OUP user menu

Heart failure and chronic obstructive pulmonary disease: the challenges facing physicians and health services

Nathaniel M. Hawkins, Sean Virani, Claudio Ceconi
DOI: http://dx.doi.org/10.1093/eurheartj/eht192 2795-2807 First published online: 5 July 2013


Pulmonary disease is common in patients with heart failure, through shared risk factors and pathophysiological mechanisms. Adverse pulmonary vascular remodelling and chronic systemic inflammation characterize both diseases. Concurrent chronic obstructive pulmonary disease presents diagnostic and therapeutic challenges, and is associated with increased morbidity and mortality. The cornerstones of therapy are beta-blockers and beta-agonists, whose pharmacological properties are diametrically opposed. Each disease is implicated in exacerbations of the other condition, greatly increasing hospitalizations and associated health care costs. Such multimorbidity is a key challenge for health-care systems oriented towards the treatment of individual diseases. Early identification and treatment of cardiopulmonary disease may alleviate this burden. However, diagnostic and therapeutic strategies require further validation in patients with both conditions.

  • Adrenergic beta-antagonists
  • Adrenergic beta-agonists
  • Heart failure
  • Chronic obstructive pulmonary disease


Heart failure (HF) and chronic obstructive pulmonary disease (COPD) are leading causes of death worldwide.1 Through shared risk factors and pathogenic mechanisms the conditions frequently coexist, presenting diagnostic and therapeutic challenges for physicians.2,3 Each is an independent predictor of morbidity, mortality, impaired functional status, and health service use. Each is also powerfully associated with socioeconomic deprivation.4,5 The conditions therefore undermine the two fundamental goals of healthcare: to improve both the overall level and distribution of health. Healthcare internationally is dominated by individual disease approaches, lacking coordination and integration.6 Multimorbidity is a key challenge for these health systems.7 We review these challenges and provide direction for future research.


The diagnoses of both HF and COPD require typical symptoms combined with objective evidence of organ dysfunction.8,9 The European Society of Cardiology mandates typical symptoms and signs resulting from any abnormality of cardiac structure or function, including systolic and diastolic dysfunction, valvular, pericardial, and heart rhythm abnormalities.8 The Global Initiative for Chronic Obstructive Lung Disease (GOLD) defines airflow obstruction by FEV1/FVC ratio < 0.70,9 a consensus definition endorsed by the European Respiratory Society and American Thoracic Society (Table 1).10 This simple definition avoids complex reference equations, is understandable, universal, generalizable, comparable, and lowers barriers to diagnosis.11 However, the fixed ratio overestimates disease in the elderly relative to lower limit of normal indices,12 as FEV1/FVC declines in healthy never smokers with advancing age.13

View this table:
Table 1

GOLD classification of airflow limitation severity based on post-bronchodilator FEV1

StageFEV1/FVCFEV1 predicted, %
I: Mild<0.70FEV1 ≥ 80
II: Moderate<0.7050 ≤ FEV1 < 80
III: Severe<0.7030 ≤ FEV1 < 50
IV: Very severe<0.70FEV1 < 30
  • FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity.

From shared risk factors to pathophysiological mechanisms

Cardiologists readily accept the ‘cardiovascular disease continuum’, the hypothesis that frames cardiovascular diseases as a chain of events initiated by risk factors and progressing through numerous physiological pathways to the development of end-stage heart disease and HF.14 Hopefully in coming years, the concept of a ‘cardiopulmonary continuum’, as a common ground of heart and lung disease,15 will be the basis for future achievements. Interest has grown in the association of chronic pulmonary and cardiac diseases; the observation that HF and COPD coexist more frequently than expected from their respective population prevalences being a major reason for this interest. These epidemiological observations encourage new pathophysiological interpretations to understand the connection between the pulmonary and cardiovascular continuum.

Apart from smoking as a common risk factor, patients with COPD share an additional determinant of cardiovascular disease: low-grade systemic inflammation.16,17 The risk of underlying ischaemic heart disease is greatest in patients with airflow obstruction and elevated C-reactive protein.18 Furthermore, almost 50% of patients with COPD present coexisting metabolic syndrome as well as increased levels of systemic inflammatory markers, independent of lung function impairment.19 Diabetes is likewise independently associated with reduced lung function, while obesity may further worsen ventilatory mechanics.20 Diabetes, metabolic syndrome (and its individual components), and physical inactivity are all major determinants of cardiovascular disease. The fact that each also acts through pro-inflammatory mechanisms strengthens the view that low-grade systemic inflammation is a common pathophysiological link between COPD and cardiovascular diseases.21

This concept is substantiated by the observation that direct pro-inflammatory agents, namely cigarette smoking, air pollution, and occupational exposures, induce systemic cellular and humoral inflammation, oxidative stress, striking changes of vasomotor and endothelial function, and enhanced circulating concentrations of several procoagulant factors.2225 Thus, ‘cardiovascular’ and ‘pulmonary’ risk factors create a ‘cardiopulmonary continuum’ through shared systemic inflammatory processes, inducing cascades of events which underpin chronic diseases including COPD, coronary disease, and HF.15,26,27 This pathophysiological relationship revolves around common inflammatory pathways (Figure 1). TNF-α plays a central role in chronic inflammatory processes of the pulmonary and cardiovascular system and is involved in the activation of secondary mediators including C-reactive protein and pro-inflammatory cytokines. These in turn mediate the phenotype of syndromes characterized by chronic comorbidities and peripheral abnormalities.26,28,29

Figure 1

Inflammatory pathways involved in the cardiopulmonary continuum. Patients with chronic obstructive pulmonary disease have peripheral lung inflammation. These patients have also increased circulating cytokines, including interleukin (IL)- 1b, IL-6, IL-18, and TNFa, as well as acute-phase proteins, such as C-reactive protein and serum amyloid A (SAA). This low grade chronic inflammation may represent the link with the increased propensity to cardiovascular, metabolic, bone, neurological diseases as well as the peripheral muscle abnormalities typical of the syndrome (From Boschetto et al., 2012).27

A challenge for the future will be to understand the role of genetic factors, their genotype–phenotype correlations, and clinical implications. The recent availability of techniques such as genome-wide association studies has increased identification of susceptibility genes for both COPD and cardiovascular disease phenotypes including hypertension, dyslipidaemia, and coronary artery disease.30,31 A functional link between susceptibility genes has been elucidated only for certain risk factors, such as the low-density lipoprotein receptor defects and other hyperlipidaemia disorders.31,32 While the association among the most annotated genes and risk factors is defined, heritability and phenotype are less clear: for COPD, several genes are associated with disease susceptibility (e.g. CHRA3–5 or FAM13A9),30,33,34 but the population attributable risk is limited (approximately below 15%) and no functional link clearly established. Notably, no potential common candidate gene for both COPD and cardiovascular disease phenotypes has yet been identified. The interplay between the two common multigenic diseases remains elusive. Thus far, our understanding of the interaction between cardiovascular and pulmonary disease derives from registries and subgroup analyses. No prospective study has addressed the specific role of pulmonary comorbidity in the treatment and outcomes of cardiovascular disease patients. Nevertheless, in large, retrospective analyses statins and/or angiotensin-converting enzyme inhibitors improve both cardiac and pulmonary outcomes in patients with COPD, with the largest benefits obtained by combining therapies.35 Cardiovascular medicine has an armamentarium of survival enhancing therapies. These need to be systematically tested in large randomized controlled trials in patients with COPD, with and without overt cardiovascular disease.

Prevalence and prognosis of heart failure and chronic obstructive pulmonary disease

The prevalence of GOLD stage II or higher COPD is ∼5–10% of adults.3639 One-year mortality in the community is relatively low (around 3%),40 but higher following hospitalization (25%).41,42 Future projections have wide uncertainty intervals, depending on statistical methodology, estimates of prevalence, and associated mortality.43 Nevertheless, the Global Burden of Disease study anticipates COPD to become the third leading cause of death globally by 2020.44

Heart failure is less common, affecting 1–3% of the general population, but carries a worse prognosis.45 Annual mortality in stable community patients approximates 5–7%,4548 while median survival following hospitalization remains just 2 years.49 Approximately half of patients have HF with preserved ejection fraction and half left-ventricular systolic dysfunction (LVSD).50 The prevalence of the latter is thus around 1% of the general population.

Prevalence of concurrent heart failure and chronic obstructive pulmonary disease

Prevalence estimates vary widely according to cohort selection, population age structure, risk factor exposure, diagnostic criteria, measurement methods, and surveillance systems.2,51 From around 10–40% of patients with HF have reported concurrent COPD.2 However, only a handful of studies employed spirometry.5255 In the largest of these, COPD was diagnosed in 36% of 532 consecutive patients hospitalized with HF, acknowledging that pulmonary oedema may have contributed to airflow obstruction.53 A similar proportion (30%) was observed in a recent prospective study of consecutive patients with stable HF.55

Estimates of HF prevalence in COPD are likewise sparse though consistent. The prevalence of unrecognized HF was 20.9% in patients with COPD or asthma presenting to the emergency department,56 20.5 and 17% in community patients with stable COPD.57,58 The latter two studies included echocardiography in all patients, detecting LVSD in 10.4 and 13.8%. These findings have significant ramifications. Given COPD is far more prevalent than LVSD in the general population, then COPD is potentially masking a large proportion of patients with LVSD.

Prognosis of concurrent heart failure and chronic obstructive pulmonary disease

Concurrent COPD independently predicts mortality in patients with reduced and preserved ejection fraction (Table 2),2,59 even following adjustment for beta-blocker utilization.5965 Furthermore, greater airflow obstruction is associated with worsening survival.52,62,66 In those studies with sufficient sample size to examine the cause of death, excess non-cardiovascular death predominated.64,65 However, a clinical diagnosis of COPD was also associated with sudden death in the Valsartan in Acute Myocardial Infarction Trial.64

View this table:
Table 2

Studies specifically examining the prognostic implications of chronic obstructive pulmonary disease

ReferencesnStudy designPrevalence COPD (%)SpirometryMean LVEF (%)OutcomeMean follow-upAdjusted risk (±95% CI)
Macchia et al.1491020Retrospective24NoMortality287 days1.42 (1.09–1.86)
Staszewsky et al.655010Prospective13No27Non-CV mortality23 months2.50 (1.58–3.96)
Mascarenhas et al.52186Retrospective39YesDeath/hospitalization433 days2.10 (1.05–4.22)
Rusinaru et al.59799Prospective20No50Mortality5 years1.53 (1.21–1.94)
Hawkins et al.6414 703Prospective9No35Mortality24.7 months1.14 (1.02–1.28)
Lainscak et al.150638Retrospective17Yes43Mortality1062 days1.38 (1.04–1.83)
Kwon et al.66184Retrospective37Yes49Mortality GOLD III731 days3.20 (1.33–7.68)
De Blois et al.634132Prospective17No32Mortality13.3 months1.19 (1.02–1.39)
Mentz et al.15120 118Prospective25No25Mortality60 days0.97 (0.68–1.38)
Mentz et al.614133Prospective10No27Mortality9.9 months1.17 (0.96–1.42)
Boschetto et al.55118Prospective30Yes40Mortality1029 daysNon-significant
Arnaudis et al.62348Prospective38Yes31Mortality GOLD II54.9 months2.27 (1.22–4.25)

A single study has examined the prognostic implications of concurrent HF in patients with COPD.67 Following extensive investigations including echocardiography and pulmonary function tests, 83 of 405 elderly community patients with COPD were diagnosed with HF (20.5%). Patients with HF had double the mortality of those without HF over a mean follow-up of 4.2 years (adjusted HR 2.1 [1.2–3.6]).

Diagnostic challenges

Patients with HF or COPD exhibit a high symptom burden including dyspnoea, orthopnoea, cough, exercise intolerance, fatigue, muscle weakness, disturbed sleep, anorexia, low mood, and anxiety.68,69 Unfortunately, neither symptoms nor signs are unique to either condition. Chest radiograph may be equally misleading, as pulmonary vascular remodelling in those with COPD either mimics (upper lobe venous diversion)70 or masks pulmonary oedema (asymmetric, regional, and reticular patterns).71,72

Pulmonary function tests

Restrictive ventilation is the hallmark of chronic stable HF,73,74 reflecting cardiomegaly,75 respiratory muscle weakness,76,77 and interstitial fibrosis.78 Backward transmission of increased left-atrial pressure disrupts vascular structural integrity and functional properties.79 Chronic pulmonary capillary stress failure induces local activation of growth stimuli, thickening of the alveolar–capillary and microvascular remodelling. The resulting reduction in alveolar–capillary membrane conductance and lung diffusion capacity impairs gas transfer.80 Simultaneously, pulmonary hypertension, right-ventricular dysfunction and failure gradually develop. This lung remodelling is compounded by COPD (Figure 2). Compared against 69 matched subjects with HF alone, 69 patients with concurrent COPD exhibited impaired cardiopulmonary exercise response. All variables were poorer, including peak oxygen consumption (12.1 vs. 16.3 mL kg−1 min−1), minute ventilation/carbon dioxide production slope (42.7 vs. 33.3), pulmonary artery systolic pressures (51.9 vs. 37.0 mmHg), and 6 min walking distance (295 vs. 367 m) (all P < 0.001).81 A significantly shorter 6 min walking distance was also recently observed in 48 of 174 elderly patients with HF and concurrent COPD (276 vs. 291 m, P < 0.05).54,81

Figure 2

Cardiopulmonary pathophysiology and interactions in patients with heart failure and pulmonary disease.

Airflow obstruction, as opposed to (or superimposed on) restriction, is typical in decompensated acute pulmonary oedema. Interstitial and submucosal oedema compress and obstruct airways, respectively, compounded by bronchial hyperresponsiveness.82,83 Both misdiagnosis and overestimation of COPD severity may ensue. FEV1 improves by 11–34% with diuresis and often normalizes (Figure 3).8387 This is particularly relevant for the elderly HF population, in whom the aforementioned fixed FEV1/FVC diagnostic criterion already over diagnoses pulmonary disease.12,13

Figure 3

Changes in forced expiratory volume in 1s following treatment of decompensated heart failure.

Natriuretic peptides

Natriuretic peptides are raised in pulmonary hypertension,88 right-ventricular failure or acute stress,89 and pulmonary disease.89,90 Typical levels overlap with compensated HF (Figure 4). This increases false positive results, thus reducing the specificity and positive predictive value for LVSD. Only one study has examined the ability to detect HF in stable patients with COPD.90 In 200 elderly patients, four natriuretic peptide assays were comparable and excluded HF with reasonable accuracy (all negative predictive values above 0.85). However, specificity and positive predictive value were much lower, around 0.6 and 0.4, respectively.90 The logical solution is to apply natriuretic peptides and a more specific test in sequence. To date, this strategy has not prospectively been tested.

Figure 4

Overlap in natriuretic peptide levels between patients with heart failure and chronic obstructive pulmonary disease.


Echocardiographic acoustic windows may be impeded by air trapping in pulmonary disease. The estimated prevalence of unsatisfactory image quality varies, from 10% in stable primary care patients with COPD91 to 35% in severe disease and 50% in very severe airflow obstruction.92,93 The generalizability of these findings is unclear, as most studies utilizing echocardiography in patients with COPD fail to report feasibility. Moreover, the incremental improvement with contrast echocardiography is unknown. Studies are needed comparing echocardiographic modalities with cardiac magnetic resonance imaging in this population.

Should patients with chronic obstructive pulmonary disease be screened for heart failure?

Asymptomatic LVSD fulfils the Wilson–Junger criteria for successful screening:94 the condition is medically important and clearly defined, with effective treatment, an established natural history and early asymptomatic stage, detectable by an acceptable and safe, valid and reliable test applicable to the target population, resulting in cost-effective screening for which adequate resources exist.95 The predictive value of screening is critically dependent on disease prevalence.96 The high prevalence of LVSD in patients with COPD thus favours success. Sequential peptide testing (sensitive to rule out disease) and then echocardiography (specific) may counter the inherent limitations of peptides or echocardiography alone in pulmonary disease. Robust randomized trials are required to determine the efficacy and cost-effectiveness of such screening. Such a trial is currently being planned.

Therapeutic challenges

Beta-blockers, angiotensin-converting enzyme inhibitors, and aldosterone antagonists reduce hospitalizations, morbidity, and mortality in HF. The cornerstone of therapy for COPD is long-acting inhaled bronchodilators, either anticholinergic, beta-agonist, corticosteroid, or combination therapy.97100 The opposing pharmacological effects of beta-blockers and beta-agonists underpin a reticence to prescribe the former in COPD or the latter in HF.3

Beta-blockers and chronic obstructive pulmonary disease

Three recent small randomized controlled trials have examined beta-blockers in patients with HF and concurrent COPD.101103 Prior to these, a Cochrane meta-analysis evaluated 20 randomized, controlled, cross-over trials of cardioselective beta-blockers in patients with COPD alone.104 The included studies were limited in size; some lacked double-blinding or placebo controls; 11 trials involved only a single treatment dose; predominantly young male patients with moderate airflow obstruction were recruited; symptoms were rarely reported; and only one study lasted longer than a month.3 Within the bounds of these significant limitations, cardioselective beta-blockade exerted minimal impact on reversible or severe airflow obstruction.104

The recent trials in patients with both conditions are consistent. In a double-blind, placebo controlled trial in 27 patients with concurrent moderate-to-severe COPD (mean FEV1 1.3L), FEV1 was reduced after 4 months treatment with bisoprolol compared with placebo (−70 vs. +120 mL).101 A subsequent open label triple cross-over trial included 35 patients with coexistent COPD.103 FEV1 was highest with bisoprolol (2.0 L), intermediate with metoprolol (1.94 L), and lowest with carvedilol (1.85 L). Finally, FEV1 increased significantly with bisoprolol but not carvedilol in another open label study recruiting 63 elderly patients with concurrent moderate-to-severe COPD.102 Without placebo control, the improvement with bisoprolol may simply reflect more intensive medical therapy during the study. Nevertheless, in all three studies, cardioselective beta-blockade was well tolerated. Of note, in all three studies, the FEV1 reduction of around 150 mL with beta-blockade (either bisoprolol vs. placebo or carvedilol vs. bisoprolol) equals or exceeds the improvement in landmark bronchodilator trials. The recent CIBIS-ELD trial reinforces these observations. Even among elderly patients largely free of pulmonary disease, carvedilol reduced FEV1 compared with bisoprolol (adjusted mean difference 50 mL).105 Given mortality was unaltered in any bronchodilator trial, this effect size appears a reasonable exchange for the unequivocal survival benefits of beta-blockade.

The protective effect of beta-blockers may extend across a broad spectrum of patients with COPD (Table 3). Ten retrospective cohort studies have demonstrated an association between beta-blockers and improved survival, in patients with COPD and concurrent HF,65,106 myocardial infarction,64,107,108 vascular disease,109 hypertension,110 and most recently unselected community patients with COPD.111,112 The pooled relative risk of mortality related to beta-blockers in a recent meta-analysis was 0.69 (95% CI 0.62–0.78).113 Though observational with obvious confounding, evidence is accumulating to support randomized controlled trials of beta-blockers in COPD. Unthinkable? No more so than the paradigm shift two decades ago with regard to beta-blockers in HF.

View this table:
Table 3

Association between beta-blockers and all-cause mortality in patients with chronic obstructive pulmonary disease in observational cohort studies

ReferencesPopulationn with COPDFollow-upAdjusted risk (±95% CI)
Sin et al.106Heart failure3834Median 21 months0.78 (0.63–0.95)
Staszewsky et al.65Heart failure628Median 23 months0.55 (0.37–0.82)
Hawkins et al.64Myocardial infarction1258Median 25 months0.74 (0.68–0.80)
Gottlieb et al.108Myocardial infarction41 8142 years0.60 (0.57–0.63)
Chen et al.107Myocardial infarction10 9881 year0.86 (0.73–1.00)
Van Gestel et al.109Vascular disease1205Median 5 years0.73 (0.60–0.88)
Au et al.110Hypertension19662 years0.57 (0.33–0.89)
Rutten et al.111COPD primary care22307.2 years0.68 (0.56–0.83)
Dransfield et al.152COPD exacerbation8250.39 (0.14–0.99)
Short et al.112COPD primary care5977mean 4.35 years0.78 (0.67–0.92)

Beta-agonists and heart failure

Paradoxically, as evidence accumulates of beta-blocker efficacy in COPD, observational studies point to adverse associations between beta-agonists and HF. Beta-agonists are associated with incident HF in patients with pulmonary disease, and with increased mortality and HF hospitalization in those with existing HF or LVSD (Table 4). These observations have sound physiological rationale. Down-regulation of beta1-receptors with the preservation of the beta2 subpopulation increases the sensitivity of the failing myocardium to beta2-agonists.114,115 The adverse effects of beta-agonists are numerous: ischaemia, arrhythmias, tachycardia, hypokalaemia, QT prolongation, disturbed autonomic modulation.116120 However, association is not causation. The poor outcomes attributed to beta-agonists may reflect the underlying pulmonary disease, smoking burden, and associated pathologies, clustering of cardiovascular risk factors and disease in patients with COPD, or residual confounding by unmeasured covariates. The most recent cohort study supports this standpoint.121 In 1294 outpatients with HF, beta2-agonist users were older, more often male, with more frequent smoking history, coronary artery disease, pulmonary disease, and higher heart rates. Although unadjusted mortality rates for beta2-agonist users were significantly higher than non-users, adjusted mortality rates were similar after comprehensive multivariable adjustment (HR 1.04 [0.77–1.41]). Nevertheless, doubts regarding the cardiovascular safety of beta-agonists have prompted calls for randomized controlled trials.3,122 The largest such trial is examining all-cause mortality in 16 000 patients with moderate COPD and established or high risk of cardiovascular disease randomized to four treatment groups: fluticasone/vilanterol combination, fluticasone alone, vilanterol alone, or placebo.123

View this table:
Table 4

Association between beta-agonists and incident heart failure, hospitalization, and mortality

ReferencesPopulationnStudy designBronchodilator and routeFollow-upOutcomeRisk associated with bronchodilator use [95% CI]Adjustment includes beta- blockade
Martin et al.153Asthma8098CohortBambuterol oralMedian 288 daysIncident HFRR 3.41 [1.99–5.86], P < 0.0001No
15 407CohortSalmeterol inhaledMedian 511 daysIncident HFRR 1.10 [0.63–1.91], P = 0.7No
Coughlin et al.154General population387Case–controlβ-Agonist oral20 monthsIncident DCMOR 3.4 [1.1–11.0]No
387Case–controlβ-Agonist inhaled/nebule20 monthsIncident DCMOR 3.2 [1.4–7.1]No
Sengstock et al.155Cardiology clinic190Case–controlβ-Agonist inhaledIncident DCMOR 1.0No
Macie et al.156COPD or asthma59 336Case–controlβ-Agonist inhaled5 yearsHF hospitalizationOR 1.74 [1.60–1.91]Yes
Au et al.157HF1121Case–controlβ-Agonist inhaled2.5 yearsHF hospitalizationOR 1.5 [0.8–2.8] 1–2 canisters
OR 2.1 [1.0–4.3] ≥3 canisters
Medical clinics13 012Case–controlβ-Agonist inhaled2.5 yearsHF hospitalizationOR 1.3 [0.9–1.8] 1–2 canisters
OR 1.1 [0.8–1.6] ≥3 canisters
Au et al.158LVSD1529Cohortβ-Agonist inhaled1 yearDeathRR 0.9 [0.5–1.6] 1 canister/month
RR 1.4 [0.9–2.2] 2 canister/month
RR 2.0 [1.3–3.2] 3 canister/month
Singer et al.159Acute HF without COPD7299CohortAny bronchodilator inhaledInpatientDeath IV vasodilator ventilationOR 1.02 (0.67–1.56)
OR 1.40 (1.18–1.67)
OR 1.69 (1.21–2.37)
Bermingham et al.121HF1294Cohortβ-Agonist inhaledMean 2.9 yearsMortalityHR 1.04 (0.77–1.41)Yes
  • CI, confidence interval; COPD, chronic obstructive pulmonary disease; DCM, idiopathic dilated cardiomyopathy; HF, heart failure; HR, hazard ratio; IV, intravenous; LVSD, left ventricular systolic dysfunction; OR, odds ratio; RR, relative risk.

Concurrent beta-blockers and beta-agonists

Beta-blockers and beta-agonists exert opposing pharmacological effects. The immediate clinical impact of this interaction largely depends on cardioselectivity. Non-cardioselective beta-blockers incontrovertibly antagonize beta-agonist-mediated bronchodilatation to varying degrees.3,104 Conversely, beta-agonist response is preserved alongside cardioselective beta-blockers.101,103,104,124 The key question is whether any interaction translates into meaningful clinical endpoints. A small number of retrospective cohort studies have tested for statistical interaction between beta-blockers and bronchodilators, with conflicting results.107,120,125 Unfortunately, the lack of stratification according to cardioselectivity, selection bias, and unavoidable confounding renders such analyses uninterpretable.


Among patients presenting with acute dyspnoea, prompt diagnosis and treatment is paramount to improved outcomes and decreased healthcare utilization.126 This is particularly challenging in patients with combined diagnoses of COPD and HF, given the overlapping clinical signs and symptoms which characterize an acute exacerbation of either disease state.2 Despite these diagnostic pitfalls, a comprehensive history and physical examination has predictive accuracy to identify HF among acutely dyspnoeic patients. In a meta-analysis of 22 publications, a history of HF was associated with a positive likelihood ratio of 5.8, while the presence of classic signs and symptoms of volume overload conveyed likelihood ratios ranging from 2.0 to 11.0 for HF as a cause of the patient's acute dyspnoea.127 In this regard, clinical gestalt remains an important diagnostic tool for the astute physician.

Natriuretic peptides provide additional diagnostic accuracy and may be the most effective tool in refining the diagnosis for this patient population given the high negative predictive value of a normal result. In the Breath Not Properly study,128 the negative predictive value of a B-type natriuretic peptide level <50 pg/mL in excluding acute HF was 96%. COPD, through the aforementioned effects on pulmonary pressures and right-ventricular function, may cause elevated circulating levels of natriuretic peptides.8890 Despite this potential confounding, the negative predictive accuracy in ruling out acute decompensated HF appears to be preserved even in cohorts of patients with a dual diagnosis.56,129 Beyond improving diagnostic accuracy, a BNP-guided treatment strategy may also improve patient care through early initiation of targeted treatments and less requirement for confirmatory diagnostic testing.126 This has significant implications both in terms of patient and system-related outcomes.

Economic costs

Individuals with chronic medical conditions, including HF and COPD, have higher rates of healthcare utilization when compared with age- and sex-matched healthy controls.130132 This observation is consistent across healthcare systems globally despite variations in clinical practice patterns, the patient population being evaluated, the cost of tests and services employed, and the underlying strategy for provider reimbursement.51,133135 Given that the natural history of both HF and COPD is characterized by frequent episodes of acute exacerbation, increased healthcare utilization is primarily attributable to unplanned admission for disease decompensation.132,136,137 This is particularly apparent in patients with more advanced symptoms and/or those with a dual diagnosis, who are at highest risk for hospitalization.138145 Diagnostic testing, medications, out-patient clinical services, and medical provider visits also contribute significantly to system related costs for both disease states.134,141

To provide granularity to this issue of increased healthcare utilization, in 2005, the US Congressional Budget Office confirmed that high-cost Medicare beneficiaries were responsible for 85% of the expenditures associated with this entitlement program but comprised only 25% of the beneficiaries. Moreover, 30% of high-cost beneficiaries had a combined diagnosis of HF and COPD and ∼75% of this group had at least one hospitalization episode over the course of a given year.146 These findings highlight the impact of co-existent HF and COPD with respect to clinical outcomes, risk of hospitalization, and healthcare expenditures.

Understanding the type and frequency of healthcare utilization among patients with multiple chronic disease diagnoses is paramount to developing strategies which improve patient- and system-related outcomes. Developing such strategies entails a robust characterization of patients, their co-morbidities, and the mechanisms through which they interact. Given the high prevalence and risk of hospitalization among individuals with a combined COPD and HF diagnosis, interventions to reduce unscheduled admission can lead to major cost savings for healthcare systems. To achieve this end, an emphasis on chronic disease management strategies which encompass patient self-management and evidence-based treatment algorithms, and which reduce the frequency of disease exacerbation must be employed to ensure efficient resource utilization.

Clinical guidance

The diagnostic and therapeutic complexity of the cardiopulmonary continuum is best addressed through simple clinical measures. In patients with suspected or established cardiopulmonary disease, clinical acumen should be reinforced by natriuretic peptides and objective measures of organ dysfunction. Pulmonary function testing when euvolaemic is mandatory, as is echocardiographic or cardiac magnetic resonance imaging. Beta-blockers markedly improve symptoms and survival across the cardiovascular spectrum, from coronary disease to HF. The European Society of Cardiology guidelines are explicit on this matter: COPD is not a contraindication to beta-blockade.8 Moreover, severe airflow obstruction is actually uncommon, mild and moderate airways disease being the norm.62 In patients with symptomatic LVSD and heart rate ≥70 beats per min receiving the maximum tolerated dose of beta-blocker (or truly intolerant), ivabradine should be considered to reduce the risk of HF hospitalization.8,147 Beta-agonists remain a cornerstone of symptomatic relief. However, it seems sensible to base treatment regimens on long-acting anti-muscarinic agents, given the superiority of tiotropium to salmeterol in reducing exacerbations.3,100

Conclusion and direction for future research

The interaction between the heart and lungs is complex and incompletely understood. As with the kidney, direct physical connection and systemic neurohormonal and inflammatory activation are responsible.148 Understanding these pathways may yield novel therapies. However, many of the challenges facing physicians and healthcare systems have potential to be addressed through existing resources. Comprehensive large-scale cohort studies and trials are required: establishing the prevalence and clinical consequences of the respective coexistent condition; prospectively testing diagnostic strategies; establishing the efficacy and cost-effectiveness of screening; undertaking randomized controlled trials of beta-blockers and beta-agonists across a spectrum of pulmonary and cardiovascular diseases, respectively; and determining the efficacy of integrated chronic disease management strategies. Such far reaching studies require closer collaboration between the cardiovascular and pulmonary communities. They also require greater financial support from governments, research funders, and healthcare payers. Translational medicine is expensive; so too are the consequences of ignoring the challenges of multimorbidity.


Conflict of interest: none declared.


View Abstract