Perioperative risk assessment for emergency general surgery in those with multimorbidity or frailty

This review explores advances in risk stratification tools and their applicability in identifying and managing high-risk emergency general surgery (EGS) patients.

Traditional risk assessment tools have several limitations when applied to complex EGS patients as comorbidities are generally treated in a binary, linear and sequential fashion. Additionally, some tools are only usable in the postoperative period, and some require multidisciplinary involvement and are not suitable in an emergency setting. Frailty in particular - for which there are multiple calculators-maladaptively influences outcomes. Artificial intelligence tools, such as the machine-learning-based POTTER calculator, demonstrate superior performance by addressing nonlinear interactions among patient factors, offering a dynamic and more accurate approach to risk prediction.

Integrating advanced, data-driven risk assessment tools into clinical practice can help identify and manage high-risk patients as well as forecast outcomes for EGS patients. Such tools are intended to trigger preoperative interventions as well as discussions that ensure goal-concordant care, align expectations with anticipated outcomes and support both facility and patient-relevant outcomes.