Non-technical errors associated with deaths in surgical care, Australia, 2012-2019, by surgical specialty (Australian and New Zealand Audit of Surgical Mortality): a retrospective cohort study.

Abstract

To compare the frequency of non-technical errors in cases of surgical care-related deaths in Australia, by surgical specialty. Retrospective cohort study; analysis of Australian and New Zealand Audit of Surgical Mortality (ANZASM) data. All surgical care-related deaths in Australia (except New South Wales), 1 January 2012 - 31 December 2019, that were flagged in ANZASM as associated with adverse events or areas of concern. Proportions of surgical care-related deaths associated with non-technical errors, overall and by domain (communication/teamwork, decision making, situational awareness, leadership); paired comparisons of likelihood of errors by specialty; change in error proportions during 2012-2019; influence of patient and admission factors on likelihood of non-technical errors. Of 30 971 surgical care-related deaths reported to ANZASM during 2012-2019, 3695 were flagged with adverse events or areas of concern, including 3422 cases (92.6%) in five surgical specialties: general surgery (1570 deaths), cardiothoracic surgery (626), orthopaedic surgery (510), vascular surgery (385), and neurosurgery (331). The proportions of surgical care-related deaths associated with non-technical errors differed by specialty (range, 52.2% to 68.5%), as did those errors in the domains decision making (range, 52.6% to 66.3%) and situational awareness errors (range, 44.4% to 62.5%). The probability of any non-technical error was greater for cardiothoracic than orthopaedic surgery (adjusted odds ratio [aOR], 1.76; 95% confidence interval [CI], 1.37-2.28), and for general than orthopaedic surgery (aOR, 1.97, 95% CI, 1.59-2.44) or neurosurgery (aOR, 1.47; 95% CI, 1.14-1.90); the probability was lower for orthopaedic than vascular surgery (aOR, 0.54; 95% CI, 0.41-0.72). The proportion of deaths associated with non-technical errors declined over time for general surgery, but not the other four specialties. Differences by patient and admission characteristics in the proportions of cases including non-technical errors were not statistically significant. At least 50% of surgical care-related deaths in five surgical specialties were associated with non-technical errors, and the proportions did not substantially change during 2012-2019. Differences between specialties in the frequency of fatal non-technical error suggest that targeted improvement strategies are needed, but the persistently high frequency for all specialties indicates that systemwide improvement is crucial.