Surgeon-specific mortality data are misleading and harmful

Br J Cardiol 2015;22:132–3doi:10.5837/bjc.2015.038 Leave a comment
Click any image to enlarge

The Government is soon to publicly disclose a league table for cardiac surgical units within National Health Service (NHS) England. While this information may be useful and raise questions as to why one unit may be better or worse than another, we are also to be made aware of surgeons who are performing significantly better or worse than expected in terms of risk-adjusted mortality. But are patient deaths following surgery caused exclusively by the surgeon, as surgeon-specific mortality data (SSMD) would imply? And is the surgeon with the lowest operative mortality the best doctor? In my opinion the answer to both these questions is a resounding no.

Angina - BJC Learning programme
For healthcare professionals only

Screen shot 2015-11-25 at 12.33.09
Dr Ravi De Silva (Papworth Hospital NHS Foundation Trust)

Many factors influence patient outcomes after surgery.1,2 Broadly speaking these may be categorised into factors present prior to hospital admission (e.g. comorbidities), factors occurring during the operation (e.g. technical error, unexpected operative findings), and factors occurring post-operatively (e.g. infections, medication error). A single individual is not able to control all of these variables, so patient deaths should not be attributed to an individual. The public may also be unaware that there exists great variation in the quantity and quality of staffing, expertise and equipment between hospitals. These infrastuctural inequalities are bound to effect patient outcomes, and are not accounted for in hospital league tables. Having moved from a safe, respected, cardiac centre to one that is soon to be acknowledged as the best by the criteria of risk-adjusted mortality, I have considerable insight into what, other than the surgeons, should make one institution have better mortality data than another.

Life-saving technology

Mechanical cardiac and/or respiratory support in the form of extra corporeal membrane oxygenation (ECMO) and ventricular assist devices (VADs) are well-established, highly effective, and may be used to support sick patients peri-operatively.3 They can provide a surgical safety net for the sickest patients. However, due to cost, very few hospitals in the NHS have been allowed to use these technologies, obviously disadvantaging patients in units where it is unavailable and potentially increasing mortality as a consequence.

Resident theatre team

Cardiac arrest following cardiac surgery invariably needs emergency re-sternotomy. Out of normal working hours many hospitals have on-call staff at home, needing travel time to attend the emergency. Other hospitals have staff on-site at all times. While the latter may be more costly and require a larger workforce, it is clearly an advantage in emergency situations.

Consultant-delivered care

Consultant delivered care is the aspiration for the modern NHS, and although this is achievable during normal working hours, in most heart units, overnight one consultant is in charge of all the patients in the critical care area, as well as, if needed, running an emergency operating theatre. Increased mortality for patients undergoing surgery out of normal working hours is well recognised.4 Other units have invested in larger workforces, allowing for multiple consultants to be on-call to cover different clinical areas, resulting in the sickest patients managed by the most experienced clinicians around the clock.

Alert teams

It is recognised that one of the main differentiators between hospitals is not how often patients encounter difficulties, but the ability to rescue patients from these situations.5,6 Many factors helping to rescue the patient have already been discussed. Another important contributor is adequate ward staff7 and the presence of an ‘alert team’, who closely manage and treat patients who become unwell, directly liaising with consultants and the intensive-care team. They are an additional tier of surveillance on the ward, and in my hospital the alert team have halved the incidence of cardiac arrest in ward patients.

The theatre team

Cardiac surgery involves a multi-skilled team of doctors and nurses. The patient may be operated on several body sites simultaneously, with junior doctors and allied health professionals (AHPs) all working under the supervision of a consultant surgeon. In the UK, the practice of AHPs performing surgical procedures was pioneered in cardiac surgery, although now commonplace in other specialties. Some hospitals do not have enough surgical AHPs to staff all theatres, this work being done by junior doctors or temporary staff from other hospitals. This can negatively impact on the quality of the operation. The surgical team is no different to any other high-performing team, consistency and familiarity with each other, are corner stones to excellent outcomes.

As a surgeon with better than expected results, working in the ‘best’ cardiac centre in the UK, I have nothing to hide. However, I feel strongly that SSMD is important for medical professionals but can be hugely misleading to the public. We must have a mechanism to identify and help failing surgeons, but unless there has been a criminal act, this information should be held by the professional bodies that are best placed to deal with it. Similarly, the ‘best’ surgeons may be very good doctors, but may also be extremely risk-averse,8 disadvantaging the sickest patients who may have the most to gain from an operation. We must appreciate that mortality is inherent in the health system, and the only way to improve it is to improve the system, using high-performing centres as benchmarks and templates. The complexity of modern medicine and why patients die in hospitals can be difficult for doctors to fathom, to throw this out to the general public is sensationalist and disingenuous. Hospital-specific data are better understood by the public, and should be used to improve infrastructure and funding, thereby reducing inequalities in the health service.

Conflict of interest

None declared.


1. Westaby S, De Silva R, Petrou M, Bond S, Taggart D. Surgeon specific mortality data disguise wider failings in delivery of safe surgical services. Eur J Cardiothoracic Surg 2015;47:341–5.

2. Shannon FL, Fazzalari FL, Theurer PF, Bell GF, Sutcliffe KM, Prager RL. A method to evaluate cardiac surgery mortality: phase of care mortality analysis. Ann Thorac Surg 2012;93:36–43.

3. Westaby S, Taggart D. Inappropriate restrictions on life saving technology. Heart 2012;98:1117–19.

4. Aylin P, Alexandrescu R, Jen MH, Bottle A. Day of week of procedure and 30 day mortality for elective surgery: retrospective analysis of hospital episode statistics. BMJ 2013;346:f2424.

5. Reddy HG, Shih T, Englesbe MJ et al. Analyzing “failure to rescue”: is this an opportunity for outcome improvement in cardiac surgery. Ann Thorac Surg 2013;95:1976–81.

6. Ahmed EO, Butler R, Novick RJ. Failure-to-rescue rate as a measure of quality of care in a cardiac surgery recovery unit: a five-year study. Ann Thorac Surg 2014;97:147–52.

7. Needleman J, Buerhaus P, Pankratz VS, Liebson CL, Steven SR. Nurse staffing an inpatient hospital mortality. N Engl J Med 2011;364:1037–45.

8. Hannan EL, Cozzens K, King SB, Walford G, Shah NR. The New York State cardiac registries: history, contributions, limitations and lessons for future efforts to assess and publically report healthcare outcomes. J Am Coll Cardiol 2012;59:2309–16.