Chronic Obstructive Pulmonary Disease and Heart Failure in Europe—Further Evidence of the Need for Integrated Care

This article refers to ‘Characteristics, treatments and 1-year prognosis of hospitalized and ambulatory heart failure patients with chronic obstructive pulmonary disease in the European Society of Cardiology Heart Failure Long-Term Registry’ by M. Canepa et al., published in this issue on pages 100–110. In this issue of the Journal, Canepa et al.1 report the characteristics, treatments and 1-year outcomes of 1322 ambulatory and 1334 hospitalized patients with heart failure (HF) and concurrent chronic obstructive pulmonary disease (COPD). The European Society of Cardiology (ESC) HF registry offers a unique and detailed exploration of contemporary practice in a large, consecutively sampled, multinational population. From the wealth of information presented some key messages are apparent. Arguably the foremost finding of the study, certainly in terms of opportunity for improvement, is the remarkably low use of spirometry. Despite being a fundamental criterion for COPD diagnosis and recommended by all pulmonology guidelines,2 only 17% of hospitalized and 31% of ambulatory patients with HF and concurrent COPD had documented spirometry. This corroborates a previous retrospective chart review in which 36% of patients discharged with HF and COPD underwent spirometry.3 Poor confirmatory practices expose the patient to double jeopardy from either under- or over-diagnosis of COPD, each with potential detrimental consequences. Up to 30% of patients with stable HF have unrecognized airflow obstruction, while a similar proportion are mislabelled with COPD but have no confirmed airflow obstruction.4 The former deprives patients of bronchodilator therapy and improvements in symptoms, quality of life and frequency of exacerbations.2 The latter is associated with underutilization of survival enhancing HF medications, unnecessary prescription of bronchodilators, and excess healthcare costs. Although evident, the well-recognized challenge of beta-blocker underutilization displayed some signs of improvement. Respectively comparing COPD vs. no COPD, beta-blocker use increased from hospital admission to discharge by 12% (51% to 63%) vs. 21% (56% to 77%), and pre- to post-ambulatory visit 5% (77% to 82%) vs. 5% (85% to 90%). Thus, in hospitalized patients beta-blocker use increased most, although levels remained lower overall and the relative treatment gap associated with COPD actually increased. However, this analysis included all patients irrespective of left ventricular ejection fraction (LVEF). Beta-blocker use was higher in those with LVEF ≤45% and definite guideline indication: at hospital discharge 70% vs. 80%, and after outpatient visit 88% vs. 94% comparing COPD vs. no COPD. This far exceeds levels observed in older cohorts, and even compares favourably to clinical trials.5, 6 The registry provides further granularity regarding choice of beta-blocker and reported intolerance. Bisoprolol was the most commonly used beta-blocker, and a small proportion (5%) of patients shifted from carvedilol to bisoprolol or nebivolol during hospitalization. These prescribing patterns suggest the message that cardio-selective beta-blockers are well tolerated in COPD is translating into clinical practice. A Cochrane library meta-analysis in 2005 evaluated pulmonary function in 20 randomized, controlled, crossover trials of cardio-selective blockers in patients with COPD.7 The studies included were limited in terms of population size, controls, blinding and follow-up duration.8 Nevertheless, cardio-selective beta-blockade appeared safe and well tolerated in COPD, including those with severe or reversible airflow obstruction. The 2008 ESC HF guidelines applicable to the ESC-HF registry in 2011 provided explicit guidance with respect to beta-blockers in pulmonary disease, advocating cardio-selective beta-blockers, low-dose initiation, gradual up-titration, and continuation even with mild deterioration in symptoms or pulmonary function.9 Approximately one third of patients with COPD in the registry were prescribed carvedilol, which is entirely reasonable in those with milder airflow obstruction. Among the more interesting findings were the reasons for not receiving a beta-blocker. Given the clustering of cardiovascular disease in COPD, it might be expected that all common ‘contraindications’ would likewise cluster. However, the converse was true, with bradyarrhythmia, hypotension and worsening HF all occurring less than half as often in patients with COPD compared to those without. Equally surprising, the primary reason for beta-blocker non-prescription was asthma/bronchospasm in approximately half of intolerant patients with COPD, compared to only 9–10% of those without COPD. Whether this represents genuine intolerance, misdiagnosis or physician's cautionary instinct to ‘do no harm’ is uncertain. A detailed chart review has never been undertaken in a sufficiently large cohort to understand the circumstances by which bronchospasm is assigned as an intolerance. Inhibition of the renin–angiotensin–aldosterone system also has recognized inequalities associated with COPD. The Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF) enrolled patients from 2003 to 2004, 25% of whom had concurrent COPD.10 Although angiotensin-converting enzyme inhibitor/angiotensin receptor blocker (ACEI/ARB) and aldosterone antagonist use was similar on admission between patients with and without COPD, those with COPD were less likely to be initiated on either class of drugs. A similar phenomenon was apparent for ACEI in the ESC-HF registry. The analysis by Canepa et al.1 reveals the ACEI underutilization was attributed to cough intolerance in the COPD population. This most likely represents mislabelling of a common COPD symptom, as ACEI induced bronchial reactivity is not appreciably worse in patients with respiratory dysfunction.11 The ACEI underutilization was only partially offset by greater ARB use. Aldosterone antagonism was fortunately more encouraging, with greater use in patients with COPD compared to those without, particularly in the ambulatory setting (62.9% vs. 58.2%). COPD is a systemic inflammatory disease with putative cardiovascular effects mediated by chronic inflammation.12 However, the analysis by Canepa et al.1 confirms findings from previous registries and clinical trials that COPD is more strongly associated with HF hospitalization than cardiovascular mortality (at least in established HF).13 There is of course no way to conclude causality from observational data. However, the attenuation of the high unadjusted risk for most outcomes to non-significance by multivariable adjustment has several implications. First, a good proportion of confounding variables are likely accounted for in this detailed dataset. Second, clustering of co-morbidities as opposed to the COPD disease state per se appears largely responsible for adverse outcomes. Finally, the persistent increased adjusted risk of HF hospitalization associated with COPD probably reflects a genuine interaction between the two conditions. The reported increased hazard (22% in hospitalized and 37% in ambulatory patients) is consistent with prior clinical trials and registries (19% to 40%).6, 13, 14 Many factors may contribute, including the aforementioned underutilization of beta-blockers and ACEI. An element of misclassification is also inevitable, whereby acute exacerbation of COPD is misdiagnosed as worsening HF. However, pulmonary infection is also a well-recognized precipitant of HF decompensation.15 The registry sheds further light on the reasons for increased hospitalization. Not only infection, but renal dysfunction, uncontrolled hypertension, bradyarrhythmia and ventricular arrhythmia were all more common in patients with COPD. Thus, to reduce the burden of HF hospitalization in patients with COPD a broad strategy to address multi-morbidity will be required. The cornerstone of any strategy should be optimal management of COPD. Pharmacological therapy reduces the frequency and severity of exacerbations, and improves symptoms, health status, and exercise tolerance.2 As with the real-world utilization of spirometry, the registry suggests our management of COPD has considerable scope for improvement. Only around half of patients classified as having COPD were prescribed one drug for COPD, with approximately one third receiving beta-2 agonists and one third anticholinergic agents in the ambulatory setting. Moreover, this exaggerates treatment levels, since anyone prescribed COPD medications was automatically assigned a diagnosis of COPD. This cohort definition, though pragmatic, significantly limits the analysis as it precludes assessment of inappropriate prescribing of bronchodilators and other agents in patients with HF but no COPD. What do these findings mean for future research and current practice? Although treatment inequalities and disparities in outcomes are improving, there is no room for complacency. We must continue our efforts in education, quality improvement, audit and registries such as the present study. However, it is our management of COPD, both diagnostic and therapeutic, which is clearly the greatest deficit. The pulmonology community increasingly recognizes the importance of HF in COPD, and the urgent need for more integrated care.16 How this should be implemented is a matter for debate. Collaboration could include screening for the respective diseases, improving access to diagnostic tests, creating combined cardiopulmonary clinics and rehabilitation, developing common remote monitoring platforms, cross-specialty registries using standardized definitions, bundled care packages with novel funding strategies, and collaborative quality improvement initiatives. The ultimate question to be answered is whether these will deliver better outcomes for patients, and value for the healthcare system. Conflict of interest: none declared.