Factors Associated With 1-Year Home Health Admission Among Older Adults Initiating Long-Term Opioid Therapy

Long-term opioid therapy (LTOT) in older adults poses risks of cognitive and functional decline. This study aimed to identify factors associated with home health admission within 1 year of LTOT initiation among Medicare-enrolled older adults.

Cohort study.

A total of 424,867 new LTOT users were identified from a 5% sample of national Medicare claims data between July 1, 2012, and December 31, 2019. After implementing the study eligibility criteria, 91,083 Medicare beneficiaries were found to be eligible for the study.

The Medicare claims database was linked with Outcome and Assessment Information Set (OASIS) data (2012-2020) to identify older adults who initiated home health care within 1 year of starting LTOT. Individuals were censored at disenrollment, hospice use, 90-day opioid discontinuation, death, or end of the 12-month follow-up period, whichever occurred first. Individual sociodemographic, clinical (time-dependent and time-independent), and medication-related variables were assessed. Extended Cox regression was used to assess the association between individual factors and 1-year home health admission.

Among 91,083 eligible beneficiaries, most were aged 65 to 74 years (51.1%), female (65.6%), non-Hispanic White (82.5%), and had chronic noncancer pain (84.6%). Overall, 8.1% had 1-year home health admission, with a restricted mean survival time of 337.9 days (SE: 0.3). Factors associated with an increased likelihood of 1-year home health admission included demographic characteristics (older age, female sex, Hispanic ethnicity), clinical conditions (eg, higher morbidity burden, chronic noncancer pain, hepatic impairment, cancer, recent surgery, hospitalization), and medication-related characteristics (hyper-polypharmacy and higher daily opioid doses).

These findings highlight the need for risk-informed care planning and stratification among older adults on LTOT requiring home health services. This may help guide better allocation of resources and facilitate targeted interventions to support this vulnerable population. Future research should explore effective strategies for early identification and tailored support within this population.