Department of Environmental and Occupational Health, Drexel University School of Public Health, Philadelphia, Pennsylvania 19102, USA.
Inj Prev. 2011 Dec;17(6):388-93. doi: 10.1136/ip.2010.029108. Epub 2011 May 5.
To evaluate the need for triangulating case-finding tools in patient safety surveillance. This study applied four case-finding tools to error-associated patient safety events to identify and characterise the spectrum of events captured by these tools, using puncture or laceration as an example for in-depth analysis.
DATA SOURCES/STUDY SETTING: Retrospective hospital discharge data were collected for calendar year 2005 (n=48,418) from a large, urban medical centre in the USA.
The study design was cross-sectional and used data linkage to identify the cases captured by each of four case-finding tools.
DATA COLLECTION/EXTRACTION METHODS: Three case-finding tools (International Classification of Diseases external (E) and nature (N) of injury codes, Patient Safety Indicators (PSI)) were applied to the administrative discharge data to identify potential patient safety events. The fourth tool was Patient Safety Net, a web-based voluntary patient safety event reporting system.
The degree of mutual exclusion among detection methods was substantial. For example, when linking puncture or laceration on unique identifiers, out of 447 potential events, 118 were identical between PSI and E-codes, 152 were identical between N-codes and E-codes and 188 were identical between PSI and N-codes. Only 100 events that were identified by PSI, E-codes and N-codes were identical. Triangulation of multiple tools through data linkage captures potential patient safety events most comprehensively.
Existing detection tools target patient safety domains differently, and consequently capture different occurrences, necessitating the integration of data from a combination of tools to fully estimate the total burden.
评估在患者安全监测中对三角定位式病例发现工具的需求。本研究应用四种病例发现工具来发现与错误相关的患者安全事件，并对这些工具所捕获的事件范围进行识别和描述，以穿刺或撕裂伤为例进行深入分析。
数据来源/研究设置：从美国一个大型城市医疗中心的 2005 年日历年度中收集回顾性住院数据（n=48418）。
研究设计为横断面研究，并使用数据链接来识别四种病例发现工具中捕获的病例。
数据收集/提取方法：应用三种病例发现工具（国际疾病分类外部（E）和性质（N）损伤代码、患者安全指标（PSI））对行政出院数据进行分析，以识别潜在的患者安全事件。第四种工具是基于网络的自愿性患者安全事件报告系统 Patient Safety Net。
检测方法之间的相互排斥程度很大。例如，在将穿刺或撕裂伤与唯一标识符相链接时，在 447 个潜在事件中，PSI 和 E 代码有 118 个相同，N 代码和 E 代码有 152 个相同，PSI 和 N 代码有 188 个相同。只有 100 个被 PSI、E 代码和 N 代码同时识别的事件是相同的。通过数据链接对多种工具进行三角定位可以最全面地捕获潜在的患者安全事件。
现有的检测工具针对患者安全领域的目标不同，因此捕获的事件也不同，需要整合来自多种工具的数据，以全面估计总负担。
