Service use after long-term care entry: low utilization and early dropout in Taiwan
Yu-Hung Chang, Chia-Ming Yen, Shuo-Chen Chien, Jian-Hong Lin, Chia-Chun Liu, Yu-Ping Hsiao, Ren-Hua Chung, Hung-Yi ChiouAbstract
Background and Objectives
Home- and community-based services (HCBS) are central to Taiwan’s long-term care (LTC) system, yet some users show limited service use after entering the system. This study examined predictors of two patterns of limited HCBS use: low utilization and early dropout.
Research Design and Methods
This study analyzed administrative LTC records from Pingtung County, Taiwan, among cases entering the LTC system during 2019–2021: 17,517 in the low-utilization analysis and 18,678 in the early-dropout analysis. Low utilization was defined as service use in the lowest quartile of service-use days per 30 follow-up days, and early dropout as case closure within 90 days. Bootstrap-enhanced least absolute shrinkage and selection operator (BoLASSO) was applied, followed by multivariable logistic regression.
Results
Low utilization was associated with care arrangements and entry pathways, including institutional residence, use of a paid migrant caregiver, co-residence with a nonfamily caregiver, and entry via the 1966 hotline. Early dropout occurred in 11.4% of cases and was associated with cancer, oxygen-related care needs, severe pain, underweight, unintentional weight loss, and inability to complete short-term memory assessment. Service-provider referral was associated with lower odds of both outcomes.
Discussion and Implications
Low utilization and early dropout showed distinct predictor patterns: the former was more closely associated with care resources and entry pathways, whereas the latter was associated with health instability and recent decline. Distinguishing these patterns may inform more tailored care management and service delivery based on users’ needs and care contexts.