2015 ACVIM Forum Research Abstract Program

Overall, 18 phyla were identified in all horses and an average of 19.9% 16S sequences remained unclassified. Diversity within subjects (alpha diversity) was not significantly altered by treatment or over time. However, community structure between samples changed with most differences observed in Proteobacteria, Firmicutes, Bacteroidetes and Verrucomicrobia (P < 0.05). The treatment duration (5 days) was sufficient to identify changes in fecal microbiota, and changes in community structure were frequently specific to the antibiotic. Ceftiofur and enrofloxacin treatment resulted in a >90% reduction in Proteobacteria compared to other treatments. Ceftiofur and oxytetracycline treatment led to the most significant changes in major phyla over time. Within treatment groups, several changes in phylum structure persisted at 30 days, however some returned to near baseline. There is scarce literature regarding the accuracy of either transabdominal ultrasound or transrectal abdominal palpation compared to surgical diagnosis in cases of equine acute abdominal pain. The aims of this study were to evaluate the level of agreement between 1) a modified FLASH protocol (mFLASH, thoracic and left middle third windows excluded) and the surgical diagnosis and 2) transrectal abdominal palpation and the surgical diagnosis. Specific diagnoses of interest included nephrosplenic entrapment, right dorsal displacement, large colon volvulus, and surgical small intestinal lesion. Sequential cases of client-owned horses presenting for evaluation of acute abdominal pain were enrolled in the study. Only horses that were ≥1 year of age and large enough for transrectal abdominal palpation were included. Transrectal abdominal palpation was performed first by the attending clinician and the descriptive findings and tentative diagnosis were recorded in a standardized recording sheet. An mFLASH was then completed and descriptive findings and ultrasound diagnosis recorded. Horses that underwent exploratory celiotomy within 6 hours of evaluation were included in the statistical analysis. Kappa two-way agreement comparisons were performed to assess agreement. Sixty-one cases fit the inclusion criteria. Results showed that mFLASH was in moderate agreement with the surgical diagnosis (Kappa = 0.42, P < 0.001), while transrectal abdominal palpation was in slight agreement with the surgical diagnosis (Kappa = 0.18, P = 0.005). In conclusion, mFLASH is a more accurate tool than transrectal abdominal palpation in predicting specific surgical diagnoses in cases of equine acute abdomen.