Research Article: Good agreement of conventional and gel-based direct agglutination test in immune-mediated haemolytic anaemia

Date Published: February 8, 2012

Publisher: BioMed Central

Author(s): Christine J Piek, Erik Teske, Martin W van Leeuwen, Michael J Day.


The aim of this study was to compare a gel-based test with the traditional direct agglutination test (DAT) for the diagnosis of immune-mediated haemolytic anaemia (IMHA).

Canine (n = 247) and feline (n = 74) blood samples were submitted for DAT testing to two laboratories. A subset of canine samples was categorized as having idiopathic IMHA, secondary IMHA, or no IMHA.

The kappa values for agreement between the tests were in one laboratory 0.86 for canine and 0.58 for feline samples, and in the other 0.48 for canine samples. The lower agreement in the second laboratory was caused by a high number of positive canine DATs for which the gel test was negative. This group included significantly more dogs with secondary IMHA.

The gel test might be used as a screening test for idiopathic IMHA and is less often positive in secondary IMHA than the DAT.

Partial Text

Immune-mediated haemolytic anaemia (IMHA) is caused by the binding of antibodies to the surface of red blood cells (RBCs). The production of such antibodies can be a primary autoimmune phenomenon or be associated with underlying neoplasia, chronic infections, inflammatory disease or be triggered by exposure to drugs or vaccines (secondary IMHA) [1]. The criteria used to define IMHA in dogs and cats vary between different studies; however, it is generally accepted that a positive direct agglutination test (DAT), marked spherocytosis or true autoagglutination are three hallmarks of canine IMHA. At least one of these changes must be present in a patient with haemolytic anaemia to warrant a diagnosis of IMHA [2].

The DAT is an accepted method of detection of anti-erythrocyte antibodies and an essential part of the diagnosis of IMHA [2]. The first aim of this study was to compare the results obtained in the DAT versus the gel test. The high number of samples, 247 canine and 74 feline samples, creates a relatively powerful study, as for such comparative investigations 40 samples has been described as sufficient [8]. Analysis of the correlation between the log titres of the DAT results and the ordinal gel test results revealed that the latter can be interpreted as binary test results. Therefore kappa statistics were chosen as the appropriate method to compare the two tests in the absence of a “gold standard” for the diagnosis of IMHA [8,9]. The kappa values for the comparison of the canine BCIDL (0.86) as well as the kappa of 0.94 for the BCIDL canine samples with Ht below 0.35% can be interpreted as “almost perfect agreement” and the kappa for the feline BCIDL samples (0.58) and kappa for the UVLD canine samples (0.48) as “moderate agreement” using the classification of Landis and Koch [9].

Overall, in the present study there was a good agreement between the results of the DAT and the gel test. Additionally, the gel test performed well in identifying dogs with idiopathic IMHA, but missed most of the samples that were from dogs with secondary IMHA. From a laboratory viewpoint these samples also differed in having lower titres in the DAT, more frequent IgM positivity and almost no spherocytosis.

(DAT): Direct Agglutionation Test; (IMHA): immune-mediated haemolytic anaemia; (RBC’s): Red Blood Cells; (UVDL): Utrecht University Veterinary Diagnostic Laboratory; (BCIDL): Bristol Clinical Immunology Diagnostic laboratory; (PBS): Phosphate Buffered Saline; (WBC): White Blood Cell Count.

The gel kits used in this study were kindly provided by Diamed Benelux BV.

ML carried out the gel tests and the DATs in the UVDL. CP carried out the statistical analysis. CP, MD, and ET participated in the design of the study and helped to draft the manuscript. All authors read and approved the final manuscript.