利用ARIMA模型预测超声科工作量变化的应用研究

Analysis of Workload Changes in the Ultrasonic Department Using ARIMA Model

目的分析某县市级医院超声科工作量变化规律,为制定科学策略,优化资源配置提供依据。方法统计该院超声科近5年工作量,导入SPSS24.0软件进行时间序列建模研究分析。结果超声科工作量呈逐年上升趋势,月工作量存在规律性波动,单日工作量周一至周日呈逐日递减趋势,且变化存在季节波动性。以ARIMA(1,1,0)(0,1,0)12模型预测数据和月工作量实际数据非常吻合,平均绝对误差百分比2.94%,最大绝对误差10.92%。ARIMA(1,0,0)(0,1,1)7模型预测数据与日工作量实际数据平均绝对误差百分比5.87%,最大绝对误差27.85%。结论采用ARIMA模型预测医院超声科工作量操作简单,模型拟合和预测效果理想,可以很好地预测超声科工作量的变化,为医疗服务资源的提前合理安排提供依据。

Objective To analyze the regulation of the workload change in the Department of Ultrasound, formulate a scientific strategy and optimize the allocation of resources. Methods We calculated the workload in the department of ultrasound in the past five years and then we used SPSS24.0 to conduct a time series analysis. Results The workload in the ultrasound department showed an increasing trend, while average monthly workload fluctuated. The average daily workload was reduced from Monday to Sunday, and showed a seasonal fluctuation. The data predicted by ARIMA （1,1,0） （0,1,0） 12 models fitted well with the actual monthly workload, with an average absolute error of 2.94% and a maximum absolute error of 10.92%. Finally, we fitted the daily workload with data predicted by ARIMA （1,0,0） （0,1,1） 7, and we found the average absolute error and the maximum absolute error of the data were 5.87% and 27.85%, respectively. Conclusion Using AR1MA model is an effective way to predict the workload in the ultrasound department because it is simple to operate and fit the actual data well. ARIMA model could enable us to predict the change of the workload and optimize the allocation of medical resources.