Article Text
Abstract
Valve repair, where feasible, rather than valve replacement is the guideline recommended treatment for severe mitral regurgitation. To characterise ‘real-world’ clinical practice data were reviewed on 12 255 mitral valve operations performed in the UK between 2004 and 2008, as reported in the 2009 UK Adult Cardiac Surgical Database Report. The data demonstrate a large variation in the use of mitral valve repair; while the national repair rate was 51%, this varied from 20% to 90% among different hospitals. Outcomes were worse in patients who had valve replacement as opposed to repair, including a higher risk of operative mortality and stroke, in all subgroups examined. Some patients were, by virtue of the hospitals they attend, therefore, less likely to survive and more likely to have complications, because of a low use of valve repair in those centres. Concentration of mitral valve surgery in designated regional reference centres should allow more equitable access to mitral valve repair.
- Surgery-valve
- delivery of care
- mitral regurgitation
Statistics from Altmetric.com
The publication in July 2009 of the sixth National adult cardiac surgical database report by the Society of Cardiothoracic Surgery (SCTS) in Great Britain and Ireland1 received substantial media coverage, and was notable for the observation of reduction in overall operative mortality rates after heart surgery in the UK. Aside from mortality, this mandatory national database captures a wide array of data on heart surgery in the UK and is a useful resource to explore patterns in healthcare provision. The most recent report includes a 10-year analysis of almost 20 000 mitral valve operations reported to the database between 1999 and 2008.1 A review of this subanalysis provides a wealth of information about the status and provision of mitral valve surgery in the UK and Ireland. For most patients with severe mitral valve regurgitation, the preferred surgical treatment is a repair of the valve, if feasible. The alternative, prosthetic valve replacement, is associated with higher operative mortality, reduced life expectancy, higher long-term risk of stroke and complications unique to valve replacement such as valve thrombosis and structural valve degeneration.2
Mitral valve surgery in the UK: mortality, repair rates and timing of surgery
What does the recent database report1 tell us about healthcare provision for patients with mitral valve disease in the UK? The report shows a lower crude mortality for mitral valve repair (2.0%) than for mitral valve replacement (6.1%), and the mortality benefit of valve repair over replacement existed in all subgroups examined. This mirrors a recent Society of Thoracic Surgeons database report on almost 60 000 North American patients operated on between 2000 and 2007, which found repair was associated with less operative mortality, even after risk adjustment (OR=0.52).3
Despite the clear advantage of repair over replacement, surprisingly the overall mitral valve repair rate in the UK was only 51.7%. This differed among the different causes of mitral regurgitation: the highest reported repair rate was for ischaemic aetiology (74.1%). For degenerative mitral valve regurgitation, the repair rate was 64.6%. Valve repair was uncommon for rheumatic heart disease (8.5% repair rate) and active endocarditis (27% repair rate). It is important to recognise that some miscoding of aetiology may have occurred in this dataset, as the importance of accurate aetiological subclassification of mitral valve disease has only recently been emphasised.3 4 The accuracy of data is also difficult to verify as the SCTS database is currently not subject to rigorous validation.
Nonetheless, there are indications from this analysis that the UK population with degenerative mitral valve disease, typically manifesting as mitral valve prolapse, is indeed underserved by current clinical practice. In the SCTS report, 24.7% of those with degenerative disease had depressed left ventricular function and 42% were reported as being in class III or IV heart failure at the time of surgery. This suggests that many patients were not being offered surgery until they developed advanced symptoms. One has to conclude that patients in the UK with minimal symptoms but declining left ventricular function (a class I indication for mitral valve surgery according to both European and North American Guidelines5 6) were either not identified by appropriate echocardiographic surveillance, or were not referred or operated on in a timely fashion. Such a high percentage of patients with severe symptoms or left ventricular dysfunction at the time of surgical intervention suggests a need for national surveillance protocols of patients known to have mitral regurgitation and also mechanisms that would facilitate earlier surgery for appropriate patients.
Mitral valve replacement in asymptomatic patients
Also concerning was the use of mitral valve replacement in asymptomatic patients with degenerative mitral valve disease. Of 709 patients in New York Heart association (NYHA) class I with degenerative aetiology over the study period, 177 (25.0%) received a valve replacement. Mitral valve replacement is, however, not recommended as treatment for asymptomatic mitral regurgitation6 because it may reduce, rather than extend, the life expectancy in asymptomatic patients. Valve replacement has not been shown to have benefit over watchful waiting in this patient subgroup and in fact may result in worse short-term and mid-term survival. This effect may be particularly pronounced in the UK considering the 10-fold higher operative mortality risk for mitral valve replacement (6.2%) than for repair (0.6%) in NYHA class I patients reported in the SCTS database.1 The 6.2% mortality risk with mitral valve replacement implies one in 17 asymptomatic patients in the UK undergoing valve replacement did not leave hospital. Additionally, patients receiving a valve replacement have been shown to be subject to a higher mid-term risk of stroke.7 Guidelines in North America recommend that surgery should only be undertaken in asymptomatic patients in designated reference valve repair centres with a very high probability of repair,6 while the European Guidelines do not endorse mitral valve surgery in asymptomatic patients.5 The current UK data suggest that neither guideline is being followed as many asymptomatic patients do receive valve replacement—the very thing the guidelines are striving to avoid. The importance of avoiding valve replacements in this patient subgroup was reinforced in a recent UK consensus statement which recommends that centres undertaking valve repair in asymptomatic patients be subject to close audit to ensure that recommended high standards of practice are achieved.8
Variation in repair rate across centres: a postcode lottery
The most notable observation, however, raised by the database report1 is the marked variation in use of mitral valve repair among heart centres in the UK. The highest ranking centre repaired almost 90% of all valves while the lowest ranking centre repaired 20% of valves. While other factors may contribute, these variations probably reflect true local and regional variations in healthcare provision—a so called ‘postcode lottery’ where the treatment received is related to where one lives or the hospital one attends. Variation was particularly marked in patients with degenerative disease: although the national repair rate for this aetiological subgroup was 66%, this varied among hospitals from 0% to 98% (figure 1). Within some towns or regions, the repair rates were very different between institutions.
Looking specifically at high-volume centres—those doing more than 30 mitral valve operations for degenerative disease per year—they did not, as one might expect, consistently deliver higher repair rates and there remained great variability (figure 1). The repair rate for degenerative mitral valve disease ranged from 36% to 98% in these high-volume centres.
The SCTS data thus demonstrate a postcode lottery in access to state of the art management of mitral valve disease in the UK, with one's chances of a repair (and by extension long-term, event-free survival) depending heavily on the hospital in which the operation is carried out. While high-volume centres were more likely to deliver repair, this was not consistent as only three of the high-volume centres reported repair rates >85% and the median repair rate for all high-volume centres was 67.5%. These high-volume centres performed 61% of all procedures for degenerative mitral valve disease in the UK. Repair rates among the low- and medium-volume centres were variable, but only one medium-volume centre had a repair rate >85%.
Implications of differences in repair rate across hospitals
These differences in repair rates are of critical importance because the SCTS report1 shows a marked difference in outcomes between patients having valve repair as opposed to replacement. For example, the mortality rate for an isolated mitral valve repair for degenerative disease was 1.3% and perioperative stroke rate 0.8%, compared with 4.3% and 3%, respectively, for mitral valve replacement.1 If concurrent coronary artery bypass grafting or other procedures were undertaken, the mortality rates were even higher (approximately 8% with replacement compared with 4% with repair). Isolated valve repair also had a substantially better 5-year survival (90.0% vs 83.6% for valve replacement). Even in the elderly age group, mitral valve repair was associated with substantial survival advantage (operative mortality for valve repair patients aged ≥80 years was 5.5% vs 15.6% for replacement). Although these data were not adjusted for patient risk or valve pathology, the observed marginal benefit of valve repair over replacement is consistent with published literature.9–11 The independent survival advantage of mitral valve repair over replacement is also supported by recent analysis of 47 126 patients reported to the Society of Thoracic Surgeons Database in the United States, where the risk adjusted mortality for valve repair was demonstrated to be lower than replacement using both multivariable logistic regression (OR=0.54, 95% CI 0.47 to 0.63) and propensity matching (standardised mortality rate for replacement 3.0% vs 1.7% for repair p<0.001).3 The unpleasant reality, therefore, is that some patients, by virtue of the hospitals they go to in the UK (and we believe the same is so for most other countries), will be less likely to survive surgery, less likely to be alive at 5 years and more likely to have a stroke, because of a higher use of valve replacement in those centres.
Studying reasons for variations in repair rates
In order to clarify the reasons for the relatively high frequency of valve replacement in the UK, future iterations of the SCTS database, and other mitral valve databases, should collect data on the type of degenerative disease, the location of leaflet prolapse (anterior, posterior, bileaflet) and presence of advanced pathology, such as leaflet calcification.12 These data are required to stratify cases according to complexity of repairs that are required. An ‘intention-to-repair’ field will also be helpful in separating valve replacements in which replacement was the preoperative plan from those cases in which a repair had been planned but not executed. These categorisations are critical to developing a means of matching individual patients to surgeons and centres most likely to achieve a successful a repair.13 It has been suggested that repair of Barlow valves—which are the most complex of degenerative valves to repair—be limited to a few super-specialised high-volume centres in the UK.8 With current National Heath Service policy allowing patients to move out of their locality for treatment, referral to regional repair centres, where appropriate, should not only be possible, but be a stated goal to improve accessibility to a higher available level of care (ie, access to complex mitral valve repair team) where necessary.
Minimising variations in care: the case for reference valve repair centres
Not surprisingly, some centres (and surgeons) are clearly more committed to mitral valve repair. The traditional view of a consultant being one who is able to tackle all problems within his specialty hinders access of patients to highly skilled subspecialists. To deliver a high repair rate across such centres, however, those surgeons and centres with less expertise or interest in valve repair should be prepared to refer more complex cases to colleagues more versed in advance repair techniques.
We would suggest designation of reference mitral valve repair centres, along the lines suggested by Enriquez-Sarano et al,2 by us8 13 14 and by others,15 that manage complex mitral valve disease. Within these centres would be identified subspecialists with a specific interest and expertise in mitral valve repair techniques. This would be similar to organisation of other specialised cardiovascular services such as transplantation and congenital heart disease. Such an approach would be likely to yield a more consistent and higher level of care, compared with the current situation where cardiothoracic centres provide the mitral valve surgery for patients in their locality, regardless of centre expertise and track record in mitral valve repair. Additionally, concentrating complex mitral valve repair in fewer centres will help to consolidate training, experience and expertise. Currently, most heart hospitals in the UK, each consisting of teams of four or more surgeons, perform fewer than 20 operations a year for degenerative mitral valve disease, a number that is too small for each to maintain expertise and skill. Indeed, a multidisciplinary consensus group in the UK suggested that surgeons undertaking mitral repair surgery should be doing more than 25 repairs each year8—a threshold currently achieved by only a minority of surgeons in the UK. Only concentration of valve repair in fewer centres can enable most centres achieve the volume threshold required to deliver high-quality care.
Conclusion
Marked variations in mitral valve surgical practice place many patients with mitral regurgitation at a disadvantaged starting point simply by virtue of their location and lack of accessibility to a specialist mitral valve repair surgeon. Being the healthcare purchaser for the vast majority of patients in the United Kingdom, the NHS is in a unique position to define what it regards as the minimum standard of surgical care for patients with mitral valve regurgitation. Surgeons and hospitals need to be more accountable and should publish and audit their valve repair rates and outcomes. Publication of the current SCTS database reports provides the impetus for implementing these recommendations, not just in the UK, but in other countries with advanced healthcare systems. As Mulley16 states in a recent article, ‘For too long, stakeholders in healthcare have avoided the inconvenient truth … The belief that science mediated by clinicians determines what is the best medical care for each patient confers authority for decision-making on the clinician and absolution from responsibility on patients and policy makers. As comfortable as this may be for all, it puts patients at risk of care they would not choose and puts policy makers at risk of presiding over the resulting inefficient healthcare … Clinicians, patients, and policy makers must find the competence, curiosity, and courage to confront the implications of practice variation …’
Resting the choice of treatment for mitral valve disease on the individual surgeon has failed many patients and the seemingly inconsequential decision of replacing the mitral valve rather than repairing a valve has left numerous patients disadvantaged as well demonstrated by this SCTS report. The postcode lottery for mitral surgery is a wake up call to change clinical practice which will improve access and care for patients with mitral valve disease (box 1).
Box 1 Eliminating the mitral repair postcode lottery
Patients
Education on natural history of disease, various treatment options and implications of various choices
Informed choice and involvement in decision-making
Healthcare purchasers and providers
Effective primary care and uniform disease surveillance
Referral pathways dependent on patient needs rather than locality
Regionalisation of specialist services
Targets and minimal care standards
Audit of results and compliance with standards as a prerequisite for accreditation and reimbursement
Clinicians
Education
Super-specialisation
Acceptance by cardiac surgeons to give up cases previously within their realm
Referral to appropriate (higher) level of specialisation as disease intensity or complexity requires
References
Footnotes
Competing interests DHA has consultancy agreements and receives royalties from Edwards Life Sciences.
Provenance and peer review Commissioned; externally peer reviewed.