When it comes to predicting surgical risk, experienced surgeons and computer statistical models are on equal footing, according to an article in Journal of the American College of Surgeons.
Researchers compared surgeon and statistical models for 30-day mortality, overall morbidity, surgical site infections and cardiac, thromboembolic, pulmonary and renal complications for nearly 1,800 general surgery patients seen between June 2010 and January 2012.
Surgeons and statistical models predicted similar risk outcomes congruous with actual outcomes. For example, a surgeon prediction of low overall morbidity risk was similar to the model prediction of 2.8 percent morbidity and actual morbidity, 4.1 percent. Surgeon prediction of high morbidity risk correlated with 10 percent model morbidity and 18 percent actual morbidity.
Researchers suggested using models to predict postoperative complications may be a good way to preemptively identify patients who may be at risk for complications and intervene to reduce that risk.
More Articles on Quality:
Study: Surgical Failure More Likely for Emergency, Post-ICU Patients
Study: Diverse Diet Predicts Lower Risk of Hospitalization for Elderly