Cubbin and Jackson量表在预测ICU休克病人压力性损伤中的应用研究

Assessing predictive validity of Cubbin and Jackson pressure injure risk scale in shock patients:A retrospective study

目的探讨Cubbin and Jackson量表对ICU休克病人压力性损伤(Pressure injure,PI)的预测效果及ICU自动评分的可行性。方法通过临床信息系统(Clinical information system,CIS),回顾我院重症医学科2016年全年收治的休克病人及其入住ICU 24h内的信息,进行Cubbin and Jackson量表评分,未能提取的数据通过查阅特别护理记录单完成收集。将其与Braden评分比较,评价Cubbin and Jackson量表预测休克病人PI时的灵敏度、特异度、预测值和受试者工作特性曲线及曲线下面积。结果共纳入休克病人448例。Cubbin and Jackson量表中11个数据可以直接从系统中获取,"大体皮肤情况""躯体活动度""失禁"和"卫生自理能力"在目前的CIS中不能直接提取;Cubbin and Jackson评分量表预测PI的临界值(Cut off point)为29分,灵敏度、特异度及曲线下面积与Braden量表之间的差异均无统计学意义(P<0.05)。结论 Cubbin and Jackson量表评分≤29分的休克病人是PI高危人群,预测效果与Braden评分比较差异无统计学意义;简单改进系统,可以实现CIS自动采集Cubbin and Jackson数据并进行评分。

Objective To assess predictive validity of Cubbin and Jackson pressure injure(PI)risk scale in shock patients,and explore the feasibility of grading automatically by clinical information system(CIS)in intensive care unit(ICU).Methods A retrospective analysis was performed to examine all shock patients who admitted in intensive care unit in Zhongda hospital from January to December 2016.The scores of Cubbin and Jackson scale were graded partly form the data of within 24 hours in ICU recorded in electronic health records(EMR),and the others were finished by looking up the specialist nursing record,while the scores of Braden scale were recorded from CIS.Sensitivity,specificity,positive predictive value,negative predictive value,receiver operator characteristic and area under the ROC were calculated.Results 448 cases were involved,11 clauses of Cubbin and Jackson could be extracted from CIS directly,while"General skin condition","Mobility","Incontinence"and"Hygiene"could not.At the cut-off score was 11 of Braden scale,and 29 of Cubbin and Jackson scale,there were no statistically significant differences about the sensitivity,specificity,and ROC.Conclusion The patients with shock have high PI risks when scores of Cubbin and Jackson scale 29,and the predictive validity of Cubbin and Jackson and Braden PI risk scale have statistically significant differences.The score of Cubbin and Jackson PI risk scale can be extracted directly and graded automatically by improving CIS.