趋势季节模型在住院人次预测中的应用  

Application of Trend-season Model in the Prediction of Hospitalized Patients Number

作　　者：饶懿[1] 

机构地区：[1]广东省东莞市长安医院信息科,东莞市523843

出　　处：《中国病案》2016年第1期77-78,92,共3页Chinese Medical Record

摘　　要：目的通过对某医院住院人次进行短期预测,为医院管理提供科学依据。方法采用趋势季节模型对2009年-2013年某医院住院人次进行建模,计算趋势方程,通过季节指数校正系数,得出各季度平均季节指数,对2014年各季度住院人次进行预测。结果以各年季平均住院人次建立直线回归方程y_c=4753.625+157.125t。季节指数的校正系数为1.0011,各季度平均季节指数分别为88.50%、105.19%、106.28%、100.03%。住院人次预测值为22 785,实际住院22 228人次,各季度住院依次为4590人次、5695人次、6109人次、5834人次,预测值与实际值相对误差分别为9.83%、5.22%、0.9%、2.33%,均小于10.00%。结论本模型预测效果较好,所预测数据能作为医院管理的参考依据。Objective Through a short-term forecasting of hospitalized patients number of a hospital to provide scientific reference for the management of the hospital. Methods To establish mode with the application of trend-season model on the hospitalized patients number of a hospital from 2009 to 2013, count trend equation, to get the average seasonal index through the seasonal index correction factor and make a prediction on the number of each quarter in 2014.Results To establish a linear regression equation y_c =4753.625＋157.125 t based on the average number of each quarter in the year. Correction factor of seasonal index was 1.0011, each quarter the average seasonal index, were 106.28%, 88.5%, and 100.03%, respectively. Hospital visitors forecast was 22785,actual inpatient was 22228, quarter of hospitalization followed by 4590, 5695, 6109, 5834, relative error of the predicted value and the actual value were 9.83%,5.22%,0.9%,2.33%,are less than 10%.Conclusion The prediction effect by the model was relatively good, the data predicted could be used as reference for the management of the hospital.

关 键 词：趋势季节模型 住院人次 预测 

分 类 号：R197.323[医药卫生—卫生事业管理]