Analysis Of Risk Factors And Health Management Strategy For Osteoarthritis Based On Zero-inflation-count Model

Objective:The zero-inflaction-count-model was applied to screen the risk factors of osteoarthritis,and compared with the results of the classic analysis method--multivariate Logistic regression analysis,so as to evaluate the fitting effect of the zero-inflaction-count-model in the epidemiological cross-sectional survey data,and provide health management and prevention strategies for the risk factors of osteoarthritis.Methods:Through the establishment of zero-inflaction-Poisson-model and zero-inflaction negative binomial regression model,risk factors screening,simulation and analysis were carried out on 3172 epidemiological field survey data of stratified cluster sampling in rural areas of Shanxi Province,and the analysis results were compared with multivariate Logistic analysis.The goodness of fit of the zero-inflaction-count-model was compared with the Poisson regression model and the negative binomial model.Results:In the zero process,whether the waist circumference is normal,digestive system diseases or ulcers,history of hypertension,diabetes,heating methods,and history of joint damage are all statistically significant factors,that is,abnormal waist circumference,digestive system diseases or ulcers People with a history of disease,a history of hypertension,diabetes,no heating equipment,and a history of joint damage are more likely to have symptoms of joint discomfort.The counting process showed that age,gender,BMI,whether waist circumference is normal,history of hypertension,history of diabetes,history of active smoking,history of passive smoking and heating methods all had a significant impact on the number of joint discomforts in patients with osteoarthritis(P<0.05),and the relationship between them is positive,that is,the older,the female,the higher the BMI,the history of hypertension,the history of diabetes,the history of active and passive smoking,and the poor heating equipment will have more severe joint discomfort,and the number of discomforts More.Among them,the coefficient of history of hypertension is 0.4549267,which has the greatest impact on joint discomfort.The AIC value of the zero-expansion Poisson model is 1112.000 and the BIC value is 1208.508,both of which are the smallest,suggesting that this model has a better fitting effect and is the best counting model that should be used in this study.Conclusion:1.For counting data,when the zero data exceeds a certain proportion,a more appropriate zero-inflation count model should be selected according to the specific structure of the original data to include more variables and avoid the exaggeration or reduction of variable impact effects due to incomplete use of data information In order to get better interpretability,the accuracy of the overall situation estimation is better,and the goodness of fit is also higher.2.Key intervention and screening should be strengthened for high-risk groups.For middle-aged and elderly people,women,overweight,people with history of hypertension,history of diabetes,history of active and passive smoking and poor heating equipment,attention should be paid to strengthen intervention;People with abnormal waist circumference,a history of digestive system diseases or ulcers,and a history of joint injuries should be monitored regularly and with emphasis on prevention and risk assessment.3.The public health department should establish and improve a three-level health management system as soon as possible,carry out classified health management and dynamic tracking for different groups(OA patients,high-risk groups,and healthy people),and take targeted intervention measures according to different risk factors.At the same time,high-risk groups and patients should also raise their awareness of self-health management and actively adopt healthy behaviors.