Risk factors analysis and prediction model construction of metabolic syndrome in erythrodermic psoriasis based on lasso regression
Objective To investigate the risk factors associated with metabolic syndrome(MS)in patients with erythrodermic psoriasis and develop a predictive model for clinical application.Methods The medical records of inpatients admitted to the dermatology ward of Beijing Hospital of Traditional Chinese Medicine from March 2014 to June 2022 were selected,including demographic data,laboratory data and outcome indicators.Logistic regression and Lasso regression were used to screen out the optimal potential predictors based on the occurrence of metabolic syndrome,and a risk prediction nomogram was constructed.The model discrimination was assessed using the receiver operating characteristic(ROC)curve,while its consistency and utility were evaluated through the Hosmer Lemeshow test and calibration curve.Results Hypertension,triglycerides,diabetes,high-density lipoprotein,low-density lipoprotein,and non-alcoholic fatty liver were used as predictors for MS in patients with erythrodermic psoriasis to establish a nomogram.The AUC value was 0.960 in the training set and 0.942 in the validation set,indicating that the model had high predictive ability.Hosmer Lemeshow test showed thatx2=0.82,P=0.9997(P＞0.1),and the AUC value of calibration curve was 0.942(AUC＞0.9),indicating that the model has excellent discrimination,accuracy and credibility.Conclusion The clinical prediction model established in this study exhibits strong discrimination,fitting and clinical effectiveness,which can be applied in clinical practice and provide a scientific basis for the prevention and treatment of MS in patients with erythrodermic psoriasis.
Psoriasis,erythrodermicMetabolic syndromeLasso regressionClinical prediction model
NETL
NSTL
万方数据
