Abstract: Objectiv eTo explore the influencing factors of drug compliance in hyperlipidemia patients treated with statins, and to establish a nomogram model to predict the risk of poor compliance. Methods One hundred and sixty patients with hyperlipidemia treated in our hospital were selected between January 2018 and September 2019. These patients were divided into the good compliance group (84 cases) and poor compliance group (76 cases) according to the results of a questionnaire survey. Logistic regression analysis was used to screen independent risk factors. A linear prediction model was established by including the selected independent risk factors before the predictability and accuracy of the model were verified. Results According to logistic regression analysis of the general clinical data and related complications of the two groups, age (OR= 6.517, 95% CI: 1.986~21.39), lifestyles (OR=0.130, 95% CI: 0.039~0.429), medical expenses (OR=0.162, 95% CI: 0.058~0.452), doctor-patient relationships (OR=15.017, 95% CI: 4.991~45.181), follow-up (OR=7.889, 95% CI: 2.668~23.323), dietary habits (OR=5.187, 95% CI: 1.839~14.634) and adverse reactions of statins (OR=0.316, 95% CI: 0.110~0.904) were independent risk factors for poor drug compliance of patients with hyperlipidemia. The predictability of this model was basically consistent with the measured value. In addition, this model was verified by the Bootstrap internal verification method. The C-index was as high as 0.934（95% CI: 0.896-0.972）, indicating the precision of the model. Conclusion It is of great clinical significance to evaluate the incidence of poor drug compliance in hyperlipidemia patients by taking into consideration age, lifestyles, medical expenses, doctor-patient relationships, follow-up, eating habits, adverse reactions of statins and other diseases.

Key words: hyperlipidemia, statins, medication compliance, nomogram, risk model