Skip Navigation
Skip to contents

JPMPH : Journal of Preventive Medicine and Public Health

OPEN ACCESS
SEARCH
Search

Articles

Page Path
HOME > Korean J Prev Med > Accepted Articles > Article
Special Article A Practical Tutorial on Semi-competing Risks in Aging Research: Long-term Care Facility Admission and Death Using SemiCompRisks in R
Jaeyeop Choi1, Jimin Yoon1, Seungyeon Chun1, Hongsoo Kim1corresp_icon, Woojoo Lee1corresp_iconorcid

DOI: https://doi.org/10.3961/jpmph.26.660 [Accepted]
Published online: October 2, 2026
  • 271 Views
  • 0 Download
  • 0 Crossref
  • 0 Scopus
Seoul National University, Seoul, Korea
Corresponding author:  Hongsoo Kim,
Email: lwj221@snu.ac.kr
Woojoo Lee,
Email: lwj221@snu.ac.kr
Received: 1 July 2026   • Revised: 2 September 2026   • Accepted: 13 September 2026

Population aging affects many aspects of health and social care, particularly the transition of older adults from home-based to facility-based long-term care. Analysis of this transition is complicated because some older adults die before facility admission can occur, and clinical factors may be associated with both admission and mortality. This tutorial introduces semi-competing risks analysis as a framework for addressing this asymmetric event structure, using long-term care facility (LTCF) admission among home-based care users as a worked example. The event process is represented by 3 transition-specific hazards: LTCF admission, death before LTCF admission, and death after LTCF admission. These hazards are estimated using the SemiCompRisks package in R, with a shared frailty linking the 3 hazards to account for residual dependence between the admission and death processes. This approach is compared with 2 analyses that define the outcome differently: a survival analysis treating LTCF admission as the event and death as censoring and a composite endpoint analysis treating LTCF admission and death as a single event. This worked example may help researchers recognize semi-competing risks structures in data from aging studies, distinguish non-terminal from terminal events, and select an analytic approach that reflects the ordering of care transition and death.

Related articles

JPMPH : Journal of Preventive Medicine and Public Health
TOP