Wallin Olof, Söderberg Johan, Van Guelpen Bethany, Stenlund Hans, Grankvist Kjell, Brulin Christine
Department of Medical Biosciences, Clinical Chemistry, Umeå University, Umeå, Sweden.
Clin Chim Acta. 2008 May;391(1-2):91-7. doi: 10.1016/j.cca.2008.02.016. Epub 2008 Feb 21.
Most errors in laboratory medicine are preanalytical in nature. In the present study, we aimed to survey preanalytical steps in venous blood sampling, prior to actual sample collection. These steps included test-request management and test-tube labelling, as well as information search procedures.
Venous blood sampling staff (n=314, response rate 94%) in hospital wards and laboratories completed a questionnaire related to clinical chemistry testing.
Instructions for test-request management and test-tube labelling were not always followed. For example, only 66% of the ward staff reported always checking the test-request if someone else completed it, compared to 90% of the laboratory staff (p=0.003). As few as 16% of the ward staff reported desirable practices regarding test-tube labelling, compared to 100% of the laboratory staff (p<0.001). Furthermore, 18% of the ward staff reported always using online manuals (the only source of updated information), compared to 63% of the laboratory staff (p<0.001).
Our results indicate a substantial risk of preanalytical error in test-request management, test-tube labelling, and information search practices, particularly in the wards. Our findings thus underscore the importance of quality control in venous blood sampling, in order to increase patient safety in modern health care.
检验医学中的大多数错误本质上是分析前的。在本研究中，我们旨在调查实际采集样本前静脉采血的分析前步骤。这些步骤包括检验申请管理、试管标记以及信息查找程序。
医院病房和实验室的静脉采血工作人员（n = 314，回复率94%）完成了一份与临床化学检测相关的问卷。
检验申请管理和试管标记的说明并非总是得到遵循。例如，只有66%的病房工作人员表示，如果是其他人填写检验申请，他们总会进行核对，而实验室工作人员的这一比例为90%（p = 0.003）。在试管标记方面，只有16%的病房工作人员报告采取了理想做法，而实验室工作人员的这一比例为100%（p < 0.001）。此外，18%的病房工作人员表示总是使用在线手册（唯一的最新信息来源），而实验室工作人员的这一比例为63%（p < 0.001）。
我们的结果表明，在检验申请管理、试管标记和信息查找实践中，尤其是在病房，存在分析前错误的重大风险。因此，我们的研究结果强调了静脉采血质量控制的重要性，以提高现代医疗保健中的患者安全。
