Date Published: May 21, 2019
Publisher: Public Library of Science
Author(s): Mohamed S. Anwar, Alan G. Japp, Nicholas L. Mills
Partial Text: Heart failure, perhaps more than any other condition, exemplifies the potential for healthcare informatics to bridge the gap between practice and evidence-based care. The prevalence of heart failure is already estimated at 1%–2% and is increasing with our ageing population . Indeed, recent work has demonstrated that incident heart failure cases exceed the four most common causes of cancer combined in the United Kingdom . Moreover, decompensated heart failure accounts for up to 5% of all acute unscheduled hospital admissions and has the longest length of stay of any cardiac condition [4–5]. Given the high rates of debilitating symptoms and death associated with heart failure, this burden to both patients and healthcare systems provides a moral and financial imperative to ensure optimal delivery of proven therapies. The treatment of chronic heart failure has one of the most robust evidence bases in clinical medicine, with multiple landmark trials leading to comprehensive guidelines [6–7]. Yet, despite effective therapies for heart failure being widely available, there is ample evidence to suggest a significant gap exists between guideline-directed practice and clinical practice [8–10].
In PLOS Medicine, Rahimi and colleagues report the results of a longitudinal analysis of diagnostic tests, drug prescriptions, and follow-up patterns in 93,000 individuals with heart failure in the UK . The authors show convincingly that some aspects of care, such as the use of diagnostic testing and initiation of evidence-based therapies, have improved over time. Between 2002 and 2014, natriuretic peptide testing was introduced, and the use of echocardiography increased from 17% to 62%. Similarly, the initiation of combination therapy with a beta blocker and angiotensin-converting enzyme (ACE) inhibitor or angiotensin receptor antagonist increased 3-fold. However, other key elements of care, such as dose titration of heart failure therapy, remained poor, and only one in five patients were followed up in primary care, with the rates declining over the study period. A notable strength of the work is the use of both primary and secondary care data within a large and relatively undifferentiated population. This approach ensures that insights from the study are applicable to a general heart failure population and reflect overall care provision. By comparing data across a range of demographic subgroups, the authors hint at factors that may contribute to suboptimal practice. They note deficiencies in a range of quality indicators for women, older patients, and those of low socioeconomic status. It is clear how such findings, if sufficiently robust, could be used to inform future targeted interventions. The study certainly provides us valuable retrospective insights into two major policy initiatives designed to improve standards in heart failure treatment focused on primary and secondary care . The data show persuasively that both the ‘quality and outcomes framework’, a primary care reporting and incentives scheme, and the ‘national heart failure audit’ initiative, a secondary care reporting programme, changed behaviour, but that the outcomes were tied too tightly to specific measures and may have led to unintended consequences. In many respects, it is not surprising that a system designed to reward initiation of heart failure treatments but not subsequent dose optimisation would generate an increase in the former and a decline in the latter. Furthermore, there was a paradoxical decline in the recording of heart failure diagnosis in primary care from 56% to 36%, whereas the diagnosis in secondary care increased steadily. It is possible that practices did not register those patients who would not achieve all management recommendations, in order to maintain high adherence rates. This is a critical lesson to bear in mind whenever policies are geared to improve surrogate markers of quality rather than actual clinical outcomes.
Our ability to positively impact on current disparities in care are limited because of the absence of comprehensive and contemporary data from across the spectrum of care settings. Insufficiently detailed data impede our ability to identify the causes of disparity in care and, crucially, to determine whether we are providing optimal care on an individualised basis. To successfully overcome these challenges, we need to collate healthcare data from across both primary and secondary care settings in real time and use robust methodology to evaluate major changes in clinical practice or policy decisions . A platform for sharing data between primary and secondary care that is linked, anonymised, and sufficiently granular to facilitate a meaningful evaluation of current practice is required (Fig 1). This approach should be adopted widely in healthcare systems such as the National Health Service to ensure we are providing the highest standards of care for all our patients and using resources most effectively.