Risk factors analysis and nomogram prediction model construction of nosocomial infection in stroke patients
Objective:To investigate the risk factors of hospital infection in stroke patients and construct a risk prediction model of nosocomial infection.Methods:A total of 300 stroke patients with nosocomial infection during hospitalization in the Department of Neurology and Neurosurgery of Affiliated Hospital of Nantong University from January 2020 to December 2023 were also selected(infection group),and another 300 stroke patients without nosocomial infection during the same period were selected(control group).The distribution of nosocomial infection and pathogenic bacteria were analyzed,and clinical characteristics of the two groups were compared.Multivariate Logistic regression was used to analyze the risk factors of nosocomial infection in stroke patients,and a nomogram model for predicting risk was established by incorporating R language,and the predictive effect of the model was evaluated.Results:The predominant nosocomial infection site of stroke patients was the respiratory system(62.67%,188/300),the predominant pathogenic bacteria of nosocomial infection were Gram-negative bacteria(59.90%,121/202),and the multidrug-resistant bacteria infection accouned for 34.16%(69/202).Compared with the control group,the infection group had higher proportion of cerebral hemorrhage,diabetes,hypertension,coronary heart disease,disturbance of consciousness,central venous catheters,ventilators,urinary catheters,and preventive use of antibiotics,body mass index(BMI)≥24 kg/m2 and hospitalization time≥14 d(P<0.05).Multivariate Logistic regression analysis showed that stroke type,hypertension,diabetes,BMI≥24 kg/m2,disturbance of consciousness,use of ventilator,indwelling urinary catheter,preventive use of antibiotics,and hospitalization time≥14 d were independent risk factors for nosocomial infection in stroke patients(P<0.05).A nomogram prediction model was constructed based on the regression results,and the area under ROC curve was 0.983(95%CI:0.975-0.991).The sensitivity and specificity was 0.940 and 0.937,respetictively.The Hosmer-Lemeshow test(χ2=5.454,P=0.708)indicated that the model had a good fit and predictive performance.Conclusion:The risk factors of nosocomial infection in stroke patients include stroke type,hypertension and diabetes,etc.The prediction model based on the risk factors could accurately predict the risk of nosocomial infection in stroke patients.
万方数据
维普
