Challenges in Modeling and Interpreting Multivariable Linear Relationships From Observational Data of Cardiac Function Utilizing Speckle Tracking Echocardiography

Abstract

Cardiac function is a complex process involving several imaging parameters for its assessment. Observational studies (longitudinal, cross-sectional, or case control) often use cardiac imaging to study cardiac diseases and their relationship with cardiac function, in particular the speckle-tracking echocardiography. The analysis of these data can be someway challenging, to avoid some bias in the estimation of linear relationships between the outcome (independent variable) and imaging parameters (dependent variables). Variables selection, heterogeneity between outcomes and dependent variables, and the interactions between cardiac parameters are all common difficulties to effort. In this case study, we present a study aimed at testing the differences in blood flow momentum and kinetic energy dissipation in a model of cardiac dyssynchrony induced by electrical right ventricular apical stimulation (pacemakers), compared with spontaneous sinus rhythm. Therefore, we, first, discuss how cardiac function can be echocardiographically assessed utilizing the speckle-tracking analysis. We, second, discuss linear and nonlinear relationships within outcomes and dependent variables—how to model observational data and how to assess interactions. Then, we lay out the problems of drawing causal inferences from correlations by illustrating the counterfactual approach to causality. Besides, we describe the application of these models by examining how right ventricular apical pacing may affect the left ventricular function.

locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles