Relationship between ejection fraction and shortening fraction in dogs, and diagnostic accuracy of shortening fraction in identifying normal and abnormal ventricular contraction.

Abstract

OBJECTIVE: To examine the relationship between shortening fraction and ejection fraction, and to calculate the diagnostic performance of shortening fraction in identifying normal, subnormal and supranormal ejection fractions in dogs. STUDY DESIGN: Reanalysis of prospective data. ANIMAL POPULATION: A group of 224 dogs; 86 healthy, 96 with mitral valve disease, and 42 with other cardiac pathologies. METHODS: Echocardiographic cineloops of right parasternal short- and long-axis (RPSA and RPLA) and left apical four-chamber (LA4C) views were analyzed by five experienced echocardiographers. Ejection fraction was measured using the Simpson's method of discs from RPLA and LA4C, and shortening fraction from RPSA with M-mode. Sensitivity, specificity, and positive and negative predictive values of shortening fraction for identifying a low, normal, or high ejection fraction for each volumetric method, and the diagnostic performance of the LA4C volumetric method against the RPLA volumetric method, were calculated. RESULTS: Sensitivity and specificity to detect low ejection fraction were 92% and 93% (RPLA) and 71% and 91% (LA4C). The positive and negative predictive values for low ejection fraction were 55% and 99% (RPLA) and 46% and 97% (LA4C). CONCLUSIONS AND CLINICAL RELEVANCE: Shortening fraction was a robust indicator of ejection fraction in dogs, with high sensitivity and specificity for the detection of decreased left ventricular systolic function. Given its simplicity and despite its limitations, shortening fraction might be a valuable tool during perianesthetic examinations of dogs.