Online Program Home
My Program

Abstract Details

Activity Number: 538
Type: Contributed
Date/Time: Wednesday, August 3, 2016 : 10:30 AM to 12:20 PM
Sponsor: Biometrics Section
Abstract #318551 View Presentation
Title: A Latent-Variable Approach to Derive a Pediatric Cardiac Inotrope Score Associated with Congenital Heart Surgery
Author(s): Mallikarjuna Rettiganti* and Avishek Chakraborty and Punkaj Gupta
Companies: University of Arkansas for Medical Sciences and University of Arkansas and University of Arkansas for Medical Sciences
Keywords: latent variable ; multivariate normal ; bayesian

In the current literature, there are multiple inotrope scores available to quantify the amount of cardiovascular support needed for children with critical illness. However, none of these scores are scientifically derived and are extrapolated from clinical practice. The aim of the present study was a model-based derivation of an inotrope score in infants undergoing cardiac surgery. We used multivariable logistic regression models to evaluate association of seven existing inotrope scores with composite poor outcome (either mortality or prolonged length of stay). We developed a new score within a hierarchical framework by representing probability models using continuous latent variables that depended on the drug dosage employed on a particular patient. We assumed a multivariate normal prior distribution for the effects of the four inotropes (epinephrine, dopamine, norepinephrine, and milrinone) on the mean of the latent variables and used Markov Chain Monte Carlo (MCMC) simulations from the resulting posterior distribution to create a score function for the outcome. The derived score has discriminative ability in predicting the outcome in children undergoing heart surgery.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association