engleski

Wrong blood in tube (WBIT) error : first study on donor blood samples

Although the error known as wrong blood in tube (WBIT) has been well documented in the hemovigilance system, it has hitherto applied only to one end of the transfusion chain, that is, to the patient. Thus, this type of error has been defined as a case where the patient identification information refers to one patient, while the blood in the tube belongs to another patient.[1] An international study from 2003 estimated the prevalence of these errors at 0.03% to 0.09% in 10 participating countries.[2] A similar frequency of WBIT errors of 0.04% has been reported in two recent studies.[3, 4] Investigating the frequency of donor sample nonconformities in a 10-year study, we found the frequency of WBIT errors in donor samples to be a considerably less frequent phenomenon relative to patients.[5] Preliminary results prompted us to embark upon a more precise analysis of the WBIT frequency in donor blood samples, the causes of their occurrence, and trends observed over years. To our knowledge, this is the first study on WBIT errors in donor blood samples which occurred during their labeling and collection. During the 12-year study period (2002-2013), 955, 218 blood donations with two samples each were collected at the Croatian Institute of Transfusion Medicine. WBIT was recorded in 19 (0.0020%) blood donations (in 12 error episodes in total). In 15 of these 19 blood donations, WBIT was detected in both test samples, whereas in four blood donations only one test sample proved questionable. Thus, WBIT was recorded in 34 (0.0018%) samples. All cases of WBIT errors recorded in this study were detectable by routine control mechanisms at the blood establishment, thus efficiently preventing the release and distribution of nonconformable blood products. In addition to blood group comparison of the respective blood donation and the historical blood group of multiple blood donors, the possible cases of mismatching donations from two new blood donors with different blood groups were prevented by additional blood group determination in the samples of all first-time donor units (forward typing from the attached plastic tube segment) and by comparison with the blood group determined from venous blood samples. As in the case of WBIT in patients, a part of “silent WBIT cases” remains unrecognized. In the working conditions at our institute, this is possible by mismatching two blood donors with identical blood groups, whereby the error is not recognized through another surveillance mechanism. In 50% of cases (6/12 error episodes), mismatching of the donation form and panel of donation identification (ID) numbers occurred after blood bag labeling. Only one case of mismatching included two first-time donors, and in case of no blood group control from the plastic tube segment may have resulted in the release or distribution of an erroneously labeled blood product (incompatible in the ABO system). In three (25%) of 12 error cases, WBIT was recorded at the institution and in the rest at mobile sessions. Comparison of the proportion of errors occurring at the institution and at mobile sessions relative to the respective number of donations collected did not yield a significant difference (p = 0.970). In three cases, inexperienced (defined as up to 1 year of work in blood collection) technicians were involved. Thus, 15.8% of 19 technicians involved in these cases could be considered to have had less experience. Comparison of the proportion of errors between these workers and more experienced workers showed no significant difference (p = 0.264). Seven (58%) of 12 blood donation sessions where errors occurred were sessions characterized by excessive workload, defined as more than eight blood donors per worker per hour (9-11). An additional two cases were evaluated as borderline workload (seven to eight blood donors/worker/hr). For comparison, in a sample of 123 donation sessions conducted in 2013 in various seasons, the mean number of blood donors per worker per hour was 5.4, with four to seven blood donors in as many as 83% of sessions. In two of 12 cases, mismatched donors had the same last name (mother and daughter and father and son), whereas in an additional case two female blood donors had the same first name. Comparison of the frequency of WBIT errors occurring during the four comparable periods (2002-2004, 2005-2007, 2008-2010, and 2011-2013) revealed half as many errors in the past 3 years relative to the previous periods, however, without reaching significance either for each period separately (p = 0.800, p = 0.801, and p = 0.671, respectively) or for the 2002 to 2010 study period as a whole (p = 0.586), using the test of proportion difference. This finding correlated with abolishing the function of the phlebotomist's assistant and full implementation of automated scale mixers. Study results pointed to a relatively low frequency of WBIT errors in donor blood samples. Owing to a series of control mechanisms, the possibility of potentially fatal outcomes of such errors is very low. Investigating these errors, however, greatly interferes with regular work, is time-consuming, and so forth. Besides its relevance in immunohematologic testing, WBIT may hamper traceback and lookback procedures and result in erroneous conclusions. This study revealed the multifactorial nature of the potential causes of WBIT errors, among which excessive workload should be noted. We believe that the results of our study will enable comparison of WBIT frequency in donor blood samples at other centers, while our experience will help them establish appropriate measures to prevent and reduce this type of error.

nije evidentirano

nije evidentirano