Proximal Versus Extensive Repair in Acute Type A Aortic Dissection Surgery

Hong Liu,Ying-yuan Zhang,Xiao-hang Ding,Si-chong Qian,Ming-yu Sun,Al-Wajih Hamzah,Ya-nan Gao,Yong-feng Shao,Hai-yang Li,Kai Wang,Bu-qing Ni,Hong-jia Zhang
DOI: https://doi.org/10.1016/j.athoracsur.2023.04.019
2023-04-27
Abstract:BACKGROUND To evaluate the impact of proximal versus extensive repair on mortality, and how this impact is influenced by patient characteristics. METHODS 5510 acute type A aortic dissection (ATAAD) patients from 13 Chinese hospitals (2016-2021) were categorized by proximal versus extensive repair: 4038 patients for model derivation using eXtreme gradient boosting and 1472 patients for model validation. RESULTS 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<0.001) with the number needed to harm (NNH) of 15 (95%CI 13-19). Seven top features of importance were selected to develop Alphabet risk model (age, body mass index, platelet–leucocyte ratio, albumin, hemoglobin, serum creatinine, urea nitrogen, and preoperative malperfusion), with an AUC of 0.767 (95%CI 0.733-0.800) and 0.727 (95%CI 0.689–0.764) in the derivation and validation cohort, respectively. The absolute rate differences in mortality between two 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 risk probability 4.5% compared with proximal repair. CONCLUSIONS 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 exceed a certain threshold. Choosing the right surgery should be based on individualized risk prediction and treatment effect.
surgery,cardiac & cardiovascular systems,respiratory system
What problem does this paper attempt to address?