Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Heart failure, once lacking in therapies, now has a rich and growing armamentarium that is available to reduce the morbidity and mortality and improve the quality of life for patients. But are we delivering this to patients and what is the evidence (beyond the randomised controlled trials (RCTs) of individual therapies) that we should do so in a systematic and auditable way via quality improvement initiatives? Analyses from England show an association of six hospital-based process measures to that of lower heart failure (HF) readmission rates (but not all-cause hospitalisations).1 Indeed, caution should be exercised as when focus is placed on one disease, and there may be missed opportunities on treating other diseases.2 Notably, individual measures were only modestly linked to better outcomes and the durability of this effect waned over the year as patients spent more time out of hospital than in.
But is association causation? To this end, Agarwal and colleagues3 provide insight into this via a systematic review of RCTs that tested the efficacy of hospital-based improvements initiatives.4 They identified 14 RCTs across five countries—all high-income countries (HIC)—that employed a variety of techniques to improve the care of patients with HF. Of interest, two very large RCTs dominated the area and the heterogeneity (both clinical and statistical) limited the ability to numerically combine the results.
The authors identified three key findings emanating from this body of work. First, there was substantial heterogeneity in the existing trials and datasets, and therefore this went from a meta-analysis to a narrative review. Second, baseline care of patients (in terms of medical therapy) in the most recent of the trials was remarkably good, making it challenging for any intervention to demonstrate a difference over the comparator for most of the outcomes. The authors point out that there may be an opportunity in regions where the quality of care may vary such as non-HIC. This has merit, although with existing therapies suboptimally dosed and newer therapies both in existence since the majority of trials were completed (and on the horizon), HIC hospitals will likely need a similar effort again. One could additionally surmise that hospital-based interventions may indeed select out only those hospitals in whom a quality improvement could work, there is willingness to participate and there is an incentive to do so.
Key to the findings of Agarwal and colleagues is the lack of high-quality interventions tested in the outpatient environment. While not the scope of the systematic review undertaken, many programmes benchmark against outpatient markers such as 30-day rehospitalisation. Why concern ourselves on this topic? We have previously estimated that the majority of patients spend a median of 347 days alive and out of hospital in the average year after their first hospitalisation,5 and that this varies based on risk and shortens with recurrent hospitalisations.6 Could we improve on the number of days outside hospital, and at the same time improve the quality of life and care of those days via quality improvement (QI)? There is little in the outpatient environment posthospital discharge that has been effectively tested in an RCT to provide us enough confidence that this effort is worthy. Indeed, outpatient interventions remain challenging but have been done successfully in other environments such as diabetes in the primary care setting.7
The connectivity of quality improvement of inpatient to the outpatient environment remains a key gap (and often chasm) that needs a deep local understanding to best improve the transmission of quality (figure 1). The outpatient environment, often more challenging due to the mixture of clinicians, clinics, interventions and perceived lack of urgency, makes for challenges to do QI. Cluster randomised trials, including those that use a stepped wedge design, may provide the ideal design strategy to best test healthcare interventions such as quality improvement for both inpatient and outpatient environments. Despite these therapies, there remains a gap and with the introduction of new therapeutic pathways, we need to collectively understand what works (and what does not) to streamline the care provided. Integration of a QI pathway that serves to better patient care in the hospital and home environment (beyond the initial 30 days postdischarge often used) would be a welcome addition to health systems focused on integrating care.
Should patients with HF concern themselves with hospital-based interventions? Absolutely. Patients (and their families and support systems) should be engaged to aid in the design of the research and QI that establishes the efficacy or effectiveness of such endeavours, as all will share in and ideally participate in the intervention themselves and patients may be the agent for change. This is a tall order for many patients, but many will realise the benefits when the partnership is fulsome.
In one of the largest trials testing whether or not early or late feedback via hospital-based publicly available report cards lead to meaningful change in either process or outcomes for acute myocardial infarction or heart failure demonstrated a meaningful difference for the former, but not the latter.8 This may in part be due to the Hawthorne effect that occurs in QI or that there are clear measurable goals with associated interventions in acute myocardial infarction (MI) that reduce mortality early whereas those in HF do not yet exist with the same validity.
Do we know what the core components of a hospital-to-home, and the best integrated outpatient heart function programme? In the systematic review of inpatient QI, many of the key elements included benchmarking, education, order sets, transition plans and medication reconciliation, often as a bundle. It is likely that a bundle of care is more aligned with the wholistic nature that patient care should entail, and targeted interventions should be designed for care that has a recognisable and proven target (eg, door to balloon time).
What are some of the keys to integrating or designing QI effectively? Establish a culture of innovation, quality and teamwork; prioritise the QI based on patient needs and outcomes; collect data and focus measurement activity; be transparent and communicate the results; share, scale and spread; establish team ownership; think globally, act locally and grow the teams organically as needed.8 Additionally, if we begin to think of ‘data’ as an individual that should be engaged as part of any change management (rather than as an onerous thing to collect or as an abstract concept), it will allow more patients and clinicians to ‘engage’ with that individual (the data) to see how they can improve care. Data aggregation and visualisation become critically important, as demonstrated by the Global Burden of Disease study, and the proliferation of dashboards. Key to this step is making data understandable and identifying actionable meaningful chunks of data to track.
In summary, the HF community needs to engage in QI in a meaningful way, look at the entire spectrum of HF inclusive of the inpatient and outpatient environment and measure as we go. Be ready to discard outdated (and often less valid) practice and embrace what is known to be efficacious. RCTs in the inpatient environment need to be replicated in the outpatient environment to further the care of the growing prevalent population of patients with HF.
I would like to acknowledge Dr N. Pannu for feedback on an earlier version and Ellen Pyear for graphics assistance.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Commissioned; internally peer reviewed.
Patient consent for publication Not required.