2015 ACVIM Forum Research Abstract Program

Horses presenting to the LSU-VTH with acute colic from 8-20-12 to 7-13-13 were studied. Clinical parameters including age, sex, breed, and rectal exam (normal or abnormal), frequency of abdominal pain (none, intermittent, or continuous), peripheral pulse character (normal or abnormal) and frequency of abdominal sounds (normal, decreased, absent, increased) were used to validate the surgical model; whereas peripheral pulse character, packed cell volume, heart rate, capillary refill time and medical vs. surgical treatment were used to validate the survival (prognosis) model. Horses that were euthanatized due to financial constraints were excluded from the survival model. Survival was defined as discharged from the VTH. Bayes theorem was utilized that calculated the post-test probability of both surgery and survival. The hosmer-Lemeshow goodness of fit chi square (GOFCS formula was used to assess if each model fit the data. A chi square value of >15.51 using 8 degrees of freedom was considered significant if it yielded a P < 0.05.