To explore the influencing factors and establishment of the prediction model for medication compliance of Helicobacter pylori(Hp)-positive patients, and find the method to improve their medication compliance.

The patients who visited Weihai municipal hospital affiliated to Shandong university and were diagnosed with Helicobacter pylori infection and treated with quadruple drugs from October 2022 to July 2023 were selected as respondents. A questionnaire survey was conducted among them. The influencing factors for their medication compliance were screened out by univariate analysis and multivariate analysis and a risk prediction model was established.

Patients not anxious due to Hp infection, no family members infected with Hp, no understanding of knowledge about Hp, failure of physicians to explain in detail the possible adverse reactions and precautions in the course of medication, long-term drinking and frequent eating fried barbecue, overnight food, pickled food and other foods were the independent risk factors for the medication compliance of patients, P< 0.05. The results of this risk prediction model for the poor medication compliance of patients showed that the area under the ROC curve (AUC) was 0.886(95%CI: 0.850-0.923), the P=0.659 in the Hosmer-Lemeshow goodness of fit test indicates that the degree of differentiation and the degree of fit of this prediction model are high, when poor medication compliance was diagnosed in case of critical value >0.252, the overall prediction accuracy of the risk prediction model was 79.9%. The sensitivity, specificity, positive predictive value and negative predictive value of poor medication compliance were 86.2%, 76.7%, 65.9% and 91.4%, respectively.

Based on this risk prediction model, the poor medication compliance of patients can be identified early and the eradication rate of Hp improved. Moreover, specific intervention measures can be adopted to generate good compliance by potentially reminding patients that their treatment is being concerned.