【摘要】：目的:探討食管癌(Esophageal carcinoma)圍術(shù)期死亡原因及危險(xiǎn)因素,建立logistics回歸模型。方法:收集2002年1月至2015年12月于新疆醫(yī)科大學(xué)第一附屬醫(yī)院進(jìn)行食管癌切除術(shù)患者臨床資料,食管癌850例,根據(jù)圍術(shù)期及住院期間結(jié)局分為死亡和非死亡各一組,將64個(gè)可能對(duì)食管癌手術(shù)死亡有相關(guān)影響的常見因素進(jìn)行回顧性分析,通過計(jì)算機(jī)用logistics回歸模型研究食管癌圍術(shù)期與死亡原因相關(guān)的危險(xiǎn)因素,進(jìn)行單因素、多因素分析,建立回歸方程,總結(jié)實(shí)際意義。結(jié)果:850例食管癌中手術(shù)死亡39例,手術(shù)死亡率4.59%。單因素logistics回歸分析,64個(gè)因素中,25個(gè)因素與食管癌切除術(shù)后圍術(shù)期死亡有關(guān),分別為民族,年齡,術(shù)前合并癥,術(shù)前血紅蛋白值,手術(shù)切除方式,吻合方式,吻合水平,手術(shù)年代,術(shù)后白蛋白值,腸內(nèi)營(yíng)養(yǎng)時(shí)間,ICU治療時(shí)間,心律失常,肺炎、肺部感染,重癥肺炎,單側(cè)胸腔積液,雙側(cè)胸腔積液,感染性休克,MODS,吻合口瘺,乳糜胸,呼吸衰竭,循環(huán)衰竭,呼吸循環(huán)衰竭,缺血缺氧綜合征等,統(tǒng)計(jì)學(xué)差異有意義(顯著性)(p0.05,具體變量定義詳見下文)。多因素logistics回歸分析提示手術(shù)年代,術(shù)后白蛋白值,腸內(nèi)營(yíng)養(yǎng)時(shí)間,心律失常,MODS,乳糜胸,呼吸衰竭等因素P0.05,有統(tǒng)計(jì)學(xué)意義,為本研究食管癌圍術(shù)期死亡原因獨(dú)立危險(xiǎn)因素并進(jìn)入進(jìn)入回歸方程,獲得預(yù)測(cè)模型為:P=1/1+e^(2.951X25+1.922X27+0.445X30-2.066X32+4.313X50+2.333X55-2.689X58-8.999),(0≤P≤1)。結(jié)論食管癌圍術(shù)期死亡原因分析中,手術(shù)年代、術(shù)后白蛋白值、心律失常、MODS、乳糜胸、呼吸衰竭等為本研究食管癌圍術(shù)期死亡原因獨(dú)立危險(xiǎn)因素并進(jìn)入回歸方程,并獲得預(yù)測(cè)模型;欲降低食管癌圍術(shù)期死亡,需改善醫(yī)療環(huán)境,提高醫(yī)療技術(shù),并在圍術(shù)期嚴(yán)密觀察和評(píng)估血清白蛋白值,科學(xué)并謹(jǐn)慎開始腸內(nèi)營(yíng)養(yǎng)時(shí)間,積極預(yù)防術(shù)后發(fā)生心律失常、MODS、乳糜胸、呼吸衰竭等并發(fā)癥,并盡早處理。
[Abstract]:Objective: to investigate the causes of death and risk factors of esophageal carcinoma during (Esophageal carcinoma) perioperative period and to establish a logistics regression model. Methods: from January 2002 to December 2015, 850 patients with esophageal carcinoma were divided into death group and non-death group according to the outcome of perioperative period and hospital stay, according to the clinical data of the patients undergoing esophageal cancer resection in the first affiliated Hospital of Xinjiang Medical University from January 2002 to December 2015. A retrospective analysis of 64 common factors related to the mortality of esophageal cancer surgery was carried out, and the risk factors related to the causes of death in the perioperative period of esophageal cancer were studied by computer logistics regression model, and the single factor and multivariate analysis were carried out. The regression equation is established and the practical significance is summarized. Results: among 850 cases of esophageal cancer, 39 cases died after operation, and the operative death rate was 4.59%. Univariate logistics regression analysis showed that 25 of 64 factors were related to perioperative death after resection of esophageal carcinoma, including nationality, age, preoperative complications, preoperative hemoglobin value, surgical resection mode and anastomosis level. Year of operation, postoperative albumin value, enteral nutrition time, ICU treatment time, arrhythmia, pneumonia, pulmonary infection, severe pneumonia, unilateral pleural effusion, bilateral pleural effusion, septic shock, MODS, anastomotic fistula, chylothorax, There were significant differences in respiratory failure, circulatory failure, respiratory and circulatory failure, ischemia and hypoxia syndrome, etc. Multivariate logistics regression analysis showed that there were significant differences in serum albumin, postoperative albumin, enteral nutrition time, arrhythmia, MODS, chylothorax, respiratory failure and other factors (P 0.05). In order to study the independent risk factors of perioperative death of esophageal cancer and enter the regression equation, the prediction model was obtained as follows: 1 / 1 e ^ (2.951X25 1.922X27 0.445X30-2.066X32 4.313X50 2.333X55-2.689X58-8.999). (0 鈮，

本文編號(hào)：2429268