控制图在医院数据质量管理中的应用实践

Application of control charts in hospital data quality management

目的研究控制图在医院数据质量管理中应用的可行性,为提升医院数据质量管理提供借鉴。方法选取某三级甲等综合医院2022年1—5月总院区的主要业务指标和分院区的部分业务指标,利用K-S检验法和χ^(2)拟合优度检验对其统计日报表数据进行检验,分析数据是否服从特定分布,然后根据数据类型和分布类型选用合适的控制图,采用Minitab 21.1软件绘制医院数据质量控制图,通过分析控制图中数据点的分布位置监测数据质量。结果经检验,总院区的入院量、彩超检查量和急诊科接诊量、分院区CT检查量的数据服从正态分布,适用单值控制图;总院区及分院区的救护车出车次数服从泊松分布,适用缺陷数控制图;总院区住院死亡人数服从几何分布,适用稀有事件控制图。总院区入院量及彩超检查量受其他因素影响较大,使用单值选控图进一步确定数据分步异常的原因。控制图分析结果显示,总院区入院量、彩超检查量、住院死亡人数的数据分布无异常点,分院区CT检查量有2个异常点,总院区急诊科接诊量、总院区和分院区救护车出车次数的控制图各有1个异常点。经核实,分院区CT检查量的1个异常点和分院区救护车出车次数的1个异常点是数据错误导致,其他数据无误。结论使用控制图监测医院数据质量具有可行性,可作为质量管理工具辅助医院数据质量管理。

Objective To study the feasibility of control charts in the quality management of hospital statistical data as a reference for improving such management.Methods Main business indicators of the main campus and some business indicators of the affiliated campus of a tertiary general hospital from January to May 2022 were selected.K-S test and chi-square goodness of fit test were used to test data in its statistical daily report,analyzing whether the data conform to the specific distributions.Then appropriat control chart were used according to the data type and distribution type.Minitab 21.1 software was used to draw the hospital data quality control chart,and data quality was monitored by analyzing the distribution of data points in the control chart.Results The test found that the number of admissions,ultrasound examinations and emergency department visits in the main campus,and CT examinations in the affiliated campus,conformed to normal distribution,and single value control charts were applied.The number of ambulance trips in the main campus and the affiliated campus conform to Poisson distribution,and the defect number control chart was applied.The number of inpatient deaths in the main campus conform to a geometric distribution,and a rare event control chart was applied.The volume of admissions and ultrasound examinations in the main campus were mostly influenced by other factors,and the single-value selective control chart was used to further determine the cause of abnormal data distribution.The results of the control chart analysis showed that,there were no abnormal points in the data distribution of admissions,color ultrasound volume and in-hospital deaths in the main campus,two abnormal points in the CT examination volume in the affiliated campus.The control charts for the number of emergency department visits in the main campus and the number of ambulance trips in the main and affiliated campuses each had one outlier.It was verified that one anomaly in the volume of CT examinations in the affiliated campus and one anomaly in the number of ambulance trips in the affiliated campus were caused by data errors,while the other data were correct.Conclusions It is feasible to use control charts to monitor the quality of hospital data,which can be used as a quality management tool to assist the quality management of hospital data.