Diagnostic performance of an in-clinic cortisol assay in dogs with suspected hypoadrenocorticism

Abstract

Objective: To assess the diagnostic accuracy of the rapid point-of-care ELISA by IDEXX (SNAP) for detecting hypoadrenocorticism (HA) in dogs. A secondary aim was to evaluate the agreement between the SNAP and the veterinary reference laboratory (VRL) using chemiluminescent enzyme immunoassay (IMMULITE 2000) cortisol measurements. We hypothesized that the SNAP would correlate well with the VRL and aid in HA diagnosis. Methods: A prospective observational diagnostic accuracy study was conducted from January 2023 through January 2024, enrolling dogs with suspected HA. Hypoadrenocorticism was diagnosed using an ACTH stimulation test, with basal serum cortisol (BSC) used for screening in some cases. Samples were analyzed by the SNAP and VRL, with the VRL serving as the gold standard. Results: 96 dogs were enrolled; 21 were diagnosed with HA. The SNAP showed correlation with the VRL (Spearman ρ = 0.923; P < .001). Median BSC and post-ACTH cortisol concentrations did not differ significantly between methods. A BSC < 2 µg/dL yielded 100% sensitivity for both, with the SNAP showing higher specificity (73.1%) than the VRL (64.2%). Post-ACTH cortisol < 2 µg/dL was 100% specific by both methods, with 95.2% sensitivity using the SNAP. A 10.5% clinical discordance was observed. Passing-Bablok regression revealed proportional and constant bias, with the SNAP test tending to overestimate cortisol concentrations, but in the low cortisol range this overestimation appears minimal. Conclusions: The SNAP correlated with the VRL for HA diagnosis; however, borderline post-ACTH SNAP results should be confirmed with the VRL. The SNAP-derived BSC may be more specific than the VRL for ruling out HA. Clinical relevance: The SNAP test aids HA diagnosis in urgent care.