摘要
目的依据某三甲医院2009年-2013年门、急诊人次和出院人数建立一元线性回归模型,预测2014年上半年的相关指标并与实际数量比较。方法利用近五年门急诊人次和出院人数与年代之间采用一元线性回归模型进行回归分析,根据回归模型系数评估2014年上半年门急诊人次和出院人数的预测值和95%置信区间,门诊人次、急诊人次和出院人数分别与年代做线图,并对人数随年代变化做相关性检测。结果 2009年-2013年门诊、急诊人次和出院人数随年代增加均具有统计学意义(p<0.05)。99%门诊人次、79%的急诊人次和99%的出院人数的变化是由年代决定的。预测2014年门诊人次2845144,急诊人次为207038,出院人数为46654。2014年上半年门急诊人次和出院人数与实际2014年门急诊人次和出院人数相比,均在预测值95%置信区间内。结论门急诊人次、出院人数与年代有较好线性关系,可以用一元线性回归模型进行回归分析并预测。
Objective To build prediction linear regression model according to the number of outpatient and emergence visits and discharged patients from 2009 to 2013 in a grade 3 and first-class hospital and predict index of correlation and Compared with the actual number in 2014.Methods According to the data nearly five years,a linear regression model for outpatient visits,emergency visits and discharge number was built with calendar year.The parameters were estimated in the model.The responding figures and 95% confidence interval in the first half year of 2014 were predicted by the model.The predictive value was assessed by comparing the estimated value with the actual value.The linear graph was prepared with statistical analysis.Results Outpatient visits,emergency visits and discharge number increasing from 2009 to 2013 were statistically significant and had a positively linear regression with calendar year （P＆lt;0.05）.99% variance of outpatient visits,79% variance of emergence visits and 99% variance of inpatient were depended on the change of years.The predictive value of outpatient visits,emergence visits and the discharge number were 2845144,207038,46654, respectively.The predictive value in the first half year of 2014 compared to the actual data was within 95% confidence interval. Conclusion The outpatient visits,emergence visits and the discharge number had a good linear relationship with calendar year.We could make regression model to predict outpatient visits,emergence visits and the discharge number.
出处
《中国病案》
2014年第11期39-40,共2页
Chinese Medical Record
关键词
回归分析
门诊人次
急诊人次
出院人数
预测
Linear regression
Outpatient visits
Emergence visits
Discharge number
Prediction
