S G Sandler1, A Langeberg, N Avery, P D Mintz. 1. Department of Laboratory Medicine, Georgetown University Medical Center, Washington, DC 20007, USA. sandlerg@gunet.georgetown.edu
Abstract
BACKGROUND: The results of routine blood bank testing by a fully automated blood typing system (ABS2000) were compared with those obtained by standard manual methods in six hospital transfusion services. STUDY DESIGN AND METHODS: The ABS2000 system uses microtiter plates for determining ABO and D types, solid-phase red cell adherence (SPRCA) assays for antibody detection, and modified SPRCA plates for IgG crossmatches. The transfusion services used their standard manual test tube methods. RESULTS: Of 3779 donors' samples tested for ABO types (red cell typings only), 3.0 percent could not be interpreted by the ABS2000 system's neural network, because of clots, hemolysis, or lipemic samples. The results for ABO types were concordant for 99.8 percent of the remaining samples. Of 3779 donors' samples tested for D types, the results were concordant for 98.7 percent. Of 7580 patients' samples tested for ABO types (red cell and plasma typings), 5.8 percent could not be interpreted by the ABS2000 system. There was 100-percent concordance of ABO typing results for the remaining 7140 samples. There was 99. 7-percent concordance of results for patients' D types. The results of 96.7 percent of antibody detection tests and 98.8 percent of crossmatches were concordant. Neither method failed to detect a serologically incompatible crossmatch that was associated with a specific, clinically significant alloantibody. The ABS2000 system performed 45 confirmatory donor ABO and D types in 115 minutes, 22 antibody detection tests in 116 minutes, 16 patients' ABO/D types in 149 minutes, and 40 crossmatches in 140 minutes. CONCLUSION: The ABS2000 blood typing system automates routine blood bank tests with accuracy comparable to that of hospital transfusion services' standard manual methods.
BACKGROUND: The results of routine blood bank testing by a fully automated blood typing system (ABS2000) were compared with those obtained by standard manual methods in six hospital transfusion services. STUDY DESIGN AND METHODS: The ABS2000 system uses microtiter plates for determining ABO and D types, solid-phase red cell adherence (SPRCA) assays for antibody detection, and modified SPRCA plates for IgG crossmatches. The transfusion services used their standard manual test tube methods. RESULTS: Of 3779 donors' samples tested for ABO types (red cell typings only), 3.0 percent could not be interpreted by the ABS2000 system's neural network, because of clots, hemolysis, or lipemic samples. The results for ABO types were concordant for 99.8 percent of the remaining samples. Of 3779 donors' samples tested for D types, the results were concordant for 98.7 percent. Of 7580 patients' samples tested for ABO types (red cell and plasma typings), 5.8 percent could not be interpreted by the ABS2000 system. There was 100-percent concordance of ABO typing results for the remaining 7140 samples. There was 99. 7-percent concordance of results for patients' D types. The results of 96.7 percent of antibody detection tests and 98.8 percent of crossmatches were concordant. Neither method failed to detect a serologically incompatible crossmatch that was associated with a specific, clinically significant alloantibody. The ABS2000 system performed 45 confirmatory donorABO and D types in 115 minutes, 22 antibody detection tests in 116 minutes, 16 patients' ABO/D types in 149 minutes, and 40 crossmatches in 140 minutes. CONCLUSION: The ABS2000 blood typing system automates routine blood bank tests with accuracy comparable to that of hospital transfusion services' standard manual methods.
Authors: Su Yeon Jo; Ju Mi Lee; Hye Lim Kim; Kyeong Hwa Sin; Hyeon Ji Lee; Chulhun Ludgerus Chang; Hyung Hoi Kim Journal: Ann Lab Med Date: 2017-03 Impact factor: 3.464