Do implementation interventions improve the delivery of evidence-based care in acute stroke units?
Key messages
Implementation interventions are designed to improve the delivery of 'evidence-based' care, which is care that has been proven in research studies to help people with a particular health condition. We do not know if implementation interventions delivered in acute stroke units lead to better delivery of evidence-based care.
More research is needed to investigate how to successfully implement evidence-based care in acute stroke settings. Future research should better describe the interventions and use consistent ways of measuring outcomes.
What did we want to find out?
We wanted to find out whether there are implementation interventions we can deliver in acute stroke settings to make sure that every patient on a stroke unit receives ‘evidence-based’ care. We were interested to look at ways to change healthcare workers' behaviour, as well as systems within hospitals, to understand what was most helpful in bringing about changes, so patients receive the best quality care.
What did we do?
We searched for research studies that were conducted in acute stroke units, where researchers compared interventions aimed at improving evidence-based care with no intervention, or different types of implementation interventions. We compared and summarised their results, and rated our confidence in the evidence, based on factors such as study methods and sizes.
What did we find?
We included seven studies that involved 42,489 acute stroke patients and an unknown number of health professionals. The studies were conducted in 129 hospitals in Australia, the UK, China and the Netherlands. The smallest study had 64 patients and the biggest study had 22,384 patients. Across the studies, over 85% of patients had ischaemic strokes, between 50% to 63% of patients were male, and their average age was between 65 tand78 years old.
Five studies compared a strategy made up of many parts (multifaceted) to no intervention and two studies compared one multifaceted strategy to another multifaceted strategy. Strategies in all studies aimed to change the behaviour of hospital staff and three studies looked at changing systems in the hospital.
We do not know if implementation strategies compared with no intervention have any effect on whether patients receive evidence-based care during their stroke unit admission. We think implementation strategies probably do not make a difference in the numbers of patients who are treated with thrombolysis (the "clot-buster" medicine), but probably do improve the number of patients who receive a swallow screen when they are first admitted to hospital. Implementation interventions compared to no intervention probably have little or no effect on the risk of patients dying or being disabled or dependent, and probably do not change how long patients stay in hospital. No studies reported economic costs or health professional knowledge.
What are the limitations of the evidence?
We are not confident in the evidence on whether patients receive evidence-based care during their stroke unit admission, because people collecting the data were aware of which patients received the interventions, the studies found very different results and there are not enough studies to be certain about the results. We are moderately confident in the evidence for the number of patients treated with thrombolysis, number of patients who receive a swallow screen, risk of patient dying or being disabled or dependent, and how long patients stay in hospital, mainly due to there not being enough studies for us to be certain.
This evidence is only relevant to acute stroke unit settings. Given that acute stroke units are expensive to set up and maintain, the evidence in this review is limited to well-funded healthcare facilities that have acute stroke units.
How up to date is this evidence?
This review includes papers that we identified from searching in April 2022.
We are uncertain whether a multifaceted implementation intervention compared to no intervention improves adherence to evidence-based recommendations in acute stroke settings, because the certainty of evidence is very low.
There is a growing body of research evidence to guide acute stroke care. Receiving care in a stroke unit improves access to recommended evidence-based therapies and patient outcomes. However, even in stroke units, evidence-based recommendations are inconsistently delivered by healthcare workers to patients with stroke. Implementation interventions are strategies designed to improve the delivery of evidence-based care.
To assess the effects of implementation interventions (compared to no intervention or another implementation intervention) on adherence to evidence-based recommendations by health professionals working in acute stroke units. Secondary objectives were to assess factors that may modify the effect of these interventions, and to determine if single or multifaceted strategies are more effective in increasing adherence with evidence-based recommendations.
We searched CENTRAL, MEDLINE, Embase, CINAHL, Joanna Briggs Institute and ProQuest databases to 13 April 2022. We searched the grey literature and trial registries and reviewed reference lists of all included studies, relevant systematic reviews and primary studies; contacted corresponding authors of relevant studies and conducted forward citation searching of the included studies. There were no restrictions on language and publication date.
We included randomised trials and cluster-randomised trials.
Participants were health professionals providing care to patients in acute stroke units; implementation interventions (i.e. strategies to improve delivery of evidence-based care) were compared to no intervention or another implementation intervention. We included studies only if they reported on our primary outcome which was quality of care, as measured by adherence to evidence-based recommendations, in order to address the review aim.
Two review authors independently selected studies for inclusion, extracted data and assessed risk of bias and certainty of evidence using GRADE. We compared single implementation interventions to no intervention, multifaceted implementation interventions to no intervention, multifaceted implementation interventions compared to single implementation interventions and multifaceted implementation interventions to another multifaceted intervention. Our primary outcome was adherence to evidence-based recommendations.
We included seven cluster-randomised trials with 42,489 patient participants from 129 hospitals, conducted in Australia, the UK, China, and the Netherlands. Health professional participants (numbers not specified) included nursing, medical and allied health professionals. Interventions in all studies included implementation strategies targeting healthcare workers; three studies included delivery arrangements, no studies used financial arrangements or governance arrangements. Five trials compared a multifaceted implementation intervention to no intervention, two trials compared one multifaceted implementation intervention to another multifaceted implementation intervention. No included studies compared a single implementation intervention to no intervention or to a multifaceted implementation intervention. Quality of care outcomes (proportions of patients receiving evidence-based care) were included in all included studies. All studies had low risks of selection bias and reporting bias, but high risk of performance bias. Three studies had high risks of bias from non-blinding of outcome assessors or due to analyses used.
We are uncertain whether a multifaceted implementation intervention leads to any change in adherence to evidence-based recommendations compared with no intervention (risk ratio (RR) 1.73; 95% confidence interval (CI) 0.83 to 3.61; 4 trials; 76 clusters; 2144 participants, I2 =92%, very low-certainty evidence). Looking at two specific processes of care, multifaceted implementation interventions compared to no intervention probably lead to little or no difference in the proportion of patients with ischaemic stroke who received thrombolysis (RR 1.14, 95% CI 0.94 to 1.37, 2 trials; 32 clusters; 1228 participants, moderate-certainty evidence), but probably do increase the proportion of patients who receive a swallow screen within 24 hours of admission (RR 6.76, 95% CI 4.44 to 10.76; 1 trial; 19 clusters; 1,804 participants; moderate-certainty evidence). Multifaceted implementation interventions probably make little or no difference in reducing the risk of death, disability or dependency compared to no intervention (RR 0.93, 95% CI 0.85 to 1.02; 3 trials; 51 clusters ; 1228 participants; moderate-certainty evidence), and probably make little or no difference to hospital length of stay compared with no intervention (difference in absolute change 1.5 days; 95% CI -0.5 to 3.5; 1 trial; 19 clusters; 1804 participants; moderate-certainty evidence). We do not know if a multifaceted implementation intervention compared to no intervention result in changes to resource use or health professionals' knowledge because no included studies collected these outcomes.