Tuesday, October 21, 2014
Poster Board # PS3-3

Jorge G. Ruiz, MD1, Allen Andrade, MD2, Ramankumar Anam, MD2, Chandana Karanam, MBBS2, Dhurga Krishnamurthy, MD2 and Joseph Sharit, PhD3, (1)University of Miami Miller School of Medicine, Miami, FL, (2)Laboratory of E-learning and Multimedia Research, Bruce W. Carter VA GRECC, Miami, FL, (3)University of Miami College of Engineering, Coral Gables, FL
Purpose: Communicating patients their numerical estimates of cardiovascular risk (CVR) can encourage risk reduction actions. Virtual humans (avatars) may enhance the ability of the CVR reduction message to improve persuasion and contribute to persistent adherence to healthy behaviors. The purpose of this study is to compare the efficacy of a computer-based aid communicating global CVR with or without animated avatars for improving intention to adhere to lifestyle changes. 

Method: Male veterans, age greater than twenty, non-depressed, cognitively intact with an intermediate to high CVR were recruited into a two-arm randomized controlled trial. Participants viewed a multimedia computer-based aid with information regarding CVR followed by the presentation of their individual CVR. The aid was presented in two versions:  1) Animated Avatars with Voice: An animated avatar delivered the information including CVR to participants using voice synchronized with text bullets; 2) Voice Alone: In this condition, the participants viewed an identical version of the aid without an avatar. We collected demographic data, comorbidities, health literacy, numeracy, lifestyle behaviors, familiarity with heart disease, attitudes towards the computer program, and risk understanding and recall.

Result: Forty-one participants completed the study (animated avatar, n=19; voice alone, n=22,). Participants were generally older (M=64, SD=7), African American 46%, White 52%. There were no baseline differences in demographics, comorbidities, health literacy, numeracy, lifestyle, familiarity with heart disease, and risk understanding and recall. Two tailed independent samples t test of intent to adhere to lifestyle change for the next 6 weeks showed a significant effect favoring the animated avatar version (M=12.63, SD=2.29) over voice alone (M=9.0, SD=4.29) demonstrating a large effect size (Cohen d=1.05). There were no differences in self-efficacy to change behavior over the next 10 years. An ANCOVA was run to determine the effect of the two different computer aids on post-risk perception and emotional response after controlling for pre-intervention scores, which showed no difference for either risk perception or emotional response.

Conclusion: A computer-based aid delivering individualized CVR significantly increase participants’ intention to adhere to lifestyle changes as compared with the same aid without avatars.