机构地区:[1]第三军医大学西南医院全军感染病研究所,重庆400038
出 处:《胃肠病学和肝病学杂志》2016年第5期495-498,共4页Chinese Journal of Gastroenterology and Hepatology
基 金:国家自然科学基金面上项目(81270525)
摘 要:目的建立糖皮质激素治疗慢加急性肝衰竭前期(acute-on-chronic pre-liver failure,pre-ACLF)的疗效预测模型与路线图。方法根据5 d时血清总胆红素(TBil)的降幅,将56例乙型肝炎pre-ACLF患者对糖皮质激素治疗的应答分为快速应答(>50%)、部分应答(30%~50%)和无应答(<30%);对快速与非快速应答组患者的基线特征进行多因素回归分析,找出影响糖皮质激素治疗快速应答的相关因素,并构建糖皮质激素疗效的预测模型与路线图。结果黄疸持续时间(P=0.013,OR=1.258,95%CI:1.050~1.508)、凝血酶原活动度(PTA,P=0.029,OR=0.014,95%CI:0.000~0.650)及丙氨酸转氨酶(ALT)/TBil比值(P=0.013,OR=0.735,95%CI:0.577~0.963)是影响糖皮质激素治疗快速应答的独立因素;推算出的预测模型为:Py=ex/(1+ex),X=0.687+0.125×黄疸时间-0.508×ALT/TBil比值-4.241×PTA;Py值<0.1、0.1~0.2、0.2~0.5和≥0.5患者的快速应答率分别为82.1%、55.6%、28.6%和0,彼此间差异有统计学意义(χ~2=24.809,P=1.69×10^(-5))。快速应答组停用激素5 d后血清TBil反弹上升者为6.7%,显著低于部分应答组的36.4%(χ~2=5.682,P=0.017)和无应答组的80.0%(χ~2=25.092,P=5.47×10^(-7)),但部分应答组与无应答组之间差异无统计学意义(χ~2=3.428,P=0.064)。结论本研究首次建立的数学模型能有效预测糖皮质激素的疗效,并提出了糖皮质激素治疗的路线图。Objective To establish a predictive model for the efficacy of glucocorticoids in treating acute-on-chronic pre-liver failure (pre-ACLF) and roadmap. Methods Fifty-six patients with hepatitis B pre-ACLF who had responses to glucocorticoids therapy were divided into three categories based on the decline extent of serum TBil at 5 days. The patients with the decline extent of serum TBil of 50% or more were defined as rapid responders, 30% -50% were as par- tial responders and less than 30% were as suboptimal responders. Multivariate analysis was carried on the baseline characteristics of rapid and non-rapid responders, in order to identify the main factors associated with rapid responses to glucocorticoids therapy. Then, a predictive model for the effect of glucocorticoids therapy and roadmap were established. Results The time of jaundice (P=0.013, 0R=1.258, 95% CI: 1.050 -1.508), prothrombin activity (PTA, P= 0.029, OR =0.014, 95% CI: 0.000 -0.650) and the ratio of ALT/TBil (P =0.013, OR =0.735, 95% CI: 0.577 ~0. 963) were independent predictive factors for the effect of glucocortieoids therapy. The created predictive model for the effect of glucocorticoids therapy was as follows : Py = eX/( 1 + ex) , X = 0. 687 + 0. 125 x Time of jaundice - 0. 508 x ALT/T-TBil ratio-4. 241 x PTA. There was significant difference with the percentage of rapid responders among the patients with Py value being〈0.1 (82.1%), 0. 1 -0.2 (55.6%), 0.2 ~0.5 (28.6%) and ≥0.5 (0) (X2 = 24. 809, P = 1.69 × 10 -5). The percentages of patients with serum bilirubin rebound was significantly lower in rapid responders (6.7 % , 2/30) than that (36.4% , 4/11 ) in partial responders (X2 = 5. 682, P = 0. 017 ) , and that (80.0% , 12/15) in suboptimal responders (X2 = 25. 092, P = 5.47 × 10-7), but there was no statistically significant difference between partial responders and suboptimal responders (X2 = 3. 428, P = 0. 064). Conclusion The logistic regression model being cr
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...