Public reporting of surgeons’ performance

In November 2012, the medical director of the United Kingdom National Health Service (NHS) announced that within 2 years, league tables showing performance data for surgeons working in the NHS in England would be published.1 The aim of this was to expose “variation and unacceptable practice” by publishing the results of “consultant-led” teams.2 While Australian national statistics in relation to diseases treated surgically (eg, cancer, trauma), associated complication rates, such as postsurgical infection, and mortality rates are generally as good as in comparable health systems, great improvements can still be made.3

However, the announcement from the UK has again raised concerns within the profession of the value and dangers of such audit programs. If reporting were introduced that focused purely on mortality then surgeons may avoid higher-risk patients. Risk adjustment of outcome can be conducted, but this is by no means a perfect science.4 Individual surgeons within hospitals or regions are often known for either excellent results or poor outcomes but, unless some form of measurement audit and feedback occurs, it is difficult to highlight problematic issues, such as poor surgical decisions or system failures, and improve patient outcomes. While mortality is a “hard” end point or outcome measure, it may not be the most important parameter to be measuring. Functional outcomes after joint replacement and tumour-free survival after bowel cancer surgery are examples of end points that may provide valuable feedback to surgeons, although they are rarely collected at present.

With regard to surgical audit in Australia, in 2005, the Royal Australasian College of Surgeons (RACS) became responsible for managing the Western Australian Audit of Surgical Mortality (WAASM), which had been established in 2001. The WAASM was modelled on the Scottish Audit of Surgical Mortality,5 which has operated successfully since 1988. The RACS has expanded the program to all states and territories, and since 2009, in conjunction with the state and territory jurisdictions, has carried out the Australian and New Zealand Audit of Surgical Mortality (ANZASM). According to the 2011 national audit, 99% of Australian public hospitals and 73% of private hospitals were participating in the audit.6

Private hospitals in New South Wales and Queensland have been slow to embrace the audit, but this is likely to be rectified over the next 12 months as funding issues are resolved.

The national audit provides feedback to surgeons on their cases, and institutes peer review of deaths by reference to the case notes of “concerning” cases in which mortality may not have been an expected outcome; for example, young patients or those in whom the management was open to question. The national annual reports for 2009–2011 allow thorough investigation of trends in areas such as deep vein thrombosis (DVT) prophylaxis, transfers, supervision and resuscitation, to name but a few.6

There are other surgical audits occurring independently. Many breast surgeons within Australia and New Zealand contribute to the National Breast Cancer Audit which enables them to compare their practice with peers.7 The Australian Orthopaedic Association, through its Joint Replacement Registry, has highlighted not only poor outcomes for certain prostheses but also a variety of outcomes obtained by different surgeons.8,9 Such data should be ideally available for other surgical procedures — for example, recurrence after hernia repair and cancer recurrence after bowel resection. However, these data are simply not available — surgeons rarely collect or report their results. It is not realistic or affordable to collect outcome data on all procedures. Nevertheless, “sentinel” procedures within specialties could be carefully monitored. The Australian and New Zealand Society for Vascular Surgery (ANZSVS) currently collects outcome data on carotid endarterectomy and has a protocol for detecting “outliers” by this approach.10 To date, there have been no serious attempts in Australia to introduce public reporting of surgeon performance data.

So what have we learned so far and what should be the way forward for Australia with respect to monitoring surgical performance?

The ANZASM has demonstrated to surgeons, from local rather than overseas experience, the value of adequate DVT prophylaxis, early patient transfer and careful resuscitation.

Careful review of all deaths should continue to be supported. More refined audits of outcome need to be established and the results provided to surgeons. Non-participants should not be allowed to maintain registration. With our national medical board registration, such sanctions are now possible and should be used for those not prepared to report on their results. Hospitals should insist on participation in such activities. The loss of even a busy surgeon with poor outcomes is a small price to pay for any public or private hospital.

There are lessons to be learned from over a decade of experience in WA since establishment of the WAASM. Better care and less enthusiasm for futile surgery appear to be leading to a real reduction in surgical mortality.

Finally, patients can ask or, indeed, insist that their surgeon participates in audit of their practice and should be more demanding in expecting results of the surgeons they visit. They should ask how the surgeon’s results are assessed, who does the assessment and how long they have engaged in such activities.11 If patients expect such information, government will need to help the profession to provide useful outcome data.

Real improvement in surgical care and performance is not really about picking the “outliers”, although this is an important benefit, but rather about improving the whole practice of surgery for the whole community.