Construction and verification of prediction model of subsyndromal delirium in adults following orthopedic surgery
Objective:To explore the risk factors of subsyndromal delirium(SSD)after orthopedic surgery in adults and construct a risk model for verification.Methods:Using a convenient sampling method,selected 333 adult patients of Luoyang Orthopedics Hospital(Henan Orthopedic Hospital),from January 1,2022,to June 30,collected clinical data and various indicators,discussed the risk factors of SSD in adults after orthopedic surgery through logistic regression analysis by univariate analysis and dichotomy surgery,constructed the risk model function and nomogram prediction model,The developed model was evaluated by both discrimination measured by the C-index and calibration evaluated by the Hosmer-Lemeshow test and calibration plot,which compares the prediction probability and the actual observation Probability.A nomogram was established based on the risk predictive model by using R for Windows with The rms package R.Results:301 cases were included in the study,including 274(non-SSD)and 27(SSD),and the incidence of SSD was 8.97%.Univariate analysis found that the influencing factors of adult orthopedic postoperative SSD are blood sodium and hemoglobin on admission,FRAIL scale score,daily living ability at admission,APACHE Ⅱscore,intraoperative use of sedative,analgesics,intraoperative blood transfusion,hemoglobin in ICU,the two groups are significant difference(P<0.05).Dichotomous logistic regression analysis found that blood sodium at admission(OR = 0.922),FRAIL scale score at admission(OR = 1.418),and daily living ability score at admission(OR =0.973)were independent risk factors for subpatients(P<0.05).Y = 11.484-0.081×cblood sodium at admission + 0.349×FRAIL scale score at admission-0.027×daily living ability score at admission.An individualized nomogram prediction model was established with the above factors,and the C index is0.788[95%CI(0.738,0.833)],The model showed a good false acceptance rate is8.97%,and the sensitivity and specificity of the model were 77.8%and 69.71%,and the correction curve was fit well(Hosmer-Lemeshow test,P = 0.705).Conclusion:The prediction model of the column graph established in this study is helpful for clinical staff to carry out early assessment and screening of these factors,intuitively and easily identify the population with subsyndromal delirium,and provide reference for early screening and intervention,as well as for optimizing the management of postoperative delirium in orthopedic patients.
subsyndromal delirium(SSD)adultorthopedic patients after surgeryrisk factorsforecasting model
NETL
NSTL
万方数据
