John Jeongseok Yang1, Suk Won Seo1, Jin Seok Kim1, Yousun Chung2, Hyungsuk Kim3, Dae-Hyun Ko4, Sang-Hyun Hwang1, Oh Heung-Bum1. 1. Department of Laboratory Medicine, University of Ulsan College of Medicine and Asan Medical Center, Seoul, South Korea. 2. Department of Laboratory Medicine, Kangdong Sacred Heart Hospital, Seoul, South Korea. 3. Department of Laboratory Medicine, Seoul National University Hospital, Seoul, South Korea. 4. Department of Laboratory Medicine, University of Ulsan College of Medicine and Asan Medical Center, Seoul, South Korea. Electronic address: daehyuni1118@amc.seoul.kr.
Abstract
BACKGROUND: Anti-blood group antibody titers (ABTs) reported in titer values are variable depending on the testing method used. The introduction of new test methods such as automated methods requires proper method comparison. In this study, the automated blood bank system and manual tube method for ABT were compared using a log-transformed dataset to evaluate the alternative statistical approach. METHODS: ABT was conducted using specimens referred for solid organ transplantation. Methods for comparison were conventional manual tube method and IH-500 automated blood bank system using column agglutination (CAT). Criteria for agreement were exact match and 1-titer match. Measured titer values were log-transformed into interval scale for Deming regression analysis. RESULTS: From the comparison of the tube and CAT methods using titer values and the two criteria, the exact and 1-titer match were 15.9-41.5 % and 65.9-97.6 %, respectively. Deming regression was used to demonstrate the presence of both proportional and constant difference between the two methods. CONCLUSION: The method comparison using conventional statistical approaches had limits due to the semi-quantitative value of the test. Log-transformed interval scale values for comparison were useful for interpretation of method comparison datasets. This alternative statistical approach could contribute to a more accurate comparison between assays and standardization of ABT testing.
BACKGROUND: Anti-blood group antibody titers (ABTs) reported in titer values are variable depending on the testing method used. The introduction of new test methods such as automated methods requires proper method comparison. In this study, the automated blood bank system and manual tube method for ABT were compared using a log-transformed dataset to evaluate the alternative statistical approach. METHODS:ABT was conducted using specimens referred for solid organ transplantation. Methods for comparison were conventional manual tube method and IH-500 automated blood bank system using column agglutination (CAT). Criteria for agreement were exact match and 1-titer match. Measured titer values were log-transformed into interval scale for Deming regression analysis. RESULTS: From the comparison of the tube and CAT methods using titer values and the two criteria, the exact and 1-titer match were 15.9-41.5 % and 65.9-97.6 %, respectively. Deming regression was used to demonstrate the presence of both proportional and constant difference between the two methods. CONCLUSION: The method comparison using conventional statistical approaches had limits due to the semi-quantitative value of the test. Log-transformed interval scale values for comparison were useful for interpretation of method comparison datasets. This alternative statistical approach could contribute to a more accurate comparison between assays and standardization of ABT testing.