[Specimen labelling: A complex process with a high error rate].

Errors associated with the labelling of laboratory specimens repeatedly occur in health facilities and are often described in Critical Incident Reporting Systems (CIRS). The analysis of these error reports and the complexity of the specimen collection process pose major challenges for responsible health care professionals. To examine which errors occur in the labelling of laboratory specimens, and what leads to their detection, the error reports of 42 Swiss hospitals within the Critical Incident Reporting & Reaching NETwork (CIRRNET) were systematically analysed.

METHOD: Within the scope of a reporting month in the year 2016, 42 hospitals were asked to pay particular attention to the issue of mislabelled laboratory specimens, to report these errors in the local CIRS system, and to forward the information to the CIRRNET database. In addition, using a systematic keyword search, a search for old error reports on the mislabelling of laboratory specimens was conducted in the CIRRNET database and the results extracted. 227 error reports were finally included in the analysis. All these error reports were analysed systematically and by content, and the problem areas described in these reports allocated to newly defined categories. The systematic analysis included: the time of error detection, the problem area described, the length of time between error occurrence and error detection, as well as possible safety barriers.

RESULTS: The majority (52 %) of labelling errors of laboratory specimens are detected in the laboratory, 21.1 % on the hospital ward before the specimens are sent to the laboratory to be processed, and a further 24.7 % at a very late stage when test results are associated with the patient on the ward. The analysis of the problem areas described showed that patient identification (7.9 %) was a key issue. The most frequent errors occurred in connection with the labelling of laboratory test tubes (45.4 %) and analysis forms or requisition slips (33 %). Numerous errors went undetected throughout the entire process, i.e. right up until the moment when the test result was finally assigned to the patient. The analysis also provided information about possible safety barriers for future prevention of laboratory specimen mislabelling.

CONCLUSIONS: A content-analysis approach is essential to the evaluation of error reports as it is the only way to identify all the problems of laboratory specimen mislabelling described. Furthermore, a deductive approach also facilitates the identification of possible safety barriers. The findings provide valuable information for mislabelling prevention approaches that is not otherwise available. Further analysis of error reports from Critical Incident Reporting Systems is necessary in order to gain more experience with the methodical approach and to draw conclusions as to whether the complex process of analysing each individual error report yields information that is more useful than the analysis of the error reports of an identified problem area.