Identification and reduction of pre-analytical errors in clinical chemistry through expert advice

Abstract

Background: Diagnostics in Clinical chemistry laboratory is a pivotal part of clinical decision-making but is not exempt from ‘human errors’. Scientific innovations such as automation and electronic order test requesting have contributed to substantial improvements in the field of laboratory science, but errors still occur. One major example of such failing is connected to the prevalence of errors occurring in pre-analytical phase of the Total Testing Process (TTP). Preanalytical errors can occur at the time of patient assessment, test order entry, patient identification, sample collection, sample transport, or sample receipt in the laboratory. Previous work and clinical insights suggest that most errors in the TTP are extra-laboratory (i.e. they occur before the samples reach the laboratory for analysis). Such errors are frequently the results of human mistakes during phlebotomy practice. Therefore to reduce these errors the pre-analytical phase of the TTP must be prioritised. Study objective: To investigate the sources of pre-analytical errors in the TTP, categorize these errors in order to identify the error prone steps, and evaluate errorreporting frequencies, with the aim of improving service. Methods: The first part of the study was a query of the laboratory information management system (LIMS) for samples rejected due to pre-analytical errors. Data collection was done retrospectively to cover two periods from 2007-2008 (manual paper test requesting) and 2012-2013 (after implementation of electronic test ordering, Anglia-ICE). Pre- and post- implementation Anglia-ICE error data were transferred to excel spreadsheets and compared by chi-squared test. The contribution of each error category to total sample error received in the laboratory was also determined. The second part of the study was a questionnaire survey of pre-analytical procedures to capture the attitudes of phlebotomists towards current practice in Sheffield Teaching Hospitals NHS Foundation Trust (STH NHS FT). Results: The results of the first part of the study indicated that of the 416,703 specimens collected pre-Anglia ICE, 2,055 (0.49%) were recorded as errors compared with 1,616 errors (0.11%) of 903,814 specimens collected post-Anglia ICE implementation, which represents a 0.31% (p<0.05) absolute error reduction rate, although more samples were received post-Anglia-ICE. The results of the second part (questionnaire survey) indicate that recommended procedure for phlebotomy practice was not strictly followed by a large percentage of the staff surveyed. Conclusion: This study is the first inquiry linking venous blood sampling (VBS) practices in phlebotomy to retrospective LIMS pre-analytical data in a UK NHS Hospital. The results suggest low compliance by staff with recommended practice, which may be responsible for the prevalence of certain categories of pre-analytical errors in the TTP and may also be associated with increased risks to attending patients. It is suggested that the development of a local guideline for VBS and compliance to this guideline will improve phlebotomy practice, improve the quality of sample testing in the clinical chemistry laboratory, reduce pre-analytical errors in TTP and consequently improve the safety of patients.