Canada Spends Logo

Explore our Budget 2026 analysis

Grant: Fellowship - Priority Announcement - HIV/AIDS and/or STBBI

Université de Montréal

$135,000.00

Evaluation of the predictive power of vascular elastography for screening of cardiovascular diseases among people living with HIV

FY 2020-21

Summary
Backgound. Cardiovascular diseases (CVD) disproportionately affect people living with HIV (PLHIV). Traditional CVD risk calculators have been shown to underestimate CVD risk among PLHIV, while imaging technologies are only moderately associated to HIV-related CVD risk, or too expensive and invasive to be performed on all PLHIV. Vascular elastography is an innovative, non-invasive, and inexpensive method that measures the rigidification of the carotid arteries occurring before any measurable thickening of the intima-media can be seen (subclinical CVD). However, it is unknown if elastography has incremental value to predict CVD in addition to traditional risk factors and intima-media thickness. Objectives. To estimate the incremental value of elastography to predict subclinical CVD and establish its value for screening of CVD among PLHIV. Methods. Using data from the prospective Canadian HIV and Aging Cohort Study (CHACS), I will sequentially build a series of 5 predictive models of subclinical CVD (defined as total coronary plaque volume > 0 as measured by coronary computed tomography angiography) including different sets of predictive variables. Model 1 will include traditional risk factors; for model 2, HIV-specific predictors will be added to model 1; for model 3, intima-media thickness will be added to model 2; for model 4, elastography parameters will be added to model 3; and model 5 will include elastography parameters only. Using a class random forests machine learning method, I will then estimate and compare each model's prediction capacity to correctly classify participants as having subclinical CVD. Using regression random forests models, I will also assess and compare the models' capacity to predict the total coronary plaque volume (continuous outcome). If my results suggest that elastography has incremental value in predicting CVD, they could rapidly be translated into practice for the benefits of PLHIV who are disproportionately affected by CVD.
Program
Fellowship - Priority Announcement - HIV/AIDS and/or STBBI
Location
Quebec, Canada
Recipient Type
Academia
Type
Grant
Source
CIHR Funding Decisions Database

As published

Id
446290
Name
Giguère Katia
Region
Québec
Country
Canada
Orgtype
Institution Paid; Research Institution
Instname2
Population and Public Health
Theme2
Biomedical
Prcname2
Fellowships - Post-PhD
Projecttitle
Evaluation of the predictive power of vascular elastography for screening of cardiovascular diseases among people living with HIV
Programtype2
Fellowship Programs
Orgnameinp2
Université de Montréal
Cihramount2
$135,000
Cihrequipment2
$0
Programname2
Fellowship - Priority Announcement - HIV/AIDS and/or STBBI
Keyworddelim
Aging; Artificial Intelligence; Cardiovascular Diseases (Cvd); Cvd Predictors; Hiv; Hiv Comorbidities; Machine Learning; People Living With Hiv; Screening Tool; Vascular Elastography
Abstract
Backgound. Cardiovascular diseases (CVD) disproportionately affect people living with HIV (PLHIV). Traditional CVD risk calculators have been shown to underestimate CVD risk among PLHIV, while imaging technologies are only moderately associated to HIV-related CVD risk, or too expensive and invasive to be performed on all PLHIV. Vascular elastography is an innovative, non-invasive, and inexpensive method that measures the rigidification of the carotid arteries occurring before any measurable thickening of the intima-media can be seen (subclinical CVD). However, it is unknown if elastography has incremental value to predict CVD in addition to traditional risk factors and intima-media thickness. Objectives. To estimate the incremental value of elastography to predict subclinical CVD and establish its value for screening of CVD among PLHIV. Methods. Using data from the prospective Canadian HIV and Aging Cohort Study (CHACS), I will sequentially build a series of 5 predictive models of subclinical CVD (defined as total coronary plaque volume > 0 as measured by coronary computed tomography angiography) including different sets of predictive variables. Model 1 will include traditional risk factors; for model 2, HIV-specific predictors will be added to model 1; for model 3, intima-media thickness will be added to model 2; for model 4, elastography parameters will be added to model 3; and model 5 will include elastography parameters only. Using a class random forests machine learning method, I will then estimate and compare each model's prediction capacity to correctly classify participants as having subclinical CVD. Using regression random forests models, I will also assess and compare the models' capacity to predict the total coronary plaque volume (continuous outcome). If my results suggest that elastography has incremental value in predicting CVD, they could rapidly be translated into practice for the benefits of PLHIV who are disproportionately affected by CVD.
Competitiondate
202012
Approvedterm2
3 yrs 0 mth
Pinamesdelim
Giguère, Katia
Approvedterm
3 ans 0 mois
Cihrcontribution2
~ $135,000~ $0~
Université de Montréal - $135,000 Grant (2020-21) | Canada Spends