RESUMO
OBJECTIVE: To estimate likelihood ratios (LRs) of correctly identifying internal Corynebacterium pseudotuberculosis infection in horses by measurement of antibody titers via serum synergistic hemolysis inhibition (SHI) testing. DESIGN: Retrospective case-control study. ANIMALS: 170 horses (171 records; 92 cases of C pseudotuberculosis infection and 79 controls). PROCEDURES: Medical records were reviewed, and horses were grouped on the basis of evidence of internal or external C pseudotuberculosis infection. The LRs and 95% confidence intervals for identification of internal C pseudotuberculosis infection by use of SHI test results were estimated. RESULTS: LRs for C pseudotuberculosis infection increased as antibody titers increased when all horses were included in analyses; LRs for detecting internal infection were significantly > 1 (null value) for reciprocal antibody titers ≥ 1,280 overall and > 160 when horses with external abscesses were excluded. Likelihood ratios for detecting internal infection did not differ from 1 (indicating no change in pretest-to-posttest odds of internal infection) when only horses with external C pseudotuberculosis infection (horses with external and internal abscesses vs those with external abscesses only) were included. The LR for detecting internal infection was 2.98 (95% confidence interval, 2.19 to 4.05) for horses with titers ≥ 512. CONCLUSIONS AND CLINICAL RELEVANCE: In the study population, higher titers were typically more indicative of active external or internal C pseudotuberculosis infection than of internal disease specifically. The SHI test was not a useful predictor of internal C pseudotuberculosis infection in horses with external abscesses but was useful in the absence of external disease.
