How the Post-Op Risk Assessments of General Surgery & Internal Medicine Residents Compare
Whether you’re an internist or a general surgeon, you want to provide surgically complex patients with the best evaluation possible for their postsurgical risk for complications. But according to a recent investigation published by the JAMA Network, you might be overestimating the risk, regardless of your medical training.
Researchers asked how internal medicine and general surgeon trainees compared when assessing post-op patient risk assessments. Seventy-six participants from each group assessed seven surgically complex scenarios: colectomy, duodenal ulcer repair, inguinal hernia repair, perforated viscus exploration, small bowel resection, cholecystectomy, and mastectomy. The researchers then compared their assessments to the American College of Surgeon’s National Surgical Quality Improvement Program’s (NSQIP) surgical risk calculator, a decision-support tool which draws upon NSQIP’s extensive surgical database, and allows surgeons to input 21 patient characteristics for 12 different types of complications over a 30-day postoperative period.
The results found that both sets of residents overestimated every type of postsurgical risk by 26-33 percent in comparison to the Surgical Risk (SRC) model. However, the study also determined that general surgery residents were more confident than internal medicine (IM) residents in their predictions and about not offering operations, as well as less likely to consult risk-adjusted models.
While surgeons and internal IM physicians develop a reliable sense of risk over time, and often collaborate around surgical decisions, the SRC has been promoted for its empirically-based risk predictions that are free of human bias. The study was the first of its kind in establishing the lack of significant difference between surgeon and IM decision-making, and indicated a need for tools like the SRC. James Healy, MD, head researcher from Yale University School of Medicine states that the study emphasizes the “importance and availability in risk-adjusted models...to provide a consistent, individualized, and evidence-based assessment of surgical risk to patients.”
Other Obstacles to the use of SRC
Even with the findings, there are two potential obstacles to a more fully adopted use of tools like the SRC: surgeons may not be swayed by objective models, and the SRC may not yet be as effective in its predictions. A 2015 study by the American College of Surgeons found that surgeons who used the calculator were equally likely to recommend surgery as those who hadn’t. Study investigator Greg Sacks, MD, from UCLA pointed to how difficult it is for surgeons to predict risk for a patient, and that “The decision to operate often involves a tradeoff between the risks and benefits of operating and not operating.”
Since the SRC was released in 2013, there have been 21 surgical research papers evaluating its use. Many concluded that the SRC was a fair to poor predictor of surgical outcomes in certain types of procedures. The studies were examined by the American College of Surgeons Division of Research and Optimal Patient Care and surgical departments at the University of California Los Angeles and Washington University, St. Louis, Mo. Their analysis also showed that many of the study designs had limited reliability due to three factors: sample size, homogeneity of case mix, and the scope of the data set.
While the review concluded that the SRC had a rate of predictive failure, researchers emphasized that this should not disqualify the SRC as a useful tool for doctors and patients. Study author Dr. M.E.Cohen said that, while imperfect, “Surgeons should not be dissuaded from relying on the calculator for determining the risk for various outcomes.”
Regardless of your background, it’s important to utilize all available resources when discussing postsurgical risks with your patient. Combining tools like the SRC with your medical expertise and personal assessment will enable you to help your patient make the best surgical choice possible.