Analysis of risk factors for liver damage in patients with hyperthyroidism and construction and validation of predictive models
Objective:To explore and analyze the risk factors for liver damage in patients with hyperthyroidism(HT),and to establish an effective column chart model for predicting liver damage in HT patients.Methods:A total of 263 HT patients admitted to Shijiazhuang Third Hospital from January 2022 to December 2023 were regarded as the study subjects,they were separated into a modeling group(184 cases)and a validation group(79 cases).The modeling group was separated into a liver damage group(78 cases)and a non liver damage group(106 cases)based on whether the patient had liver damage.Logistic regression analysis was applied to screen for risk factors for liver damage in HT patients;R software and rms package were applied to construct a column chart risk prediction model for liver damage in HT patients;ROC curve and calibration graph were applied to verify the discrimination and consistency of the model.Results:Compared with the non liver damage group,the age,proportions of course of disease ≥ 1 year,married,high school education or below,history of alcohol consumption,biased disease cogni-tion,and adverse drug reactions were obviously increased in liver damage group(P<0.05).The age(95%CI 1.220-1.521),disease duration(95%CI 1.222-8.320),marital status(95%CI 1.067-7.011),educational level(95%CI 1.389-9.664),alcohol consumption history(95%CI 1.013-6.849),disease awareness(95%CI 1.278-10.196),and adverse drug reactions(95%CI 1.234-10.019)of patients with HT were independent risk factors for liver damage(P<0.05).A column chart model was established to predict the risk of liver damage in HT patients based on the above risk factors,the area under the ROC curve(AUC)was 0.936(95%CI 0.900-0.972),with good discrimination and good fitting between the correction curve and the ideal curve.Conclusion:The age,course of disease,marital status,educational level,drinking history,disease cognition,and adverse drug reactions of HT patients are risk factors for liver damage.Based on the above factors,a column chart model can be constructed to accurately predict the risk of liver damage in HT patients,and targeted treatment plans can be devel-oped for patients,which is conducive to reducing the incidence of liver damage.
hyperthyroidismliver damagerisk factorsprediction model