Research On The Influence Factors And Forecast Analysis Of Hospital Stay Time Of Renal Carcinoma Surgery Patients

The patient's hospitalization period is an important comprehensive evaluation index that is key to the overall service level of hospital and the utilization ratio of equipment resources.Reducing the patient's hospitalization period can not only improve the turnover of hospital beds and key equipment resources to serve more patients,but also will enhance the social and economic benefits of a hospital.The prediction of hospitalization period is of great significance for the hospital to do a good job in the hospital admission plan and the hospital bed.In this paper,based on the clinical theory,combined with the results of the characteristics of patients with radical nephrectomy and the results of expert interviews,the factors affecting the hospitalization period are established,which are reflected by the 4 factors.These 4 factors were: patient factors,pathological factors,operative factors and concomitant factors.At the same time,the corresponding observed variables are found to be the observed indicators.Under the guidance of influencing factors of system,using the research method of structural equation model,the exploratory factor analysis is first carried out to verify the explanatory power of observation index,and then the first-order factor analysis model is established to analyze the correlation between the four factors,and based on the second-order factor analysis,the model is built to reflect radical renal resection surgery patients hospitalized period.Selecting 243 patients with pathological data in three years from 2013 to 2015 as samples to fit,correct and check the model,significantly better structural equation model is eventually determined including four latent variables and 14 variables.The advantages of structural equation model optimization and simplified structure make it more concise and objective to reflect the influencing factors of hospitalization period.On the basis of the study of factors affecting hospitalization cycle,this paper makes the forecast analysis on hospitalization period.A prediction model of BP neural network with hospital cycle related factors is set,three kinds of improved BP algorithm(adaptive BP algorithm,the flexible gradient descent method and the method of L-M)are used to forecast the radical renal resection patients hospitalized period.Through the comparison and analysis,a L-M method which is the fastest and of the highest prediction accuracy is obtained.