Skip to Main Content

HIGHLIGHTS

HIGHLIGHTS

  • Preoperative risk stratification allows surgeons to set patient and family expectations for postoperative outcomes.

  • Artificial intelligence (AI) allows risk stratification calculators to accurately capture the complex interactions of preoperative risk factors in surgical patients.

  • Newly developed preoperative risk calculators use different technologies within AI, each with its own unique set of capabilities.

  • Although AI improves the accuracy of preoperative risk stratification, surgeons should remain aware of the technology’s limitations and pitfalls.

INTRODUCTION

Picture a 45-year-old gentleman with known moderate cirrhosis presenting to the emergency department with acute cholecystitis. On evaluation, the patient appears frail and fatigued, complaining of fever and abdominal pain. Should the surgeon book the patient for an emergent cholecystectomy? Will the patient survive the postoperative course? Would percutaneous cholecystostomy be a more appropriate temporizing measure while the patient is clinically optimized for surgery? How should the risks, benefits, and alternatives to surgery be explained to the patient and his family?

Clinical scenarios that carry a significant risk of postoperative complications, like the one just described, are common. Primum non nocere1 is a philosophy embodied in the Hippocratic oath made by doctors worldwide. The maxim is intuitive and theoretically unproblematic. However, contemporary medical practice is complex, and despite best intentions, some patients are inadvertently harmed. The Institute of Medicine estimates that almost 98,000 people die every year from medical errors.2 The world of surgery is no exception to this reality: errors happen. Even when a surgeon’s decision to operate was sound and a technically safe procedure was performed, postoperative adverse events and complications are still common. Postoperative complications occur in 15% of the 19 million surgeries performed yearly in the United States,3 incurring a total cost of 31.35 billion dollars.4 As such, many preoperative risk stratification systems have been created over the past 2 decades to support surgeon experience and gestalt in estimating a patient’s postoperative course. Throughout the years, these systems have proved to be invaluable, allowing providers to set patient and family expectations and to guide preoperative decision making.

This chapter tackles the introduction of artificial intelligence (AI) into preoperative risk stratification by providing an overview of the AI technologies used in state-of-the-art preoperative surgical risk calculators. The chapter also addresses the limitations and pitfalls of AI in risk stratification, shedding light on the future of preoperative risk stratification using AI methodologies.

BACKGROUND ON LINEAR RISK STRATIFICATION

Prior to delving into the world of AI, some background knowledge of existing non-AI risk stratification models can help readers better understand the advances offered by this new technology.

One of the earliest non-AI preoperative risk stratification tools that is still in use today is the American Society of Anesthesiologists (ASA) Physical Classification System. Originally conceived in 1941, the scale allows the simple albeit subjective classification of preoperative patients into ...

Pop-up div Successfully Displayed

This div only appears when the trigger link is hovered over. Otherwise it is hidden from view.