Standards for heart valve surgery in a ‘Heart Valve Centre of Excellence’

Developing expertise

Surgeons performing mitral repair surgery should undergo focused training in mitral repair, in atrial fibrillation ablation,9 and in tricuspid annuloplasty8 as part of their basic local board certified surgical training. Surgeons should also be mentored through new repair techniques and ideally in the implantation of new types of replacement heart valve to minimise the effects of the ‘learning curve’ on adverse results.9 ,31 ,32 There is a tendency for new products to require increasingly technique-sensitive implantation procedures, and regulatory bodies have insisted that manufacturers implement training and accreditation programmes where this is necessary to reduce procedural risk.33 The International Standard for transcatheter valves33 requires manufacturers to “establish a structured training program for the physician and staff who will be involved in the periprocedural care of the patient”. This principle should be seen as the rule rather than the exception although there is a strong argument for the processes to be supervised by organisations independent of the manufacturers for example national and international cardiovascular professional societies. Effective training measures could be accomplished by collaborative working within units, visits to other units or by visiting external experts.

Surgeon and hospital volumes

The relationship between case volume and outcomes for surgery and percutaneous interventions is complex, but volume recommendations already exist or are being discussed for PCI,34 vascular surgery35 and percutaneous valve techniques.13 ,14 What constitutes sufficiently high individual surgeon or hospital volumes to maintain good results for repair of mitral valve prolapse is controversial. A large retrospective review33 showed that mitral repair rates were higher and mortality rates lower in large volume hospitals. A high hospital volume partly reflects high individual surgeon volumes7 ,31 but may also be a surrogate for excellent facilities and processes (box 1).

Annual thresholds of 20–40 mitral valve repair procedures for individual surgeons7 ,9 ,10 ,36 have been suggested based on expert consensus or the observation from retrospective analyses that repair rates are on average higher among higher volume surgeons. Similarly hospital volumes of >50/year have been suggested,9 although studies in the USA36 have not confirmed that centre volume alone reliably predicts outcome.

For mitral valve replacement, no individual surgeon thresholds have been identified. Lower mortality rates are shown in higher volume centres,37 including one study with centre volumes >80/year for repairs and replacements.38

Minimal access mitral surgery has great appeal for patients and cardiologists. The cosmetic advantage is obvious, and when successful allows patients a quicker recovery. Data from one center32 suggests that the learning set for minimal access mitral repair is 75–125 operations and that a minimum of one operation each week is necessary to maintain good results.32 The authors noted a marked variation in the contours of learning curves among individual surgeons. The technique is challenging for the surgeon and the entire theatre team, and may not be reliably achieved by all centres.

These few studies are all retrospective making it difficult to control for the characteristics of individual patients or surgeons and repair rates were not matched with the preoperative likelihood of repair. Within the populations there were low-volume surgeons with high repair rates and high volume surgeons with low repair rates.7 ,36 Furthermore, these publications did not record non-fatal complications including stroke and bleeding, rates of residual mitral regurgitation or need for redo surgery. In view of these limitations, we propose that the ability to demonstrate good results (table 1) is a more important standard than the attainment of volume targets.

Assignment for mitral valve repair

In patients with mitral regurgitation, surgical assignment differs according to the likelihood of repair. Repairable primary disease should be operated by surgeons with special expertise in valve repair8 ,9 and acceptable results (table 1). Some types of P2 repair in degenerative disease may be relatively straightforward, but it is not always possible to guarantee this during preoperative assessment.8 ,9 A surgeon without specialist expertise might then find him or herself forced to implant a replacement valve while a more specialised surgeon might have successfully repaired it.

Annuloplasty rings are a part of the regular work of all cardiac surgeons, but good results are not guaranteed. Choosing the size of the ring and when to use adjuvant treatments (eg, cutting chords or left ventricular reshaping) may be difficult. Repair procedures for functional mitral regurgitation should therefore be performed under the auspices of a surgeon with specialised training and interest in valve repair.

If an MDT decides that repair is not possible, there is an argument for cardiac surgeons without competencies in repair to be free to replace the mitral valve. This would allow surgeons to remain competent to perform emergency mitral valve surgery on-call for example, for acute endocarditis or papillary muscle rupture. No consensus exists and arrangements should be made locally acknowledging the difficulty of balancing on–call rosters and the need for specialisation.

Developing expertise

Models for collaborative working in aortic valve intervention already exist. Conventional aortic valve replacement is part of basic surgical training. However, as for the mitral position, new designs of replacement valve having implantation techniques different from established valves should initially be used under guidance to minimise adverse effects from the learning curve. The Ross operation is complex and must be learned at a specialist centre.

Repair techniques are feasible in bicuspid or prolapsing aortic valves but remain underutilised. Approximately 60% of regurgitant aortic valves may be repaired at a specialist center40 ,49 while repair is uncommon in the UK database.5 Aortic root remodelling may be necessary as part of these repair procedures.39 Specialist experience and expertise is needed for the whole team since careful preoperative assessment and intraoperative imaging are essential.40 ,41 Repair can avoid the complications of a replacement valve although a proportion require revision in the immediate postoperative period.42 Thereafter the reoperation rate is approximately 2–3% p.a.42 ,49 although results beyond 10 years are not established.

Minimal access aortic valve surgery is increasingly accepted.43 There are ongoing trials to define if there is an advantage to mini-sternotomy (down to the 3rd or 4th space) and manubrium only (2nd space).44 These techniques which may present special challenges and require careful training and proctorship for the surgeon and theatre team. These minimal access techniques may be further progressed by the availability of sutureless valves which are also in the process of evaluation.45

Volumes

The effect of procedural volume has been observed for aortic valve replacement50 with a mortality rate of 9.1% for surgeons with annual volumes of less than 22 cases compared with a rate of 6.5% for those with annual volumes exceeding 42 cases. No volume data are available for aortic valve repair. As for the mitral position, individual surgeon and hospital volumes are unlikely to be perfect surrogates for outcome data and the ability to demonstrate good results is more important than working to volume targets.