Reporting surgeon-specific outcome data: panacea for everything?

It isn't that they can't see the solution. It is that they can't see the problem. Patient safety is the fundamental and primary goal of all efforts of physicians. Especially cardiothoracic surgeons have pioneered the development of quality assessments and data reporting for many decades, which has been widely recognized and enabled the rapid and safe implementation of many innovative surgical techniques over the years [ 1 ]. Our surgical results have improved significantly in all aspects over the last decades. Nevertheless, in medicine—as in all other complex organizations and systems—errors and preventable events continue to result in suboptimal outcomes [ 1 , 2 ], as it is well known that ‘to err is human’ [ 3 ]. High-tech medicine such as cardiothoracic surgery is a very complex sociotechnical system [ 3 ] in which many different professionals with varying levels of skills under diverse leadership work on high-risk patients [ 1 ]. Performance depends on individual, technical and organizational factors and their interactions [ 4 ]. Where and when do surgical adverse events occur? Two-thirds in in-patients and approximately half of these are preventable [ 5 ]. In addition, we know that 21% of all reported incidents related to cardiac procedures occur in the operation theatre [ 6 ]. These have to be diminished. But this also shows that 79% of all incidents occur outside the operating room. The incidence of pre-, intra- and postoperative errors varies among the different surgical treatments (e.g. staged palliation of hypoplastic left heart syndromes versus coronary artery bypass grafting). Moreover, one type of error increases the likelihood of another, e.g. an intraoperative and/or technical error, will increase the risk of postoperative errors [ 3 ]. Medicine in general but cardiothoracic surgery in particular has to be counted to the so-called High-Reliability Organizations (HROs) where complex processes are in use to manage our surgical work. Examples of HROs are aircraft carriers, air traffic control systems, fire incident command systems, commercial aviation and nuclear power operations. In a complex environment, such as cardiothoracic surgery, where the coordinated efforts of many individuals working in teams are needed to achieve high levels of cognitive and technical performance, one has to question if single parameters, such as mortality rates from a single surgeon, are all that is needed for quality control. While the surgeon is the captain of the ship and he/she is responsible for the main part of the treatment process, the operation or the surgical treatment is not a one-man show, but a team effort. In addition, we also know that there are talented and less talented surgeons—differences as they exist in all other professions as well. In our information era, where everybody seems to know everything about everybody, the main question still is how to report medical outcome data. Surgeons have always been at the forefront of quality assessment of their results. However, after decades of quality control, we know that even though it sounds easy (i.e. documenting the outcome of a specific treatment), it is not so easy after all. We know that even mortality rates need the right risk scores and other variables like postoperative infections, statements about who looks after the wounds (surgeon and infectious disease specialist), which person defines the degree of infection (A1 superficial infections, A2 deep infections and A3 infections of organs or body cavity) and how complete the postoperative follow-up is (>97%?), because infected patients may go to other hospitals, etc. Therefore, the publication of surgeon-specific data has been very controversial since its introduction. Nevertheless, public reporting of single surgeon outcome data is a fact and is being done in Great Britain and in some states in the USA (e.g. New York and Pennsylvania). The reason for public reporting of individual surgeon's mortality rate is to identify the technically unsafe surgeon. With the complex infrastructure in which operations are being performed, it remains unproved that this can be achieved reliably by public reporting. This was the reason that the European Association for Cardio-Thoracic Surgery (EACTS) organized a forum with the title ‘Safer surgery for whom?’ at their 29th Annual Meeting in Amsterdam, Netherlands (2015). The results of a questionnaire survey of all attendees of the forum have now been published by Jahangiri et al. [ 7 ] in this issue of the journal. In their paper, the pros and cons of surgeon-specific public outcome reporting have been well described and discussed, including non-surgical causes of outcome, rejecting high-risk patients, favouring simpler operations over more complex reconstructions with better long-term outcomes, etc. In fact, the present questionnaire also shows conflicting results of those attending the forum. Crude mortality reporting was seen as insufficient as a quality marker [ 1 , 7 , 8 ]. Because cardiothoracic surgery is a team effort, it is not surprising to find a clear high volume–improved outcome relationship in many studies for different cardiothoracic operations [ 7 , 9 ]. Many other variables also influence the final result, e.g. applying the health benefits of a process of care (e.g. adherence to Guidelines) improves outcomes [ 10 ]. High-volume surgeons in high-volume hospitals seem to have the best results because the surgeon and his/her team are well trained and are familiar even with rare events. The development of programmes with subspecialization for complex procedures within a large unit will further improve outcomes, which again underscores the importance of high-volume hospitals. It also has to be remembered that talented surgeons operate more than less talented ones. In systems where the head of the Department assigns individual surgeons to certain patients, he will assign more patients to talented surgeons in order to have good results at his clinic. Likewise, in private clinics, talented surgeons get more patients referred from other physicians. Therefore, not only do high-volume surgeons have better results, but the opposite is true as well: because they have good results they operate more. Risk-adjusted results have to be reported to make medical results transparent. However, oversimplification may lead to more harm than benefit. Everything should be made as simple as possible, but not simpler.