灰色GM（1,1）模型在门急诊、出院人次预测中的应用

Application of GM（1,1） model in predicting the number of outpatients and inpatients

目的 探讨灰色预测方法在医院管理中的应用,对医院门急诊量、出院人次进行预测,有助于管理者掌握医院门急诊量、出院人次的变动趋势及规律,为医院在管理、决策中的合理应用提供量化的理论依据方法 选用某医院2000-2010年门急诊人次、出院人次统计年报数据,基于灰色系统理论的建模方法建立GM（1,1）预测模型,采用Excel表进行计算,预测2011-2013年医院门急诊量、出院人次结果 根据后验差比值C和小误差概率P值的计算及预测精度等级判定,所建预测模型拟合精度高,结果较理想,可用于外推预测.预测评价结果显示：2011-2013年门急诊人次预测值为256 267、27 666、300 852人次,评价等级为合格之上（C=0.33＜0.35,P=1＞0.9）;出院人次预测值为17 739、19 044、20 445人次,评价等级为优秀（C=0.29＜0.35,P=1）.医院门急诊人次、出院人次的年增长率分别为为8.35%、7.36%.通过预测医院门急诊量、出院人次有缓慢上升趋势,且趋于平稳,是一种呈稳步上升发展态势结论 灰色预测有一定优点,不要求统计变量具有正态分布,而是在杂乱无章的、有限的、离散的数据中找出规律,按系统发展趋势预测分析,提高了预测和分析的精确度.通过预测结果相互印证,理论值接近实际值,此方法在医院管理中有实用性.

Objective To explore the application of The Gray Model and forecast the trend of outpatients and inpatients from 2011 to 2013 , and provide the references for hospital managers. Methods Based on the annual statistical report of outpatients and inpatients in a hospital during 2000 to 2010, we build GM（ 1,1 ） model using gray prediction system and Microsoft Excel to forecast the number of outpatients and inpatients. Results Analyzed by posterior error ratio C and small error probability P, the fitting result of GM（ 1,1 ） model is satisfied, which can be applied for extrapolation prediction. Predicted number of outpatients from 2011 to 2013 is followed by 25 6267, 27 666, 300 852. Evaluate results is qualified （C=0. 33〈0. 35 ,P= ≥0.9）, predicted number of inpatients from 2011 to 2013 is followed by 17 739,19 044, 20 445, evaluate results is excellent（ C=0. 29 〈0. 35 ,P= 1 ） ）. 8.35% and 7.36% are the annual growth rate of outpatients and inpatients respectively. The number of outpatients and inpatients may rise gradually and stably. Conclusion The gray predication has some advantages and the statistical variable may not be normal distribution. Using it, we can study the regulation by analyzing the chaotic discrete and limit data. The model can improve the accuracy of predication and analysis. Theoretical value is in accord with actual value. The GM（ 1, I ） model is practical in hospital management.