Grant: DRA : Canada Graduate Research Scholarship - Doctoral Program (CGRS D)
University of Calgary
$120,000.00
Developing innovative methods to account for population heterogeneity in longitudinal patient-reported health outcomes
FY 2025-26
Summary
Patient-reported outcome measures (PROMs) are instruments that allow patients to describe their health, including symptoms, quality of life, and functional abilities. These measures are essential tools for healthcare providers to understand treatment effectiveness or changes in patients' symptoms from the patient's perspective and make informed care decisions. However, patients' interpretations of health questions often change over time due to psychological adaptation and evolving priorities. This variation introduces systematic measurement bias, leading to inaccurate conclusions about treatment effectiveness and suboptimal care decisions. Conventional methods for analyzing PROM data fail to account for heterogeneity across patient subgroups, assuming all patients interpret and respond to questions similarly over time. These assumptions often lead to biased results and reduced ability to detect meaningful treatment effects in clinical studies. This research aims to develop innovative statistical methods that identify and account for subgroup differences when analyzing PROM data collected over multiple time points. Specifically, this research will integrate tree-based machine learning algorithms with item response theory models-statistical models that account for how individual questions contribute to overall health measurement-to identify patient subgroups with distinct patterns of response over time. Using computer simulations and real-world data from over 300,000 heart disease patients in Alberta, this research will develop and test these novel approaches. The expected outcomes include new statistical methods and open-source software that enable health researchers and decision-makers to analyze PROM data more accurately. This research will improve the estimation of treatment effectiveness in clinical studies and enhance personalized care delivery by helping healthcare providers better interpret patient-reported information to support shared decision-making.
Department
Program
DRA : Canada Graduate Research Scholarship - Doctoral Program (CGRS D)
Recipient
Location
Alberta, Canada
Recipient Type
Academia
Type
Grant
Source
CIHR Funding Decisions Database
As published
Id
559837
Name
Arimoro Olayinka I
Region
Alberta
Country
Canada
Orgtype
Institution Paid; Research Institution
Instname2
Health Services and Policy Research
Theme2
Clinical
Prcname2
DT2 (Health services and population health research) CGRS Doctoral
Projecttitle
Developing innovative methods to account for population heterogeneity in longitudinal patient-reported health outcomes
Programtype2
Studentship Programs
Orgnameinp2
University of Calgary
Cihramount2
$120,000
Cihrequipment2
$0
Programname2
DRA : Canada Graduate Research Scholarship - Doctoral Program (CGRS D)
Keyworddelim
Clinical Registries; Computer Simulation; Item-Response Theory Model; Longitudinal Data Analysis; Machine Learning; Measurement Bias; Patient-Reported Outcome Measures; Repeated Measures
Abstract
Patient-reported outcome measures (PROMs) are instruments that allow patients to describe their health, including symptoms, quality of life, and functional abilities. These measures are essential tools for healthcare providers to understand treatment effectiveness or changes in patients' symptoms from the patient's perspective and make informed care decisions. However, patients' interpretations of health questions often change over time due to psychological adaptation and evolving priorities. This variation introduces systematic measurement bias, leading to inaccurate conclusions about treatment effectiveness and suboptimal care decisions. Conventional methods for analyzing PROM data fail to account for heterogeneity across patient subgroups, assuming all patients interpret and respond to questions similarly over time. These assumptions often lead to biased results and reduced ability to detect meaningful treatment effects in clinical studies. This research aims to develop innovative statistical methods that identify and account for subgroup differences when analyzing PROM data collected over multiple time points. Specifically, this research will integrate tree-based machine learning algorithms with item response theory models-statistical models that account for how individual questions contribute to overall health measurement-to identify patient subgroups with distinct patterns of response over time. Using computer simulations and real-world data from over 300,000 heart disease patients in Alberta, this research will develop and test these novel approaches. The expected outcomes include new statistical methods and open-source software that enable health researchers and decision-makers to analyze PROM data more accurately. This research will improve the estimation of treatment effectiveness in clinical studies and enhance personalized care delivery by helping healthcare providers better interpret patient-reported information to support shared decision-making.
Competitiondate
202511
Approvedterm2
3 yrs 0 mth
Pinamesdelim
Arimoro, Olayinka I
Approvedterm
3 ans 0 mois
Cihrcontribution2
~ $120,000~ $0~