Clinical inertia and medical therapy for heart failure: the unintended harms of 'first, donoharm'.

Abstract

This article refers to ‘Treatment patterns and clinical outcomes among patients <65 years with a worsening heart failure event’ by J. Butler et al., published in this issue on pages 1334–1342. Primum non nocere (‘First, do no harm’) Conceptually embodied within the Hippocratic Oath, this phrase is taught to students in healthcare across the world as one of the core principles of medical ethics and clinician responsibility. In part, the phrase is meant to prepare clinicians for a career where on a daily basis, they will be forced to weigh the risks and benefits of almost everything they do for patients, whether diagnostic or therapeutic. However, like many maxims, ‘first, do no harm’ serves only as a rudimentary piece of advice.1 Receiving much less emphasis is the reality that nearly all attempts to benefit a patient involve the potential risk of harm (e.g. risk of side effects with any prescribed medication). Rather, well intentioned clinicians make decisions with the expectation that potential benefits outweigh potential harms, but there are rarely any absolute guarantees and complications inevitably occur with virtually any diagnostic test or therapy. Indeed, practicing under a literal interpretation of ‘first, do no harm’ may lead a clinician to delay indicated tests or treatments to the point of doing nothing at all.1 In the care of patients with heart failure (HF) with reduced ejection fraction (HFrEF), multiple lines of evidence suggest a strong culture of hesitancy, needlessly prolonged sequencing of beneficial therapies, and clinical inertia towards making medication changes in routine outpatient practice.2-5 Among patients eligible for therapy, longitudinal follow-up shows most outpatients have no alteration in medical therapy, and initiation and dose escalation of guideline-directed medical therapy (GDMT) remain relatively rare.3, 6 As a consequence of this ‘get to the next therapy or dose titration one day in the future’ approach, many patients eligible for therapy never receive medical therapies proven to extend survival, receive them with significant delay, or receive them at perpetual sub-target doses. Multiple potential barriers to achieving optimal GDMT have been proposed, including medical (e.g. true medication intolerance), clinician-level (e.g. education, awareness, needlessly delayed sequencing algorithms), patient-level (e.g. patient preference), and system-level barriers (e.g. drug formularies, out-of-pocket costs). Although likely multifactorial to a degree, no specific explanation has been proven dominant or definitive, and the central question remains: what explains clinician hesitancy to initiate and escalate medications that are proven to improve survival and quality of life? In routine practice, ‘first, do no harm’ is generally interpreted as a stern reminder of the perils of side effects and adverse events with ‘overmedication’ or aggressive therapy.1 Indeed, such ‘risks of commission’ with regard to potential side effects or intolerance are inherently present with any medication change for any medical condition. Such risks of initiating new medications are also very tangible for clinicians, and may engender a feeling of direct responsibility or guilt when patients contact them to report a side effect or when a new laboratory abnormality arises. For HFrEF, these sentiments may permeate medical decision-making via assuming patient intolerance and anecdotal clinical experiences with hypotension, worsening renal function, or hyperkalaemia with escalation of GDMT. However, in the management of patients with HFrEF, it is also critical to consider the risks of not prescribing medications of proven benefit (Figure 1).7 Although vital to informed decision-making, these ‘risks of omission’ may receive much less emphasis. Instead, ‘stable’ symptoms are easily misconstrued as ‘low risk’ and justification to not escalate GDMT as tolerated. Likewise, for patients with a recent HF hospitalization, many clinicians consider the recent clinical instability as justification for not starting beneficial therapies at that time, preferring to wait for some future date. Both of these practices ignore the natural history of HF and risks of sudden death. In sharp contrast, when such patients are subsequently hospitalized or die from HFrEF, it is perhaps then convenient and human nature for clinicians and families to reference the progressive and high-risk nature of HFrEF, rather than the lack of optimized lifesaving therapy upstream of the event, or the potential for the event to have been prevented. In this context, in this issue of the Journal, Butler et al.8 present the prescription patterns of GDMT before and after a worsening HF event (WHFE) among younger adults <65 years old. The US claims-based analysis leveraged the IBM® MarketScan® database to identify patients with a diagnosis of chronic HFrEF who had both longitudinal medication data and a WHFE, defined as a HF hospitalization or outpatient intravenous diuretic use. The authors then assessed use of GDMT both before and after the WHFE, as well as outcomes (e.g. readmission, healthcare costs) following the WHFE. Longitudinal examination of the cohort showed overall persistently low rates of GDMT over time with minimal change in relation to the WHFE. Specifically, both before and after the WHFE, few patients were treated with triple therapy with an angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker/angiotensin receptor–neprilysin inhibitor, beta-blocker, and mineralocorticoid receptor antagonist, with rates peaking at only 17.4% in the first 3 months post-WHFE. In contrast, across all time points before and after the WHFE, 41–59% of patients were treated with either 0 or 1 evidence-based therapy. The implications of these gaps in GDMT were then further contextualized with the assessment of patient outcomes and healthcare costs. Following a WHFE, one in eight patients were hospitalized for HF within 30 days, and almost one in two were hospitalized for HF in the following 1 year. Costs directly attributed to HF were high throughout, and peaked at a mean US$4813 per patient per month during the first 3 months after the WHFE. At each time point, >70% of HF costs were attributed to in-hospital care. Butler and colleagues should be congratulated on an elegant analysis that further highlights the clinical inertia and gaps in GDMT that overlie US outpatient care for HFrEF.8 These data are consistent with an earlier study of older US adults with a WHFE, confirming that the unfortunate lack of urgency with medical therapy upstream and downstream of a WHFE extends to younger patients, with gaps in use and titration of GMDT being remarkably similar. Nonetheless, as acknowledged by the authors, limitations of the current study should be noted. First, specific confirmatory data on ejection fraction were not available. Instead, this HFrEF population was defined by ICD codes for systolic HF, which can have varying reliability. Likewise, other key data to confirm treatment eligibility were not available. Specifically, data regarding contraindications, prior intolerance to each GDMT, vital signs, and laboratory results were not reported. Second, longitudinal medication data should be interpreted in the context of patients with complete data, thus excluding patients with interval death who may have different GDMT patterns. Nonetheless, this survival bias may actually make the low rates of GDMT use in this study an overestimate, since ‘sicker’ patients may be at higher risk of under-treatment or medication intolerance. Lastly, this study does not address specific clinician-reported reasons for medication underuse, or the specific reasons or appropriateness of medication discontinuations. Nevertheless, these important data from Butler and colleagues highlight yet again the missed opportunities and potential for patient harm associated with contemporary outpatient management for HFrEF.8 How many of these WHFE could have been avoided with optimal implementation of maximally tolerated or target doses of GDMT? Instead, each year, outpatient HFrEF care embeds patient harm through routine omissions – clinical risk is underappreciated, medication changes are deferred, time is lavished, and patients die or suffer hospitalizations without receiving the therapies proven to prevent these events. Particularly alarming, the current analysis suggests that even the presence of a WHFE is unable to meaningfully break the cycle of clinical inertia. Although specifics surrounding medication eligibility, tolerability, and treatment decision-making were not available, it is notable that rather than triggering a sense of urgency and sparking a meaningful and sustained increase in use of lifesaving therapy, the authors observed minimal overall change in GDMT after a WHFE. Nonetheless, further research is needed to understand the degree to which lack of GDMT escalation following a WHFE is due to persistent clinical inertia and suboptimal quality of care, vs. true medication intolerance and contraindications for which medications with different tolerability profiles may be needed. In summary, the current analysis should serve as an important reminder of the substantial potential for harm associated with not initiating GDMTs simultaneously or in rapid sequence, and not escalating dosing as tolerated.9 With the recent addition of sodium–glucose co-transporter 2 inhibitors as the fourth pillar of comprehensive disease-modifying medical therapy, the opportunity cost of suboptimal implementation of evidence-based therapy has never been higher. Indeed, in randomized clinical trials, each of the four pillars of quadruple therapy offer significant, incremental, and additive reductions in all-cause mortality.10-13 While patients at higher clinical risk (e.g. older patients, multiple comorbidities, recent hospitalization, long durations of HF) are potentially at greater risk of medication intolerance, these patients also generally reap the largest reductions in absolute risk with appropriate GDMT.14, 15 Thus, it is paramount for clinicians and patients to understand the risks of not initiating each therapy proven to significantly improve patient-centred outcomes.7 Likewise, it is important to recognize the harm associated with assuming intolerance to a lifesaving medication and denying the patient any opportunity to benefit. Although shared decision-making and individual patient preferences should always remain central, these conversations must be anchored in scientific evidence and awareness of the ‘harms of omission’ with not trialling a new or higher-dose medication. As another saying goes, ‘change is hard’. Yet, breaking clinical inertia, embracing initiation and titration of multiple medications, and striving to achieve comprehensive disease-modifying medical therapy, as tolerated, remain foundational requirements for improving clinical outcomes in HFrEF. Conflict of interest: S.J.G. has received research support from the American Heart Association, Amgen, AstraZeneca, Bristol-Myers Squibb, Cytokinetics, Merck, Novartis, and Pfizer; has served on advisory boards for Amgen, AstraZeneca, and Cytokinetics; and serves as a consultant for Amgen, Bayer, Merck, and Vifor. G.C.F. serves as a consultant for Abbott, Amgen, AstraZeneca, Bayer, Cytokinetics, Janssen, Medtronic, Merck, and Novartis.