海南医院急诊患者就诊人数的统计建模

Statistical Modelling for the Number of Hospital Emergency Patients in Hainan

从宏观面与微观面研究海南某医院2012-2014年急诊患者就诊人数的统计建模,宏观面根据月,周数据建立时间序列模型,进行了相关预测,发现存在负的"一月效应"和负的"周二,周四效应",正的"周日效应".微观面根据日数据及分时段数据,研究就诊人数的概率分布,借助Box-Cox变换和Matlab软件编程,把非正态分布的序列变换成正态分布,估计其中的有关参数,确定密度函数,从而进行统计推断,给定显著性水平α(如α=0.05),在置信水平(如95%)下,得到患者在一日内及不同时段就诊人数的置信区间,为医院管理部门合理安排医护人员提供科学依据.

This paper studies on statistical modeling for the number of emergency patients of a Hainan hospital from macroscopic and microscopic view. From macroscopic view, we establish time series model for monthly data and weekly data, make relative prediction, find existence of 'January effect' and 'Thursday, Sunday effect'. From microscopic view, we study the probability distribution of the number of emergency patients by daily data and time interval data.With the aid of Box-Cox transform and MATLAB software programming,the sequence of the non normal distribution is transformed into normal distribution,we get the estimated values of parameters in density function, so that we can make the statistical inference, given the significant level(as α = 0.5), under the confidence level(such as 95%),obtain The confidence interval of the number of patients in a day and in different periods,so that the number of doctors or medical staff can be arranged scientifically and reasonably.