Abstract Number:
2846
Submission Type:
Contributed Abstract
Contributed Abstract Type:
Paper
Participants:
Ruilie Cai (1), Jiajia Zhang (1), Wenbin Lu (2), Pulak Ghosh (3)
Institutions:
(1) University of South Carolina, Columbia, SC, (2) North Carolina State University, Raleigh, NC, (3) Indian Institute of Management, Bangalore, India
Co-Author(s):
First Author:
Presenting Author:
Abstract Text:
Motivated by the hospitalized COVID-19 patient data from West Bengal in India, we aim to dynamically predict the chance of discharge or death using the landmark competing risk model based on longitudinal observations of laboratory measurements and medication usages. The data included 147805 patients with 1091322 observations. To address challenges in dynamic prediction such as the uncertainty in longitudinal functional form and computational difficulties in jointly estimates, we propose a two-step approach: initially extracting functional principal components from laboratory measurements, followed by fitting the landmark competing risk model. We evaluated our approach against conventional methods which handle the longitudinal observations via baseline, last value carry forward, mean, and linear regression. The proposed method outperformed others regarding to weighted Harrell's C-Index and Brier score. The proposed model could dynamically predict competing risks and depict the association between them and COVID-19 medication. This can assist clinicians in understanding patient prognosis at different stages and guide medication strategies, thereby enhancing patient care for COVID-19.
Keywords:
landmark model|competing risk|dynamic prediction|longitudinal data|survival|
Sponsors:
Section on Nonparametric Statistics
Tracks:
Statistical Methods for Functional Data
Can this be considered for alternate subtype?
Yes
Are you interested in volunteering to serve as a session chair?
No
I have read and understand that JSM participants must abide by the Participant Guidelines.
Yes
I understand that JSM participants must register and pay the appropriate registration fee by June 1, 2024. The registration fee is non-refundable.
I understand