Application and verification of moving average seasonal index method in predicting emergency depart-ment visits
10.3969/j.issn.1671-332X.2025.09.019
- VernacularTitle:移动平均季节指数法在门急诊量预测中的应用与验证
- Author:
Yi ZHANG
1
;
Guilian WU
1
;
Yaoke MAO
1
;
Yuejiao TUO
1
Author Information
1. 南部战区空军医院 广东 广州 510602
- Publication Type:Journal Article
- Keywords:
Time series;
Moving average method;
Seasonal index;
Outpatient and emergency department visit forecasting
- From:
Modern Hospital
2025;25(9):1386-1390
- CountryChina
- Language:Chinese
-
Abstract:
Objective This study aims to apply the moving average method of time series analysis to forecast outpatient and emergency department visits at our hospital for 2024.Additionally,we will validate the trend prediction model against actual visit data from January to April 2024,assessing the accuracy of the time series fitting.The insights generated will serve as a sci-entific foundation for the hospital to allocate resources effectively,formulate work plans,and meet annual objectives.Methods We collected data on outpatient and emergency visits from 2020 to 2023(n=16 periods)and employed the four moving average method for time series decomposition.We calculated the adjusted seasonal index values and developed a linear trend prediction model(Y=7 3847+568.08t)that incorporates seasonal factors.We then computed the predicted monthly outpatient emergency visits for 2024 and compared these forecasts with actual values from the first four months of 2024 to test the model's reliability.Results The predicted visits for the first four months of 2024 were 84 396,80 633,88 244,and 84 158 respectively.The rela-tive errors compared to actual figures ranged from 1.13％to 8.81％,with an average relative error of 4.22％.The seasonal indi-ces revealed that the third quarter represents the peak period(104.26％),while the second quarter is the low point(95.91％).Conclusion The moving average seasonal index method effectively captures the seasonal variations in outpatient and emergency department visits,offering high prediction accuracy.This methodology can assist hospitals in dynamically adjusting their schedu-ling and optimizing resource allocation.
