Substantial variability between different antibody titration methods has been identified since the development and introduction of the uniform procedure in 2008.
To determine whether more recent methods or techniques decrease interlaboratory and intralaboratory variation measured using proficiency testing.
Proficiency test data for antibody titration between 2014 and 2018 were obtained from the College of American Pathologists. Interlaboratory and intralaboratory variations were compared by analyzing the distribution of titer results by method and phase, comparing the results against the supplier’s quality control titer, and by evaluating the distribution of paired titer results when each laboratory received a sample with the same titer twice.
A total of 1337 laboratories participated in the antibody titer proficiency test during the study period. Only 54.1% (5874 of 10 852) of anti-D and 63.4% (3603 of 5680) of anti-A reported responses were within 1 titer of the supplier’s intended result. Review of the agreement between laboratories of the same methodology found that 78.4% (3139 of 4004) for anti-A and 89.0% (9655 of 10 852) of laboratory responses for anti-D fell within 1 titer of the mode response. When provided with 2 consecutive samples of the same titer (anti-D titer: 16), 85% (367 of 434) of laboratories using the uniform procedure and 80% (458 of 576) using the other method reported a titer difference of 1 or less.
Despite advances, interlaboratory and intralaboratory variance for this assay remains high in comparison with the strong reliance on titer results in clinical practice. There needs to be a reevaluation of the role of this test in clinical decision-making.
Antibody titration in transfusion medicine is a commonly used clinical assay to detect the strength of reactivity of red cell antibodies present in a patient’s plasma.1 There are a number of current uses for this assay. Some of the common uses include screening and monitoring for the risk of hemolytic disease of the fetus and newborn,2 preventing certain kinds of hemolytic reactions,3 evaluating the safety of incompatible blood transfusions,4,5 and assessing and monitoring incompatible solid organ or hematopoietic stem cell transplants.6,7 Inherent with all of these uses is that there is some critical titer whereby an important clinical decision is made about a patient, blood product, or stem cell product.
It has been known for quite some time that the antibody titer assay has challenges with assay variability, especially as it pertains to comparing results between different laboratories.8 Even in 2022, the assay methods used to determine antibody titer require a semiquantitative approach in which the laboratory evaluates serial dilutions by a serologic technique. Because of the fact that the testing conditions are known to differ over time, many laboratories still require that previously tested samples be retested alongside a new/current sample to ensure that an apparent change in result is not due to differing applications of technique.9,10 Although duplicate testing mitigates some of the assay imprecision, variation in techniques makes application of uniform guidelines challenging and contributes to the difficulty in predicting the clinical impact of an antibody titer of a particular strength.
In addition to historical sample testing, other approaches have been developed to reduce antibody titer variability. AuBuchon et al8 developed a technique known as the uniform procedure and found that by using a weak positive (instead of a 1+) end point as the reportable titer, they could statistically reduce the variance of titers reported between laboratories for tube-based titrations of anti-D and of anti-A with 37°C incubation and reading at the antihuman globulin (AHG) phase. Unfortunately, a subsequent retrospective study with a larger sample size revealed that the uniform procedure did not impact the real-world variability of this assay.11
Because patients often seek care at different institutions, it remains the goal of the laboratory to not only identify the appropriate antibody and its exact strength, but also to make sure the testing methodology can obtain reproducible results in different settings. Hence, the College of American Pathologists (CAP) Transfusion, Apheresis, and Cellular Therapy committee endeavored to critically evaluate the extent to which operator and laboratory factors influence the variability of antibody titer results, and to reassess the impact of the uniform procedure on the current versions of antibody titer methods. To accomplish these goals, we analyzed data from 10 consecutive CAP proficiency tests (PTs) during 5 years (2014–2018).
MATERIALS AND METHODS
The CAP’s Antibody Titer PT survey is conducted biannually (surveys A and B) for anti-A and anti-D antibody titration. Laboratories report their results based on their primary testing method. Each survey contained 2 separate plasma samples (1 with anti-A and 1 with anti-D) and group A1 RhD-positive (D+) red cells for laboratories to perform antibody titration. For anti-A, titers were performed both at immediate spin and/or room temperature and at AHG phase, and for anti-D, at AHG phase as per the laboratories’ primary standard procedure. The uniform procedure or other procedure could be used for reporting anti-A and anti-D titers, and tube and gel techniques could also be used for both titer types. Users of the antibody titer PT were specifically instructed to state whether they use a uniform procedure or other procedure based on the following definitions:
Uniform procedure: a titer procedure specifically described by AuBuchon et al.8 Those using the tube technique would need to report the highest titer that gives a weak (barely visible agglutination) serologic reaction (as opposed to a 1+ reaction). Those using the gel technique would need to report the highest titer that gives a 1+ serologic reaction (small agglutinates).
Other procedure: any titer methodology or procedure that does not strictly adhere to the uniform procedure as described in AuBuchon et al.8 Those using the tube and gel technique would need to report the highest titer that gives a 1+ serologic reaction (small agglutinates).
The PT results were submitted on standardized result forms that collected information on the laboratory’s methodology and titer results. For both anti-A and anti-D, titers were categorized as: (1) uniform procedure/tube technique (UT); (2) uniform procedure/gel technique (UG); (3) other procedure/tube technique (OT); (4) other procedure/gel technique (OG).
Laboratory institution type and overall annual test volume were extracted from the CAP’s demographics database.
We examined CAP PT results from 2014 through 2018 (5 years). In order to critically evaluate the interlaboratory and intralaboratory variability of the titer assay, we performed 3 key analyses on these 10 PT results: (1) We evaluated the ability of laboratories to match the supplier’s quality control (QC) titer for each sample. For this analysis, accuracy was defined as a laboratory reporting a titer that was +/− 1 titer of (1 titer greater or 1 titer lower than) the intended titer as defined by the PT supplier. For anti-D titers, the supplier specifically quality checked its PT samples using the other procedure, tube technique with a saline diluent at 37°C, and with anti–immunoglobulin G (anti-IgG) AHG. For anti-A, the supplier used the other procedure, tube technique, saline diluent, and immediate spin or at 37°C with anti-IgG AHG. Given that the supplier only used these methods to determine the intended titer value, we performed a subset analysis that evaluated the ability of laboratories to be +/− 1 titer when only using the same titer methodology and technique as the supplier. (2) We further evaluated the interlaboratory variability of reported titers during the study interval by using the mode value of participating laboratories. For this analysis, accuracy was defined as a laboratory reporting a titer that was +/− 1 titer of the mode for a specific technique and procedure. (3) Lastly, we evaluated the intralaboratory reproducibility of an anti-D antibody titer in the 2018 A and B antibody surveys. The anti-D plasma titer was the same in both mailings and had an intended titer, as defined by the supplier, of 16.
Performance results were represented as frequencies with a percentage if applicable. To evaluate the variability of the anti-D and anti-A antibody titer results, we evaluated whether a responding laboratory was within +/− 1 titer of the supplier’s QC result or the mode titer for a specific procedure and platform. A χ2 analysis was used to detect significant differences in the proportion of those who were within +/− 1 titer of the mode by antibody detected (anti-A versus anti-D). Logistic regression models were additionally used to test for the factors that were associated with laboratory response performance and included comparisons of uniform versus other procedure, gel versus tube technique, the intended titer value from the supplier’s QC value, and the differing diluents and reagents as applicable.
For the reproducibility of the anti-D antibody titer for the samples included in both 2018 A and B mailings, a paired match for a laboratory’s results was defined as a difference within 1 titer. For this analysis, logistic and multivariate logistic regression models were used to test for factors associated with the paired-match performance. Test method was the practice characteristic evaluated for the uniform procedure analysis, whereas test platform, dilution, and titer technique were analyzed for the other procedure. A significance level of .05 was used for all analyses.
RESULTS
Demographics
From 2014 to 2018, 1337 US and international laboratories participated in the antibody titer PT (1148 [85.9%] domestic; 189 [14.1%] international). Complete demographic results were not available for about 100 laboratories, and so details of 1234 laboratories regarding institution type and 1224 laboratories regarding test volume were reported (Table 1). Most of the laboratories were from academic centers (526 of 1234; 42.6%), followed by private medical centers (477 of 1234; 38.7%). Most of the laboratories participating in this PT reported performing more than 250 000 laboratory tests annually (1069 of 1224; 87.3%).
Titer Performance Compared With the Intended Titer
For anti-D, 10 852 total titer responses were evaluated, with 4885 responses (45.0%) using the uniform procedure (tube, 3517 of 4885 [72.0%]; gel, 1368 of 4885 [28.0%]), and 5967 responses (55.0%) using the other procedure (tube, 5644 of 5967 [94.6%]; gel, 323 of 5967 [5.4%]; Tables 2 and 3). For anti-A, 5680 total titer responses were evaluated, with 3017 responses (53.1%) using the uniform procedure (tube, 2497 of 3017 [82.8%]; gel, 520 of 3017 [17.2%]), and 2663 responses (46.9%) using the other procedure (tube, 2437 of 2663 [91.5%]; gel, 226 of 2663 [8.5%]; Tables 2 and 3).
The Number of Laboratory Responses That Were Within 1 Titer of the Intended Anti-D and Anti-A Titer Reported by the Supplier Overall and When the Laboratory Used the Same Technique and Method

The Number of Laboratory Responses That Were Within 1 Titer of the Intended Anti-D and Anti-A Titer Reported by the Supplier Separated by Procedure, Technique, and Intended Titer

For the 10 evaluated anti-D PT samples, 3 had an intended titer of 32, 2 had a titer of 16, 2 had a titer of 128, and the remaining had titers of 8, 64, and 256, respectively. For anti-A, 2 had an intended titer of 32, 2 had a titer of 16, 2 had a titer of 128, and the remaining mailings had titers of 8, 64, 256, and 512, respectively. Overall, only 5874 anti-D (54.1%) and 3843 anti-A (63.4%) reported responses that were within 1 titer of the supplier’s intended result (Table 2). For anti-D, there was a higher percentage of responses within 1 titer when using the other procedure (P < .001) and when using a tube technique (P < .001; Table 3). Additionally, there were statistically significant performance differences (P < .001) based on the titer result (QC titer), with the lowest accuracy for the titers of 16 and 32 for both uniform and other procedure (Table 3, Supplemental Figures 1 and 2, see supplemental digital content, available at https://meridian.allenpress.com/aplm in the December 2023 table of contents). There were no other significant performance differences based on the technique, incubation temperature, or diluent used (data not shown, P = .29).
For anti-A, there were more responses within 1 titer when using the other procedure (P < .001) compared with the uniform procedure (P < .001). Tube testing gave more responses than gel testing within 1 titer when using the other procedure (P < .001), but tube and gel were equivalent when using the uniform procedure (P = .65; Table 3). There also were statistically significant performance differences based on the titer results, with the lower titers again generally performing worse, although the lowest accuracy rates differed slightly by procedure (other procedure, titers 8 and 64; uniform procedure, titers 8 and 128; Table 3 and Supplemental Figures 1 and 2). Room temperature incubation techniques tended to less closely reflect the supplier’s titer than testing with other techniques, that is, 37°C with anti-IgG or polyspecific AHG (data not shown, P = .03).
Given that it is well established that titer variability differs by test methodology,11 we further evaluated the accuracy of reported titers by comparing only those laboratories that used the same procedure and technique as the supplier. We identified that 4796 of 10 852 laboratory titer results (44.2%) used the same method for anti-D, and 1303 (22.9%) used the same method as the PT supplier for anti-A (Table 2). Overall, 3159 (65.9%) of anti-D and 991 (76.1%) of anti-A reported responses were within 1 titer of the vendor’s intended result when using the same methodology (P < .001; Table 2). For anti-D, there were statistically significant performance differences based on the titer specification (QC titer), with the lowest accuracy rates for the titers of 16 and 32 (P < .001; Table 4). The percentages of laboratories differing from the supplier’s intended titer by 2 or more dilutions ranged from 11.7% (most accurate: titer 256, 52 of 445) to 54.5% (least accurate: titer 16, 519 of 952; Figure 1). For anti-A, there also was a statistically significant performance difference based on the titer, with the lowest accuracy rates at titers of 8 and 64. The percentage of laboratories differing from the supplier’s intended titer by 2 or more dilutions ranged from 10.7% (most accurate: titer 16, 13 of 121) to 43.2% (least accurate: titer 64, 131 of 303). Immediate spin techniques tended to more closely reflect the supplier’s QC titer than testing at 37°C with anti-IgG AHG (P < .001).
The Number of Laboratory Responses That Were Within 1 Titer of the Intended Anti-A and Anti-D Result by the Supplier When the Laboratory and the Supplier Used the Same Technique, Diluent, and Method

The percent of laboratory responses that were identical to (light blue), and within 1 titer (dark blue), 2 titers (gray), 3 titers (dark red), and 4 titers (light red) from the anti-D (left) and anti-A (right) result reported by the supplier when the laboratory and the supplier used the same technique and procedure.
The percent of laboratory responses that were identical to (light blue), and within 1 titer (dark blue), 2 titers (gray), 3 titers (dark red), and 4 titers (light red) from the anti-D (left) and anti-A (right) result reported by the supplier when the laboratory and the supplier used the same technique and procedure.
Titer Performance Compared With the Mode Reported Value (Interlaboratory Comparison)
For anti-D titers, 10 852 total titer responses were evaluated, with 4885 responses (45.0%) using the uniform procedure (tube, 3517 of 4885 [72.0%]; gel, 1368 of 4885 [28.0%]), and 5967 responses (55.0%) using the other procedure (tube, 5644 of 5967 [94.6%]; gel, 323 of 5967 [5.4%]). For anti-A, 4004 total titer responses were evaluated for the mode analysis, with 3017 responses (75.3%) using the uniform procedure (tube, 2497 of 3017 [82.8%]; gel, 520 of 3017 [17.2%]), and 987 responses (24.7%) using the other procedure (tube, 918 of 987 [93.0%]; gel, 69 of 987 [7.0%]).
Overall, 3139 (of 4004; 78.4%) of all evaluable laboratory responses for anti-A fell within 1 titer of the mode value for each procedure and technique, and 9655 (of 10 852; 89.0%) of laboratory responses for anti-D fell within 1 titer of the mode value (P < .001). The procedure and technique again significantly impacted the proportions observed. For anti-D, laboratories that used the uniform procedure (4415 of 4885; 90.4%) had a greater proportion of labs falling within 1 titer of the mode than those that used the other procedure (5240 of 5967; 87.8%; P < .001). Gel technique users (1551 of 1691; 91.7%) also performed generally better than those who used tube techniques (8104 of 9161; 88.5%; P < .001). The method with the most laboratories reporting titers within 1 titer of the mode was the uniform procedure in gel (1266 of 1368; 92.5%), and the method with the least was a subset of those using the other technique using a tube with 22% albumin as the diluent (102 of 128; 79.7%). For anti-A titers, other procedure users (841 of 987; 85.2%) performed generally better than those that used the uniform procedure (2298 of 3017; 76.1%; P < .001). However, gel (469 of 589; 79.6%) and tube (2670 of 3415; 78.2%) technique users performed similarly (P = .43). For anti-A, the method with the most laboratories reporting titers within 1 titer of the mode value was the other procedure in gel (66 of 69; 95.7%), and the method with the least was those using the uniform procedure using a tube technique (1895 of 2497; 75.9%; Figure 2).
The percent of laboratory responses for anti-A and anti-D antibody titers that were within 1 titer (dark blue), 2 titers (gray), and 3 titers (dark red) of the mode value for each procedure and technique. Abbreviations: AHG, antihuman globulin; OG, other method, gel technique; OT, other method, tube technique; Poly, polyspecific antibody; UG, uniform method, gel technique; UT, uniform method, tube technique.
The percent of laboratory responses for anti-A and anti-D antibody titers that were within 1 titer (dark blue), 2 titers (gray), and 3 titers (dark red) of the mode value for each procedure and technique. Abbreviations: AHG, antihuman globulin; OG, other method, gel technique; OT, other method, tube technique; Poly, polyspecific antibody; UG, uniform method, gel technique; UT, uniform method, tube technique.
Anti-D Reproducibility Between 2018 A and B Mailings (Intralaboratory Comparison)
For the 439 laboratories using the uniform procedure, 434 (98.9%) used the same method for both mailings and were included in this intralaboratory comparison. For the 610 laboratories using the other procedure, 595 (97.5%) used the same method for both mailings, including dilution and titer technique. Nineteen results were excluded because of 1 or more missing practice characteristics or a low-frequency characteristic, so 576 laboratories (94.4%) were ultimately included in this analysis.
Between the A and B anti-D mailing, only 174 (40.1%) of those using the uniform procedure and 214 (37.2%) of those using the other procedure reported the same titer. More than 15% (67 of 434) of laboratories using the uniform procedure and more than 20% (118 of 576) of laboratories using the other procedure reported a titer difference of 2 or more (Figure 3). There was no significant difference in the paired results between the uniform and other procedure users (P = .26). That said, laboratories using the uniform procedure had a greater proportion of responses falling within 1 titer between the 2 mailings (uniform procedure, 367 of 434 [84.6%]; other procedure, 458 of 576 [79.5%]; P = .04). When evaluating laboratories using the uniform procedure, the technique was significantly associated (P < .001) with the result match rate. The 136 laboratories using a gel technique had statistically higher match rates (126 of 136; 92.6%) than ones using tube techniques (241 of 298; 80.9%). For those using the other procedure, none of the 3 differing test characteristics (technique, dilution, or type of AHG) were significantly associated with the match rate (Table 5).
Paired laboratory performance results for anti-D titers when provided with the same sample between A and B mailings overall (left) and by procedure (right).
Paired laboratory performance results for anti-D titers when provided with the same sample between A and B mailings overall (left) and by procedure (right).
DISCUSSION
To our knowledge, this study represents the largest and most comprehensive evaluation of antibody titer assay performance to date. This 5-year study of PT results demonstrated continued significant challenges with assay accuracy and reproducibility. We identified that only 54.1% of anti-D and 63.4% of anti-A reported responses were within 1 titer of a supplier’s QC result. When evaluating only those laboratories that used the same procedure and technique as the supplier, accuracy improved, but only by 11.8 percentage points (anti-D) and 12.7 percentage points (anti-A), respectively. When evaluating responses by using the mode value for each PT, nearly 22% of responses for an anti-A titer were greater than 2 or more titers from the reported mode value, and 11.4% of responses for an anti-D titer were greater than 2 or more titers from the mode value. Multiple factors appear to impact the results obtained, because performance differences emerged between the antibody titer result recorded, the procedure (uniform versus other), and the technique used (tube versus gel).
The antibody titer assay is relatively complex and operator dependent. Consistent with our own findings, significant variation in titer reporting between different laboratories has been documented previously.8,11,12 Reasons for this variability are well established. First, the grading of the strength of agglutination (ie, distinguishing between weak+ and 1+ agglutination) is done manually. Based on the varying expertise of individuals participating in these PT surveys, interpreting titration end points could be subject to variation (operator bias). Prior studies also have shown that even in a single laboratory, titer reports can substantially vary, giving support for operator-dependent bias as one of the factors contributing to variance.13 Our findings further support this hypothesis, because a sizable number of laboratories were not able to obtain a similar titer, even when challenged with two consecutive samples of the same titer. Second, the equipment (centrifuge machines, speed/duration of centrifugation, test tube size, diluents, etc) or key reagents (homozygous or heterozygous cells) used in the titer assay differ between institutions participating in PT surveys, thereby introducing additional bias. Findings from our PT review support the idea that the variation in titer results is likely to be multifactorial and involves a complex combination of operator, equipment, and procedure variability. As a consequence of this variability, procedures and techniques that aim to reduce this variability become critical.
Previous studies are contradictory regarding whether using the uniform procedure improves the variability seen with the antibody titer assay.8,11 In this large sample size, we identified that the uniform procedure did not help laboratories attain titers that reflected the intended supplier QC titer. This was anticipated, however, because the supplier did not use the uniform procedure, and thus differences in titer were to be expected. The uniform procedure did, however, appear to mildly but significantly increase test comparability for anti-D titers between laboratories when using the mode as the measure of interlaboratory variability. This observation was further confirmed by the paired anti-D titer analysis, because uniform procedure users had a significantly greater proportion of responses that were within 1 titer between the 2 separate PT mailings of the same titer. The procedure, however, did not appear to be as helpful for anti-A, and this observation has been reported previously.8,11
Titration using gel column technology is distinct in many ways from traditional tube-based techniques and may also help diminish assay variability. Similarly to previous studies,11 gel platforms remain less commonly used than tube techniques. This difference may continue to reflect the fact that, despite many advantages, such as increased automation, greater test sensitivity, reduced sample volume requirements, reduced subjectivity of titer strength grading, and faster turnaround time, using gel platforms is not without disadvantages.13–18 One key reason for its limited use may be from its known differences in titer strength and comparability with tube platform results. Estimates vary, but many comparative studies have demonstrated that gel platforms report multiple-fold higher titer results than tube-based methodologies.19–21 The current study supports these observations, because we generally confirmed both that gel platforms resulted in an improved intralaboratory and interlaboratory variability, especially for anti-D, but gel techniques generally did not have titers that were within 1 titer of our supplier QC, which used a tube technique. Another issue remains an absence of robust evidence linking gel titer levels with clinical outcomes.1,20 Well-designed future clinical studies adequately powered for statistical and clinical relevance will be needed if gel-based platforms are to become more common.14
The finding that interlaboratory performance differed by titer, with more laboratories being within 1 titer of the expected value when evaluating higher titers in comparison with lower titers, was a novel but understandable observation. Titer results are determined by the dilution factor at which an antibody is no longer is able to agglutinate test red cells. Given that the dilutions used in the assay become exponentially more dilute, it would logically make sense that laboratories would have greater difficulty with result consistency at lower titers, where the differences in dilution are relatively smaller (1:4 versus 1:8, for example), than at higher titers, where each fold dilution is proportionally much greater (ie, 1:1024 versus 1:2048).
Unfortunately, the most clinically useful titers are generally low. The American College of Obstetricians and Gynecologists suggests using an anti-D titer of 8 as the critical titer for pregnant females, where a result above that titer or a titer increase greater than 1 dilution would require increased monitoring and fetal testing.2,22,23 Regarding transfusion safety for incompatible blood products, avoiding products with critical titers of 100 or greater for group A plasma,24 50 or greater for group O whole blood,4 and 50 to 250 or greater for ABO-incompatible platelets3 has been proposed as a useful method for limiting the risk of hemolytic reactions in patients. For ABO-incompatible solid organ transplants, a titer threshold of 8 has been suggested to determine candidate eligibility,7 and the use of plasma exchange desensitization has been guided by the initial antibody titer.6 For ABO-incompatible bone marrow transplants, titers of 16 to 32 have been suggested as thresholds for additional graft processing or therapeutic apheresis.25–27 The titer methods used to derive these clinical practice guidelines are generally not discussed, leading to the concern that universally applied titer cutoffs may be inappropriate. Given the noted variability observed by this current study, it is highly possible that some patients receive treatments that they should not have, and, perhaps of greater concern, some do not receive treatments that they should have. To this point, Stussi et al28 identified that when titer reduction techniques were used for their ABO-incompatible bone marrow transplants, 45% of their cohort continued to show evidence of hemolysis after transplantation, and when titer reduction techniques were not used, 31% still had some evidence of hemolysis. Although the cause of these findings cannot be determined with certainty, it is possible that titer assay variability may have played some role in these reported outcomes.
Novel techniques and methods are going to be required to improve the accuracy and reliability of the red blood cell (RBC) antibody titer assay. One possibility would be the development of a nationally accepted calibration standard. In the United Kingdom, national guidelines recommend quantifying anti-D and anti-c in obstetric patients as International Units (IU)/mL in comparison with validated standards available from the National Institute for Biological Standards and Control, and other antibodies should be titered in parallel with the anti-D standard as an internal control to improve result reproducibility.2 If such standards were widely used and available for commonly tested RBC antibodies, intralaboratory and interlaboratory variability might improve. Another possibility would be the increased development and use of quantitative titer methodologies, such as flow cytometry. Flow-based techniques have improved the reliability of fetal hemoglobin detection assays, which has itself been challenged with the variability associated with an operator-dependent, high-complexity assay (Kleihauer-Betke test).29,30 In early studies, application of flow cytometry for RBC antibody titers has been shown to accurately measure IgM and IgG, with relatively comparable results and with perhaps a higher sensitivity, compared with those of tube and gel techniques.31–35 Fully automated titer platforms are also increasing in availability, and these platforms may also improve test reliability by reducing operator bias. Lastly, improved scoring methods may reduce interlaboratory variability. Standardized methods have been published to report titers as a score, which incorporates agglutination strength across the dilutions, rather than simply reporting the last sample dilution that had agglutination.36 Very few laboratories report a titer as a score, but with a greater range of possible values, interlaboratory variability may be muted.
This study had some limitations that are worth noting. First, there remains no gold standard assay to determine antibody titers for anti-A or anti-D, and so using the supplier’s QC titer as the “correct” value from which to compare all other laboratories is not ideal. That said, given that we identified comparable results when evaluating only those laboratories whose procedure and technique matched those of the supplier, we hold that the observations made using the larger cohort demonstrate the variability of the titer assay well. Second, our evaluation of intralaboratory variability was imperfect. Although we endeavored to create a perfectly identical sample between the 2 sample mailings, this was not possible because of the volume of sample required for the PT, and the changes expected with these samples over storage time. It is thus possible that factors outside of our control could have confounded the results that we obtained. Still, we hold that this evaluation has reasonable validity, because the intended titer was the same between mailings as confirmed by the supplier QC, and laboratories were only compared with themselves in the paired analysis.
CONCLUSIONS
This study reflects the current state of the antibody titer assay across different laboratories. Standardization of antibody titration techniques aimed at improving interlaboratory and intralaboratory variability has not occurred, and this continues to be a concerning issue because of the clinical expectations for titer results as accurate, reproducible, and actionable values. Tube-based titer assays continue to be the most commonly used technique, despite their challenges, and despite suggestions that the gel technique and the uniform procedure may be helpful in reducing assay variability. The role of this assay in clinical decision-making should be critically reevaluated, and future research should focus on incorporating novel emerging technologies that minimize the established operator- and instrument-specific biases noted with our current methods.
References
Author notes
Supplemental digital content is available for this article at https://meridian.allenpress.com/aplm in the December 2023 table of contents.
The authors have no relevant financial interest in the products or companies described in this article.
All authors are current or past members for the College of American Pathologists Transfusion, Apheresis, and Cellular Therapy Committee. Dvorak and Souers are employees of the College of American Pathologists.