Failure to Rescue: A Quality Metric for Cardiac Surgery and Cardiovascular Critical Care

Caroline R. Gross, David H. Adams, Parth Patel, Robin Varghese

Research output: Contribution to journalReview articlepeer-review

4 Scopus citations


Failure to rescue, defined as mortality after a surgical complication, is a widely accepted quality metric across many specialties and is becoming an important metric in cardiac surgery. The failure to rescue metric provides a target for improvements in patient outcomes after complications occur. To be used appropriately, the failure to rescue metric must be defined using a prespecified set of life-threatening and rescuable complications. Successful patient rescue requires a systematic approach of complication recognition, timely escalation of care, effective medical management, and mitigation of additional complications. This process requires contributions from cardiac surgeons, intensivists, and other specialists including cardiologists, neurologists, and anaesthesiologists. Factors that affect failure to rescue rates in cardiac surgery and cardiovascular critical care include nurse staffing ratios, intensivist coverage, advanced specialist support, hospital and surgical volume, the presence of trainees, and patient comorbidities. Strategies to improve patient rescue include working to understand the mechanisms of failure to rescue, anticipating postoperative complications, prioritizing microsystem factors, enhancing early escalation of care, and educating and empowering junior clinicians. When used appropriately, the failure to rescue quality metric can help institutions focus on improving processes of care that minimize morbidity and mortality from rescuable complications after cardiac surgery.

Original languageEnglish
Pages (from-to)487-496
Number of pages10
JournalCanadian Journal of Cardiology
Issue number4
StatePublished - Apr 2023


Dive into the research topics of 'Failure to Rescue: A Quality Metric for Cardiac Surgery and Cardiovascular Critical Care'. Together they form a unique fingerprint.

Cite this