Research On Prediction Of Chronic Disease Risk In Medical Service Supply Chain

With the rapid development of the times,how to live a healthy life has become the focus of people’s attention,and medical service resources are the most basic guarantee for maintaining health.Due to the characteristics of multi-sector ownership and multi-subject jurisdiction,the medical service supply chain is currently in a state of mismatch between supply and demand and uneven distribution,with obvious regional disparities.In some areas,the distribution of medical institutions is too concentrated,and the service capacity is seriously surplus.In other areas,it is difficult and expensive to see doctors,which results in scarcity and waste of medical resources,and the overall efficiency of medical services is low.Since the demand for medical services is highly uncertain and the flexibility of the medical service supply chain is relatively weak,it is particularly important to minimize the uncertainty of demand.Based on this,the purpose of this research is to proceed from the demand flow in the medical supply chain,and use the information and data generated by each subject,especially the patient medical record data in the medical information system,to predict health risks.Through risk prediction,from previous treatment to early prevention,reduce the potential demand in the medical service supply chain,reduce the conversion rate from potential demand to real demand,promote the rational allocation of medical resources,and thus improve the medical service supply chain.Performance.In this study,by collecting more than 7,800 inpatient departments of a top three hospital in Guangzhou,after data preprocessing,logistic regression modeling and fitting,finally selected 10 statistically significant risk factors and constructed a multi-layer perceptron neural network based on this Model,build the model through Python and test the effect.Results Identified body mass index,total cholesterol,stroke,leukocytes,age,triglyceride as six important risk variables for detecting early onset of early diabetes.In the modeling process,by adjusting the number of hidden layer neurons,it is found that when the number of hidden layer neurons is 5,the prediction accuracy of the model is better,and the prediction accuracy of the model is 81.5%.The research in this paper is based on the prediction of chronic disease risk in the medical service supply chain.In medical services,risk prediction can effectively identify key risk factors,predict patient morbidity risk in advance and intervene to reduce potential demand conversion rate,reduce waste of medical resources,and optimize Medical service resource allocation improves the overall performance of the medical service supply chain.This study also has certain limitations.This article mainly controls the demand side from the perspective of chronic disease risk and reduces the demand conversion rate of the medical service supply chain,without further planning and management of demand.At the same time,in the risk prediction process,due to the many risk factors of chronic diseases,more risk factors need to be included in the later research and a variety of machine learning algorithms are selected for fusion modeling,and the better fitting model is selected through comparative analysis to look forward to better effect.