The complexities of benchmarking surgical complications in gynaecological oncology.

Abstract

Burnell et al., in an attempt to establish benchmarks for potential use as future standards when evaluating quality of care provided by the individual physician, subspecialties, or hospital services, have undertaken a significant statistical endeavour using the data previously reported by Iyer et al. (Br J Cancer 2015;112:475–84). The authors’ use of complex statistical analyses has been tempered by their noting the shortcomings of the methods used and more importantly, their care to apply this analysis in a clinically meaningful context. Iyer et al. compiled intra-operative and postoperative complications as well as co-morbidities associated with surgery undertaken by gynaecological cancer services in ten major UK centres. These raw data have been used to construct benchmarks and to identify, with some limitations, institutions with higher and lower rates of complications, while not neglecting the institutional degree of surgical complexity. Given the relative paucity of data the authors were not able to extend this analysis to the individual surgeon. At first glance, establishment of benchmarks seems logical and important, as it contributes to the patients’ right to transparency regarding healthcare and a potential pathway to improve physician and institutional performance. In the USA, Gynecologic Oncology Group (GOG) trial results are not separated by institution, much less by the individual physician. Potentially this is problematic because these are publically funded trials and it is argued that the public has the right to know outcomes associated with specific institutions. Burnell et al. begin to address this issue. Complete transparency in healthcare sounds like a panacea, but there are potential risks associated with it. Specifically, it could lead to surgeons undertaking only low-risk procedures so as to lower their individual and institutional complication rates. This could become problematic in patients with ovarian cancer because rendering them free of disease requires highly complex surgery, which accordingly is associated with increased peri-operative complications, but also with improved overall survival. The authors do not address this most important of issues. Perhaps that institution with the highest rate of complications also has the highest overall survival; these two issues are inseparable. Using these data to identify individual surgeons or institutions in need of remedial training or instruction might be even more complicated. The premise of such benchmarking requires underlying honesty on the part of the surgeon and the institution and if the response to such reporting is seen as punitive, as opposed to educational or constructive, then it is doubtful that reporting will be forthright. This potential scenario requires a built-in mechanism for remedial training for surgeons who may fall below acceptable standards of care. Similarly, hospitals must be given an opportunity to correct any deficiencies. Accordingly, as these benchmarks are established, reporting of substandard performance should probably be limited to those failing to correct identified deficiencies. The last area that may be adversely affected by such reporting is resident education. Undoubtedly, if complications are going to be publically associated with individual surgeons, consultants will probably be less inclined to ‘turn cases over to the trainee’ and the long-term impact of this potentially outweighs any benefit associated with the transparency of reporting. “for there is nothing either good or bad, but thinking makes it so.” None declared. Completed disclosure of interests form available to view online as supporting information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.