Intended for healthcare professionals

Research News

Reporting bowel surgeons’ death rates did not lead to “gaming”

BMJ 2018; 361 doi: (Published 03 May 2018) Cite this as: BMJ 2018;361:k1931
  1. Nigel Hawkes
  1. London

Fears that publishing bowel surgeons’ patient death rates would lead to risk averse practices or to exaggerating how ill their patients were to justify poor performance can be set aside, a study in The BMJ suggests.1

It found no evidence of any change in the proportion of patients in England having an elective major resection of the bowel since publication of results began, and death rates have fallen substantially, concluded a team from the Royal College of Surgeons, the London School of Hygiene and Tropical Medicine, and other UK centres.

Publishing surgical data has strong professional backing on transparency grounds, but evidence that it improves outcomes was limited.

Derek Alderson, president of the Royal College of Surgeons, said that the college supported publication of surgeons’ results as a way of improving quality and safety.

“There has, however, been concern in some quarters that reporting the outcomes of individual surgeons could discourage some from offering surgery to high risk patients,” he said. “Today’s study is reassuring, as it did not find any evidence of risk averse patient selection following the introduction of public reporting of outcomes after bowel cancer surgery.

“The fact that surgical mortality decreased significantly following the public reporting of individual colorectal cancer surgeons’ outcomes further underlines the importance of a culture of transparency and openness in improving the overall care of patients.”

Heart surgeons were the first to publish outcomes in 2006, and bowel surgeons followed in 2013, providing data to the National Bowel Cancer Audit, which has been running since 2010. This allowed comparison of data from before and after publication began. The characteristics and outcomes of 111 000 patients were included by linking the audit data to the hospital episode statistics database.

If publication had made surgeons nervous the most likely effects would have been fewer surgical resections, a tendency to refer the highest risk patients on to specialised centres, and an increase in surgery offering a lower risk than resection. Alternatively, surgeons might have sought to convey a better picture of their outcomes by classifying more of their patients as urgent or emergency cases, a process known as “gaming.”

None of these effects was found in the before/after comparison. For example, the proportion of major resections classified as urgent or emergency did not change significantly: 15.5% before publication of outcomes and 15.6% after.

The study did find a change in 90 day mortality, however. In elective patients it fell from 2.8% before publication to 2.1% after, a statistically significant difference over and above the long term downward trend. No change was seen in mortality among the smaller number of patients treated as urgent or emergency cases.

The team interpreted these results as suggesting that surgical teams prepared better for operations and managed patients better once they knew that outcomes would be published: the “heightened responsibility” among surgeons affected the whole team.

James Hill, president of the Association of Coloproctology of Great Britain and Ireland and lead clinician on the study, said, “The publication of surgical outcomes turns the spotlight on individual surgeons. The improvements that we have seen demonstrate that surgeons have an important role in galvanising the entire team involved in managing patients before and after this major surgical procedure.”

An alternative explanation is that the patients were improving their chances by scanning published outcomes and choosing the best surgeons. But the team considers this unlikely because the fall in mortality happened so rapidly.


View Abstract

Log in

Log in through your institution


* For online subscription