Statistical Modeling Analysis Of The Number Of Inpatients

In recent years,with the gradual penetration of information technology into the medical field,there are many kinds of predictions about the medical field.The development speed of large hospitals in China is very fast,and the number of inpatients is also increasing year by year.Therefore,doing a good job of statistical modeling and prediction analysis of the number of inpatients can provide a more scientific reference for the next hospital management work.The hospital can prepare the storage of medical materials and arrange the work of medical staff reasonably according to the change trend of the number of inpatients in the hospital.At the same time,the hospital can also arrange the number of beds according to the prediction of various diseases in the number of inpatients and the correlation between disease groups,and reasonably adjust the distribution of beds,so as to make full use of resources.In this paper,the disease code ICD-10 is divided into the most extensive categories to classify and count the number of inpatients,sex ratio and age distribution of corresponding disease categories.The number of inpatients in a hospital in Sichuan Province from 2004 to 2018 is mainly studied from the following aspects:(1)The total number of inpatients,the number of men and women,the sex ratio of inpatients,the inpatient situation of different age groups,the average age of inpatients and the change trend of the average duration of hospitalization with time were analyzed.(2)According to the classification and statistics of inpatients' diseases,the changes of disease types with time,the proportion of inpatients of different ages,and the distribution of diseases of inpatients of different genders were observed.(3)According to the disease categories of the inpatients,the number of inpatients of all kinds of diseases is classified and counted by week.The structural vector autoregressive model(SVAR model)is used to select the characteristic diseases of each disease group and group the diseases.Because the interaction among the disease groups is difficult to see in the same period,we choose the third-order optimal lag to group.Then,the multi response linear regression model based on group Lasso is used to model the prediction analysis,so as to observe the correlation between various diseases.