天冬氨酸转氨酶/丙氨酸转氨酶比值与缺血性心脏病患者28天死亡率之间的关联:一项MIMIC-IV数据库分析

Association between aspartate transaminase/alanine transaminase ratio and 28-day mortality in patients with ischemic heart disease:a MIMIC-IV database analysis

摘要:

目的]探讨天冬氨酸转氨酶(AST)/丙氨酸转氨酶(ALT)比值(De Ritis比值)对缺血性心脏病患者28天死亡率的影响及其关联性,并分析该比值在合并不同疾病人群中的具体预测价值。 [方法]采用回顾性研究方法,经过缺失值筛选,最终提取了7 513例缺血性心脏病患者的数据,所有数据均来源于MIMIC-IV数据库。根据缺血性心脏病ICD编码进行提取,对时间和实验室缺失数据采用均值插补方法。对于缺失率超过20%的变量,在模型中将其剔除。采用包含限制性立方样条(4个节点)的多变量Cox模型评估De Ritis比值,并调整SOFA评分、APACHEⅡ评分、器官支持和生物标志物。LASSO回归用于变量选择。通过ROC-AUC和校准指标评估模型性能。预先设定的亚组检验合并症对De Ritis比值影响的调节作用。 [结果]De Ritis比值与缺血性心脏病患者28天死亡率显著相关。Kaplan-Meier生存分析显示,De Ritis比值最高的Q4组在28天内的死亡率显著高于Q1组(20.52%比15.08%,P＜0.0001)。单因素Cox回归分析显示,死亡风险随着De Ritis比值的增加而显著增加,Q4组的死亡风险比Q1组高42.1%(P＜0.001)。多因素Cox回归分析显示,年龄、呼吸频率、脓毒症状态等被确认为独立风险因素,模型的AUC值高达0.821。决策曲线分析表明,De Ritis比值在25%～75%阈值概率范围内具有临床净效益。限制性立方样条回归模型显示,调整混杂因素后,De Ritis比值与缺血性心脏病患者28天死亡率的非线性关系不显著(P＝0.143)。亚组分析显示,年龄、肝硬化及机械通气治疗对De Ritis比值与死亡风险关联的交互作用P值小于0.05,尤其在年轻、未接受机械通气及重度炎症患者中,其预测价值更为显著；年轻组的风险增加幅度较老年组高出24%；在未接受机械通气的患者中,De Ritis比值与死亡风险呈现强相关性；而对于接受机械通气的患者,则未发现显著相关性(P＝0.213)。 [结论]De Ritis比值与缺血性心脏病患者28天死亡率呈现显著正相关关系,因此可作为评估该类患者短期死亡风险的预警指标。

Abstract:

Aim To explore the association between aspartate transaminase (AST)/alanine transaminase (ALT) ratio (De Ritis ratio) and 28-day mortality in patients with ischemic heart disease, and to analyze its specific predictive value in populations with different comorbidities. Methods A retrospective study was conducted, and 7 513 patients with ischemic heart disease were finally included after screening for missing values, and all data were obtained from MIMIC-IV database. Extracting based on the ICD codes for ischemic heart disease, mean imputation was used for missing time-related and laboratoriy data. Variables with a missing rate exceeding 20% were excluded from the model. The De Ritis ratio was assessed using a multivariate Cox model with restricted cubic splines with four knots and adjusted for SOFA score, APACHEⅡ score, organ support, and biomarkers. LASSO regression was used for variable selection. Model performance was evaluated by the ROC-AUC (DeLong method) and calibration metrics. Prespecified subgroup analysis were performed to explore the modifying effect of comorbidities on the association between the De Ritis ratio and mortality. Results The De Ritis ratio was significantly associated with the 28-day mortality in patients with ischemic heart disease.Kaplan-Meier survival analysis showed that the Q4 group with the highest De Ritis ratio had a significantly higher 28-day mortality compared to the Q1 group (20.52% vs. 15.08%, P＜0.0001). Univariate Cox regression analysis showed that the risk of death increased significantly with the increase of De Ritis ratio, the risk of death was 42.1% higher in Q4 group than that in Q1 group (P＜0.001). Multivariate Cox regression analysis showed that age, respiratory rate, and sepsis status were identified as independent risk factors, and the AUC value of the model was 0.821. The decision curve analysis showed that the De Ritis ratio had a net clinical benefit within the threshold probability range of 25%～75%. The restrictive cubic spline regression model showed that there was no significant nonlinear relationship between the De Ritis ratio and mortality after adjusting for confounding factors (P＝0.143). Subgroup analysis showed that age, cirrhosis, and mechanical ventilation had significant interactions between the association between this ratio and the risk of death (P＜0.05), and the predictive value was more pronounced in young patients, non-mechanically ventilated patients and those with severe inflammatory. The risk increase in the young group was 24% higher than in the elderly group; in non-mechanically ventitated patients, the De Ritis ratio was strongly associated with mortality risk. In patients receiving mechanical ventilation, no significant correlation was observed (P＝0.213). Conclusion The De Ritis ratio is significantly positively correlated with the 28-day mortality of patients with ischemic heart disease, and therefore can be used as a warning indicator to evaluate the short-term mortality risk of such patients.