• Albert Wu, Daniel E Rinewalt, Robert W Lekowski, and Richard D Urman.
• Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women's Hospital, Boston, MA.
• J. Cardiothorac. Vasc. Anesth. 2017 Jun 1; 31 (3): 810-815.

ObjectivesTo test whether a model using a historical average of a surgeon's surgical times for primary aortic valve replacements is a more accurate predictor of actual surgical times than solely relying on a surgeon's estimate.DesignRetrospective review.SettingSingle university hospital that serves as a tertiary referral center.ParticipantsAll patients undergoing primary aortic valve replacement between October 2008 and September 2014.InterventionsNone.Measurements And Main ResultsEstimation biases, calculated as the difference between actual and predicted surgical time, were compared between the surgeon and the model, which included between 2 and 20 cases in the historical average. Kruskal-Wallis analysis of variance was used to compare all values. Pairwise comparisons were made using the Steel-Dwass test to determine whether using more cases in the model resulted in smaller estimation biases. Using the historical model reduced mean overestimation bias from 55.30 minutes to 0.90-to-4.67 minutes. No significant difference was seen based on the number of cases used.ConclusionsAn uncomplicated model can assist in providing comparatively unbiased estimations of surgical time for aortic valve replacements. The model can rely on a fewer number of cases (eg, 5) and does not benefit from including more cases (eg, 20).Copyright © 2017 Elsevier Inc. All rights reserved.

22 keywords...

hide…