The increasing burden of valvular heart disease

The epidemiology of valvular heart disease (VHD) has changed dramatically over the past 50 years in developed nations. While rheumatic disease is now uncommon, a steady rise in life expectancy has been accompanied by a progressively increasing frequency of degenerative valve disease. Although an accurate overall picture of the epidemiology of VHD has yet to be established, the overall prevalence in the USA is 2.5% with wide age-related variation from 0.7% in those aged 18–44 years to 13.3% in those over the age of 75.1 Similar age dependency has been shown in hospital-based surveys in Europe.2 3 Population statistics worldwide project an imminent increase in the elderly population—in the UK, for example, there will be an estimated 4 million people aged between 75 and 84 in the UK by 2018, and the population beyond the age of 85 is set to double by 2028. VHD is associated with significant morbidity and mortality,2 and treatment, particularly by valve surgery, is expensive. Furthermore, research concerning the epidemiology, pathophysiology and clinical management of VHD is limited. Mechanisms underlying aortic stenosis are poorly elucidated and those involved in mitral regurgitation are virtually unknown.

Accurate data concerning the contemporary prevalence and natural history of VHD are required to inform economists and policy makers responsible for healthcare planning and the allocation of resources to newer developments, such as percutaneous valve implantation and repair. The recent National Adult Cardiac Surgery Database report for Great Britain and Ireland provides important detail not available in more general surveys—including a significant increase in all types of valve surgery and variation in the type of surgery over time and between hospitals—and highlights limitations and geographical variation in the current provision of healthcare for VHD.4

The care pathway for VHD

VHD may be detected as an incidental finding in the asymptomatic patient or manifest at a later stage in the natural history with symptoms of dyspnoea, chest pain or arrhythmia. Many patients present initially to primary care, but others access secondary or tertiary care directly as an urgent or emergency admission. Timely detection, referral and assessment are essential and facilitated by the availability and universal implementation of clearly articulated and adopted clinical pathways. Variation highlighted in the recent surgical report suggests this is currently not the case—public education and modification of access routes are sorely needed to remove inequity and improve outcomes.

Current international guidelines for the management of VHD assume the ready availability of specialist clinical assessment and accurate quantitative echocardiography. This is particularly important in determining the timing of surgery for degenerative valve disease where ‘watchful waiting’ is only safe with meticulous follow-up, requiring the availability of expert echocardiography in a non-invasive department with tight protocols, low interobserver variability and quality assessment. These are not universally available, but are part of the accepted global standards for echocardiographic departmental accreditation.

Early symptoms and signs of left ventricular decompensation are easily overlooked, particularly in elderly patients. Despite a doubling of the proportion of asymptomatic patients with good left ventricular function undergoing mitral valve surgery in recent years, the Great Britain and Ireland report shows that 42% still undergo surgery with low functional status (New York Heart Association (NYHA) class III/IV) and impaired left ventricular function (figure 1).4 These characteristics are associated with increased short- and medium-term mortality, prolonged hospital stay and less postoperative symptomatic improvement.5 Similar failure of adherence to guidelines reported from the USA and Europe6 is a source of concern and strategies to improve surveillance and ensure prompt surgical referral are urgently required.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Degenerative mitral valve disease and New York Heart Association (NYHA) class. Reproduced by kind permission of Dendrite Clinical Systems, Henley-on-Thames, UK.

A similar situation exists for aortic stenosis where about one half of diagnoses are made at post mortem and 5% of patients undergoing valve replacement are in NYHA class IV.7 These findings are replicated in the recent UK data, which suggest that 36% of patients undergoing aortic valve replacement already have severe symptoms (NYHA class III/IV 47%, class IV 8%) or impaired left ventricular function at the time of surgery, features which are associated with increased mortality during the perioperative period and subsequent follow-up. The onset of symptoms in severe aortic stenosis is often subtle and easily missed without a careful history, which can often be supplemented by treadmill exercise testing. However, this investigation is only performed in 6% of those in whom it is potentially indicated. These observations suggest that both initial diagnosis and subsequent follow-up are currently inadequate and that patients are routinely referred late in the natural history of the condition, beyond the window where surgery is of maximum benefit. Recent evidence demonstrating the adverse prognosis of very severe aortic stenosis (even in the absence of symptoms),8 the beneficial effect of surgery in such patients9 and the availability of transcatheter aortic valve implantation for high-risk surgical candidates should drive a significant increase in referrals for specialist assessment with a view to intervention.

Variations in access to surgery

The number of men undergoing aortic valve replacement between 2001 and 2008 was twofold greater than the number of women.4 While this may be partly explained by the increased frequency of aortic stenosis in men, similar discrepancies have been demonstrated in European series. Likely explanations include inter-gender variations in expectations of health and health care, occult physician and surgeon bias, cultural factors and the increased life expectancy of women.1

Indeed, the elderly appear to be systematically disadvantaged. Risk prediction in this group is fraught with difficulties but operative mortality rates continue to improve despite significant comorbidities affecting many patients.2 By 2008, the majority of patients undergoing aortic valve surgery were over the age of 70 (and 1 in 10 over the age of 80) (figure 2) with an accompanying predictable increase in the proportion with an elevated logistic EuroSCORE.4 Importantly, these statistics only relate to those elderly patients who underwent surgery after selective referral and case selection; several contemporary series demonstrate substantial unmet need in this group.10 Extrapolation of US data suggests that there are around 120 000 patients aged ≥75 with moderate or severe aortic stenosis in the UK and around 60 000 aged 65–74.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Age categories in isolated aortic valve replacement (AVR). Reproduced by kind permission of Dendrite Clinical Systems, Henley-on-Thames, UK.

Comparison with surgical figures suggests that patients over the age of 75 years with moderate to severe AS are half as likely to undergo valve replacement (table 1) as their younger counterparts.4 The rapid emergence of transcatheter aortic valve implantation reflects the size of this patient pool and the preponderance of elderly patients in published transcatheter aortic valve implantation (TAVI) registries11 should help to redress this underprovision.

Risk prediction in valve surgery is an imprecise art and current scoring systems, originally designed for use in patients undergoing coronary artery surgery, consistently overestimate risk in the elderly.12 Advancing age is often used to justify the decision to withhold surgery but suitably selected patients may derive considerable improvement in symptomatic burden and overall quality of life following successful intervention. Indeed, from the patient's perspective these benefits often vastly outweigh concerns over the risk of mortality. The most commonly used risk prediction score in Europe (the logistic EuroSCORE) significantly overestimates observed mortality,13 and the recent UK data demonstrate that this ‘failure of calibration’ is particularly marked for patients undergoing valve surgery (by a factor of four for isolated aortic valve replacement). New, more sophisticated, models specific for valve surgery are required to support the clinical decision-making process, particularly in the emerging era of alternative transcatheter techniques.

There are also unacceptable variations in practice between and within countries which seem to be based on non-clinical factors. In a mature healthcare system, optimal care of patients with VHD should be accessible to all and independent of geographical location and socioeconomic class. Mapping of the Great Britain and Ireland data according to Strategic Health Authority and primary care trust of residence shows wide geographical variation in rates of access to aortic valve surgery that are not explicable on known demographic patterns.4 The reasons underlying this variation are likely to be multifactorial but demonstration of this inequality with feedback to relevant healthcare providers should stimulate constructive discussion and political solutions.

While rising overall, surgical procedural volume varies widely between centres and individual surgeons—for example, the majority of UK centres perform fewer than 20 mitral valve repair procedures a year, an insufficient volume to develop appropriate expertise and ensure the reproducibility and predictability of results. An association between procedural volume and outcome is well recognised and proposed UK guidance recommends a minimum caseload of 50 mitral valve repairs per institution and 25 per individual operator each year.14 Some high-volume UK centres still have relatively low rates of mitral valve repair, suggesting insufficient inter-surgeon referral. Current international guidelines recommend that early mitral valve surgery in asymptomatic patients with good left ventricular function should only be contemplated in specialist units where there is a greater than 90% likelihood that a repair procedure will be performed. Despite this, rates of mitral valve replacement in NYHA class I patients undergoing surgery for degenerative mitral valve disease remain high (25%) in Great Britain and Ireland.4 Cardiologists and cardiac surgeons need to work together to heighten awareness of the benefits of early valve repair and encourage earlier referral of suitable patients to teams offering a very high probability of successful repair associated with very low rates of mortality and morbidity.15

Other trends, including the increased use of new-generation biological rather than mechanical valves in younger and younger patients without clear evidence of benefit (other than the avoidance of anticoagulation),16 lie outside current guidelines and require ongoing analysis of longer-term outcomes to guide future practice.