The prevalence of pre-analytical errors in the laboratory of the Cantonal Hospital Zenica in Bosnia and Herzegovina.

Aim To identify rates of most common pre-analytical errors and to document possible (different) error rates between inpatients and outpatients. Methods This retrospective study was conducted at the Department of Medical Biochemistry and Immunology Diagnostics, Cantonal Hospital Zenica, from December 2016 until March 2017. Data on rejected blood samples in the laboratory information system were analysed. Results During the 3-month period 35,343 patient blood samples (25,545 inpatients and 9,798 outpatients) were received in the laboratory. The study identified 602 (1.70%) rejected samples because of pre-analytical errors, mostly due to haemolysis, 292 (48.50%), and clotted samples, 240 (39.87%). The remaining 70 (11.63%) samples were rejected because of inappropriate sample volume, inappropriate container and identification errors (7.81%, 2.16% and 1.66%, respectively). The proportion of inpatient rejected samples was 8.7-fold higher than in the outpatient samples. The proportion of inpatient rejected samples because of haemolysis, clotted samples, inappropriate sample volume and inappropriate containers were higher than in the outpatient samples (20.5-, 12.1-, 2.3- and 1.3-fold higher, respectively); proportion of rejected samples because of identification errors was 8.0-fold higher in the outpatient (collection sites outside the hospital) than in the inpatient samples. Conclusion Higher pre-analytical sample error rates were connected with inpatient samples, while higher identification error rates were connected with outpatient samples. Establishment of periodic stuff training and introduction of information technology could reduce pre-analytical errors.