不同评分系统预测高脂血症性急性胰腺炎严重程度的比较研究

发布时间：2018-12-15 10:07

【摘要】：目的：研究高脂血症性急性胰腺炎的临床特点，比较BISAP、Ranson评分、CTSI、SIRS、HAPS五种评分系统对早期预测高脂血症性急性胰腺炎患者的病情严重程度、局部并发症、器官衰竭、死亡的能力，探讨BISAP、Ranson、CTSI、SIRS、HAPS评分系统在早期预测高脂血症性急性胰腺炎患者病情严重程度的优缺点。方法：收集2008年9月至2014年3月收治于上海市第一人民医院南院和上海市第十人民医院的909例急性胰腺炎患者的临床资料，应用回顾性分析研究方法，比较高脂血症性急性胰腺炎、非高脂血症性急性胰腺炎的临床特点；利用BISAP、Ranson、CTSI、SIRS、HAPS对所有急性胰腺炎患者进行评分，，比较各评分在轻度、中度、重度急性胰腺炎的分布差异；通过受试者工作特征曲线（receiver operating characteristiccurve，ROC curve）分析BISAP、Ranson、CTSI、SIRS对高脂血症性急性胰腺炎及非高脂血症性急性胰腺炎组预测重度胰腺炎、局部并发症、器官衰竭、死亡的能力。对比BISAP、Ranson、CTSI评分系统在高脂血症性急性胰腺炎组及非高脂血症性急性胰腺炎组预测重度急性胰腺炎的能力；通过一致性检验，评估无害性胰腺炎预测高脂血症性急性胰腺炎组及非高脂血症性急性胰腺炎组轻症胰腺炎的能力。结果：909例次急性胰腺炎患者中，高脂血症性急性胰腺炎129例（14.2%），其中轻度急性胰腺炎68例，中度急性胰腺炎41例，重度急性胰腺炎20例，并发假性囊肿8例，胰腺包裹性坏死9例，胸腔积液30例，全身炎症反应综合征33例，持续性器官衰竭14例，死亡1例。高脂血症性急性胰腺炎患者中五种评分系统的评分分值在轻度、中度、重度急性胰腺炎三组中存在统计学差异（P0.05）。BISAP、Ranson、SIRS、CTSI在预测重度急性胰腺炎时的ROC曲线下面积（the area under receiveroperating characteristic curve，AUC）分别为0.905、0.938、0.812、0.834，两两比较差异无统计学差异（P0.05），最佳阈值（cutoff）分别为1、2、2、4，约登指数分别为0.604、0.608、0.640、0.668；在预测局部并发症时AUC分别为0.874、0.726、0.668、0.848，AUC两两比较时仅BISAP、Ranson与SIRS之间AUC差异存在统计学差异(P0.05)，约登指数分别为0.631、0.402、0.294、0.540；在预测器官衰竭时分别为AUC分别为0.904、0.917、0.758、0.849，AUC两两比较时，Ranson、SIRS两组的AUC差异具有统计学意义(P0.05)，约登指数分别为0.634、0.661、0.509、0.616；在预测死亡时四种评分系统无统计学意义(P0.05)。无害性胰腺性胰腺炎与高脂血症性急性胰腺炎组按亚特兰大分级的轻度急性胰腺炎一致性检验中，一致性差(P0.05)。结论：高脂血症性急性胰腺炎具有发病年龄轻，复发率高的特点，并发胰腺坏死、SIRS、器官衰竭的发病率较非高脂血症性急性胰腺炎组高。BISAP、Ranson、SIRS、CTSI对预测重度急性胰腺炎均具有较好的准确性，且能力相似；BISAP、Ranson、CTSI预测局部并发症的能力相似，优于SIRS；SIRS评分在预测器官衰竭方面不如BISAP、Ranson、CTSI；四种评分预测死亡率无明显意义。HAPS并不适用预测轻度高脂血症性急性胰腺炎。
[Abstract]:Objective: To study the clinical characteristics of acute pancreatitis with hyperlipidemia, compare the five scoring systems of BISAP, Ranson, CTSI, SIRS and HAPS to the early prediction of the severity, local complications, organ failure and death of the patients with acute pancreatitis. The advantages and disadvantages of the CTSI, SIRS and HAPS scoring system in the early prediction of the severity of the patients with hyperlipidemic acute pancreatitis. Methods: The clinical data of 909 patients with acute pancreatitis in the first People's Hospital of Shanghai and the 10th People's Hospital of Shanghai were collected from September 2008 to March 2014. The clinical data of 909 patients with acute pancreatitis were analyzed retrospectively. The clinical characteristics of acute pancreatitis associated with inflammation and non-hyperlipoidemia were assessed by using BSAP, Ranson, CTSI, SIRS, and HAPS in all patients with acute pancreatitis. The distribution of each score in mild, moderate and severe acute pancreatitis was compared. e. ROC curve analysis of the ability of BSAP, Ranson, CTSI, SIRS to predict severe pancreatitis, local complications, organ failure, and death in patients with acute pancreatitis and non-hyperlipidemia acute pancreatitis The ability of the BSAP, Ranson, and CTSI scoring systems to predict severe acute pancreatitis in the group of hyperlipidemic acute pancreatitis and the non-hyperlipidemic acute pancreatitis group; and by consistency test The effects of acute pancreatitis in the treatment of hyperlipidemic acute pancreatitis and non-hyperlipidemic acute pancreatitis Results: Of the 909 patients with acute pancreatitis, 129 (14. 2%) of the patients with severe acute pancreatitis, including 68 cases of mild acute pancreatitis, 41 with moderate acute pancreatitis, 20 cases of severe acute pancreatitis, 8 cases of concurrent pseudocyst, 9 cases of pancreatic inclusion necrosis, and pleural effusion 30 cases, 33 cases of systemic inflammatory response syndrome, 14 cases of persistent organ failure, and death The scores of the five scoring systems in the patients with hyperlipidemic acute pancreatitis had a statistical difference in the three groups of mild, moderate and severe acute pancreatitis (P0.05). The area under the ROC curve of the BSAP, Ranson, SIRS and CTSI in the prediction of severe acute pancreatitis was 0.905, 0.938, 0.812, and 0, respectively. There was no statistical difference between the two groups (P0.05). The best threshold (coutoff) was 1, 2, 2, 4, and the ~ (n) index was 0.604, 0.608, 0.640, and 0.668 respectively. The AUC in the predicted local complications was 0. 874, 0. 726, 0. 668, 0. 848, and the AUC was 0.631, 0.402, 0.294, 0, respectively. The AUC of the two groups was 0.904, 0.917, 0.758, 0.849 and AUC, respectively. The difference of AUC in the two groups was 0.634, 0.661, 0.509, and 0.616, respectively. 05). The consistency of the non-benign pancreatic pancreatitis and the acute pancreatitis group of the hyperlipidemic acute pancreatitis in the consistency test of the mild acute pancreatitis in Atlanta (P0. Conclusion: Hyperlipidemic acute pancreatitis has the characteristics of light weight and high recurrence rate, and the incidence of pancreatic necrosis, SIRS and organ failure is higher than that of non-hyperlipidemic acute pancreatitis. The high. BSAP, Ranson, SIRS and CTSI in the group with severe acute pancreatitis had good accuracy in predicting the severe acute pancreatitis, and the ability was similar; the ability of the BSAP, Ranson, CTSI to predict local complications was similar to that of the SIRS; the SIRS score was inferior to that of the BSAP, Ranson, and CTSI in predicting organ failure; and the death rate was predicted by four scores. There is no significant difference. HPS does not apply to the prediction of mild hyperlipoidemia
【学位授予单位】：苏州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R589.2;R576

【参考文献】