Abstract
Bacterial endocarditis (BE) is the most common valvular disease in cattle but diagnosis in the living animal remains a challenge for clinicians. The objective of the study was to report evidence-based veterinary medicine data concerning the clinical presentation and results of ancillary tests of necropsy-confirmed cases of bovine BE. A systematic review and subsequent meta-analysis was performed using Medline and CAB abstracts of every article on bovine BE published in English, Japanese, German and French. The clinical criteria that were specifically assessed for diagnosis of BE were: tachycardia, heart murmur, signs of congestive heart failure, presence of fever, evidence of lameness/polyarthritis, one or more positive blood cultures and positive echocardiograms for BE. A total of 34 studies (460 cases of BE) satisfied the inclusion criteria for the systematic review. The sensitivity (Se), specificity (Sp) and 95% confidence interval (CI) were obtained using a random-effect meta-analysis for studies reporting five or more cases.
The Se (95% CI) were 86.9% (39.1–98.6%) for positive haemoculture, 84.3% (60.4–95.0%) for echocardiography, 79.7% (70.1–86.8%) for the presence of tachycardia, 60.3% (51.8–68.3%) for the presence of a murmur, 45.7% (32.5–59.5%) for the presence of fever, 43.5% (25.6–63.3%) for the presence of lameness/polyarthritis, and 37.3% (21.6–57.0%) for the presence of clinical signs of heart failure. The Sp (95% CI) was 95.3% (93.3–96.8%) for lameness, 72.6% (45.8–89.2%) for the presence of a murmur, 67.0% (55.5–76.7%) for the presence of fever, and 27.1% (14.3–45.2%) for the presence of tachycardia.
This meta-analysis confirmed that the diagnosis of BE is a difficult process. Echocardiography seems to be a sensitive diagnostic tool despite the absence of any consensus on the ultrasonographic definition of the disease. However, from these results, it was impossible to determine whether multiple positive findings or clinical tests increase the sensitivity for the diagnosis of bovine BE.