Perioperative Risk Models: A Narrative Review

Articles selected for review were based on a MEDLINE/PubMed search utilizing Boolean logic and medical subject heading terms as outlined in Supplementary Appendix A, as well as author's personal experience. There was a focus on including randomized controlled trials and observational studies published in the past ten years, and studies of lower-quality evidence, e.g. retrospective studies, are specifically noted when discussed. Older models beyond the ten-year search period were reviewed when deemed appropriate for historical context or if still in common clinical use. Studies examining multiple rather than single variable predictors of risk were sought, and we specifically excluded single-variable models. Our aim was to focus on patients undergoing noncardiac surgical interventions, since there is already a wealth of published data on risk models and predictors of outcomes following cardiac surgery Thus, cardiac specific models were excluded. Additionally, non-surgical risk models were excluded from this review. See Supplementary Appendix A for more details