摘要
目的分析重症患儿转诊死亡的危险因素,建立重症患儿转诊死亡风险模型并评价模型的预测效果。方法选取医院2022年1-12月收治的从外院转诊至本院的重症患儿204例为研究对象,根据转入本院后1周的转归情况分为死亡组(n=23)和生存组(n=181)。多因素Logistic分析影响死亡的因素,依据分析结果构建风险预测模型,通过ROC曲线及灵敏度、特异度、Kappa值等指标评价预测模型的预测效果。结果多因素Logistic分析结果显示,血糖水平、营养不良、严重脱水和专业转诊团队是影响重症患儿转诊死亡的主要因素。根据回归分析结果构建重症患儿转诊死亡风险模型:Y=2.022+1.497低血糖+1.587营养不良+10.392严重脱水-2.366专业转诊团队,其ROC曲线下面积为0.899,95%CI为0.806~0.975,说明本模型具有较好的风险预测能力。模型预测的灵敏度和特异度均较高,分别为86.96%,97.24%,表明模型预测结果具有较高的真实性;并且预测结果与实际结果具有非常高的一致性(Kappa=0.79),表明预测模型具有临床应用价值。结论本研究在调查重症患儿转诊死亡的危险因素基础上,建立重症患儿转诊死亡风险模型,该模型预测结果与实际结果具有较高的真实性和一致性,可为临床早期预测转诊死亡风险提供参考。
Objective To analyse the risk factors of referral death of critically ill children,construct a referral death risk model for critically ill children and evaluate the predictive effect of the model.Methods 204 cases of critically ill children admitted to the hospital from January to December 2022 who were referred from outside hospitals to our hospital were selected for the study,they were divided into the death group(n=23 cases)and the living group(n=181 cases)according to the regression status 1 week after transferring to the hospital.Multifactorial Logistic analysis was performed to analyse the factors affecting death,and a risk prediction model was constructed based on the results of the analysis.The predictive effect of the prediction model was evaluated by the ROC curve and the indexes of sensitivity,specificity and Kappa value.Results The results of multifactorial logistic analysis showed that blood glucose level,malnutrition,severe dehydration and professional referral team were the main factors affecting the referral of critically ill children to death.The risk model for referral death of critically ill children was constructed based on the results of regression analysis:Y=2.022+1.497 hypoglycaemia+1.587 malnutrition+10.392 severe dehydration-2.366 professional referral team,with an area under the ROC curve of 0.899,and a 95%CI of 0.806~0.975,which indicates that this model had a good risk prediction ability.The sensitivity and specificity of the model prediction were high,which were 86.96% and 97.24%,respectively,indicating that the model prediction results had a high degree of authenticity;and the prediction results had a very high consistency with the actual results(Kappa=0.79),indicating that the prediction model had clinical application value.Conclusion This study based on the investigation of the risk factors of referral death of critically ill children,a referral death risk model of critically ill children was established,and the predicted results of the model have high authenticity and consistency with the actual results,which could provide a reference for the early clinical prediction of the risk of referral death.
作者
陈亮
李佳蓉
肖兰兰
袁远宏
CHEN Liang;LI Jiarong;XIAO Lanlan;YUAN Yuanhong(Hunan Children's Hospital,Changsha,410007,China)
出处
《护理实践与研究》
2023年第21期3196-3201,共6页
Nursing Practice and Research
基金
湖南省科技厅临床医疗技术创新引导项目(编号:2021SK50501)。
关键词
重症疾病
转诊死亡风险
患儿
预测模型
影响因素
Critical illness
Referral mortality risk
Ill children
Prediction model
Influencing factors
作者简介
通信作者:袁远宏,硕士,急诊科主任。
