基于大数据平台的高血糖危象患者的临床特征和预后影响因素分析

Clinical characteristics and prognostic influencing factor of hyperglycemic crisis patients based on the big data platform

目的分析高血糖危象(HC)患者的临床特征及预后影响因素。方法基于某医科大学医学数据研究院大数据平台,回顾性收集重庆市6家综合医院2015-2020年急诊就诊的HC患者临床资料,将住院期间全因死亡作为观察终点,分为存活组和死亡组。采用描述性分析方法分析两组HC患者入院时的临床特征、实验室检查等,单因素和二元logistics回归分析HC患者预后的影响因素。结果共纳入研究对象1668例,其中死亡121例(7.3%)。两组患者在入院时生命体征,高血压、冠心病、老年痴呆、慢性阻塞性肺疾病、糖尿病慢性并发症、肺部感染比例,住院时间、住院费用等比较差异有统计学意义(P<0.05)。存活组入院时症状以烦渴多饮、多尿为主,死亡组中意识障碍等非典型症状占比较大。二元logistics回归分析显示,年龄≥65岁(OR=2.65,95%CI:1.39~5.03,P=0.003)、CCI≥4分(OR=2.30,95%CI:1.24~4.26,P=0.008)、入院时休克(OR=6.01,95%CI:2.46~14.67,P<0.001)、意识障碍(OR=5.75,95%CI:2.95~11.22,P<0.001)、pH≤7(OR=3.24,95%CI:1.53~6.87,P=0.002)、HC史(OR=5.52,95%CI:2.85~10.67,P<0.001)、低白蛋白血症(OR=2.26,95%CI:1.25~4.09,P=0.007)、并发急性肾衰竭(OR=9.69,95%CI:3.75~25.06,P<0.001)、低血糖(OR=7.95,95%CI:3.63~17.41,P<0.001)是HC患者死亡的独立危险因素。结论入院症状表现为意识障碍、胸痛的HC患者应多加关注,对与预后相关的因素进行风险分层,识别高危患者。

Objective To analyze the clinical characteristics and prognostic influencing factors of patients with hyperglycemic crisis(HC).Methods Clinical data of HC patients who admitted into emergency department in six hospitals from January 2015 to December 2020 based on the data science academy of a medical university was retrospectively collected.All-cause death during hospitalization was taken as the observation endpoint,and the patients were divided into the survival group and the death group.Descriptive analysis was used to analyze the clinical features and laboratory examination of the two groups of HC patients at admission.Univariate analysis and binary logistics regression were used to analyze the factors influencing the prognosis of HC patients.Results A total of 1668 subjects were included.Among the 1668 patients,a total of 121 patients died(7.3%).There were statistically significant differences in vital signs,the proportion of hypertension,coronary hea rt disease,senile dementia,chronic obstructive pulmonary disease,chronic complications of diabetes,pulmonary infection,hospitalization time and hospitalization expenses between the two groups(P<0.05).In the survival group,polydipsia,polydipsia and polyuria were the main symptoms at admission,while atypical symptoms such as disorder of consciousness were more common in the death group.Binary logistics regression analysis showed that age≥65(OR=2.65,95%CI:1.39-5.03,P=0.003),CCI≥4(OR=2.30,95%CI:1.24-4.26,P=0.008),shock(OR=6.01,95%CI:2.46-14.67,P<0.001),disorder of consciousness(OR=5.75,95%CI:2.95-11.22,P<0.001),pH≤7(OR=3.24,95%CI:1.53-6.87,P=0.002),history of hyperglycemic crisis(OR=5.52,95%CI:2.85-10.67,P<0.001),hypoglycemia(OR=2.26,95%CI:1.25-4.09,P=0.007),acute renal failure(OR=9.69,95%CI:3.75-25.06,P<0.001),hypoglycemia(OR=7.95,95%CI:3.63-17.41,P<0.001)were independent influencing factors of death of patients with HC.Conclusion HC patients with admission symptoms such as disturbance of consciousness and chest pain should be paid more attention to the risk stratification of factors related to prognosis and identify high-risk patients.