Proximal vs Extensive Repair in Acute Type A Aortic Dissection Surgery.

Abstract

This purpose of this study was to evaluate the impact of proximal vs extensive repair on mortality and how this impact is influenced by patient characteristics. Of 5510 patients with acute type A aortic dissection from 13 Chinese hospitals (2016-2021) categorized by proximal vs extensive repair, 4038 patients were used for for model derivation using eXtreme gradient boosting and 1472 patients for model validation. Operative mortality of extensive repair was higher than proximal repair (10.4% vs 2.9%; odd ratio [OR], 3.833; 95% CI, 2.810-5.229; P < .001) with a number needed to harm of 15 (95% CI, 13-19). Seven top features of importance were selected to develop an alphabet risk model (age, body mass index, platelet-to-leucocyte ratio, albumin, hemoglobin, serum creatinine, and preoperative malperfusion), with an area under the curve of 0.767 (95% CI, 0.733-0.800) and 0.727 (95% CI, 0.689-0.764) in the derivation and validation cohorts, respectively. The absolute rate differences in mortality between the 2 repair strategies increased progressively as predicted risk rose; however it did not become statistically significant until the predicted risk exceeded 4.5%. Extensive repair was associated with similar risk of mortality (OR, 2.540; 95% CI, 0.944-6.831) for patients with a risk probability < 4.5% but higher risk (OR, 2.164; 95% CI, 1.679-2.788) for patients with a risk probability > 4.5% compared with proximal repair. Extensive repair is associated with higher mortality than proximal repair; however it did not carry a significantly higher risk of mortality until the predicted probability exceeded a certain threshold. Choosing the right surgery should be based on individualized risk prediction and treatment effect. (ClinicalTrials.gov no. NCT04918108.).