Surgeon-specific mortality data disguise wider failings in delivery of safe surgical services

Objectives: Feedback of clinical outcome data to clinicians can promote and enhance patient safety. Surgeon-specific mortality data (SSMD) have been released to the public for a number of specialties. This implies that one individual is culpable for all deaths. Debate continues about SSMD because of risk-averse behaviour. In the USA, improved outcome measures derived from phase of care mortality analysis (POCMA) and the failure to rescue (FTR) are replacing SSMD, but they have not been tested in Europe.

Methods: Using POCMA and FTR analysis, we studied hospital deaths in 1558 cardiac surgical patients between 2009 and 2013. Comorbidity and urgency status were used to calculate modified logistic EuroSCORE (MLE). The circumstances of death were critically reviewed by a panel of four experienced surgeons. Death certificate information and autopsy were taken into account. Deaths were then classified: Class 1 surgeon dependent, Class 2 FTR or Class 3 where multiple factors conspired to cause death.

Results: There were 51 deaths providing 3.3% mortality, as predicted by MLE. In the 86% who underwent autopsy, no surgical error was identified. Most deaths in each group were related to high-risk status, age, frailty, comorbidity and urgency. FTR was the predominant factor occurring in 45%. Though difficult operations were implicated in 37%, no deaths occurred in the operating theatre. Some FTR deaths occurred in low-risk patients. Scrutiny of FTR deaths provided important information that could be used for quality improvement.

Conclusions: The study showed that most deaths cannot be prevented by the operating surgeon. They occurred through issues of patient comorbidity, lack of process or infrastructure. This casts doubt on SSMD publication alone as a tool for quality improvement. In contrast, POCMA and FTR highlight problems of process, and are more likely to promote advances in surgical care.