医学
倾向得分匹配
优势比
内科学
糖尿病
置信区间
心肌梗塞
冲程(发动机)
围手术期
动脉瘤
外科
心脏病学
机械工程
工程类
内分泌学
作者
Ivancarmine Gambardella,Berhane Worku,Christopher Lau,Robert F. Tranbaugh,Sheida Tabaie,Natalia S. Ivascu,Leonard N. Girardi
出处
期刊:Annals of Surgery
[Ovid Technologies (Wolters Kluwer)]
日期:2022-07-15
卷期号:278 (2): e382-e388
被引量:3
标识
DOI:10.1097/sla.0000000000005572
摘要
To discern the impact of diabetes mellitus (DM) on spinal cord injury (SCI) after open descending thoracic and thoracoabdominal aneurysm repair (DTAAAR).Compared with euglycemia, hyperglycemia, and ketosis make neurons respectively more vulnerable and more resilient to ischemia.During the study period (1997-2021), patient who underwent DTAAAR were dichotomized according to the presence/absence of DM. The latter was investigated as predictor of our primary (SCI) and secondary [operative mortality (OM), myocardial infarction, stroke, need for tracheostomy, de novo dialysis, and survival] endpoints. Two-level risk-adjustment employed maximum likelihood conditional regression after 1:2 propensity-score matching.DTAAAR was performed in 934 patients. Ninety-two diabetics were matched to 184 nondiabetics. All preoperative variables had a standardized mean difference <0.1 between the matched groups. Patients with DM had higher SCI (6.5% vs. 1.6%, P 0.03) and OM (14.1% vs. 6.0%, P =0.01), while the other secondary endpoints were similar between groups in the matched sample. DM was an independent predictor for SCI in the matched sample (odds ratio: 5.05, 95% confidence interval: 1.17-21.71). Matched patients with DM presented decreased survival at 1 (70.2% vs. 86.2%), 5 (50.4% vas 67.5%), 10 years (31.7% vs. 36.7%) ( P =0.03). The results are summarized in the graphical abstract.DM is associated to increased OM and decreased survival, and it is an independent predictor of SCI after open DTAAAR. Strict perioperative glycemic control should be implemented, and exogenous ketones should be investigated as neuroprotective agents to reduce such adverse events.
科研通智能强力驱动
Strongly Powered by AbleSci AI