Consistency assessment of a canine heartworm point-of-care antigen test using fresh whole blood and archived sera

Dirofilariosis, caused by infection with Dirofilaria immitis, relies on point-of-care (POC) testing as a rapid and affordable diagnostic method. Considerable investigation into the diagnostic accuracy of heartworm POC tests has been conducted, however, there are no available reports for the precision of those tests, typically investigated under repeatability and reproducibility conditions. The repeatability of one canine heartworm POC test was estimated by comparing the agreement between results of two test runs of the same 63 whole blood specimens. Its reproducibility was estimated by comparing the results of an extended run of 331 fresh blood specimens with a run of their archived sera after up to seven months of frozen (-20˚C) storage. Results were analysed using Bowker's symmetry test, and exact binomial confidence intervals of proportional agreement were calculated. We observed perfect test repeatability with fresh blood (100%, 95%CI: 94.3-100.0%). Reproducibility assessment of a binary test result revealed marginal non-significant imperfections (96.1%; 95%CI: 93.4%-97.9%; p = 0.41). However, POC tests were observed to return less intense test bands after freezing and storing sera, and a three-scale ordinal classification system revealed significant disagreement between fresh and archived specimens (p = 0.008). The observed imperfect reproducibility suggests that the relative diagnostic sensitivity of heartworm POC tests may have been underestimated in reports utilising archived sera.