Augmentation index in the assessment of wave reflections and systolic loading.

Background: Augmentation index (AI) is used to quantify the augmented systolic aortic pressure that impedes ventricular ejection. Its use as an index of wave reflections is questionable. We hypothesize that AI is quantitatively different from the reflection coefficient under varied physiological conditions.

Methods: 42 datasets of aortic pressure and flow waveforms were obtained during induced hypertension (methoxamine infusion) and vasodilation (nitroprusside infusion) in our mongrel dog experiments (n = 5) and from Mendeley data during various interventions (vasoconstrictors, vasodilators, pacing, stimulation, hemorrhage and hemodilution). Wave reflections and principal components of reflection coefficients were computed for comparison to AI and heart rate normalized AI RESULTS: Principal reflection coefficient, Γ, increased in hypertension and decreased in vasodilation, hemorrhage and hemodilution. AI followed the trend in many cases but was consistently lower than Γ in almost all the subjects. The Bland-Altman analysis also showed that both AI and normalized AI underestimated Γ. The relationship between augmentation index and reflection coefficient was explained by a linear regression model (r = 0.23, p < 0.01) in which AI followed directional changes in Γ and the normalization of AI resulted in a linear model that explained less variation in the relationship between AI and Γ.

Conclusion: AI is a reasonable clinical trend indicator, albeit not an accurate surrogate measure of the amount of wave reflections.