80, 0.32]
11 Proporation of participants with secondary stroke156Odds Ratio (M-H, Random, 95% CI)5.0 [0.23, 109.01]
Analysis 1.1.

Comparison 1 Educational or behavioural interventions for patients versus usual care, Outcome 1 Mean systolic blood pressure.

Analysis 1.2.

Comparison 1 Educational or behavioural interventions for patients versus usual care, Outcome 2 Mean diastolic blood pressure.

Analysis 1.3.

Comparison 1 Educational or behavioural interventions for patients versus usual care, Outcome 3 Blood pressure target achievement.

Analysis 1.4.

Comparison 1 Educational or behavioural interventions for patients versus usual care, Outcome 4 Mean total cholesterol.

Analysis 1.5.

Comparison 1 Educational or behavioural interventions for patients versus usual care, Outcome 5 Total cholesterol target achievement.

Analysis 1.6.

Comparison 1 Educational or behavioural interventions for patients versus usual care, Outcome 6 Mean low density lipoprotein.

Analysis 1.7.

Comparison 1 Educational or behavioural interventions for patients versus usual care, Outcome 7 Mean high density lipoprotein.

Analysis 1.8.

Comparison 1 Educational or behavioural interventions for patients versus usual care, Outcome 8 Mean triglycerides.

Analysis 1.9.

Comparison 1 Educational or behavioural interventions for patients versus usual care, Outcome 9 HbA1C target achievement.

Analysis 1.10.

Comparison 1 Educational or behavioural interventions for patients versus usual care, Outcome 10 Mean BMI.

Analysis 1.11.

Comparison 1 Educational or behavioural interventions for patients versus usual care, Outcome 11 Proporation of participants with secondary stroke.

Comparison 2. Organisational interventions versus usual care
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Mean systolic blood pressure91514Mean Difference (IV, Random, 95% CI)-2.57 [-5.46, 0.31]
2 Mean diastolic blood pressure71384Mean Difference (IV, Random, 95% CI)-0.90 [-2.49, 0.68]
3 Blood pressure target achievement71346Odds Ratio (M-H, Random, 95% CI)1.24 [0.94, 1.64]
4 Mean total cholesterol5648Mean Difference (IV, Random, 95% CI)-0.08 [-0.37, 0.20]
5 Total cholesterol target achievement4757Odds Ratio (M-H, Random, 95% CI)1.56 [0.67, 3.66]
6 Mean low density lipoprotein3245Mean Difference (IV, Random, 95% CI)-0.42 [-1.12, 0.28]
7 Low density lipoprotein target achievement136Odds Ratio (M-H, Random, 95% CI)1.0 [0.27, 3.72]
8 Mean high density lipoprotein3247Mean Difference (IV, Random, 95% CI)0.06 [-0.06, 0.19]
9 High density lipoprotein target achievement136Odds Ratio (M-H, Random, 95% CI)1.27 [0.33, 4.97]
10 Mean triglycerides3246Mean Difference (IV, Random, 95% CI)-0.05 [-0.26, 0.16]
11 Triglyceride target achievement136Odds Ratio (M-H, Random, 95% CI)0.25 [0.05, 1.18]
12 Mean HbA1C3234Mean Difference (IV, Random, 95% CI)-0.13 [-1.33, 1.08]
13 HbA1C target achievement3553Odds Ratio (M-H, Random, 95% CI)3.93 [0.30, 51.33]
14 Mean BMI3423Mean Difference (IV, Random, 95% CI)-0.68 [-1.46, 0.11]
15 BMI target achievement2234Odds Ratio (M-H, Random, 95% CI)1.73 [0.93, 3.25]
16 Mean Framingham cardiovascular risk score136Mean Difference (IV, Random, 95% CI)-6.5 [-10.22, -2.78]
17 Proportion of participants with secondary stroke or TIA4791Odds Ratio (M-H, Random, 95% CI)0.66 [0.23, 1.86]
18 Number of secondary strokes3 Odds Ratio (Random, 95% CI)1.19 [0.70, 2.03]
19 Number of secondary strokes1 Odds Ratio (Random, 95% CI)0.72 [0.12, 4.32]
20 Number of secondary TIAs1 Odds Ratio (Random, 95% CI)2.49 [1.18, 5.22]
21 Proportion of participants with secondary cardiovascular events1324Odds Ratio (M-H, Random, 95% CI)1.48 [0.79, 2.77]
22 Number of secondary cardiovascular events2 Odds Ratio (Random, 95% CI)1.24 [0.40, 3.84]
23 Number of myocardial infarctions3 Odds Ratio (Random, 95% CI)0.47 [0.15, 1.48]
24 Number of vascular deaths1 Odds Ratio (Random, 95% CI)0.75 [0.17, 3.35]
25 Proportion of participants with vascular death1324Odds Ratio (M-H, Random, 95% CI)1.75 [0.41, 7.46]
Analysis 2.1.

Comparison 2 Organisational interventions versus usual care, Outcome 1 Mean systolic blood pressure.

Analysis 2.2.

Comparison 2 Organisational interventions versus usual care, Outcome 2 Mean diastolic blood pressure.

Analysis 2.3.

Comparison 2 Organisational interventions versus usual care, Outcome 3 Blood pressure target achievement.

Analysis 2.4.

Comparison 2 Organisational interventions versus usual care, Outcome 4 Mean total cholesterol.

Analysis 2.5.

Comparison 2 Organisational interventions versus usual care, Outcome 5 Total cholesterol target achievement.

Analysis 2.6.

Comparison 2 Organisational interventions versus usual care, Outcome 6 Mean low density lipoprotein.

Analysis 2.7.

Comparison 2 Organisational interventions versus usual care, Outcome 7 Low density lipoprotein target achievement.

Analysis 2.8.

Comparison 2 Organisational interventions versus usual care, Outcome 8 Mean high density lipoprotein.

Analysis 2.9.

Comparison 2 Organisational interventions versus usual care, Outcome 9 High density lipoprotein target achievement.

Analysis 2.10.

Comparison 2 Organisational interventions versus usual care, Outcome 10 Mean triglycerides.

Analysis 2.11.

Comparison 2 Organisational interventions versus usual care, Outcome 11 Triglyceride target achievement.

Analysis 2.12.

Comparison 2 Organisational interventions versus usual care, Outcome 12 Mean HbA1C.

Analysis 2.13.

Comparison 2 Organisational interventions versus usual care, Outcome 13 HbA1C target achievement.

Analysis 2.14.

Comparison 2 Organisational interventions versus usual care, Outcome 14 Mean BMI.

Analysis 2.15.

Comparison 2 Organisational interventions versus usual care, Outcome 15 BMI target achievement.

Analysis 2.16.

Comparison 2 Organisational interventions versus usual care, Outcome 16 Mean Framingham cardiovascular risk score.

Analysis 2.17.

Comparison 2 Organisational interventions versus usual care, Outcome 17 Proportion of participants with secondary stroke or TIA.

Analysis 2.18.

Comparison 2 Organisational interventions versus usual care, Outcome 18 Number of secondary strokes.

Analysis 2.19.

Comparison 2 Organisational interventions versus usual care, Outcome 19 Number of secondary strokes.

Analysis 2.20.

Comparison 2 Organisational interventions versus usual care, Outcome 20 Number of secondary TIAs.

Analysis 2.21.

Comparison 2 Organisational interventions versus usual care, Outcome 21 Proportion of participants with secondary cardiovascular events.

Analysis 2.22.

Comparison 2 Organisational interventions versus usual care, Outcome 22 Number of secondary cardiovascular events.

Analysis 2.23.

Comparison 2 Organisational interventions versus usual care, Outcome 23 Number of myocardial infarctions.

Analysis 2.24.

Comparison 2 Organisational interventions versus usual care, Outcome 24 Number of vascular deaths.

Analysis 2.25.

Comparison 2 Organisational interventions versus usual care, Outcome 25 Proportion of participants with vascular death.

Appendices

Appendix 1. CENTRAL search strategy

1. MeSH descriptor Cerebrovascular Disorders explode tree 1

2. ((cva or stroke or poststroke or (post NEXT stroke) or (transient NEXT isch*mic NEXT attack) or TIA or ministroke or (mini NEXT stroke)) NEAR/6 (people or patient or outpatient or adult or survivor or victim or individual or client or population or community or subject)):ti,ab,kw

3. (cerebrovascular* or cerebral vascular):ti,ab,kw

4. (cerebral or cerebellar or brain* or vertebrobasilar):ti,ab,kw

5. (infarct* or isch*mi* or thrombo* or apoplexy or emboli*):ti,ab,kw

6. (#4 AND #5)

7. (cerebral or intracerebral or intracranial or brain* or cerebellar or subarachnoid):ti,ab,kw

8. (accident* or h*morrhag*):ti,ab,kw

9. (#7 AND #8)

10. (#1 OR #2 OR #3 OR #6 OR #9)

11. MeSH descriptor Child, this term only

12. MeSH descriptor Infant explode all trees

13. MeSH descriptor Pediatrics explode all trees

14. (child* or neonat* or p?ediatric* or infant*):ti,ab,kw

15. (#11 OR #12 OR #13 OR #14)

16. MeSH descriptor Patient Care Management, this term only

17. MeSH descriptor Comprehensive Health Care, this term only

18. MeSH descriptor Nursing Process, this term only

19. MeSH descriptor Nursing Assessment explode all trees

20. MeSH descriptor Patient Care Planning, this term only

21. MeSH descriptor Case Management, this term only

22. MeSH descriptor Delivery of Health Care, this term only

23. MeSH descriptor Delivery of Health Care, Integrated, this term only

24. MeSH descriptor Managed Care Programs explode tree 2

25. MeSH descriptor Disease Management, this term only

26. MeSH descriptor Patient Care Team explode all trees

27. MeSH descriptor Primary Health Care explode all trees

28. MeSH descriptor Reminder Systems, this term only

29. MeSH descriptor Guideline Adherence, this term only

30. MeSH descriptor Home Care Services, this term only

31. MeSH descriptor Home Nursing, this term only

32. MeSH descriptor Nursing Services explode all trees

33. MeSH descriptor Professional Role explode all trees

34. MeSH descriptor Community Health Services, this term only

35. MeSH descriptor Medical Records, this term only

36. MeSH descriptor Medical Records Systems, Computerized, this term only

37. MeSH descriptor Patient Education as Topic, this term only

38. MeSH descriptor Patient Compliance explode tree 1

39. MeSH descriptor Life Style, this term only

40. MeSH descriptor Health Promotion, this term only

41. MeSH descriptor Health Services Administration, this term only

42. MeSH descriptor Education, Medical, Continuing, this term only

43. MeSH descriptor Marketing of Health Services, this term only

44. MeSH descriptor Patient Participation, this term only

45. MeSH descriptor Quality of Health Care, this term only

46. MeSH descriptor Quality Assurance, Health Care, this term only

47. MeSH descriptor Exercise, this term only

48. MeSH descriptor Physical Fitness, this term only

49. MeSH descriptor Smoking Cessation, this term only

50. MeSH descriptor Diet, this term only

51. MeSH descriptor Diet, Fat-Restricted, this term only

52. MeSH descriptor Diet, Carbohydrate-Restricted, this term only

53. MeSH descriptor Diet, Reducing, this term only

54. MeSH descriptor Caloric Restriction, this term only

55. MeSH descriptor Alcohol Drinking, this term only with qualifier: PC

56. MeSH descriptor Health Education, this term only

57. MeSH descriptor Community Health Planning, this term only

58. MeSH descriptor Communication, this term only

59. MeSH descriptor Communication Barriers, this term only

60. MeSH descriptor Information Dissemination, this term only

61. MeSH descriptor Interdisciplinary Communication, this term only

62. MeSH descriptor Nurse Clinicians, this term only

63. MeSH descriptor Nurse Practitioners, this term only

64. MeSH descriptor Risk Reduction Behavior, this term only

65. MeSH descriptor Pamphlets, this term only

66. MeSH descriptor Health Behavior, this term only

67. MeSH descriptor Health Knowledge, Attitudes, Practice, this term only

68. MeSH descriptor Secondary Prevention, this term only

69. MeSH descriptor Preventive Health Services, this term only

70. (manag* NEAR/3 care):ti,ab,kw

71. (management NEAR/3 program*):ti,ab,kw

72. (case NEAR/3 manag*):ti,ab,kw

73. (patient NEAR/3 management):ti,ab,kw

74. (home NEAR/3 intervention):ti,ab,kw

75. (home NEXT visit*):ti,ab,kw

76. (discharg* NEAR/3 program*):ti,ab,kw

77. (practice NEXT guideline*):ti,ab,kw

78. (discharg* NEAR/3 plan*):ti,ab,kw

79. (comprehensive NEAR/3 care):ti,ab,kw

80. (treatment NEAR/3 plan*):ti,ab,kw

81. (nurse NEAR/3 led):ti,ab,kw

82. (disease NEXT management):ti,ab,kw

83. (multi NEXT disciplin*):ti,ab,kw

84. (multidisciplin*):ti,ab,kw

85. (secondary NEXT prevention NEXT clinic):ti,ab,kw

86. (reminder):ti,ab,kw

87. (recall*):ti,ab,kw

88. (nurse NEAR/3 clinic):ti,ab,kw

89. (secondary NEXT prevention NEAR/3 intervention):ti,ab,kw

90. (secondary NEXT prevention NEAR/3 program*):ti,ab,kw

91. MeSH descriptor Appointments and Schedules, this term only

92. (appointment):ti,ab,kw

93. (outreach NEXT nurs*):ti,ab,kw

94. (outreach NEXT visit*):ti,ab,kw

95. (lifestyle NEAR/3 intervention*):ti,ab,kw

96. (physical NEXT (activity or exercise)):ti,ab,kw

97. (aerobic):ti,ab,kw

98. (fitness):ti,ab,kw

99. (exercise NEAR/3 (train* or intervention or program* or activity or regim*)):ti,ab,kw

100. (nurs* NEXT intervention*):ti,ab,kw

101. (education* NEXT program*):ti,ab,kw

102. ((risk NEXT factor*) NEAR/5 (modif* or reduc* or manage* or monitor* or self-manage*)):ti,ab,kw

103. (#16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38 OR #39 OR #40 OR #41 OR #42 OR #43 OR #44 OR #45 OR #46 OR #47 OR #48 OR #49 OR #50 OR #51 OR #52 OR #53 OR #54 OR #55 OR #56 OR #57 OR #58 OR #59 OR #60 OR #61 OR #62 OR #63 OR #64 OR #65 OR #66 OR #67 OR #68 OR #69 OR #70 OR #71 OR #72 OR #73 OR #74 OR #75 OR #76 OR #77 OR #78 OR #79 OR #80 OR #81 OR #82 OR #83 OR #84 OR #85 OR #86 OR #87 OR #88 OR #89 OR #90 OR #91 OR #92 OR #93 OR #94 OR #95 OR #96 OR #97 OR #98 OR #99 OR #100 OR #101 OR #102)

104. (#10 NOT #15)

105. (#103 AND #104)

Appendix 2. MEDLINE (Ovid) search strategy

1. exp Cerebrovascular Disorders/

2. ((cva$ or stroke$ or poststroke$ or post-stroke$ or post stroke$ or transient isch?emic attack$ or TIA$ or ministroke$ or mini-stroke$ or mini stroke$) adj6 (people or patient$ or inpatient$ or outpatient$ or adult$ or survivor$ or victim$ or individual$ or client$ or population$ or community or subject$)).tw.

3. (cerebrovascular$ or cerebral vascular).tw.

4. (cerebral or cerebellar or brain$ or vertebrobasilar).tw.

5. (infarct$ or isch?emi$ or thrombo$ or apoplexy or emboli$).tw.

6. 4 and 5

7. (cerebral or intracerebral or intracranial or brain$ or cerebellar or subarachnoid).tw.

8. (accident$ or h?emorrhag$).tw.

9. 7 and 8

10. 1 or 2 or 3 or 6 or 9

11. Child/

12. exp Infant/

13. exp pediatrics/

14. (child$ or neonat$ or p?ediatric$ or infant$).tw.

15. 11 or 12 or 13 or 14

16. 10 not 15

17. Patient Care Management/

18. Comprehensive Health Care/

19. Nursing Process/

20. exp Nursing Assessment/

21. Patient Care Planning/

22. Case Management/

23. delivery of health care/

24. "Delivery of Health Care, Integrated"/

25. exp Managed Care Programs/

26. Disease Management/

27. exp Patient Care Team/

28. exp Primary Health Care/

29. Reminder Systems/

30. Guideline Adherence/

31. Home Care Services/

32. Home Nursing/

33. exp Nursing Services/

34. exp Professional Role/

35. Community Health Services/

36. Medical Records/ or Medical Records Systems, Computerized/

37. Patient Education as Topic/

38. exp Patient Compliance/

39. Life Style/

40. Health Promotion/

41. Health Services Administration/

42. Education, Medical, Continuing/

43. Marketing of Health Services/

44. Patient Participation/

45. Quality of Health Care/

46. Quality Assurance, Health Care/

47. Exercise/ or Physical Fitness/

48. Smoking Cessation/

49. Diet/ or Diet, Fat-Restricted/ or Diet, Carbohydrate-Restricted/ or Diet, Reducing/ or Caloric Restriction/

50. Alcohol, Drinking/pc

51. Health Education/

52. Community Health Planning/

53. Communication/ or Communication Barriers/ or Information Dissemination/ or Interdisciplinary Communication/

54. Nurse Clinicians/

55. Nurse Practitioners/

56. Risk Reduction Behavior/

57. Pamphlets/

58. Health Behavior/

59. Health Knowledge, Attitudes, Practice/

60. Secondary Prevention/

61. Preventive Health Services/

62. (manag$ adj3 care).tw.

63. (management adj3 program$).tw.

64. (case adj3 manag$).tw.

65. (patient adj3 management).tw.

66. (home adj3 intervention$).tw.

67. (home adj visit$).tw.

68. (discharg$ adj3 program$).tw.

69. (practice adj guideline$).tw.

70. (discharg$ adj3 plan$).tw.

71. (comprehensive adj3 care).tw.

72. (treatment adj3 plan$).tw.

73. (nurse$ adj3 led).tw.

74. (disease adj management).tw.

75. multi-disciplin$.tw.

76. multidisciplin$.tw.

77. secondary prevention clinic$.tw.

78. reminder$.tw.

79. recall$.tw.

80. (nurse adj3 clinic$).tw.

81. (secondary prevention adj3 intervention$).tw.

82. (secondary prevention adj3 program$).tw.

83. "Appointments and Schedules"/

84. appointment$.tw.

85. (outreach adj nurs$).tw.

86. (outreach adj visit$).tw.

87. (lifestyle adj3 intervention$).tw.

88. (nurs$ adj intervention$).tw.

89. (education$ adj program$).tw.

90. (physical adj (activit$ or exercise$)).tw.

91. (exercise adj3 (train$ or intervention$ or program$ or activit$ or regim$)).tw.

92. aerobic.tw.

93. fitness.tw.

94. (risk factor$ adj5 (modif$ or reduc$ or manage$ or monitor$ or self-manage$)).tw.

95. 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 38 or 39 or 40 or 41 or 42 or 43 or 44 or 45 or 46 or 47 or 48 or 49 or 50 or 51 or 52 or 53 or 54 or 55 or 56 or 57 or 58 or 59 or 60 or 61 or 62 or 63 or 64 or 65 or 66 or 67 or 68 or 69 or 70 or 71 or 72 or 73 or 74 or 75 or 76 or 77 or 78 or 79 or 80 or 81 or 82 or 83 or 84 or 85 or 86 or 87 or 88 or 89 or 90 or 91 or 92 or 93 or 94

96. randomized controlled trial.pt.

97. controlled clinical trial.pt.

98. randomized.ab.

99. randomised.ab.

100. placebo.ab.

101. clinical trials as topic.sh.

102. randomly.ab.

103. trial.ti.

104. 96 or 97 or 98 or 99 or 100 or 101 or 102 or 103

105. 16 and 95 and 104

106. exp animals/ not humans.sh.

107. 105 not 106

Appendix 3. EMBASE (Ovid) search strategy

1. exp *cerebrovascular disease/

2. ((cva$ or stroke$ or poststroke$ or post-stroke$ or post stroke$ or transient isch?emic attack$ or TIA$ or ministroke$ or mini-stroke$ or mini stroke$) adj6 (people or patient$ or inpatient$ or outpatient$ or adult$ or survivor$ or victim$ or individual$ or client$ or population$ or community or subject$)).tw.

3. (cerebrovascular$ or cerebral vascular).tw.

4. (cerebral or cerebellar or brain$ or vertebrobasilar).tw.

5. (infarct$ or isch?emi$ or thrombo$ or apoplexy or emboli$).tw.

6. 4 and 5

7. (cerebral or intracerebral or intracranial or brain$ or cerebellar or subarachnoid).tw.

8. (accident$ or h?emorrhag$).tw.

9. 7 and 8

10. 1 or 2 or 3 or 6 or 9

11. child/

12. exp infant/

13. exp pediatrics/

14. (child$ or neonat$ or p?ediatric$ or infant$).tw.

15. 11 or 12 or 13 or 14

16. 10 not 15

17. (child$ or neonate$ or p?ediatric$ or infant$).tw.

18. 16 not 17

19. patient care planning/

20. case management/

21. health care delivery/

22. integrated health care system/

23. disease management/

24. reminder system/

25. *medical record/

26. health education/

27. patient education/

28. *patient compliance/

29. lifestyle modification/ or lifestyle/

30. health promotion/

31. medical education/

32. patient participation/

33. *exercise/ or aerobic exercise/ or fitness/ or *physical activity/

34. *smoking cessation/

35. *diet/ or low calory diet/ or low carbohydrate diet/ or low fat diet/ or diet restriction/

36. alcohol consumption/

37. health care planning/

38. interdisciplinary communication/

39. information dissemination/

40. risk reduction/

41. health behavior/

42. secondary prevention/

43. preventive medicine/

44. risk management/

45. medical specialist/

46. medical information/

47. (manag$ adj3 care).tw.

48. (management adj3 program$).tw.

49. (case adj3 manag$).tw.

50. (patient adj3 management).tw.

51. (home adj3 intervention$).tw.

52. (home adj visit$).tw.

53. (discharg$ adj3 program$).tw.

54. (practice adj guideline$).tw.

55. (discharg$ adj3 plan$).tw.

56. (comprehensive adj3 care).tw.

57. (treatment adj3 plan$).tw.

58. (nurse$ adj3 led).tw.

59. (disease adj management).tw.

60. multi-disciplin$.tw.

61. multidisciplin$.tw.

62. secondary prevention clinic$.tw.

63. reminder$.tw.

64. recall$.tw.

65. (nurse adj3 clinic$).tw.

66. (secondary prevention adj3 intervention$).tw.

67. (secondary prevention adj3 program$).tw.

68. appointment$.tw.

69. (outreach adj nurs$).tw.

70. (outreach adj visit$).tw.

71. (lifestyle adj3 intervention$).tw.

72. (nurs$ adj intervention$).tw.

73. (education$ adj program$).tw.

74. (physical adj (activit$ or exercise$)).tw.

75. (exercise adj3 (train$ or intervention$ or program$ or activit$ or regim$)).tw.

76. aerobic.tw.

77. fitness.tw.

78. (risk factor$ adj5 (modif$ or reduc$ or manage$ or monitor$ or self-manage$)).tw.

79. 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 38 or 39 or 40 or 41 or 42 or 43 or 44 or 45 or 46 or 47 or 48 or 49 or 50 or 51 or 52 or 53 or 54 or 55 or 56 or 57 or 58 or 59 or 60 or 61 or 62 or 63 or 64 or 65 or 66 or 67 or 68 or 69 or 70 or 71 or 72 or 73 or 74 or 75 or 76 or 77 or 78

80. random$.tw.

81. placebo$.tw.

82. placebo/

83. double-blind$.tw.

84. randomized controlled trial/

85. randomization/

86. double blind procedure/

87. single blind procedure/

88. triple blind procedure/

89. 80 or 81 or 82 or 83 or 84 or 85 or 86 or 87 or 88

90. 18 and 79 and 89

Appendix 4. CINAHL (EBSCO) search strategy

1. exp CEREBRAL ISCHEMIA/ OR exp INTRACRANIAL HEMORRHAGE/ OR *STROKE/

2. *STROKE PATIENTS/

3. (cva* OR stroke* OR poststroke* OR post-stroke* OR "transient ischemic attack*" OR "transient ischaemic attack*" OR TIA* OR ministroke* OR mini-stroke*).ti,ab

4. (cerebrovascular* OR "cerebral vascular").ti,ab

5. (cerebral OR cerebellar OR brain* OR vertebrobasilar).ti,ab

6. (infarct* OR ischemi* OR ischaemi* OR thrombo* OR apoplexy OR emboli*).ti,ab

7. 5 AND 6

8. (cerebral OR intracerebral OR intracranial OR brain* OR cerebellar OR subarachnoid).ti,ab

9. (accident* OR hemorrhag* OR haemorrhag*).ti,ab

10. 8 AND 9

11. 1 OR 2 OR 3 OR 4 OR 7 OR 10

12. NURSING INTERVENTIONS/

13. NURSING PRACTICE/

14. ADVANCED NURSING PRACTICE/

15. HEALTH CARE DELIVERY/

16. HEALTH CARE DELIVERY, INTEGRATED/

17. DISEASE MANAGEMENT/

18. CASE MANAGEMENT/

19. MULTIDISCIPLINARY CARE TEAM/

20. exp CONTINUITY OF PATIENT CARE/

21. PATIENT EDUCATION/

22. LIFE STYLE CHANGES/

23. BEHAVIOR MODIFICATION/

24. exp PATIENT COMPLIANCE/

25. EDUCATION, MEDICAL, CONTINUING/

26. EDUCATION, NURSING, CONTINUING/

27. (manag* ADJ3 care).ti,ab

28. (management ADJ3 program*).ti,ab

29. (case ADJ3 manag*).ti,ab

30. (patient ADJ3 management).ti,ab

31. (home ADJ3 intervention*).ti,ab

32. "home visit*".ti,ab

33. (discharg* ADJ3 program*).ti,ab

34. "practice guideline*".ti,ab

35. (discharg* ADJ3 planning).ti,ab

36. (comprehensive ADJ3 care).ti,ab

37. (treatment ADJ3 plan*).ti,ab

38. (nurse* ADJ3 led).ti,ab

39. "disease management".ti,ab

40. multi-disciplin*.ti,ab

41. multidisciplin*.ti,ab

42. "secondary prevention clinic*".ti,ab

43. reminder*.ti,ab

44. recall*.ti,ab

45. (nurse ADJ3 clinic*).ti,ab

46. ("secondary prevention" ADJ3 intervention*).ti,ab

47. ("secondary prevention" ADJ3 program*).ti,ab

48. appointment*.ti,ab

49. "outreach nurs*".ti,ab

50. "outreach visit*".ti,ab

51. (lifestyle ADJ3 intervention*).ti,ab

52. "nurs* intervention*".ti,ab

53. "education* program*".ti,ab

54. ("physical activit*" OR "physical exercise*").ti,ab

55. (exercise ADJ3 train*).ti,ab

56. (exercise ADJ3 intervention*).ti,ab

57. (exercise ADJ3 program*).ti,ab

58. (exercise ADJ3 activit*).ti,ab

59. (exercise ADJ3 regim*).ti,ab

60. aerobic.ti,ab

61. fitness.ti,ab

62. ("risk factor*" ADJ5 modif*).ti,ab

63. ("risk factor*" ADJ5 reduc*).ti,ab

64. ("risk factor*" ADJ5 manage*).ti,ab

65. ("risk factor*" ADJ5 monitor*).ti,ab

66. ("risk factor*" ADJ5 self-manage*).ti,ab

67. 12 OR 13 OR 14 OR 15 OR 16 OR 17 OR 18 OR 19 OR 20 OR 21 OR 22 OR 23 OR 24 OR 25 OR 26 OR 27 OR 28 OR 29 OR 30 OR 31 OR 32 OR 33 OR 34 OR 35 OR 36 OR 37 OR 38 OR 39 OR 40 OR 41 OR 42 OR 43 OR 44 OR 45 OR 46 OR 47 OR 48 OR 49 OR 50 OR 51 OR 52 OR 53 OR 54 OR 55 OR 56 OR 57 OR 58 OR 59 OR 60 OR 61 OR 62 OR 63 OR 64 OR 65 OR 66

68. CLINICAL TRIALS/

69. RANDOM ASSIGNMENT/

70. PLACEBOS/

71. DOUBLE-BLIND STUDIES/ OR SINGLE-BLIND STUDIES/ OR TRIPLE-BLIND STUDIES/

72. random*.ti,ab

73. placebo.ti,ab

74. 68 OR 69 OR 70 OR 71 OR 72 OR 73

75. 11 AND 67 AND 74

Appendix 5. AMED (Ovid) search strategy

1. CEREBRAL HEMORRHAGE/ OR CEREBRAL INFARCTION/ OR CEREBRAL ISCHEMIA/ OR CEREBROVASCULAR ACCIDENT/ OR STROKE/

2. (cva* OR stroke* OR poststroke* OR post-stroke* OR "transient ischemic attack*" OR "transient ischaemic attack*" OR TIA* OR ministroke* OR mini-stroke*).ti,ab

3. (people OR patient* OR outpatient* OR inpatient* OR adult* OR survivor* OR victim* OR individual* OR client* OR population* OR community OR subject*).ti,ab

4. 2 AND 3

5. (cerebrovascular* OR "cerebral vascular").ti,ab

6. (cerebral OR cerebellar OR brain* OR vertebrobasilar).ti,ab

7. (infarct* OR ischemi* OR ischaemi* OR thrombo* OR apoplexy OR emboli*).ti,ab

8. 6 AND 7

9. (cerebral OR intracerebral OR intracranial OR brain* OR cerebellar OR subarachnoid).ti,ab

10. (accident* OR hemorrhag* OR haemorrhag*).ti,ab

11. 9 AND 10

12. 1 OR 4 OR 5 OR 8 OR 11

13. "DELIVERY OF HEALTH CARE"/

14. PATIENT CARE MANAGEMENT/

15. PROGRAM EVALUATION/

16. PATIENT EDUCATION/

17. LIFE STYLE/

18. PREVENTION/

19. PATIENT COMPLIANCE/

20. PATIENT CARE TEAM/

21. COMMUNITY HEALTH SERVICES/

22. HEALTH PROMOTION/

23. EXERCISE/

24. DIET/

25. SMOKING CESSATION/

26. HEALTH BEHAVIOR/

27. (manag* ADJ3 care).ti,ab

28. (management ADJ3 program*).ti,ab

29. (case ADJ3 manag*).ti,ab

30. (patient ADJ3 management).ti,ab

31. (home ADJ3 intervention*).ti,ab

32. "home visit*".ti,ab

33. (discharg* ADJ3 program*).ti,ab

34. "practice guideline*".ti,ab

35. (discharg* ADJ3 planning).ti,ab

36. (comprehensive ADJ3 care).ti,ab

37. (treatment ADJ3 plan*).ti,ab

38. (nurse* ADJ3 led).ti,ab

39. "disease management".ti,ab

40. multi-disciplin*.ti,ab

41. multidisciplin*.ti,ab

42. "secondary prevention clinic*".ti,ab

43. reminder*.ti,ab

44. recall*.ti,ab

45. (nurse ADJ3 clinic*).ti,ab

46. ("secondary prevention" ADJ3 intervention*).ti,ab

47. ("secondary prevention" ADJ3 program*).ti,ab

48. appointment*.ti,ab

49. "outreach nurs*".ti,ab

50. "outreach visit*".ti,ab

51. (lifestyle ADJ3 intervention*).ti,ab

52. "nurs* intervention*".ti,ab

53. "education* program*".ti,ab

54. ("physical activit*" OR "physical exercise*").ti,ab

55. (exercise ADJ3 train*).ti,ab

56. (exercise ADJ3 intervention*).ti,ab

57. (exercise ADJ3 program*).ti,ab

58. (exercise ADJ3 activit*).ti,ab

59. (exercise ADJ3 regim*).ti,ab

60. aerobic.ti,ab

61. fitness.ti,ab

62. ("risk factor*" ADJ5 modif*).ti,ab

63. ("risk factor*" ADJ5 reduc*).ti,ab

64. ("risk factor*" ADJ5 manage*).ti,ab

65. ("risk factor*" ADJ5 monitor*).ti,ab

66. ("risk factor*" ADJ5 self-manage*).ti,ab

67. 13 OR 14 OR 15 OR 16 OR 17 OR 18 OR 19 OR 20 OR 21 OR 22 OR 23 OR 24 OR 25 OR 26 OR 27 OR 28 OR 29 OR 30 OR 31 OR 32 OR 33 OR 34 OR 35 OR 36 OR 37 OR 38 OR 39 OR 40 OR 41 OR 42 OR 43 OR 44 OR 45 OR 46 OR 47 OR 48 OR 49 OR 50 OR 51 OR 52 OR 53 OR 54 OR 55 OR 56 OR 57 OR 58 OR 59 OR 60 OR 61 OR 62 OR 63 OR 64 OR 65 OR 66

68. RANDOMIZED CONTROLLED TRIALS/

69. CLINICAL TRIALS/

70. PLACEBOS/

71. DOUBLE BLIND METHOD/

72. random*.ti,ab

73. placebo*.ti,ab

74. 68 OR 69 OR 70 OR 71 OR 72 OR 73

75. 12 AND 67 AND 74

Appendix 6. BNI (Ovid) search strategy

1. BNI STROKE/

2. BNI (cva* OR stroke* OR poststroke* OR post-stroke* OR "transient ischemic attack*" OR "transient ischaemic attack*" OR TIA* OR ministroke* OR mini-stroke*).ti,ab

3. BNI (people OR patient* OR outpatient* OR inpatient* OR adult* OR survivor* OR victim* OR individual* OR client* OR population* OR community OR subject*).ti,ab

4. BNI 2 AND 3

5. BNI (cerebrovascular* OR "cerebral vascular").ti,ab

6. BNI (cerebral OR cerebellar OR brain* OR vertebrobasilar).ti,ab

7. BNI (infarct* OR ischemi* OR ischaemi* OR thrombo* OR apoplexy OR emboli*).ti,ab

8. BNI 6 AND 7

9. BNI (cerebral OR intracerebral OR intracranial OR brain* OR cerebellar OR subarachnoid).ti,ab

10. BNI (accident* OR hemorrhag* OR haemorrhag*).ti,ab

11. BNI 9 AND 10

12. BNI 1 OR 4 OR 5 OR 8 OR 11

13. BNI PATIENTS : EDUCATION/

14. BNI NURSING : ROLE/

15. BNI CARE PLANS AND PLANNING/

16. BNI EVIDENCE BASED PRACTICE/

17. BNI MULTIDISCIPLINARY TEAMS/

18. BNI CONTINUITY OF CARE/

19. BNI PATIENTS : COMPLIANCE/

20. BNI (manag* ADJ3 care).ti,ab

21. BNI (management ADJ3 program*).ti,ab

22. BNI (case ADJ3 manag*).ti,ab

23. BNI (patient ADJ3 management).ti,ab

24. BNI (home ADJ3 intervention*).ti,ab

25. BNI "home visit*".ti,ab

26. BNI (discharg* ADJ3 program*).ti,ab

27. BNI "practice guideline*".ti,ab

28. BNI (discharg* ADJ3 planning).ti,ab

29. BNI (comprehensive ADJ3 care).ti,ab

30. BNI (treatment ADJ3 plan*).ti,ab

31. BNI (nurse* ADJ3 led).ti,ab

32. BNI "disease management".ti,ab

33. BNI multi-disciplin*.ti,ab

34. BNI multidisciplin*.ti,ab

35. BNI "secondary prevention clinic*".ti,ab

36. BNI reminder*.ti,ab

37. BNI recall*.ti,ab

38. BNI (nurse ADJ3 clinic*).ti,ab

39. BNI ("secondary prevention" ADJ3 intervention*).ti,ab

40. BNI ("secondary prevention" ADJ3 program*).ti,ab

41. BNI appointment*.ti,ab

42. BNI "outreach nurs*".ti,ab

43. BNI "outreach visit*".ti,ab

44. BNI (lifestyle ADJ3 intervention*).ti,ab

45. BNI "nurs* intervention*".ti,ab

46. BNI "education* program*".ti,ab

47. BNI ("physical activit*" OR "physical exercise*").ti,ab

48. BNI (exercise ADJ3 train*).ti,ab

49. BNI (exercise ADJ3 intervention*).ti,ab

50. BNI (exercise ADJ3 program*).ti,ab

51. BNI (exercise ADJ3 activit*).ti,ab

52. BNI (exercise ADJ3 regim*).ti,ab

53. BNI aerobic.ti,ab

54. BNI fitness.ti,ab

55. BNI ("risk factor*" ADJ5 modif*).ti,ab

56. BNI ("risk factor*" ADJ5 reduc*).ti,ab

57. BNI ("risk factor*" ADJ5 manage*).ti,ab

58. BNI ("risk factor*" ADJ5 monitor*).ti,ab

59. BNI ("risk factor*" ADJ5 self-manage*).ti,ab

60. BNI 13 OR 14 OR 15 OR 16 OR 17 OR 18 OR 19 OR 20 OR 21 OR 22 OR 23 OR 24 OR 25 OR 26 OR 27 OR 28 OR 29 OR 30 OR 31 OR 32 OR 33 OR 34 OR 35 OR 36 OR 37 OR 38 OR 39 OR 40 OR 41 OR 42 OR 43 OR 44 OR 45 OR 46 OR 47 OR 48 OR 49 OR 50 OR 51 OR 52 OR 53 OR 54 OR 55 OR 56 OR 57 OR 58 OR 59

61. BNI random*.ti,ab

62. BNI placebo*.ti,ab

63. BNI trial.ti,ab

64. BNI 61 OR 62 OR 63

65. BNI 12 AND 60 AND 64

Appendix 7. Web of Science Conference Proceedings Citation Index - Science search strategy

Stroke* OR TIA OR "transient isch*mic attack" OR "cerebral infarct*" OR "brain infarct*" OR cerebrovascular IN TITLE

AND

("secondary SAME prevention") OR ("recurrent stroke") OR (risk SAME reduc*) IN TOPIC

AND

intervention or program or service* or management IN TOPIC

AND

Proceedings paper IN DOCUMENT TYPE

What's new

Last assessed as up-to-date: 9 April 2013.

DateEventDescription
2 May 2014AmendedAn updated 'Declaration of Interest' statement has been added for one author.

Contributions of authors

Dr Kate Lager contributed to the conception and design of the review. She was principally responsible for data collection, analysis of data, interpretation of data and writing the review.

Professor Andrew Wilson guided the conception and design of the review. He contributed to data collection, interpretation of data and revising the review.

Dr Amit Mistri guided protocol development, and also contributed to data collection, interpretation of data and revising the review.

Professor Kamlesh Khunti guided protocol development, and contributed to interpretation of the data and revising the review.

Dr Victoria Haunton and Dr Aung K Sett selected studies according to the review criteria, and contributed to data extraction and revising the review.

Declarations of interest

Dr Amit Mistri has received speaker fees for talks on stroke from various companies manufacturing drugs for vascular disease including Boehringer-Ingelheim, Bayer, Bristol-Myers Squibb, Astellas Pharma, Pfizer and Astra Zeneca, and a travel grant for conference attendance from Boehringer-Ingelheim.

Professor Kamlesh Khunti has acted as a consultant and speaker for Novartis, Novo Nordisk, Sanofi-Aventis, Lilly and Merck Sharp & Dohme. He has received grants in support of investigator and investigator initiated trials from Novartis, Novo Nordisk, Sanofi-Aventis, Lilly, Pfizer, Boehringer Ingelheim and Merck Sharp & Dohme. He has received funds for research, honoraria for speaking at meetings and has served on advisory boards for Lilly, Sanofi-Aventis, Merck Sharp & Dohme and Novo Nordisk.

Professor Andrew Wilson: none known.

Dr Kate Lager: none known.

Sources of support

Internal sources

  • Department of Health Sciences, University of Leicester, UK.

External sources

  • No sources of support supplied

Differences between protocol and review

The following authors have contributed to the review but were not authors of the review protocol: Dr Victoria J Haunton and Dr Aung K Sett.

The title for the review has been changed following recommendations made by the Cochrane Stroke Group Editorial Team.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Adie 2010

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital stroke clinic and hospital neurovascular clinic

Numbers randomised: total: 56; I: 29; C: 27

% Completing final follow-up: 100%

Inclusion criteria: < 1 month since minor stroke or TIA; > 18 years; clinic SBP ≥ 140 mmHg; living at home at time of follow-up

Exclusion criteria: known dementia, "significant disability or co-morbidity which would impair ability to consent or cause undue distress"

Type of stroke: minor stroke (57%); TIA (43%)

Mean age (SD): 72.5 (8.9)

Gender (% male): 50%

Ethnicity: not reported

Socioeconomic or sociodemographic status: not reported

Interventions

Intervention details (components, length, frequency): motivational telephone follow-up intervention based on social cognitive theory. Participants received a 20 minute telephone call at 7 days, 1, 2 and 4 months to review risk factors, medication and goal setting; participants provided with tailored educational material; participants with high blood pressure encouraged to visit their GP

Location: community

Mode of delivery: telephone follow-up

Personnel responsible for delivery: 1 researcher

Timing post-stroke: < 1 month

Control: usual care (patients received instructions for follow-up with their general practitioner; no follow-up visits arranged in secondary care)

Outcomes6 months: SBP (clinic and ambulatory); DBP (clinic and ambulatory); total cholesterol; BP ≤ 130/80 mmHg; total cholesterol ≤ 4 mmol/L
General Information

Country of origin: UK

Publication language: English

Notes

Analysis method: not stated

Risk of bias: low

Comments: definition of minor stroke not stated

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskEnvelope method: "participants were randomised … at the end of their first study visit (baseline; month 0) by sequential opaque envelopes stratified by stroke or TIA"
Allocation concealment (selection bias)Low riskEnvelope method
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskNo blinding reported, but the review authors judge that the outcomes are not likely to be influenced by lack of blinding
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskProtocol available and outcomes are reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Allen 2002

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital acute stroke department

Numbers randomised: total: 96; I: 47; C: 46

% Completing final follow-up: 76%

Inclusion criteria: ischaemic stroke or TIA; discharged to home or short-term rehabilitation facility (for < 1 month); no other illnesses that would dominate post-discharge care; Rankin Scale score ≤ 3;

Exclusion criteria: Rankin score of 4 or 5; discharged to long-term care facility

Type of stroke: ischaemic stroke ( I: 70%; C: 71%); TIA (I: 30%; C: 29%)

Mean age (SE): I: 69 (1.7); C: 72 (1.5)

Gender (% female): I: 57; C: 54

Ethnicity (% African-American): I: 30%; C: 20%

Socioeconomic or sociodemographic status: not reported

Interventions

Intervention details (components, length, frequency): APN telephoned patients 3 to 7 days post-discharge to assess needs and deliver education; APN conducted home assessment within 1 month post-discharge; individualised patient care plans developed by interdisciplinary team using evidence-based recommendations; APN implemented treatment plan and conducted follow-up assessments; primary care physicians provided with care plans/evidence-based recommendations

Location: community

Mode of delivery: home visits

Personnel responsible for delivery: advanced practice nurse and interdisciplinary team

Timing post-stroke: discharge home

Control: usual care provided by primary care physician

Pre-discharge care (I and C): interdisciplinary care and stroke education

Outcomes3 months: BP: mean mmHg BP > 140/90; proportion of participants rehospitalised for stroke
General Information

Country of origin: US

Publication language: English

Notes

Analysis method: not stated

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Patients were assigned to the intervention or to usual postdischarge care by drawing consecutive concealed tickets that were randomised within permuted blocks of 10"
Allocation concealment (selection bias)Unclear riskNot reported
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low risk

Rehospitalisation for stroke: Recording of events was blinded to group assignment

Blood pressure: "Some of the outcome measures were obtained by nurses who were not blinded to patient status"

Review authors judge that objective outcomes included in this review are not likely to be affected by lack of blinding

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Missing data not reported by group

Attrition: 1 became cognitively impaired; 2 moved out of state; 3 moved to nursing home; 5 died; 12 refused follow up visit

Judgement: not enough information to permit judgement (missing data not reported by group)

Selective reporting (reporting bias)Unclear riskInsufficient information (protocol not obtained)
Other biasLow riskThe study appears to be free of other sources of bias

Allen 2009

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital acute stroke department

Numbers randomised: total: 380; I: 190; C: 190

% Completing final follow-up: 84% to 100% depending on outcome measure

Inclusion criteria: ischaemic stroke; NIHSS ≥ 1; discharged to home or short-term rehabilitation/nursing facility (for < 8 weeks); no other illnesses that would dominate post-discharge care; English-speaking; no planned carotid endarterectomy

Type of stroke: ischaemic (100%)

Mean age (SE): I: 68 (1); C: 69 (1)

Gender (% male): I: 48%; C: 52%

Ethnicity (% African American): I: 17%; C: 15%

Socioeconomic or sociodemographic status (% married): I: 47%; C: 46%

Interventions

Intervention details: participant received home assessment at 1 week from APN; individualised patient care plans developed by interdisciplinary team using evidence-based recommendations; ongoing care management provided by APN for 6 months (telephone contact every week for first month and monthly thereafter; home visits as needed; physical therapist visits arranged as needed; liaison with social services; patients provided with personalised health record and pill organisers for risk factor management); primary care physicians provided with care plans/evidence-based recommendations

Location: community

Mode of delivery: home visits and telephone follow-up

Personnel responsible for delivery: APN and interdisciplinary team

Timing post-stroke: discharge home

Control: usual care provided by primary care physician; received postal stroke-related educational materials every 2 months

Usual care before discharge (I and C): organised stroke department care with enhanced discharge planning. Involved physical and psychological evaluation using standardised assessment tools; initiation of appropriate medication; development of individualised discharge plan; discharge summary sent to primary care physician

Outcomes6 months: SBP > 140 mmHg; DBP > 90 mmHg; total cholesterol > 180 mg/dL; Hb1Ac > 6.5%; proportion of participants on anticoagulant; proportion of participants using method for medication compliance
General Information

Country of origin: US

Publication language: English

Notes

Analysis method: stated intention-to-treat

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"The randomisation sequence was by permuted blocks of fixed size (10) generated by study biostatisticians"
Allocation concealment (selection bias)Low risk"Group assignment was made by a research assistant using the sealed envelope method"
Blinding of outcome assessment (detection bias)
All outcomes
Low risk"Outcome measurements were performed at a home visit (when possible) by a research nurse blinded to group assignment at 6 months post-discharge … some measurements were confirmed by review of hospital and PCP records"
Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Missing data reported by group but reasons not fully described

Attrition (dependent on outcome): I: range 0/90 to 25/190 (reasons unclear); C: range 0/190 to 36/190 (reasons unclear)

Judgement: not enough information to permit judgement (reasons for missing data not provided)

Selective reporting (reporting bias)Low riskExamination of study reports suggests that all outcomes were reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Boter 2004

MethodsUnit of randomisation: patient
Participants

Place of recruitment: 2 university hospitals; 10 general hospitals

Numbers randomised: total: 536; I: 263; C: 273

% Completing final follow-up: 91%

Inclusion criteria: TIA, ischaemic stroke, primary intracerebral haemorrhage, or subarachnoid haemorrhage; Dutch-speaking; ≥ 18 years; first admission for stroke or TIA; hospitalisation within 72 hours after onset of symptoms; life expectancy > 1 year; Rankin grade 0 to 3; discharged home

Type of stroke: TIA (I: 9%; C: 8%); ischaemic stroke (I: 53%; C: 55%); haemorrhagic stroke (I: 10%; C: 9%); subarachnoid haemorrhage (I: 19%; C: 19%)

Median age (IQR): I: 66 (52 to 76); C: 63 (51 to 74)

Gender (% female): I: 51%; C: 52%

Ethnicity (% African American): I: 17%; C: 15%

Socioeconomic or sociodemographic status:

  • Education level: I: primary school or less - 24%, secondary school - 60%, higher education or university – 15, unknown - 1%; C: primary school or less - 27%, secondary school - 58%, higher education or university - 15%, unknown < 1%

  • Living alone: I: 30%, C: 26%

Interventions

Intervention details (components, length, frequency): participants and their carers received 3 telephone calls from a stroke nurse at 1 to 4, 4 to 8 and 18 to 24 weeks; participants received 1 home visit from a stroke nurse at 10 to 14 weeks; checklists used to address stroke risk factors, stroke consequences and unmet needs in terms of stroke services; nurses supported participants and carers according to their individual needs

Location: community

Mode of delivery: home visits and telephone follow-up

Personnel responsible for delivery: stroke nurses trained for 2 days on "secondary prevention of stroke, rehabilitation, therapies, prognosis and knowledge of local care facilities"

Timing post-stroke: post-discharge

Control: standard care

Outcomes6 months: proportion of participants using secondary prevention drugs (anticoagulants or antiplatelets)
General Information

Country of origin: Netherlands

Publication language: English

Notes

Analysis method: stated intention-to-treat

Risk of bias: low

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Allocation was done by means of a central telephone service"
Allocation concealment (selection bias)Low risk"Allocation was done by means of a central telephone service"
Blinding of outcome assessment (detection bias)
All outcomes
Low risk"Outcome assessors blinded to group allocation"
Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 32/263 (7 died; 25 declined follow-up); C:18/273 (5 died; 13 declined follow-up)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskStudy protocol available and all outcomes are reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Boysen 2009

MethodsUnit of randomisation: patient
Participants

Place of recruitment: stroke units

Numbers randomised: total: 314; I: 157; C: 157

% Completing final follow-up: 88%

Inclusion criteria: ischaemic stroke; aged > 40 years; able to walk

Exclusion criteria: contraindications to exercise; modified Rankin scale of 4 or 5 pre-stroke; cognitive impairment; discharge to nursing home; severe neurological deficit

Type of stroke: ischaemic (100%)

Median age (IQR): I: 69.7 (60.0 to 77.7); C: 69.4 (59.6 to 75.8)

Gender (% female): I: 43%; C: 44%

Ethnicity: not reported

Socioeconomic or sociodemographic status:

  • Years of education (%): I: ≤ 8 (45%), 9 to 12 (34%), ≥ 13 (21%); C: ≤ 8 (47%), 9 to 12 (40%), ≥ 13 (13%)

Interventions

Intervention details (components, length, frequency): repeated verbal instructions about physical activity over 2 years; first meeting (30 to 60 minutes) to develop individualised plan for physical activity; follow-up visits (20 to 30 minutes) every 3 months for the first year and every 6 months thereafter to provide repeated instructions and readjust physical activity plan; between-visit reminder telephone calls

Location: community

Mode of delivery: home visits and telephone follow-up

Personnel responsible for delivery: physiotherapist in 8 centres, neurologist in 1 centre

Timing post-stroke: beginning < 90 days post-stroke

Control: received information about physical activity; received follow-up visits at same frequency as intervention group but without instructions about physical activity

Outcomes24 months: number of secondary strokes; number of myocardial infarctions; number of vascular deaths
General Information

Country of origin: Denmark, China, Poland and Estonia

Publication language: English

Notes

Analysis method: stated intention-to-treat; per protocol

Risk of bias: low

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskCentral randomisation: "generation of allocation sequences was computer based"
Allocation concealment (selection bias)Low risk"Allocation concealment was achieved through centralised randomisation by telephone or email."
Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Outcome assessors were blinded to group allocation

"All events were adjudicated by an independent adjudication committee, which was blinded to the intervention group of the patient"

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 24/157 (11 died; 3 withdrawn due to severe neurological deficits caused by recurrent stroke; 10 lost to follow-up); C: 14/157 (9 died; 2 withdrawn due to severe neurological deficits caused by recurrent stroke; 2 lost to follow-up)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskStudy protocol available and all outcomes are reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Brotons 2011

MethodsUnit of randomisation: general practice
Participants

Place of recruitment: 42 primary care centres in 8 regions of Spain

Numbers randomised: total: 1224 (414 stroke/TIA); I: 624 (203 stroke/TIA); C: 600 (211 stroke/TIA)

% Completing final follow-up: 70%

Inclusion criteria: cardiovascular disease (ischaemic heart disease, stroke /TIA and peripheral arterial disease); ≤ 80 years

Exclusion criteria: cardioembolic stroke or subarachnoid haemorrhage as a result of valvulopathy; serious disease or terminal illness; bedbound

Type of stroke (%): not stated

Mean age (SE): I: 68 (11); C: 69 (11)

Gender (% male): I: 64%; C: 64%

Ethnicity: not reported

Socioeconomic or sociodemographic status:

  • Employment status: employed - 11%, unemployed - 2%, sick leave/invalidity - 10%, retired 61%, Other - 16%

  • Education level: illiterate - 4%, uneducated, literate - 36%, primary education - 39%, secondary education - 13%, higher education - 6%, university 3%

Interventions

Intervention details (components, length, frequency): comprehensive secondary prevention program including tailored patient education, promotion of medication adherence and review of secondary prevention medication; participants attended appointment every 4 months for 2.75 years; health professionals delivering the intervention followed protocols for patient care and attended training sessions on secondary prevention of cardiovascular disease

Location: primary care

Mode of delivery: outpatient appointment

Personnel responsible for delivery: nurses with specific training in the secondary prevention of cardiovascular disease

Timing post-stroke: < 1 year 

Control: usual care

Outcomes3 years: SBP; DBP; total cholesterol; LDL; HDL; triglycerides; BMI; BP < 140/90 in non-diabetics or BP < 130/80 in diabetics/ patients with chronic renal failure; cardiovascular readmissions; cardiovascular fatal events
General Information

Country of origin: Spain

Publication language: English

Notes

Analysis method: intention-to-treat

Risk of bias: low

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandom numbers generated using a validated computer program
Allocation concealment (selection bias)Low riskCentral allocation service, stratified by region ("the randomisation sequence was not revealed until the intervention was assigned")
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskNo blinding reported, but the review authors judge that objective outcomes are not likely to be influenced by lack of blinding
Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 11 died; 51 lost to follow-up (reasons provided); 6 unknown; C: 13 died; 69 lost to follow-up (reasons provided); 41 unknown*

*study authors explain that it was difficult to recover reasons for losses in control group because they were visited only at baseline and at end of follow-up

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskStudy protocol available and outcomes are reported in pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Chanruengvanich 2006

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital (centre specialising in neurology)

Numbers randomised: total: 72; I: 36; C: 36

% Completing final follow-up: 86%

Inclusion criteria: > 6 weeks since TIA or minor stroke; energy expenditure < 1000 Kcal/week; age > 45 years; no cognitive impairment; able to exercise; BP ≤ 180/100 mmHg; fasting blood sugar ≤ 150 mg%

Exclusion criteria: complications e.g. heart attack or chest pain

Type of stroke (%): not reported

Mean age (SD): I: 62.8 (7.4); C: 63.1 (7.1)

Gender (% female): I: 68%; C: 68%

Ethnicity: not reported

Socioeconomic or sociodemographic status:

  • Marital status: single – 11%, couple – 63%, separated – 26%

  • Educational level: elementary – 53%, high school – 21%, vocational/college – 15%, bachelor degree – 10%, master degree – 1.6%

  • Income (Baht): < 5000 – 63%, 5001 to 10,000 – 16%, 10,001 to 15,000- 8%, 15,001 to 20,000 - 8%, > 20,000 - 5%

Interventions

Intervention details (components, length, frequency): 12 week self-regulated exercise program; first week - educational meeting (topics included disease management, diet, exercise and stress management); second week - instruction in self-regulation techniques and recommended exercises (using group demonstration and video); third week -  home visit from researcher to identify problems; second to twelfth weeks – moderate exercise for a minimum of 15 minutes 2 to 3 times per day (recorded in exercise diary) with energy expenditure target 1000 kcal per week; researcher made weekly telephone calls to encourage participants to adhere to the exercise program

Location: community

Mode of delivery: patient education, home visit and telephone follow-up

Personnel responsible for delivery: researcher/investigator

Timing post-stroke: > 6 weeks

Control: usual care

Outcomes12 weeks: SBP; DBP; total cholesterol; HDL
General Information

Country of origin: Thailand

Publication language: English

Notes

Analysis method: not stated (per protocol)

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk"each patient was randomly assigned" - method not reported
Allocation concealment (selection bias)Unclear riskNot reported
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskNurses conducting outcome assessments were blinded to group allocation (information provided by trialists)
Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 5/36 (1 withdrew; 4 illness prohibited exercise); C: 3/36 (3 withdrew)

Excluded from analysis: I: 0; C: 2/36 (2 excluded to balance the groups)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskStudy protocol available and all outcomes are reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Chiu 2008

MethodsUnit of randomisation: patient
Participants

Place of recruitment: tertiary referral hospital (outpatients)

Numbers randomised: total: 160; I: 80; C: 80

% Completing final follow-up: not reported

Inclusion criteria: ischaemic stroke; national health insurance (coverage: 95%); attending outpatient clinics for > 12 months

Exclusion criteria: currently enrolled in other trials; terminal illness

Type of stroke: ischaemic stroke (100%)

Mean age (SD): I: 65.7 (10.0); C: 64.8 (10.6)

Gender (% female) I: 50%; C: 50%

Ethnicity: not reported

Socioeconomic or sociodemographic status:

  • Education (%): I: illiterate - 45%, educated – 55%; C: illiterate – 46%, educated - 54%

Interventions

Intervention details (components, length, frequency): monthly 1 hour pharmacist-led educational program conducted over 6 months; topics included drug effects, treatment goals, lifestyle modification, compliance and adverse effects; no scheduled monitoring of modifiable risk factors

Location: hospital

Mode of delivery: outpatient appointment

Personnel responsible for delivery: pharmacist

Timing post-stroke: > 12 months

Control: usual care (attendance at outpatient clinics)

Outcomes6 months: SBP; DBP; total cholesterol; LDL; triglycerides; BP < 140/90 mmHg; LDL < 100 mg/dL or TC < 160 mg/dL; HbA1c < 7% or fasting blood glucose < 126 mg/dL or random postprandial blood glucose < 200 mg/dL
General Information

Country of origin: Taiwan

Publication language: English

Notes

Analysis method: not stated

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk"Simple random sampling"
Allocation concealment (selection bias)Unclear riskNot stated
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskNo blinding reported, but the review authors judge that the outcomes are not likely to be influenced by lack of blinding
Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Missing data not reported

 

Selective reporting (reporting bias)Unclear riskInsufficient information (protocol not obtained)
Other biasLow riskThe study appears to be free of other sources of bias

Eames 2013

MethodsUnit of randomisation: patient
Participants

Place of recruitment: 2 acute stroke units in metropolitan hospitals

Numbers randomised: total: 77; I :37; C: 40

% Completing final follow-up: 86%

Inclusion criteria: ischaemic stroke, haemorrhagic stroke or TIA; admitted to hospital for stroke or TIA; living in a residential care facility prior to admission and it was not a planned discharge destination; adequate spoken English, cognition, communication and corrected vision and hearing to complete the outcome measures

Exclusion criteria: poor medical prognosis (i.e. medically unstable patients or those undergoing palliative treatment)

Type of stroke: ischaemic (I: 73%; C: 84%); haemorrhagic (I: 25%; C: 14%), TIA (I :3%, C: 0%)

Mean age (SD): I: 57.0 (16.6); C: 64.1 (14.3)

Gender (% male) I: 55%; C: 51%

Ethnicity: not reported

Socioeconomic or sociodemographic status: not reported

Interventions

Intervention details (components, length, frequency): tailored written stroke information (stroke booklet) and verbal reinforcement of this information by a health professional (verbal reinforcement was offered face-to-face up to 3 times prior to discharge and over the telephone up to 3 times following discharge). Participants could tailor the content of the information booklet and the verbal sessions

Location: acute stroke unit (prior to discharge) and community/inpatient rehabilitation ward (post-discharge)

Mode of delivery: outpatient appointment

Personnel responsible for delivery: occupational therapist

Timing post-stroke: approximately 1 week prior to acute stroke unit discharge

Control: usual care (stroke unit care included usual medical, nursing, and allied health management)

Outcomes3 months: adherence to secondary prevention medications
General Information

Country of origin: Australia

Publication language: English

Notes

Analysis method: unknown

Risk of bias: low

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Concealed, random allocation was achieved via sequentially numbered envelopes containing computer-generated random numbers prepared by a person not involved in the study"
Allocation concealment (selection bias)Low risk"Baselines outcome measures were obtained prior to randomisation and therefore by a blinded assessor. Administration of outcome measures at the follow-up interview was undertaken by a blinded assessor. Once completed, the assessor opened a sealed section of the form to determine group allocation and asked intervention group participants additional questions regarding the intervention." (Unpublished information provided by trialists)
Blinding of outcome assessment (detection bias)
All outcomes
Low risk"Outcome measures were administered prior to hospital discharge and at 3-month follow-up by blinded assessors"
Blinding of outcome assessment (detection bias)
Self-reported outcomes (e.g. medication adherence)
Low risk 
Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 5/40 (4 unable to be contacted; 1 cognition impairment too severe for interview follow-up; C:6/37 (2 withdrew; 3 unable to be contacted; 1 admitted to residential care

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskProtocol is available and outcomes are reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Ellis 2005

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital TIA clinic or geriatric medical day hospital

Numbers randomised: total: 205; I: 100; C: 105

% Completing final follow-up: 94%

Inclusion criteria: < 3 months since stroke, TIA or amaurosis fugax; ambulant patients; one of more cardiovascular risk factor (high BP, history of current smoking, high cholesterol, diabetes)

Exclusion criteria: cognitive impairment (AMT < 5 on screening)

Type of stroke: TIA (I: 29%; C: 26%); stroke (I: 61%; C: 65%)

Mean age (95% CI): I: 64.3 (62.4 to 66.1), C: 65.8 (64.0 to 67.5)

Gender (% male): I: 54%; C: 50%

Ethnicity: not reported

Socioeconomic or sociodemographic status: not reported

Interventions

Intervention details (components, length, frequency): monthly reviews (approximately 3) with a stroke nurse specialist; participants received tailored verbal and written information addressing medication compliance, lifestyle modification, interaction with medical services, risk factor status and risk factor targets; participants advised to visit their GP if risk factors poorly controlled

Location: hospital outpatient setting

Mode of delivery: outpatient appointment

Personnel responsible for delivery: stroke nurse specialist

Timing post-stroke: first review at 3 months

Control: usual care (1 review in hospital outpatient setting where patients received standard outpatient advise on risk factors and secondary prevention; discharged to general practice care)

Outcomes

5 months (per protocol): SBP; DBP; total cholesterol; Hb1Ac; combined risk factor control

3.6 years (additional follow-up): SBP; DBP; total cholesterol; Hb1Ac; persistence with therapy; self-reported adherence; recurrent cardiovascular events; percentage of patients meeting target for combined risk factor control

General Information

Country of origin: UK

Publication language: English

Notes

Analysis method: stated intention-to-treat

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Patients were randomly allocated to treatment or control groups using a computer-generated random sequence"
Allocation concealment (selection bias)Low risk"concealed in sequentially numbered opaque sealed envelopes"
Blinding of outcome assessment (detection bias)
All outcomes
Low riskOutcome assessors were blinded to group allocation
Blinding of outcome assessment (detection bias)
Self-reported outcomes (e.g. medication adherence)
Unclear risk 
Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 6 lost to follow-up (reasons unclear); C: 7 lost to follow-up (reasons unclear)

Excluded from analysis: I: 3 patients entered twice by error: duplicate results excluded from the analysis; C: 1 patient found to be ineligible: results included in the analysis (intention-to-treat)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Unclear riskInsufficient information (protocol not obtained)
Other biasLow riskThe study appears to be free of other sources of bias

Evans 2010

MethodsUnit of randomisation: patient
Participants

Place of recruitment: primary care medical clinic

Numbers randomised: total: 176 (8 stroke/TIA); I: 88 (4 stroke/TIA); C: 88 (4 stroke/TIA)

% Completing final follow-up: 89%

Inclusion criteria: Framingham risk score ≥ 15% or coronary artery disease risk equivalent (coronary artery disease, peripheral artery disease, cerebrovascular disease, diabetes mellitus)

Exclusion criteria: severe psychiatric conditions or demential symptomatic heart failure; terminal illness

Type of stroke (%): not stated

Mean age (SD): 62.5 (10.5)

Gender (% male): 87.5%

Ethnicity: not reported

Socioeconomic or sociodemographic status: not reported

Interventions

Intervention details (components, length, frequency): pharmacist-delivered secondary prevention program involving cardiovascular risk stratification, monitoring of cardiovascular risk factors and drug adherence support; participants were contacted approximately every 8 weeks for minimum of 6 months (telephone call, appointment, mailed letters); mean duration of follow-up was 380 days; participants and their primary care physicians were informed if risk factors were uncontrolled

Location: primary care medical clinic

Mode of delivery: primary care appointment

Personnel responsible for delivery: pharmacist (intervention designed for non-specialist pharmacists in order to facilitate collaborative partnerships without the need for advanced training)

Timing post-stroke: unknown

Usual care (I and C): general counselling about cardiovascular disease (1 hour pharmacist appointment)

Outcomes12 months: SBP; DBP; total cholesterol; LDL; HDL; triglycerides; HbA1C; 10 year Framingham risk score
General Information

Country of origin: Canada

Publication language: English

Notes

Analysis method: stated intention-to-treat

Risk of bias: low

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Randomisation lists were stratified by each physician and were created by using a table of random numbers in permuted blocks of four"
Allocation concealment (selection bias)Low risk"Randomisation codes were kept in individually sealed envelopes and opened by the study pharmacist at the end of the initial visit"
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskNo blinding reported, but the review authors judge that objective outcomes are not likely to be influenced by lack of blinding
Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 11/88 (9 laboratory data not available; 1 moved; 1 died); C: 9/88 (8 laboratory data not available; 1 withdrew due to unrelated illness)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskProtocol available and outcomes reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Flemming 2013

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital

Numbers randomised: total: 41; I: 20; C: 21

% Completing final follow-up: 88%

Inclusion criteria: ischaemic stroke or TIA and at least one uncontrolled stroke risk factor (hypertension, hyperlipidaemia, diabetes or tobacco use); > 55 years old

Exclusion criteria: NIHSS > 7; prior enrolment in cardiovascular prevention clinic; life expectancy < 1 year

Type of stroke (%): TIA (I:40%, C:52%); ischaemic stroke (I:60%, C:48%)

Mean age (SD): I: 70 (13); C: 71 (9)

Gender (% male): I: 50%; C: 66%

Ethnicity: not reported

Socioeconomic or sociodemographic status: not reported

Interventions

Intervention details (components, length, frequency): nurses were trained in stroke risk factors and motivational interviewing; participants attended nurse-led appointments for risk factor review (baseline, 6 weeks, 6 months and 1 year) and received additional nurse-led telephone follow-up; nurses followed standardised protocols for the assessment and management of stroke risk factors; participants attended consultations with dietician and exercise physiologist; secondary stroke prevention recommendations and participants’ risk factor assessments were sent to their GP/neurologist

Location: outpatient clinic

Mode of delivery: outpatient appointment and telephone follow-up

Personnel responsible for delivery: nurses

Timing post-stroke: < 12 weeks

Usual care (I and C): usual care: baseline risk factor assessment and follow-up appointment (1 year); usual follow-up by primary care/neurology

Outcomes12 months: change in cardiovascular risk factors (SBP; LDL; HDL; triglycerides; HbA1c; BMI; Framingham cardiovascular risk score); achievement of targets for cardiovascular risk factors; number of vascular events; adherence to secondary prevention medication
General Information

Country of origin: US

Publication language: English

Notes

Analysis method: available case analysis

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskShuffling envelopes
Allocation concealment (selection bias)Low riskEnvelope method
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskNo blinding of outcome assessors, but the review authors judge that objective outcomes are not likely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)
Self-reported outcomes (e.g. medication adherence)
Unclear risk

Assessed via patient interview - outcome assessors were not blinded to group allocation

The review authors judge that non-blinding may have affected assessment of self-reported outcomes

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 2/20 (1 died; 1 lost to follow-up); C:3/21 (2 died; lost to follow-up)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskProtocol available and outcomes reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Hornnes 2011

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital

Numbers randomised: total: 349; I: 172; C: 177

% Completing final follow-up: 87%

Inclusion criteria: ischaemic stroke, intracerebral haemorrhage or TIA

Exclusion criteria: discharged to a nursing home; cognitive deficits prohibiting informed consent; life expectancy < 2 years

Type of stroke (%): ischaemic (I: 71%; C: 73%); intracerebral haemorrhage (I: 3%; C: 5%); TIA: (I: 26%; C: 22%)

Mean age (SD): I: 70.2 (13.7); C: 68.5 (12.2)

Gender (% female): I: 48%; C: 50%

Ethnicity: not reported

Socioeconomic or sociodemographic status:

  • Living alone (%): I: 52%; C: 52%

  • Educational level (%): I: low – 31%, medium – 26%, high – 43%; C: low – 32%, medium – 26%, high – 42%

Interventions

Intervention details (components, length, frequency): 4 home visits from a nurse at 1, 4, 7 and 10 months; each visit included blood pressure monitoring, tailored lifestyle counselling and promotion of medication compliance; hypertensive patients encouraged to visit their GP

Location: community

Mode of delivery: home visits

Personnel responsible for delivery: nurse

Timing post-stroke: randomised at time of discharge

Control: usual care (neurologist outpatient visit 3 months post-stroke)

Outcomes12 months: SBP; DBP; proportion of participants meeting BP targets; proportion of participants adhering antihypertensive therapy
General Information

Country of origin: Denmark

Publication language: English

Notes

Analysis method: not reported

Risk of bias: low

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"used a computer-generated, block randomisation procedure"
Allocation concealment (selection bias)Low risk"the allocation sequence was concealed…the study nurses who administered the intervention had access to a computer program…entering the patient’s Central Person Registry number, BP value, and hospital yielded a printout of the patient’s randomisation number and allocation"
Blinding of outcome assessment (detection bias)
All outcomes
Low riskOutcome assessors were blinded to the allocation of the patients
Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 27/172 (13 dropped out; 3 diagnosis revised; 10 died; 1 too ill); C: 19/177 (9 dropped out; 5 died; 2 too ill; 2 diagnosis revised; 1 other reason)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskOutcomes reported in a pre-specified way (trial registry: http://www.clinicaltrials.gov/ct2/show/NCT00253097)
Other biasLow riskThe study appears to be free of other sources of bias

Johnston 2010

MethodsUnit of randomisation: hospital
Participants

Place of recruitment: 12 hospitals

Numbers randomised: total: 3361; I: 1464; C: 1897

% Completing final follow-up: 80%

Inclusion criteria: ischaemic stroke; Kaiser Permanente Medical Care Plan members with pharmacy benefits; age ≥ 40 years; acute hospitalisation for stroke

Exclusion criteria: haemorrhagic stroke; discharged to hospice

Type of stroke: ischaemic (100%)

Mean age (SD): 72.9 (12.6)

Gender (% female): 53%

Ethnicity: non-Hispanic white 66%; African American 14%; Asian/Pacific Islander 11%; Hispanic 7%; other/unknown 1%

Socioeconomic or sociodemographic status: members of Kaiser Permanente Medical Care Plan with "under-representation of the very poor and wealthy"

Interventions

Intervention details (components, length, frequency): hospitals received support from a central co-ordinator in the development and implementation of standardised stroke discharge orders (discharge orders based on American Heart Association recurrent stroke prevention guidelines and included (1) statin prescription for all patients irrespective of cholesterol levels; (2) antihypertensive prescriptions for hypertensive patients; (3) warfarin prescription for patients with atrial fibrillation); two physician “champions” (from neurology and hospital-based medicine) from each hospital tailored discharge order and supervised implementation; two educational presentations delivered to healthcare providers (timing: development of discharge orders and 3 months post-implementation)

Location: Kaiser Permanente Medical Care Plan hospitals

Mode of delivery: health provider education and pre-printed stroke discharge orders

Personnel responsible for delivery: central co-ordinator and 2 physicians supervised implementation

Timing post-stroke: discharge from hospital

Control: usual care without contact from study staff; some hospitals implemented their own discharge orders

Outcomes6 months: BP < 140/90 mmHg; combined cardiovascular risk factor control; adherence to secondary prevention medications
General Information

Country of origin: US

Publication language: English

Notes

Analysis method: stated intention-to-treat

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Participating hospitals were paired based on characteristics that could have impacted the success of the intervention, including patient demographics, hospital size, number of enrollees, and presence of a motivated stroke expert. Then, using a random number generator, 1 hospital in each pair was randomised to receive the intervention, whereas the other was randomised to usual care."
Allocation concealment (selection bias)Low risk"Participating hospitals were paired based on characteristics that could have impacted the success of the intervention, including patient demographics, hospital size, number of enrollees, and presence of a motivated stroke expert. Then, using a random number generator, 1 hospital in each pair was randomised to receive the intervention, whereas the other was randomised to usual care."
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low risk

Blood pressure: "blood pressures were obtained for routine clinical purposes by personnel unaware of the study goals"

Filled prescription data: statins and warfarin: "statin and warfarin data were gathered from linked pharmacy records"

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 1149/1464 (237 died; 78 lost to follow-up); C: 1533/1897 (277 died; 87 lost to follow-up)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Unclear riskProtocol available and primary outcomes are reported in the pre-specified way; some secondary outcomes not reported
Other biasLow riskThe study appears to be free of other sources of bias

Joubert 2009

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital

Numbers randomised: total: 233; I: 123; C: 110

% Completing final follow-up: 80%

Inclusion criteria: ischaemic stroke, parenchymal haemorrhage or TIA; aged ≥ 20 years

Exclusion criteria: not managed by GP; discharged to nursing home; serious co-morbidities; non-English speaking; serious cognitive impairment; significantly aphasic

Type of stroke (%): ischaemic (I: 73%; C: 80%); haemorrhagic (I: 10%: C: 7%); TIA (I: 17%; C: 13%)

Mean age (SD): I: 63.4 (13.7); C: 68.2 (12.7)

Gender (% male): I: 58%; C: 52%

Ethnicity: not reported

Socioeconomic or sociodemographic status: not reported

Interventions

Intervention details (components, length, frequency): "shared care" program; risk factor targets derived from National guidelines and consensus statements; medication initiated in hospital; lifestyle education provided by nurse co-ordinator; GP appointments pre-arranged for 2 weeks, 3 months, 6 months, 9 months and 12 months post-discharge; recommendations and evidence-based guidelines sent to GP; nurse co-ordinator telephoned participants before and after every GP visit to screen for depression; risk factor data collected at each GP visit and faxed to nurse co-ordinator; nurse co-ordinator facilitated transfer of information and recommendations between stroke specialists and GPs; GPs able to telephone stroke specialist for advice

Location: community

Mode of delivery: telephone follow-up; information management

Personnel responsible for delivery: stroke specialists, a nurse co-ordinator and patients’ GPs

Timing post-stroke: intervention initiated before hospital discharge

Control: standard care from GP

Outcomes12 months: SBP; DBP, total cholesterol, BMI, systolic BP < 140 mmHg; total cholesterol < 5.18 mmol/L; proportion of AF patients taking warfarin
General Information

Country of origin: Australia

Publication language: English

Notes

Analysis method: not stated

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk

"Computer-generated process"

"At a later stage, the coordinator checked the patient’s GP, and if this GP was also responsible for a different patient already in the trial, the current patient was assigned to the same group as the previous patient"

Allocation concealment (selection bias)Low risk"The allocation to group was undertaken after consent, so the coordinator was unaware of treatment allocation prior to consent"
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskNo blinding reported, but the review authors judge that objective outcomes are not likely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)
Self-reported outcomes (e.g. medication adherence)
Unclear risk

Assessed via patient interview - blinding of outcome assessors was not reported

The review authors judge that non-blinding may have affected assessment of self-reported outcomes

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 32/123 (7 unwilling to participate; 2 withdrew due to other medical problems, 2 changed GP; 11 withdrew for unknown reasons; 3 did not have stoke; 3 not contactable; 2 died; 1 moved to nursing home; 1 GP refused); C: 15/110 (2 unwilling to participate; 1 left country; 3 withdrew for unknown reasons; 2 did not have stroke; 1 not contactable; 6 died)

Judgement: imbalances in missing data between the groups; however the review authors judge that this is unlikely to be related to study outcomes

Selective reporting (reporting bias)Unclear riskInsufficient information (protocol not obtained)
Other biasLow riskThe study appears to be free of other sources of bias

Kerry 2013

MethodsUnit of randomisation: patient
Participants

Place of recruitment: outpatient and inpatient stroke clinics

Numbers randomised: total: 381; I: 187; C: 194

% Completing final follow-up: 88%

Inclusion criteria: ≤ 9 months since stroke or TIA and hypertension (BP > 140/85 mmHg or treatment with antihypertensive medications)

Exclusion criteria: enrolled in another trial; severely ill or too frail; already using a blood pressure monitor; severe cognitive impairment; non-English speaking

Type of stroke (%): ischaemic (I: 58%; C: 64%); haemorrhagic (I: 7%; C: 5%); TIA (I: 34%; C: 30%); both types of stroke or unknown (I: 1%; C: 2%)

Mean age (SD): I: 71.1 (12.6); C: 72.6 (11.4)

Gender (% male): I: 59%; C: 56%

Ethnicity: White (I: 80%; C: 73%); Black (I: 11%; C: 15%); Asian (I: 4%; C: 7%); other (I: 5%; C: 5%)

Socioeconomic or sociodemographic status:

  • Index of Multiple Deprivation score* (mean ± SD): I: 17.5 ± 10.7; C: 19.3 ± 10.1

Interventions

Intervention details (components, length, frequency): participants provided with a home blood pressure monitor, brief training and ongoing nurse-led telephone support targeting BP reduction (average of 3.8 telephone calls over 12 months); participants with consistent blood pressure readings ≥ 130/80 mmHg advised to consult their GP and received intensified nurse-led telephone follow-up until the target was reached (i.e. implementation of protocols for BP reduction)

Location: community

Mode of delivery: home visits and telephone follow-up

Personnel responsible for delivery: nurse

Timing post-stroke: ≤ 9 months

Control: baseline assessment conducted during home visit and all patients with BP > 150/90 mmHg were advised to see their GP; usual care provided by GP (all GPs sent information about the study and a recommended target for home blood pressure of < 130/80 mmHg); participants in the control group received telephone calls after 3 and 9 months to check on their well-being

Outcomes12 months: SBP; DBP, proportion of participants with recurrent stroke
General Information

Country of origin: UK

Publication language: English

Notes

Analysis method: available case analysis

Risk of bias: low

*Trialists state that “the Index of Multiple Deprivation 2007 scale is a measure of poverty and is based on postal codes and ranges from 0.37 to 85.46. A higher score indicates higher deprivation. Further information can be found at www.communities.gov.uk/communities/research/indicesdeprivation/deprivation10/”

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"a computer-generated randomisation sequence to implement stratified randomisation … with a 1:1 allocation using random block sizes of 4 and 6"
Allocation concealment (selection bias)Low risk"Allocation to the intervention or control group was contained within a sealed, numbered envelope and assigned to the participant by the trial administrator before the baseline visit. The research nurse opened the envelope after she completed the home baseline assessment."
Blinding of outcome assessment (detection bias)
All outcomes
Low risk"measured by an independent researcher unaware of group allocation"
Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 18/187 (9 died, 5 lost contact, 1 moved away, 3 declined); C: 25/194 (10 died, 6 lost contact, 5 withdrew because of illness, 2 moved away, 2 declined)

Excluded from analysis: I: 1/187 (reason not provided); C: 0

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskProtocol is available and outcomes are reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Kim 2013

MethodsUnit of randomisation: patient
Participants

Place of recruitment: neurology clinic

Numbers randomised: total: 36; I: 18; C: 18

% Completing final follow-up: 94%

Inclusion criteria: < 12 months since ischaemic stroke; visited a neurology clinic for stroke treatment; normal cognitive function (Mini Mental State Examination > 19); living at home; Internet access

Exclusion criteria: n/a

Type of stroke (%): ischaemic (100%)

Mean age (SD): I: 67.4 (7.3); C: 63.9 (7.4)

Gender (% male): I: 73%; C: 56%

Ethnicity: not stated

Socioeconomic or sociodemographic status (% graduated the middle school): I: 61%; C: 56%

Interventions

Intervention details (components, length, frequency): 9-week web-based education program focusing on secondary prevention (9 weekly sessions involving video lectures/quizzes, website links to stroke-related information, automated feedback about self-reported health behaviours and the opportunity to email health professionals); guidebook for the programme was provided to participants; research assistant provided telephone-based technical support for the Internet program

Location: participants’ homes

Mode of delivery: internet-based education

Personnel responsible for delivery: web-based education program was developed by healthcare professionals

Timing post-stroke: < 12 months

Control:usual care provided by physicians

Outcomes3 months: total cholesterol, triglycerides, medication adherence
General Information

Country of origin: South Korea

Publication language: English

Notes

Analysis method: stated intention-to-treat

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"The participants were randomly assigned to an experimental or control group in a 1:1 ratio, using a computer-generated random code"
Allocation concealment (selection bias)Unclear riskNot stated
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskNo blinding reported, but the review authors judge that objective outcomes are not likely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)
Self-reported outcomes (e.g. medication adherence)
Unclear risk

Assessed via patient interview - blinding of outcome assessors was not reported

The review authors judge that non-blinding may have affected assessment of self-reported outcomes

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 1/18 (1 lost to follow-up as a result of poor health); C: 1/19 (1 declined to complete follow-up assessment)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Unclear riskInsufficient information (protocol not obtained)
Other biasLow riskThe study appears to be free of other sources of bias

Lowe 2007

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital stroke unit

Numbers randomised: total: 100; I: 50; C: 50

% Completing final follow-up: 84%

Inclusion criteria: stroke; discharged home; able to complete questionnaire or had carer who could complete questionnaire

Exclusion criteria: severe cognitive impairment or communication difficulties; discharged to institutional care

Type of stroke (%): ischaemic (I: 96%; C: 94%)

Median age (IQR): I: 68 (62 to 74); C: 73 (65 to 80)

Gender (% male): I: 58%; C: 62%

Ethnicity: not reported

Socioeconomic or sociodemographic status: not reported

Interventions

Intervention details (components, length, frequency): information book (CareFile) containing general information about stroke and tailored information about stroke risk factors; researcher explained contents of book to participants/carers during 15 to 20 minute discussion; participants advised to take the CareFile to GP and stroke review clinic appointments

Location: hospital

Mode of delivery: educational materials

Personnel responsible for delivery: researcher (stroke research registrar)

Timing post-stroke: before discharge 

Control: usual care ("usual stroke information leaflets (Stroke Association leaflets) provided by the stroke unit and follow-up in a stroke review clinic")

Outcomes3 months and 6 months: SBP; DBP
General Information

Country of origin: UK

Publication language: English

Notes

Analysis method: not stated

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskShuffling envelopes
Allocation concealment (selection bias)Low risk"When a diagnosis of stroke was confirmed, eligible patients were randomised by the researcher into the control or intervention group (using sealed opaque envelopes containing blocks of 10 names, in a one-to-one ratio)."
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low risk

Measured by an "independent assessor (research nurse)" although blinding of group allocation could not always be performed

The review authors judge that objective outcomes are not likely to be influenced by lack of blinding

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 6/50 (2 could not be contacted; 4 died); C: 10/50 (4 could not be contacted; 6 died)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Unclear riskInsufficient information (protocol not available)
Other biasLow riskThe study appears to be free of other sources of bias

Lowrie 2010

MethodsUnit of randomisation: general practice
Participants

Place of recruitment: 31 general practices

Numbers randomised: total: 4040 (461 stroke/TIA); I: 2373 (289 stroke/TIA); C: 1667 (172 stroke/TIA)

% Completing final follow-up: information only provided for patients with baseline and follow-up data

Inclusion criteria: previous diagnosis associated with vascular disease ("myocardial infarction, coronary artery bypass graft/angioplasty, angina, angiographic coronary artery disease, stroke/transient ischaemic attack, peripheral ischaemic arterial disease/intermittent claudication or, diabetic patients aged over 45 years")

Type of stroke among participants with a history of stroke/TIA (%): stroke (66%); stroke only (56%); TIA (44%); TIA only (34%); stroke and TIA (10%)

Mean age (SD): I: 68 (11); C: 72 (11)

Gender (% male): I: 47%; C: 47%

Ethnicity: not reported

Socioeconomic or sociodemographic status:

  • Mean Modified Scottish Index of Multiple Deprivation (SD): I: 46.8 (15.1); C: 35.3 (12.4)

Interventions

Intervention details (components, length, frequency): "pharmacist-led educational outreach directed at general practices, aiming to improve statin prescription for community dwelling patients with vascular disease”; pharmacists received specific training relevant to the delivery of the intervention (5.5 training days); pharmacists delivered 3 educational outreach meetings at each general practice at 4 monthly intervals; pharmacists worked in practices on 1 day per week for 44 weeks to identify patients who were eligible to receive Simvastatin 40 mg and encourage GPs/nurses to systematically contact/follow-up patients

Location: general practices

Mode of delivery: pharmacist-led outreach visits

Personnel responsible for delivery: pharmacists

Timing post-stroke: not reported

Control: practices did not receive pharmacist-led prescribing support

Outcomes5 to 13 months (mean 8.8 months): total cholesterol; total cholesterol < 5.0 mmol/L
General Information

Country of origin: UK

Publication language: English

Notes

Analysis method: n/a (available case data used in this review)

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"random number table"
Allocation concealment (selection bias)Low riskN/A: all clusters were randomised at once
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low risk

 

No blinding reported, but the review authors judge that objective outcomes are not likely to be influenced by lack of blinding

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskReasons for missing data not available since results were only presented for patients with baseline and follow-up data (confirmed via correspondence with trialists)
Selective reporting (reporting bias)Low riskStudy protocol available and outcomes are reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Maasland 2007

MethodsUnit of randomisation: patient
Participants

Place of recruitment: TIA service ("provides a rapid diagnostic work-up of patients with TIA or minor stroke in a single day")

Numbers randomised: total: 65; I: 33; C: 32

% Completing final follow-up: 88%

Inclusion criteria: < 3 months since TIA or minor ischaemic stroke; ≥ 18 years; fluent in spoken and written Dutch; modified Rankin score < 4

Exclusion criteria: involved in cardiovascular health education; aphasia, dementia (diagnosis based on DSM-Iv criteria); visual impairment that would affect health education

Type of stroke: TIA (I: 57%; C: 52%); minor stroke (I: 43%; C: 46%)

Mean age (SD): I: 65 (12); C: 63 (13)

Gender (% male): I: 57%; C: 63%

Ethnicity: not reported

Socioeconomic or sociodemographic status:

  • Educational level (%): I: primary school - 27%, secondary school - 37%, college - 20%, university - 17%; C: primary school - 15%, secondary school - 41%, college - 26%, university - 19%

Interventions

Intervention details (components, length, frequency): 20 to 25 minute computerised education program about TIA and stroke, antiplatelet and anticoagulant medication and modifiable risk factor control; information tailored according to the impact of each risk factor on secondary prevention (calculated using algorithm) and each patient’s current risk factor status, treatment status, educational level and age; participants received a printed summary of the information

Location: TIA service

Mode of delivery: computer-based education

Personnel responsible for delivery: n/a

Timing post-stroke: acute TIA or minor stroke

Control: usual care (health education by a neurologist as part of the TIA service)

Outcomes12 weeks: SBP; DBP; total cholesterol; LDL, triglycerides; BMI; compliance with anticoagulants; compliance with lipid-lowering medication; compliance with antihypertensive medication
General Information

Country of origin: Netherlands

Publication language: not stated

Notes

Analysis method: available case analysis

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Treatment allocation was random, and based on computer-generated random numbers"
Allocation concealment (selection bias)Low risk"The randomisation was blocked in lots of 10; block size was unknown to the investigators at the time of the trial"
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskNo blinding reported, but the review authors judge that objective outcomes are not likely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)
Self-reported outcomes (e.g. medication adherence)
Unclear risk

Assessed via patient interview - outcome assessors were not blinded to group allocation

The review authors judge that non-blinding may have affected assessment of self-reported outcomes

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 2/33 lost to follow-up; C: 5/32 lost to follow-up

Excluded from analysis: I: 1/33 professional health worker (ineligible); C: 0/32

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskProtocol available and primary outcomes are reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

MacKenzie 2013

MethodsUnit of randomisation: patient
Participants

Place of recruitment: 4 urban stroke prevention clinics

Numbers randomised: total: 56; I: 29; C: 27

% Completing final follow-up: 100%

Inclusion criteria: probable TIA or confirmed stroke; aged > 18 years; psychological/cognitive deficits (Montreal Cognitive Score < 26) OR < 100% medication self efficacy or self-reported medication non-adherence; uncontrolled hypertension (BP > 140/90 mmHg or > 130/80 mmHg for individuals with diabetes or chronic renal insufficiency

Exclusion criteria: inability to speak/read English; reliant on others to administer medications

Type of stroke: stroke (64%); TIA (36%)

Age: > 65 years: 59%

Gender (% male): 68%

Ethnicity: not reported

Socioeconomic or sociodemographic status:

  • Living alone (21%);

  • Education < 9 years (16%)

Interventions

Intervention details (components, length, frequency): nurse-led intervention targeting participants at high risk of suboptimal BP control or non-adherence to antihypertensive medication: involved medication counselling, provision of home BP monitoring equipment and medication Dosette, and nurse-led telephone calls (monthly intervals for 6 months) to deliver motivational interviewing for secondary prevention behaviours (nurses responsible for delivering the intervention received training in motivational interviewing techniques)

Location: community

Mode of delivery: outpatient appointment and telephone follow-up

Personnel responsible for delivery: nurse practitioner/clinical nurse specialist

Timing post-stroke: not reported

Control: usual care - "stroke physician specialist assessment, initiation and titration of BP medication, adherence and risk factor counselling at clinic visits and follow-up by family physicians"

Outcomes6 months: stroke recurrence, SBP, DBP, BP < 140/90 mmHg; adherence to antihypertensive medication
General Information

Country of origin: Canada

Publication language: not stated

Notes

Analysis method: intention-to-treat

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"centralized telephone randomization system"
Allocation concealment (selection bias)Low risk"centralized telephone randomization system"
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskNo blinding reported, but the review authors judge that objective outcomes are not likely to be influenced by lack of blinding
Blinding of outcome assessment (detection bias)
Self-reported outcomes (e.g. medication adherence)
Unclear risk

Assessed via patient interview - blinding of outcome assessors was not reported

The review authors judge that non-blinding may have affected assessment of self-reported outcomes

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskExamination of study reports suggests that all outcomes were reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Markle-Reid 2011

MethodsUnit of randomisation: patient
Participants

Place of recruitment: community care access centre

Numbers randomised: total: 101; I: 52; C: 49

% Completing final follow-up: 81%

Inclusion criteria: < 18 months since stroke or TIA; living in community; newly referred (< 2 weeks) to home care services; competent to give informed consent or substitute decision maker available; competent in English or with an interpreter available

Type of stroke (%): not reported

Mean age (SD): I: 75.8 (12.4); C: 70.6 (14.5)

Gender (% male): I: 49%; C: 62%

Ethnicity: not reported

Socioeconomic or sociodemographic status:

  • Married (%): I: 40%; C: 51%

  • Living with others (%): I: 54%; C: 64%

Interventions

Intervention details (components, length, frequency): usual home care services plus organised home visits from an interprofessional team (care co-ordinator, nurse, physiotherapist, occupational therapist, speech language pathologist, dietician, social worker, physiotherapist, personal support worker) over a 12 month period; rehabilitation followed evidence-based rehabilitation protocols addressing community reintegration and stroke prevention; use of standardised screening tools e.g. stroke risk assessment tool; members of interdisciplinary team met at monthly case conferences and attended training sessions delivered by the study investigators;

Location: community

Mode of delivery: home visits; healthcare provider meetings

Personnel responsible for delivery: interprofessional team

Timing post-stroke: < 18 months

Control: usual home care services (follow-up by a care co-ordinator who provided in-home assessments and co-ordinated home support services)

Outcomes12 months: number of secondary strokes
General Information

Country of origin: Canada

Publication language: English

Notes

Analysis method: not stated

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Randomly generated numbers constructed by a biostatistician who was not involved in the recruitment process"
Allocation concealment (selection bias)Low risk"Consecutively numbered, sealed, opaque envelopes"
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskOutcome assessors blind to group assignment
Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 9/52 (I: 4 died; 4 refused; 1 unable to contact); C: 10/49 (C: 3 died; 7 refused)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Unclear riskInsufficient information (protocol not obtained)
Other biasLow riskThe study appears to be free of other sources of bias

O'Carroll 2011

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital stroke clinic and stroke unit

Numbers randomised: total: 62; I: 31; C :31

% Completing final follow-up: 87%

Inclusion criteria: first stroke or TIA; discharged home; prescribed secondary prevention antihypertensive medication; suboptimal medication adherence score

Exclusion criteria: requirement for help with taking medications; using a Dosette box; cognitive difficulties that precluded participation in the study

Type of stroke (%): not reported

Mean age (SD): I: 68 (11); C: 71 (11)

Gender (% male): 65%

Ethnicity: not reported

Socioeconomic or sociodemographic status (Scottish Index of Multiple Deprivation Quintile): 1 (highest deprivation) – 2%, 2 – 10%, 3 – 19%, 4 – 19%, 5 (lowest deprivation) – 51%

Interventions

Intervention details (components, length, frequency): 2 intervention sessions (approximately 30 minutes each) conducted 2 weeks apart: session 1 helped participants to establish a better medication-taking routine through completing individualised worksheets; session 2 reviewed participants’ plans and addressed barriers to implementation; electronic recording of pill-taking for a duration of 3 months (researcher made monthly home visits to refill the electronic pill bottle)

Location: participants’ homes or a research facility

Mode of delivery: home visits

Personnel responsible for delivery: researcher

Timing post-stroke: < 3 months post-discharge

Control: participants attended 2 sessions with a researcher who "engaged the patient in non-medication related conversation in an attempt to provide some control for non-specific effects of attention/social contact"; electronic recording of pill-taking for 3 months

Outcomes3 months: medication adherence; SBP; DBP
General Information

Country of origin: UK

Publication language: English

Notes

Analysis method: stated intention-to-treat

Risk of bias: low

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Participants were randomised to either the Intervention or Control group using web-based software set up by the Edinburgh Clinical Trials Unit."
Allocation concealment (selection bias)Low riskWeb-based randomisation
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low risk

Adherence to medication - assessed via electronic recording of pill-taking: "The main limitation of the current study was that the Research Fellow delivered the intervention and conducted the analysis and so was not blind to treatment arm of patients. However, the main outcome measure was electronically-recorded MEMS readings".

The review authors judge that objective outcomes are not likely to be influenced by lack of blinding

Blinding of outcome assessment (detection bias)
Self-reported outcomes (e.g. medication adherence)
Low risk

Assessed via a previously validated questionnaire - administered to patients by a researcher who was not blinded to group allocation

The review authors judge that outcomes are not likely to be influenced by lack of blinding

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition I: 2/31 (2 hospitalised for non-stroke reasons); C: 2/31 (1 hospitalised for non-stroke reasons; 1 relocated)

Excluded from the analysis: (did not receive intervention): I: 2/31 (1 declined to use electronic pill bottle; 1 hospitalised for non-stroke reasons); C: 2/31 (2 hospitalised for non-stroke reasons)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskProtocol available and outcomes reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Slark 2013

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital (inpatient)

Numbers randomised: total: 96; I: 47; C: 49

% Completing final follow-up: 98%

Inclusion criteria: Ischaemic stroke

Exclusion criteria: cognitive or memory difficulties that precluded participation in the intervention

Type of stroke: ischaemic (100%)

Mean age (SD): I: 65 (12); C: 66 (13)

Gender: I: 64%; C: 53%

Ethnicity: White: I: 62%; C: 67%; "Black Ethnic Minority (BME) groups made up 13% of the total cohort"

Socioeconomic or sociodemographic status:

  • University education: I: 40%; C: 18%

  • Married: 57%; C: 55%

Interventions

Intervention details (components, length, frequency): 30 minute risk awareness session: involved tailored information provision on the topics of stroke aetiology, risk factors and secondary prevention medications; participants were informed of their individual risk scores for secondary stroke

Location: hospital

Mode of delivery: inpatient appointment

Personnel responsible for delivery: researcher

Timing post-stroke: initiated prior to hospital discharge

Control: usual care (no additional risk awareness information)

Outcomes3 months: recurrent stroke; SBP, DBP, total cholesterol, adherence to secondary prevention medications
General Information

Country of origin: UK

Publication language: English

Notes

Analysis method: available case analysis

Risk of bias: unclear

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Subjects were randomised using computer-generated random codes"
Allocation concealment (selection bias)Low risk"The researcher was blind to randomisation until after recruitment of each participant to avoid selection bias….this was achieved through sealing each random code in an envelope prior to commencing the trial, which was only selected after the participant had been recruited."
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low risk

"The researcher conducted all the assessment and interventions as well as the follow-up assessments therefore it is important to discuss the limitation of lack of blinding of the researcher"

The review authors judge that objective outcomes are not likely to be influenced by lack of blinding

Blinding of outcome assessment (detection bias)
Self-reported outcomes (e.g. medication adherence)
Unclear risk

Assessed via patient self-report - blinding of outcome assessors was not reported

The review authors judge that non-blinding may have affected assessment of subjective outcomes

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 0/47; C: 2/47 (2 lost to follow-up)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskExamination of study reports suggests that all outcomes were reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

Wang 2005

MethodsUnit of randomisation: patient
Participants

Place of recruitment: hospital

Numbers randomised: total: 198; I: 146; C: 52

% Completing final follow-up: unknown

Inclusion criteria: stroke in internal carotid artery; first stroke

Exclusion criteria: none stated

Type of stroke (%): not stated

Mean age (SD): I: 63.24 ± 7.35; C: 60.94 ± 9.87

Gender (% male): I: 54%; C: 50%

Ethnicity: not reported

Socioeconomic or sociodemographic status: not reported

Interventions

Intervention details (components, length, frequency): follow-up by a neurologist within one week post-discharge and then every at 1, 2 or 3 months; patients and caregivers educated about nursing care, home rehabilitation, neuropsychology and modifiable risk factors

Location: community

Mode of delivery: visits, lectures, leaflets, multimedia teaching

Personnel responsible for delivery: neurologists

Timing post-stroke: < 1 week post-discharge

Control: usual care

Outcomes3 years: time to first stroke relapse; stroke relapse rate; proportion of participants meeting targets for blood pressure, blood fats, blood sugar and BMI
General Information

Country of origin: China

Publication language: Mandarin

Notes

Analysis method: not stated

Risk of bias: high

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)High risk

Not stated

Unexplained imbalances in numbers allocated to intervention and control groups

Allocation concealment (selection bias)Unclear riskNot stated
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low riskNo blinding reported, but the review authors judge that objective outcomes are not likely to be influenced by lack of blinding
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNot stated
Selective reporting (reporting bias)Unclear riskNo protocol available
Other biasLow riskThe study appears to be free from other sources of bias

Welin 2010

  1. a

    AF: atrial fibrillation
    AMT: Abbreviated Mental Test
    APN: advanced practice nurse
    BMI: body mass index
    BP: blood pressure
    C: control
    DBP: diastolic blood pressure
    GP: general practitioner
    HDL: high density lipoprotein
    I: intervention
    IQR: interquartile range
    LDL: low density lipoprotein
    NIHSS: National Institutes of Stroke Scale
    SBP: systolic blood pressure
    SD: standard deviation
    SE: standard error
    TIA: transient ischaemic attack

MethodsUnit of randomisation: patient
Participants

Place of recruitment: rural hospital

Numbers randomised: total: 163; I: 81; C: 82

% Completing final follow-up: 71%

Inclusion criteria: ischaemic or haemorrhagic stroke; first stroke; < 85 years; living at home before the stroke

Exclusion criteria: previous stroke; severe dementia; severe stroke (Rankin score > 5); severe cardiovascular disease; life expectancy < 1 year

Type of stroke (%): haemorrhagic I:9%, C:16%

Mean age (SD): I: 71.2 (9.9); C: 69.6 (11.7)

Gender (% female): I: 41%; C: 37%

Ethnicity: not reported

Socioeconomic or sociodemographic status: not reported

Interventions

Intervention: follow-up appointments with a stroke nurse at 1.5, 6 and 12 months post-discharge (included assessment of handicap and depression, measurement of blood pressure, provision of health information and referral to physiotherapist or occupational therapist if necessary); appointments with a stroke physician at 3 and 9 months (included a review of medication and medical problems with referral to other specialists if necessary)

Location: hospital stroke clinic

Mode of delivery: outpatient appointment

Personnel responsible for delivery: stroke nurse and stroke physician

Timing post-stroke: 1.5 to 12 months post-discharge

Control: usual care involved follow-up with GP; GPs were sent discharge summaries; "the quality of follow-up care by general practitioners varies in Sweden from non follow-up at all to regular visits every third or fourth month"

Usual care before discharge (I AND C): initiation of secondary prevention medications and referral to continuous physiotherapy or occupation therapy, if necessary

OutcomesSBP (12 months); DBP (12 months); recurrent stroke (3.5 years)
General Information

Country of origin: Sweden

Publication language: English

Notes

Analysis method: not stated

Risk of bias: low

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskShuffling sealed envelopes
Allocation concealment (selection bias)Low riskShuffling sealed envelopes
Blinding of outcome assessment (detection bias)
Objective outcomes (e.g. physiological risk factors, recurrent cardiovascular events)
Low risk

Blood pressure: no blinding reported

The review authors judge that outcomes are not likely to be affected by lack of blinding

Recurrent stroke: outcome assessors were blinded ("outcome data were presented and all information about which group the patient belonged to was concealed")

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Missing data reported by group

Attrition: I: 18/81 (5 died, 13 did not attend follow-up visit); C: 30/82 (9 died, 21 did not attend follow-up visit)

Judgement: reasons for missing data reported and review authors judge that they are unlikely to be related to study outcomes

Selective reporting (reporting bias)Low riskStudy protocol available and outcomes are reported in the pre-specified way 
Other biasLow riskThe study appears to be free from other sources of bias

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
  1. a

    TIA: transient ischaemic attack

Amariles 2012Outcomes not reported separately for stroke/TIA patients
Banet 1997No relevant outcomes
Bokemark 1996No relevant outcomes
Gillham 2010No relevant outcomes
Goessens 2006Outcomes not reported separately for stroke/TIA patients
Green 2007No relevant outcomes
Harrington 2007Not intended to improve modifiable risk factor control
Johnston 2000Not a stroke service intervention
Joshi 2012Outcomes not reported separately for stroke/TIA patients
Ma 2009Outcomes not reported separately for stroke/TIA patients
Middleton 2004No relevant outcomes
Nir 2006No relevant outcomes
Ornstein 2004Not a stroke service intervention
Palanco 2011Outcomes not reported separately for stroke/TIA patients
Powers 2011Not a stroke service intervention
Rimmer 2000Contained exercise training program
Ross 2007Not intended to improve modifiable risk factor control
Sides 2012Not a randomised controlled trial
Strandberg 2006Outcomes not reported separately for stroke/TIA patients
UMIN000001865Contained exercise training program
Vernooij 2012Outcomes not reported separately for stroke/TIA patients

Characteristics of studies awaiting assessment [ordered by study ID]

Feld-Glazman 2012

MethodsParallel RCT
ParticipantsStroke
InterventionsStroke education program; motivational interviewing to facilitate behaviour change for secondary stroke prevention
Outcomes12 weeks: risk factor behaviour
NotesStatus: completed

ISRCTN23868518

MethodsParallel RCT
ParticipantsStroke or TIA
InterventionsNurse-led telephone follow-up to improve risk factor control
Outcomes12, 24 and 36 months: total cholesterol, LDL, SBP, DBP, BMI
NotesStaus: completed

ISRCTN63816609

MethodsParallel RCT
ParticipantsTIA
InterventionsNurse-led care pathway
Outcomes6 months: ambulatory 12-hour systolic blood pressure
NotesStatus: completed

ISRCTN95662526

MethodsParallel RCT
ParticipantsMild stroke
InterventionsTelephone support addressing secondary prevention and adaption; use of written information and "StrokEngine" website
Outcomes12 months: use of health services and reasons (e.g. recurrent stroke)
Notes

Status: completed (June 2012), in press

No study reports available (correspondence April 2013)

NCT00172484

MethodsParallel RCT
ParticipantsStroke
InterventionsNursing programme: education/rehabilitation
OutcomesUnknown
NotesStatus: completed

NCT00211731

MethodsRCT
ParticipantsStroke or TIA
InterventionsChronic disease self-management course
OutcomesAdherence to secondary prevention measures
Notes

Status: completed

No study reports available (correspondence: April 2013)

NCT00664846

MethodsParallel RCT
ParticipantsIschaemic stroke or TIA
InterventionsInteractive education program
Outcomes12 months: recurrent cardiovascular events
Notes

Status: completed (31 December 2010)

No study reports available (correspondence April 2013)

NCT00687869

MethodsParallel RCT
ParticipantsIschemic stroke or intra-cerebral bleeding
InterventionsCase management involving web portal, telephone hotline, individual counselling sessions and home visits
Outcomes12 months: recurrent stroke/TIA
Notes

Status: completed; in press

No study reports available (correspondence April 2013)

NCT00703274

MethodsRCT
ParticipantsIschaemic stroke or TIA
InterventionsLay persons ('stroke navigators') trained to help participants reduce their risk of secondary stroke
Outcomes12 months: LDL; SBP; HbA1c; pill count (antiplatelet medication)
Notes

Status: completed; analysing data

No study reports available (correspondence April 2013)

NCT01027273

MethodsParallel RCT
ParticipantsStroke or TIA
InterventionsPeer-led education program to reduce risk factors for recurrent stroke
Outcomes6 months: BP; LDL; use of anti-thrombotic medication; medication adherence
Notes

Status: completed; analysing data

No study reports available (correspondence April 2013)

NCT01071408

MethodsRCT
ParticipantsStroke, TIA
InterventionsOutpatient stroke prevention program involving group clinics, patient self-management and telephone care co-ordination
Outcomes3 months and 7 months: BP; lipids; medication adherence
Notes

Status: completed (31 May 2012); analysing data

No study reports available (correspondence April 2013)

NCT01466907

MethodsRCT
ParticipantsStroke
InterventionsAssessement of risk factors three months after stroke and referral if indicated
Outcomes3 months and 12 months: risk factors
Notes

Status: completed

No study reports available (correspondence April 2013)

Nguyen 2011

MethodsParallel RCT
ParticipantsStroke patients
InterventionsPharmacist telephone intervention to deliver secondary stroke prevention education and promote medication adherence
Outcomes6 months: medication adherence, blood pressure, cholesterol, HbA1c
NotesStatus: completed

Wolfe 2010

  1. a

    BMI: body mass index
    BP: blood pressure
    DBP: diastolic blood pressure
    GP: general practitioner
    LDL: low density lipoprotein
    RCT: randomised controlled trial
    SBP: systolic blood pressure
    TIA: transient ischaemic attack

MethodsCluster RCT
ParticipantsStroke
InterventionsIndividualised evidence-based secondary prevention plans provided to patients/caregivers ("keeping well plans") and GPs ("secondary prevention plans") on a maximum of 3 occasions (10 weeks, 5 months and 8 months post-stroke); structured approach to risk factor monitoring
Outcomes12 months: modifiable risk factors for stroke: blood pressure, total cholesterol, HbA1c, BMI
Notes

Status: completed (2007) and study reports available

Outcome data relevant to the review not available (correspondence June 2012)

Characteristics of ongoing studies [ordered by study ID]

ACTRN12608000166370

Trial name or titleShared team approach between nurses and doctors for improved risk factor management for stroke patients
MethodsParallel RCT
ParticipantsIschaemic/haemorrhagic stroke or TIA
InterventionsCo-ordinated team approach for risk factor management in primary care setting
Outcomes12 months and 24 months: Framingham cardiovascular disease risk score; use of secondary prevention medications; BP
Starting date

Start: April 2008

Estimated completion: unknown

Contact information

Professor Amanda Thrift

Stroke and Ageing Research Centre (STARC)

Monash University

Australia

Email: amanda.thrift@monash.edu

NotesStatus: recruiting participants (correspondence April 2013)

ACTRN12611000792921

Trial name or titleMulti-centre randomised controlled trial comparing the effect of general practitioner utilisation of a TIA electronic decision support tool in the management of TIA/minor stroke patients versus usual care on the subsequent 90 day stroke risk, degree of guideline adherence, and overall treatment costs
MethodsParallel RCT
ParticipantsIschaemic stroke or TIA
InterventionsElectronic decision support tool for GPs: intended for use during the management of patients with possible minor stroke/TIA
Outcomes3 months: cardiovascular events (stroke, myocardial infarction)
Starting date

Start: February 2012

Estimated completion: unknown

Contact information

Dr Annemarei Ranta

Department of neurology Private Bag 11036 MidCentral Health Palmerston North 4442

New Zealand

Email: anna.ranta@midcentraldhb.govt.nz

NotesStatus: ongoing (correspondence April 2013)

ChiCTR-TRC-12002127

Trial name or titleEffects of clinical pharmacist interventions on the secondary prevention in the ischaemic stroke patients
MethodsParallel RCT
ParticipantsIschaemic stroke
InterventionsPharmacist-led individualised pharmaceutical care
OutcomesStroke recurrence, myocardial infarction, vascular death, medication compliance, body weight, blood pressure, serum glucose, serum lipids
Starting date

Start: April 2012

Estimated completion: unknown

Contact information

Xu Huimin

88 Jiefang Road, Hangzhou

310009

Email: haibindai@163.com

NotesStatus: recruiting participants (correspondence April 2013)

ISRCTN07607027

Trial name or titlePromoting Adherence to a Regimen of risk factor modification by Trained Non-medical personnel Evaluated against Regular practice Study PARTNERS
MethodsRCT
ParticipantsTIA or non-disabling stroke; hypertension
InterventionsSupport from a trained volunteer for risk factor reduction
Outcomes12 months and 24 months: DBP; medication adherence; BMI; cardiovascular risk score; LDL; total cholesterol/HDL ratio; HbA1c
Starting date

Start: April 2009

Estimated completion: October 2013

Contact information

Richard Chan

339 Windermere Rd

Rm B10-118

University Hospital

N6A 5A5

London

Canada

NotesStatus: ongoing/recruiting (correspondence April 2013)

ISRCTN08913646

Trial name or titleThe effect of a Health Empowerment Intervention for Stroke Self-management (HEISS) on the self-management behaviour and health outcomes of stroke rehabilitation patients
MethodsParallel RCT
ParticipantsStroke
InterventionsStroke self-management intervention (involves group education and nurse-led telephone follow-up)
OutcomesStroke recurrence, self-management behaviour
Starting date

Start: May 2012

Estimated completion: May 2014

Contact information

Dr Janet Sit

The Nethersole School of Nursing
Faculty of Medicine
The Chinese University of Hong Kong

NotesStatus: ongoing (correspondence April 2013)

ISRCTN35701810

Trial name or titleSecondary prevention after first stroke
MethodsMulti-centre clustered RCT
ParticipantsStroke
InterventionsElectronic prompts to promote GP adherence to guidelines during primary care consultations
Outcomes12 months: SBP; DBP; cholesterol; prescription adherence; recurrent vascular events
Starting date

Start: April 2010

Estimated completion: October 2011

Contact information

Professor Martin Gulliford

King's College London

6th Floor Capital House

42 Weston Street

London

SE1 SQD

Email: martin.gulliford@kcl.ac.uk

NotesStatus: ongoing but no longer recruiting (correspondence: April 2013)

ISRCTN97412358

Trial name or titleECG monitoring to detect atrial fibrillation after stroke
MethodsRCT
ParticipantsIschaemic stroke or TIA
InterventionsContinuous ECG monitoring to detect atrial fibrillation after acute stroke or TIA
Outcomes12 months: recurrent stroke
Starting date

Start: May 2010

Estimated completion: December 2016

Contact information

Professor Kennedy R Lees

Acute Stroke Unit & Cerebrovascular Clinic

Western Infirmary

Glasgow

Email: k.r.lees@clinmed.gla.ac.uk

NotesStatus: ongoing

JPRN-UMIN000007808

Trial name or titleConstruction of a self-management patient education program and evidences for preventing recurrence of stroke in community settings
MethodsParallel RCT
ParticipantsStroke or TIA
InterventionsNurse-led patient education
OutcomesFramingham Risk Score (CVD), recurrence rate, behaviour modification, physiological indicators (e.g. BP, HbA1c)
Starting date

Start: September 2010

Esimated completion: unknown

Contact information

Yasuko Fukuoka

Hiroshima University, Institute of Biomedical & Health Sciences Division of Nursing Science

Kasumi 1-2-3 Minami-ku,

Hiroshima

734-8553 Japan

Email: yasukofukuoka@hotmail.com  

NotesStatus: ongoing but no longer recruiting (correspondence April 2013)

NCT00355147

Trial name or titleAdapting tools to implement stroke risk management to veterans (TOOLS)
MethodsRCT
ParticipantsIschaemic stroke; TIA
InterventionsStroke prevention tools e.g. written materials; videos; training guides for doctors; home blood pressure machines; blood sugar monitors; messaging devices for contact between patient and healthcare provider
OutcomesRisk factor screening
Starting date

Start: January 2009

Estimated completion: December 2013

Contact information

Teresa M Damush, PhD

Roudebush VA Medical Center

Indianapolis

United States

NotesStatus: ongoing but not recruiting participants

NCT00574808

Trial name or titleImproved Delivery of Cardiovascular Care Through Outreach Facilitation (IDOCC)
MethodsRCT
ParticipantsCoronary artery disease; cerebrovascular disease; peripheral vascular disease; diabetes mellitus; chronic kidney disease; high risk of CVD (presence of at least 3 established cardiovascular risk factors)
InterventionsOutreach facilitator implementing chronic care model in primary care practices
Outcomes5 years: recommended targets reached
Starting date

Start: April 2007

Estimated completion: April 2012

Contact information

Clare E Liddy, MD, MSc

Email: cliddy@scohs.on.ca

NotesStatus: unknown

NCT00861081

Trial name or titleIntervention to enable stroke survivors in Los Angeles County hospitals to "Stay Within the Guidelines" SUSTAIN
MethodsParallel RCT
Participants

Included: ischaemic stroke or TIA

Excluded: haemorrhagic stroke

InterventionsChronic care program aimed at improving secondary stroke prevention (intervention includes "group clinics, self-management support, report cards, decision support through care guides and protocols, and coordination of ongoing care")
Outcomes

3 months and 12 months: BP; lipid levels

8 months: medication adherence

Starting date

Start: January 2010

Estimated completion: June 2013

Contact information

Barbara Vickrey, MD, MPH

University of California, Los Angeles

United States

NotesStatus: ongoing but no longer recruiting (correspondence: April 2013)

NCT00931788

Trial name or titlePreventing recurrent vascular events in patients with stroke or transient ischaemic attack (PREVENTION)
MethodsParallel RCT
ParticipantsIschaemic stroke or TIA
InterventionsPharmacist case management to improve risk factor control
Outcomes6 months: optimal BP and lipid control; SBP; LDL cholesterol
Starting date

Start: January 2009

Estimated completion: November 2013

Contact information

Miriam Fradette, MSc Pharm; Finley A McAlister, MD, MSc

University of Alberta Hospital

Edmonton, Alberta,

Canada, T6G 2B7

NotesStatus: ongoing but no longer recruiting (correspondence: April 2013)

NCT01122394

Trial name or titleReducing Risk of Recurrence (RRR)
MethodsParallel RCT
ParticipantsStroke or TIA
InterventionsTelephone intervention to reduce behavioural risk factors for secondary stroke
Outcomes6 months: BP; total cholesterol/HDL ratio; antihypertensive/lipid-lowering medication adherence
Starting date

Start: January 2010

Estimated completion: May 2014

Contact information

Julia Levine, MPA

New York Harbor HCS

New York

United states, 10010

Email: ulia.levine@va.gov

NotesStatus: recruiting participants (correspondence April 2013)

NCT01517542

Trial name or titleEvaluation of effectiveness of nutritional counselling in patients after stroke
MethodsParallel RCT
ParticipantsStroke
InterventionsNutritional counselling (participants received written guidance to promote adherence to 'DASH' diet recommendations )
Outcomes30 days; 3, 6, 9 and 12 months: body weight, blood glucose, blood pressure, lipid profile
Starting date

Strart: February 2010

Estimated completion: February 2012

Contact information

Sheila CO Martins, PI; Vanessa A Piper, SI

Hospital de Clínicas de Porto Alegre

Porto Alegre, RS, Brazil, 90035903

Email: smartins@portoweb.com

Email: vanalves001@gmail.com

NotesStatus: recruiting participants (correspondence April 2013)

NCT01550822

Trial name or titleHEALS (Healthy Eating And Lifestyle After Stroke): a pilot trial of a multidisciplinary lifestyle intervention program
MethodsRCT
ParticipantsStroke
InterventionsLifestyle intervention delivered via group education
OutcomesUnknown
Starting date

Start: April 2012

Estimated completion: unknown

Contact information

Dr Amy Towfighi

Rancho Los Amigos National Rehabilitation Center

Downey, California, United States, 90242

NotesStatus: recruiting participants

NCT01586702

Trial name or titleIntensified Secondary Prevention Intending a Reduction of Recurrent Events in TIA and Minor Stroke Patients (INSPiRE-TMS). A randomized trial comparing a patient centered support program versus conventional care
MethodsParallel RCT
ParticipantsTIA or minor stroke
Interventions"Stepwise intensified patient support program" delivered in outpatient clinics over two years (patients are provided with individualised risk factor data and supported in finding physical activities/smoking cessation programs)
Outcomes3.5 years and 6 years: major vascular events (including stroke, TIA and major coronary events)
Starting date

Start: September 2011

Estimated completion: June 2017

Contact information

Heinrich J Audebert, MD

Department of Neurology, Charité Universitätsmedizin Berlin

Berlin, Germany, 12200

Email: heinrich.audebert@charite.de

NotesStatus: recruiting participants (correspondence April 2013)

NCT01684176

Trial name or titlePharmacist intervention programme to improve medication adherence in stroke patients
MethodsParallel RCT
ParticipantsStroke or TIA
InterventionsTailored pharmacist intervention to increase adherence to secondary prevention medications (includes motivational interviewing, telephone follow-up and medication review)
Outcomes12 months: adherence to secondary prevention medications, composite endpoint (stroke, myocardial infarction or cardiovascular death)
Starting date

Start: August 2012

Estimated completion: March 2014

Contact information

Ulla Hedegaard, PhD student, MSc (Pharm)

Odense University Hospital

NotesStatus: ongoing but no longer recruiting (correspondence April 2013)

NCT01721538

Trial name or titleSecondary prevention of stroke through participation in the non-drug therapeutic weight reduction program: a single-blinded randomized controlled multicenter trial
MethodsParallel RCT
ParticipantsIschaemic stroke
InterventionsWeight reduction program
Outcomes18 months: major vascular events (stroke, TIA myocardial infarction), hypertension, diabetes mellitus and hyperlipidaemia
Starting date

Start: October 2012

Estimated completion: April 2015

Contact information

Yaroslav Winter, MD

Department of Neurology, Philipps-University Marburg, Germany

Email: wintery@med.uni-marburg.de

Notes

Status: recruiting participants

Unclear whether intervention is an exercise training program

NCT01763203

Trial name or titleSecondary stroke prevention by Uniting Community and Chronic care model teams Early to End Disparities: the SUCCEED Trial
MethodsParallel RCT
ParticipantsStroke
InterventionsCare manager (nurse practitioner or physician assistant) to implement protocols for secondary prevention; group education sessions on chronic disease self-management; home visits from a community health worker; patients provided with blood pressure monitors
Outcomes12 months: SBP, dyslipidaemia, HbA1c, BMI, vascular events, medication adherence
Starting date

Start: September 2013

Estimated completion: August 2017

Contact information

Barbara G Vickrey, MD, MPH; Amytis Towfighi, MD

Rancho Los Amigos National Rehabilitation Center

Downey, California, United States, 90242

NotesStatus: participants are not yet being recruited (correspondence April 2013)

NCT01776034

Trial name or titleHealth promotion and wellness program for stroke survivors
MethodsParallel RCT
ParticipantsStroke
InterventionsHealth promotion program to reduce body weight (involving lifestyle counselling delivered through group education and telephone follow-up)
Outcomes3 months and 6 months: body weight, biomarkers (cholesterol, triglycerides, HbA1c)
Starting date

Start: January 2013

Estimated completion: January 2014

Contact information

Corey McDaniel, BA

Cleveland Clinic

Cleveland, Ohio, United States, 44195

Email: mcdanic3@ccf.org   

NotesStatus: recruiting participants

NCT01807793

  1. a

    BMI: body mass index
    BP: blood pressure
    CVD: cardiovascular disease
    DBP: diastolic blood pressure
    ECG: electrocardiogram
    GP: general practitioner
    LDL: low density lipoprotein
    RCT: randomised controlled trial
    SBP: systolic blood pressure
    TIA: transient ischaemic attack

Trial name or titleTargeted management intervention for African-American men with TIA or stroke
MethodsParallel RCT
ParticipantsStroke or TIA
InterventionsPsychoeducation (individual and group sessions)
Outcomes3 months and 6 months: adherence to secondary prevention medications, blood pressure, HbA1c, BMI, cholesterol, triglycerides
Starting date

Start: May 2012

Estimated completion: April 2014

Contact information

John Harvis-Bey,

Case Western Reserve University

Cleveland, Ohio

United States, 44109

216-368-5723        

Prinicipal Investigator

Martha Sajatovic, MD, Professor

University Hospitals of Cleveland

NotesStatus: recruiting participants (correspondence April 2013)

Ancillary