Construction of acquired weakness risk prediction model in intensive care unit
Objective To construct a risk prediction model for intensive care unit-acquired weakness(ICU-AW)to guide clinical prevention and treatment strategies.Methods The correlation between gender,age,age-adjusted Charlson Comorbidity Index(aCCI),injecting albumin,heart failure,rehabilitation treatment,midazolam dosage,norepinephrine dosage and mechanical ventilation duration in 1 063 patients admitted to the intensive care unit was analyzed.Independent risk factors were identified to establish the prediction model,and the predictive ability of the model was analyzed.Results Among 1 063 patients,370 developed ICU-AW.Logistic regression analysis identified advanced age,higher aCCI,prolonged mechanical ventilation duration,increased midazolam and norepinephrine dosages,and heart failure as independent risk factors for ICU-AW,while rehabilitation treatment and injecting albumin were identified as independent protective factors.The regression equation of the prediction model was:Logit(P)= 0.017×age + 0.008×mechanical ventilation duration + 0.006×norepinephrine dosage-0.832×rehabilitation treatment-0.648×injecting albumin + 1.224×aCCI + 0.017×midazolam dosage + 1.834×heart failure-6.806.The area under the curve(AUC)of the model was 0.908(0.890-0.925),with the sensitivity of 82.20%and specificity of 82.40%.Conclusion The model constructed using these variables demonstrates good predictive efficiency and can provide new insights for clinical prevention and treatment of ICU-AW.
intensive care unit-acquired weaknessclinical prevention and treatmentindependent risk factorspredictive model
NETL
NSTL
万方数据
