Search

BackgroundAlcohol use disorder （AUD） is a common chronic and relapsing psychiatric disorders. Identifying severe AUD early and intervening promptly is crucial to prevent irreversible harm. Currently， the assessment of AUD severity primarily relies on psychiatric examination by clinicians， and there is limited research on the factors influencing AUD severity and the development of prediction models.ObjectiveTo analyze the factors influencing AUD severity， and construct a risk prediction model to aid in the assessment of disease progression in AUD patients.MethodsA retrospective analysis was conducted on 1 358 first-time hospitalized patients admitted to Nanning Fifth People's Hospital from January 1， 2017 to December 31， 2022. These patients met the Diagnostic and Statistical Manual of Mental Disorders， fifth edition （DSM-5） criteria for AUD. Basic patient data was collected， and the patients were divided into two groups based on disease severity： mild-moderate group （n=330） and severe group （n=1 028）. The patients were randomly divided into training and test sets in a 7∶3 ratio. A Logistic regression model was constructed in the training set， and the predictive ability of the model for disease severity was evaluated using the receiver operating characteristic （ROC） curve in the test set.ResultsCompared with the mild-moderate group， the severe group had a higher proportion of patients living in urban areas （χ2=7.804）， were farmers （χ2=17.991）， had a higher frequency of alcohol consumption （more than 1 to 2 drinks/day） （χ2=35.267）， had a higher age at first drinking （t=-3.858）， had a greater number of comorbid somatic disorders （Z=-22.782）， and had higher proportions of γ-Glutamyl transpeptidase （χ2=259.940） and total bilirubin abnormalities （χ2=148.552） （P