Thrombolysis ImPlementation in Stroke (TIPS): evaluating the effectiveness of a strategy to increase the adoption of best evidence practice – protocol for a cluster randomised controlled trial in acute stroke care

Stroke is a major cause of death and disability internationally, with over 5 million deaths from stroke worldwide [1]. Stroke accounted for approximately 1 in every 19 deaths in the United States in 2009 [2], with US projections indicating that an additional 4 million people will have had a stroke by 2030 [3], a 22% increase in prevalence from 2013 [4]. The mean cost per person for stroke care in the United States in 2009 was estimated at $6,018 [5]. In 2009, The Australian Bureau of Statistics reports an estimated 381,400 Australians (1.8% of the total population) had suffered a stroke [6], with 35% reporting at least one impairment as a result of that stroke [6] and 62% of these reporting their stroke as their main disabling condition [6]. Stroke was the underlying cause of death for 11,220 Australians in 2010 [6] and is the second leading cause of death and a leading cause of long-term disability [7]. In 2012, the total financial cost of stroke was around $5 billion [8] with an estimated burden of disease cost of $49.3 billion [8]. In 2012, there were nearly 50,000 strokes in Australia, with approximately 12,000 people dying as a result of their stroke [8]. In 2009 to 2010, 35,345 hospitalisations were recorded in Australia with a principal diagnosis of stroke [9] and an average length of stay of 9 days [9]. The majority of strokes (89%) are admitted to hospital [10], and approximately 50% of sufferers are left either deceased or dependent [11].

Improved outcomes for acute stroke can be achieved by: use of aspirin within 48 hours to provide a modest absolute risk reduction of 1% [12]; Stroke Care Units (SCU) [13]; and thrombolytic therapy using intravenous tissue plasminogen activator (tPA) administered to appropriate patients according to guidelines and within 4.5 hours of symptom onset [14, 15]. A systematic review of randomised trials of intravenous tPA within 6 hours of acute ischaemic stroke onset found intravenous tPA significantly increased the odds of being alive and independent (modified Rankin Scale of 0 to 2) at 3-month follow-up for those treated within 3 hours of onset [16]. Sustained improvements in functional outcome and health-related quality of life have been identified at up to 18-month follow-up [17]. Treatment benefits for tPA are highest if administered within 3 hours of stroke onset [16, 18, 19] but can extend beyond the 3-hour window [20], with tPA considered safe for eligible patients up to 4.5 hours after stroke onset [21]. The ‘number needed to treat’ with tPA to prevent death and dependency is between 6 and 14, but may be higher depending on the favourable outcome definition and method of calculation [15, 22, 23]. Although there has been some debate about the strength of the evidence provided by thrombolysis [24, 25], the treatment is endorsed via national and international guidelines [26‐29].

Cost-effectiveness analyses have indicated that high-cost treatments for acute stroke can provide major public health benefits even when minor reductions in disability are achieved [30]. A number of studies have identified thrombolysis as being a cost-effective treatment when used appropriately [31‐34]. A UK study estimated that over a lifetime, tPA was associated with cost-savings of £96,565 per quality-adjusted life year [33]. The Model of Resource Utilization, Costs, and Outcomes for Stroke (MORUCOS) study found that tPA saved 155 disability-adjusted life years in a group of 256 eligible patients and concluded that tPA is more effective and cost-saving than treatment with aspirin [32].

Despite evidence for the cost-effectiveness and safety of thrombolysis along with international recommendations being in place for a number of years [21, 35, 36], there has been little overall growth in thrombolysis rates. Thrombolysis rates have remained relatively low in a number of countries over the past decade: In the US, tPA administration rates increased from 0.87% in 2001 to 2.40% in 2006 [37]. In the UK in 2008, while an estimated 26% of acute stroke patients were eligible for thrombolysis, only 1.4% were administered the treatment [38]. Only 4.3% of ischemic stroke patients in the California Acute Stroke Pilot Registry received thrombolysis [3]. In Australia, despite making major progress in terms of wide establishment of stroke care units, tPA rates are approximately 7% [30].

Changing practice is an acknowledged challenge in a number of settings [39]. Improving rates of thrombolysis delivery, even after medical treatment has been sought, involves challenges at several levels, including paramedic recognition of stroke prior to hospital arrival, attendance at a tPA-capable hospital, prompt triaging from the Emergency Department to a Stroke Care Team, rapid access to CT or MRI scan to establish ischaemic stroke, and the availability of appropriately-trained clinicians to provide tPA. A description of the series of tasks involved in the delivery of timely thrombolysis (see Table 1) illustrates the number of tasks, complexity, multidisciplinary nature, and time urgency involved in successful delivery of thrombolysis for acute stroke. An Australian study at a single large teaching hospital (Pre-hospital Acute Stroke Triage [PAST]) developed and implemented a pre-hospital stroke assessment tool for ambulance officers and a pre-hospital notification system following assessment of system barriers [40]. The PAST study demonstrated an increase in tPA rates from 4.7% to 21.4% [40].

Achieving multi-site change in thrombolysis rates for acute stroke poses an even more complex challenge, with very few published trials. One study looked at the benefits and challenges of establishing stroke units [41], and two trials have explored thrombolysis in particular. The US Get-With-The Guidelines stroke project – a national quality improvement program – demonstrated an increase in adherence to guideline recommendations for stroke from a baseline rate for intravenous thrombolytics of 42.09% to 72.84% in the fifth year of the project [42]. The PRomoting ACute Thrombolysis in Ischemic StrokE (PRACTISE) cluster randomised trial in The Netherlands used the Breakthrough approach – a quality improvement methodology using Plan-Do-Study-Act (PDSA) cycles – to change clinical practice [43]. The PRACTISE trial found that a significantly higher proportion of intervention group ischemic stroke patients presenting within four hours of stroke onset were thrombolysed (44.5%) compared to ischemic stroke patients in the control group (39.3%) [43]. However, patients in the intervention group experienced higher post-care dependency [43]. The PRACTISE trial authors suggested a need for a more intense intervention with continuous process measures to achieve more substantial improvements in thrombolysis rates [43]. Studies of the Breakthrough approach in other settings have produced mixed results [44, 45], which may suggest the need for additional strategies or a more intensive approach in order to achieve sustained change in complex systems. The INSTINCT cluster randomised controlled trial [46] of 24 acute-care community hospitals in the US used a multilevel, barrier assessment-interactive educational intervention. The intervention included educational elements, assessment of barriers, audit and feedback. Trial results indicated a modest increase in the proportion of admitted stoke patients receiving thrombolysis between pre- and post-intervention periods (1.25% and 2.79% at intervention hospitals, compared to 1.25% and 2.10% at control hospitals using Intention to Treat analysis) but showed non-significant differences between groups. The authors concluded that additional strategies are needed in order to increase acute stroke treatment.

Therefore, a need remains for a rigorous trial of usual approaches versus a comprehensive and intensive strategy for increasing thrombolysis rates in a manner suitable for large scale implementation. The Thrombolysis ImPlementation in Stroke (TIPS) study aims to provide a robust test of a theory and evidence-driven multi-component and multidisciplinary collaborative approach designed to achieve sustained change in complex health systems over multiple sites.

Pen and paper staff surveys will be conducted at all participating hospitals regarding perceived barriers and facilitators to tPA implementation. Survey items include: perceptions of tPA efficacy, local policies about tPA, knowledge of tPA, skills relating to assessment for and administration of tPA, challenges in tPA implementation, training relating to tPA, expertise and confidence in relation to tPA, contingencies relating to tPA, participant’s gender, age, experience, staff role, and estimated rates of tPA administration. For intervention hospitals, engagement with each intervention activity will be assessed, including number and type of staff from each site attending each workshop or meeting. Assessment of baseline overall readiness for change will also be assessed by a member of the research team using a readiness for change checklist based on the work of Warrick et al. (2009), Kasurinen et al. (2002), and the National Institute for Health and Clinical Excellence (2007). The full version of the process measure tools can be obtained from the corresponding author.

The TIPS intervention components described below are grounded in behavioural theory and accord with the Behaviour Change Wheel (BCW) method [47]. The BCW emphasises the importance of ensuring that proponents have the capability, opportunity and motivation to perform the desired behaviour through interventions such as education, persuasion, incentivisation, coercion, training, restriction, environmental restructuring, modelling and enablement. A situational analysis based on the ‘patient journey’ is a foundational intervention component in that it delineates the steps and associated staff roles involved in successful delivery of thrombolysis. The TIPS intervention components are delivered via a number of intervention activities, the timing and content of which is presented in Figure 2.