Perioperative and anaesthetic-related mortality in developed and developing countries: a systematic review and meta-analysis

Background The magnitude of risk of death related to surgery and anaesthesia is not well understood. We aimed to assess whether the risk of perioperative and anaesthetic-related mortality has decreased over the past five decades and whether rates of decline have been comparable in developed and developing countries. Methods We did a systematic review to identify all studies published up to February, 2011, in any language, with a sample size of over 3000 that reported perioperative mortality across a mixed surgical population who had undergone general anaesthesia. Using standard forms, two authors independently identified studies for inclusion and extracted information on rates of anaesthetic-related mortality, perioperative mortality, cardiac arrest, American Society of Anesthesiologists (ASA) physical status, geographic location, human development index (HDI), and year. The primary outcome was anaesthetic sole mortality. Secondary outcomes were anaesthetic contributory mortality, total perioperative mortality, and cardiac arrest. Meta-regression was done to ascertain weighted event rates for the outcomes. Findings 87 studies met the inclusion criteria, within which there were more than 21.4 million anaesthetic administrations given to patients undergoing general anaesthesia for surgery. Mortality solely attributable to anaesthesia declined over time, from 357 per million (95% CI 324-394) before the 1970s to 52 per million (42-64) in the 1970s-80s, and 34 per million (29-39) in the 1990s-2000s (p<0.00001). Total perioperative mortality decreased over time, from 10 603 per million (95% CI 10 423-10 784) before the 1970s, to 4533 per million (4405-4664) in the 1970s-80s, and 1176 per million (1148-1205) in the 1990s-2000s (p<0.0001). Meta-regression showed a significant relation between risk of perioperative and anaesthetic-related mortality and HDI (all p<0.00001). Baseline risk status of patients who presented for surgery as shown by the ASA score increased over the decades (p<0.0001). Interpretation Despite increasing patient baseline risk, perioperative mortality has declined significantly over the past 50 years, with the greatest decline in developed countries. Global priority should be given to reducing total perioperative and anaesthetic-related mortality by evidence-based best practice in developing countries.