急性生理功能和慢性健康状况评分系统Ⅱ和Ⅳ对192例重度急性胰腺炎死亡风险预测的比较  

Comparison between acute physiology and chronic health evaluation Ⅱ and Ⅳ in mortality risk prediction of 192 servere acute pancreatitis patients

作　　者：韩莉[1] 康焰[1] 

机构地区：[1]四川大学华西医院重症医学科,成都610041

出　　处：《中华消化杂志》2016年第3期177-181,共5页Chinese Journal of Digestion

摘　　要：目的比较急性生理功能和慢性健康状况评分系统（APACHE）Ⅱ和APACHEⅣ对重度急性胰腺炎（SAP）患者死亡风险的预测效能。方法单中心、回顾性分析2013年1月至2014年12月ICU收治的SAP患者。采集患者进入ICU后第1个24h的临床数据，计算每个患者的改良Marshall评分、APACHEII和APACHE1V评分，以及APACHEⅡ和APACHEⅣ死亡风险预测值。根据患者最后临床结局，采用Hosmer—Lemeshow拟合优度检验比较实际和预测病死率的结果评估APACHEⅡ和APACHEⅣ预测模型对患者死亡风险的校准度，绘制ROC曲线比较两个评分系统的分辨度。结果最终纳入192例SAP患者，存活152例（79．2％），死亡40例（20．8％）。死亡组患者改良Marshall评分、APACHEⅡ和APACHEⅣ评分结果分别为（6．30±0．36）、（21．3±8．0）和（88．1±30．2）分，而存活组分别为（3．70±0．20）、（12．3±5．6）和（53．4±19．0）分，两组间比较差异均有统计学意义（t=-6．436、-6．683、-6．913，P均d0．01）。经Hosmer-Lemeshow拟合优度检验，APACHEⅡ与APACHEIV评分预测模型的校准度差异无统计学意义（P〉0．05），两者均能较好地预测患者的死亡风险。APACHEⅡ评分（截断值≥26分）和APACHEIV评分（截断值≥91分）的AUC值分别为0．81（95％C10．74～0．89）和0．83（95％C10．75～0．90），差异无统计学意义（X^2=0．21，P=0．644），两者的分辨度无差异。结论APACHEⅣ对SAP患者的预后预测并不优于APACHEⅡ，对SAP患者使用APACHEⅡ即可较准确评估患者预后。[Abstract] Objective To compare the efficiency of acute physiology and chronic health evaluation （APACHE）Ⅱ and Ⅳ in mortality risk prediction of severe acute pancreatitis （SAP）. Methods From January 2013 to December 2014, SAP patients admitted to intensive care units （ICU） were retrospectively analyzed in single center. The clinical data of the first 24 hours since the patients admitted into ICU were collected. The modified Marshall score, APACHE Ⅱ and APACHE Ⅳ score were calculated. The mortality risk predictive value of each patient was calculated by APACHE Ⅱ and APACHE Ⅳ. According to the final clinical outcome of patients, Hosmer-Lemeshow was performed to compare real mortality rate with predictive mortality rate, and calibration of APACHE Ⅱ and APACHE Ⅳ in the mortality risk of each patients was evaluated. The resoluation of the two scoring systems was compared by the area under the receiver operator characteristic curve （AUC）. Results Intheend, 192 patients （152（79. 2%） survivors and 40（20. 8%） dead） were enrolled. Modified Marshall score, APACHE Ⅱ score and APACHE Ⅳ score of patients in dead group was 6.30±0.36, 21.3±8.0 and 88.1±30.2, respectively; and those of survival group was 3.70 ±0. 20, 12.3±5. 6 and 53. 4±19.0, respectively, and the differences between two groups were statistically significant （t = 6. 436,- 6. 683, -6. 913,all P〈 0. 01）. The results of Hosmer-Lemeshowin calibration of APACHE Ⅱ and APACHE Ⅳ indicated that both two systems could predict mortality risk of SAP patients well （P〉0. 05）. The AUC of APACHE Ⅱ score （cut-off ≥26） and APACHE Ⅳ score （cut off≥91） was 0.81（95%CI 0.74 to 0.89） and O. 83（95%CI 0. 75 to 0.90）, respectively, and the difference was not statistically significant （Zz 0.21, P=0. 644）, which indicated that there was no statistically significant difference in calibration. Conclusions APACHE Ⅳ scoring system is not better than APACHE Ⅱ scoring system in prognosis predicti

关 键 词：APACHEⅡ APACHE IV 重症急性胰腺炎 预后 

分 类 号：R576[医药卫生—消化系统]