期刊文献+

基于Logistic回归的住院精神障碍患者医院感染危险因素预测

Prediction of nosocomial infection risk factors based on Logistic regression in hospitalized mental patients
下载PDF
导出
摘要 目的 运用Logistic回归模型和受试者工作特征(ROC)曲线探讨住院精神障碍患者发生医院感染危险因素并分析预测价值,为精神卫生专科医疗机构临床实践和感染控制及预防提供参考依据。方法 回顾性调查收集2022年1月~12月间某三级甲等精神卫生专科医院9 989例住院患者病历资料,采用Logistic回归分析医院感染危险因素并进行预测,应用ROC曲线进行预测价值评价。结果 9 989例患者中发生医院感染201例,感染发病率2.01%,医院感染203例次,感染例次发病率2.03%,感染部位分布主要以上、下呼吸道为主,共175例次,构成比占86.21%。单因素分析结果显示,性别、年龄、住院天数、精神障碍类型、住院季节、精神类药物、患者管理模式和保护性约束8个因素与住院精神障碍患者医院感染的发生相关,组间相关数据比较差异有统计学意义(P<0.05或0.01)。多因素非条件Logistic回归分析结果显示,患者封闭管理模式、精神类药物≥2种、住院天数>30 d、男性、年龄≥60岁和重型精神障碍患者共6个因素是住院精神障碍患者医院感染的主要独立危险因素,联合预测因子的AUC值为0.735,95%CI为0.701~0.769,其预测价值均优于各单一变量的预测价值。结论 住院精神障碍患者易发生医院感染,应重点关注患者封闭管理模式、精神类药物≥2种、住院天数>30 d、男性、年龄≥60岁和重型精神障碍患者,开展定向目标监测,有针对性地制定相关防控措施,以减少住院精神障碍患者医院感染的发生。 Objective Logistic regression model and Receiver Operating Characteristic(ROC)curve were used to explore the risk factors of nosocomial infection and analyze predictive values in hospitalized mental patients so as to provide reference for clinical practice as well as control and prevention of infection at mental health specialist medical institutions.Methods Medical record data of 9989 inpatients from Jan.to Dec.2022 in a tertiary class A mental hospital were retrospectively investigated,the risk factors of nosocomial infection analyzed using Logistic regression analysis and predicted,and predictive values evaluated using the ROC curve.Results Of 9989 patients 201 developed nosocomial infections,the incidences of infections was 2.01%,nosocomial infections were 203 cases/times,the incidences of infection cases/times 2.03%.The infection locations were mainly distributed in upper and lower respiratory tract,with 175 cases/times in total,and constituent ratio accounted for 86.21%.Univariate analysis showed that factors related to the occurrences of nosocomial infections included gender,age,length of hospital stay,type of mental disorder,hospitalization season,antipsychotic drugs,patients'management mode,and protective constraints,differences in intergroup relevant data comparison were statistically significant(P<0.05 or 0.01).Multivariate unconditional logistic regression analysis showed that such factors were main independent risk factors for nosocomial infection in hospitalized mental patients as closed management mode,medication of≥2 antipsychotics,lengths of stay>30 days,male,ages≥60 years,and severe mental disorders,the AUC value of the combined predictive factors was 0.735,with a 95%CI of 0.701-0.769,and which was superior to single variable.Conclusion Hospitalized mental patients are prone to nosocomial infections,attention should be paid to the closed management mode,medication of≥2 antipsychotics,lengths of stay>30 days,male,ages≥60 years,and severe mental disorders,directional target monitoring carr
作者 张红云 郑霁瑜 董玮 张红梅 Zhang Hongyun;Zheng Jiyu;Dong Wei;Zhang Hongmei(The Second Affiliated Hospital of Xinxiang Medical University,Xinxiang 453002,Henan,China;不详)
出处 《临床心身疾病杂志》 CAS 2023年第4期92-97,共6页 Journal of Clinical Psychosomatic Diseases
基金 河南省医学科技攻关计划软科学项目(编号RKX202202038) 上海市杨浦区精神卫生中心项目(编号20MC2020005)。
关键词 精神障碍 住院患者 医院感染 危险因素 LOGISTIC回归 受试者工作特征曲线 预测 mental disorders inpatient nosocomial infection risk factor Logistic regression ROC curve prediction
  • 相关文献

参考文献10

二级参考文献97

共引文献5921

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部