急性胃肠损伤分级系统对重症患者适应性及临床预后的评估

Clinical application of Acutegastrointestinal injury grading system assocaited with clinical severity outcome in critically ill patients： a multi-center prospective, observational study

目的观察急性胃肠损伤（AGI）分级系统在重症医学科（ICU）患者中临床应用可行性和AGI严重程度对临床预后的评估意义。方法2014年3至8月在浙江省14个综合医院ICU内连续招募550例预期入住ICU〉24h的患者，前瞻性采集患者临床、实验室和生存数据，患者入ICU后第1周每日进行胃肠道症状、肠内喂养状况以及合并器官功能障碍评估，采用AbViser系统测定腹腔内压（IAP），综合评估患者AGl分级。结果入选患者平均年龄为（64．9±17．2）岁；APACHEⅡ评分为（19．5±7．4）分，456例（82．9％）患者接受机械通气治疗；470例患者发生AGI，其中，人ICU第1天AGI分级分别为50。6％（I级，n=238），34．2％（Ⅱ级，n=161），12．4％（Ⅲ级，n=58）和2．8％（Ⅳ级，n=13），而综合患者ICU7d内最高AGI分级分别为24．5％（I级，n=115），49．4％（Ⅱ级，n=232），20．6％（Ⅲ级，n=97）和5．5％（Ⅳ级，n=26）；28d和60d病死率分别为29．3％和32．5％。与非AGI患者相比，AGI患者28d（31．1％比18．8％，P=0．025）和60d病死率（34．7％比20．0％，P=0．01）显著增加，AGl分级严重程度与患者28d和60d病死风险增加呈显著正相关。单因素Cox回归分析示：年龄、入ICU内科来源、脓毒症、2型糖尿病、冠心病，缩血管药物使用、血乳酸和肌酐、接受机械通气、入ICU综合AGI分级和APACHEⅡ评分与60d病死预后显著相关（P≤0．02）；多因素Cox回归分析示：ICU内科来源（x2=4．34，P=0．04）、2型糖尿病（x2=3．96，P=0．015）、血管活性药物使用（x2=6．55，P=0．01）、血乳酸（x2=4．73，P=0．03）、入ICU综合AGI分级（x2=7．10，P=0．008）和APACHEII评分（x2=12．1，P〈0．001）是预测重症患者60d病死独立危险因素。此外，402例7d存活患者亚组分析显示：在入ICU第1天AGI分级和临床因素预测死亡的基础上，7d喂养不耐受能提供独立额外地预测其60d病死价值（x2=52．2与41．9，P=0．007）。结论AGI分级系统能有效识别重症患者胃肠功能障碍严重程度，并能预测其临床预后；本研究证据支持人ICU7d内喂养不耐受是死亡风险独立危险因素。

Objective To investigate the feasibility of utilizing the current acute gastrointestinal injury（AGI） grading system, and explore the association of severity of AGI grade with clinical outcome in critically ill patients. Methods The adult patients from 14 general ICUs in Zhejiang Province with an expected admission to ICU for at least 24 h were recruited, and all clinical, laboratory, and survival data were prospectively collected. The AGI grade was daily assessed based on GIsymptoms, feeding details and organ dysfunctionon the first week of admission to ICU. The intra-abdominal pressures （IAP） was measured using AbViser device. Results Of 550 patients enrolled, mean values for age and APACHE II score were （64. 9 ± 17.2 ） years and （ 19. 5 ±7. 4）, respectively. 456 patients （82. 9% ） took mechanical ventilation, and 470 patients were identified for AGI. The distribution of AGI grade on the frist day of ICU admission were 50. 6%（ I grade, n=238）, 34. 2%（ II grade, n=161）, 12.4%（ III grade, n=58） and 2. 8% （IV, n = 13 ）, respectively, while the distribution of the global AGI grade based on the 7-day AGI assessment of ICU admission were 24. 5% （ I grade, n = 115）, 49.4% （ II grade, n = 232）, 20. 6% （ Ill grade, n =97） and 5.5% （ IV, n =26）, respectively. 28- and 60-day mortality rate was 29. 3% （n = 161） and 32. 5% （n = 179）, respectively. The patients with AGI had a higher 28-（31.1% vs 18.8%, P= 0. 025） and 60-day survival rate（34. 7% vs 20. 0% ,P = 0, 01 ） than those with non-AGI, and also there were positive correlations between AGI grade and 28- and 60-day mortality （ P 〈 0. 001 ）. Univariate Cox regression analysis showed that age, the source of medicial admission, diabetes mellitus, coronary heart disease, the use of vasoactive drugs, serum creatinine and lactate, mechanical ventilation, APACHE II score, the AGI grade in the first day of ICU admission and feeding intolerance within the first week of ICU stay were significantly （P ≤ 0. 02 ） associated with mortality. In multivariate analysis including all these variables, the source of medical admission（ X2 = 4. 34, P = 0. 04 ）, diabete mellitus （ X2 = 3.96, P = 0. 05 ）, the use of vasoactive drugs （ X2 = 6. 55, P = 0. 01 ） , serum lactate （ X2 = 4. 73, P = 0. 03 ）, the global AGI grade in the 7-day of ICU admission（x2 =7.10,P=0.008）, and APACHE II score（x2 =12.1,P〈 0. 001 ） remained independent predictors for 60-day mortality. In the further subgroup analysis including 402 patients with 7-day survival, the feeding intolerance within the first week of ICU stay could provide independent and incremental prognostic value of 60-day mortality wtih increased x2 value of Cox regression model（ X2 = 52. 2 vs 41.9, P = 0. 007 ） . Conclusion The AGI grading system is useful for identifying the severity of gastrointestinal dysfunction, and could be used as a strong predictor of impaired outcome. The results provide evidence to support that feeding intolerance within 7 days of admission to ICU was an independent determinant of mortality.