基于线性回归模型的医疗设备故障预测

Medical equipment failure prediction based on linear regression model

目的：通过建立合适的数学模型,预测某三甲医院医疗设备故障发生趋势,为医疗设备管理保障部门决策提供理论依据。方法：采用最小二乘法建立线性模型,并预测近2 a设备发生故障台次。结果：某医院2005—2012年医疗设备发生故障数量与时间呈线性关系,直线回归方程为Y赞=1 962.335＋699.01X。预测的2013年医疗设备发生故障数量为8 253台次,95%概率的波动区间为6 576~9 930台次,实际2013年医疗设备发生故障数量为7 873台次;预测的2014年医疗设备发生故障数量为8 952台次,95%概率的波动区间为7 156~10 748台次。结论：通过数学模型分析预测医疗设备故障发生台次完全可行,预测结果可用于指导医疗设备管理保障部门做好医疗设备维修计划,为医院设备管理工作提供了科学的理论依据。

Objective To predict medical equipment failure trend of some class A tertiary hospital by building appropriate mathematical models. Methods A linear model was established by using the method of least squares, and medical equipment failure times in the past two years were predicted. Results The hospital had the number of medical equipment failure from2005 to 2012 occur linearly with time, and the linear regression equation was =1 962.335 ＋699.01 X. It＇s predicted that there were 8 253 times of medical equipment failures occurred in 2013, and the range of fluctuation 95% probability was between 6 576 and 9 930 times, with the real number being 7 873. It＇s predicted there were 8 952 times of failures in 2014,and the range of fluctuation 95% probability was between 7 156 and 10 748 times. Conclusion It is completely feasible to predict medical equipment failure times through mathematical models. The prediction results can be used to guide medical equipment management security department to prepare medical equipment repair plan and provide a scientific theoretical basis for hospital equipment management.