Not Just a Label – Preoperative Risk Assessment Can Change Patient Outcomes

Young May Cha, MD

Anxiety about the risks of surgery is common among patients and their families. Despite decades of research and a plethora of risk assessment tools, few surgeons and anesthesiologists regularly use these tools. In the session, “Does Preoperative Risk Prediction Improve Perioperative Outcome? An Update,” moderator Joshua Bloomstone, MD, MSc, FASA, D.ABA, SSGB, clinical professor of anesthesiology at the University of Arizona College of Medicine – Phoenix, introduced three speakers to review the role for these tools in improving patient outcomes.

Rafael Noriega, MD, MPH, assistant clinical professor of anesthesiology at the University of Arizona College of Medicine – Phoenix, provided an overview of the available risk prediction tools. Individualized risk assessment tools, such as the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) risk calculator, provide data-driven and procedure-specific risk prediction. General tools, such as frailty tools, are not specific for a procedure but can capture risk for poor outcomes not covered by other risk calculators. The Physiological and Operative Severity Score for the enUmeration of Mortality and morbidity (POSSUM) and its modification Portsmouth-POSSUM (P-POSSUM) are validated risk-scoring systems to predict postoperative mortality that incorporate intraoperative data. However, since they require intraoperative data, they are not as useful for assessment of preoperative risk. Other commonly used risk assessment tools, such as the ASA Physical Status classification system and the Revised Cardiac Risk Index (RCRI) are global assessments that can be easily calculated at bedside, but they are not procedure-specific. The RCRI predicts major cardiac events but only predicts cardiac risk. It can underestimate risk in our modern population that is older with more comorbid conditions. Biomarkers, such as B-type natriuretic peptide (BNP), troponin, cystatin C, and procalcitonin, can also be used to assess preoperative risk but these are unfortunately not readily available at every institution. Each of these tools can help inform the anesthesiologist of the various risks, which can alter management, and hopefully change outcomes.

Patients want to be informed about their risk for surgery, and many would consider prehabilitation to lower their risk. Benjamin Houseman, MD, PhD, FASA, Program Director at Memorial Healthcare System, discussed the importance of considering surgical risk prediction through the lens of shared decision-making. There are many patient-specific barriers, such as recall failure due to cognitive load issues during the consent process and comprehension deficits that extend beyond language barriers. Moreover, regulatory frameworks mandate individualized risk communication, but fewer than 1 in 20 anesthesiologists document organ-specific morbidity or mortality risks. Dr. Houseman referenced yesterday’s session, “Shared Decision-Making in Perioperative Care” to emphasize that true shared decision-making is a dynamic, longitudinal process that begins preoperatively and continues after discharge. Other barriers include time, the downstream position of the anesthesiologist in surgical planning, the task-focused culture of the perioperative environment, and the truncated format of the average anesthesia consultation. Recent studies have incorporated technologies, such as virtual reality (VR) and artificial intelligence (AI) models. VR-assisted consents were well-received by patients and decreased time needed for the specialist. In a comparison of AI models, ChatGPT and Gemini performed well to explain general complications of anesthesia. To improve documentation, individualized risk calculators should be embedded in the electronic health record and tracked as a clinical quality metric. Such metrics can then influence behavior. Patient-centered outcomes and experience measures should also be included so there can be better clarity on what a successful outcome looks like for our patients.

With the many risk assessment tools available, all anesthesiologists should be using them in their day-to-day clinical decision-making. But surgical risk assessment as a standalone risk is insufficient. The ethical imperative of beneficence calls on us to improve patient outcomes and respect patient autonomy for a truly informed decision. Kamal Maheshwari, MD, MPH, FASA, anesthesiologist at Kaiser South Sacramento Hospital, discussed how we can take the next step and leverage all the data available to build even more comprehensive predictive models. With the inclusion of data from wearable devices and genetic information, for example, large language models can be used to synthesize complex patient information into individualized decision support. This personalization can optimize perioperative care with prehabilitation, individualized care pathways in the hospital, and individualized postoperative monitoring at home. Risk assessments are now considering additional outcomes, such as acute kidney injury, and more procedure-specific outcomes. Generative AI can be used to create an intelligent health record to better tailor a patient’s perioperative course and incorporate patient-specific goals to improve quality of life. Surgeries cannot always be delayed, but investing in resources like preoperative risk assessment can provide better returns with improved patient care.