The Impact of Hip Fracture on Geriatric Care and Mortality Among Older Swedes: Mapping Care Trajectories and Their Determinants

Abstract

In this study, we examined the impact of hip fractures on trajectories of home care, nursing home residence, and mortality among individuals aged 65 years or more and explored the impacts of living arrangements, cohabitation, frailty, and socioeconomic position on these trajectories. Based on a linkage of nationwide Swedish population registers, our study included 20,573 individuals with first hip fracture in 2014-2015. Care trajectories during the 2 years following the fracture were visualized and compared with those of 2 hip-fracture-free control groups drawn from the general population: age- and sex-matched controls and health-matched controls identified through propensity score matching. Multistate modeling was employed to identify sociodemographic and health-related factors associated with care trajectories among hip fracture patients. We found that hip fracture patients already had worse health than the general population before their fracture. However, when controlling for prefracture health, hip fractures still had a considerable impact on use of elder-care services and mortality. Comparisons with the health-matched controls suggest that hip fractures have an immediate, yet short-term, impact on care trajectories. Long-term care needs are largely attributable to poorer health profiles independent of the fracture itself. This emphasizes the importance of adequate comparison groups when examining the consequences of diseases which are often accompanied by other underlying health problems.

Keywords: Sweden; aging; elder care; hip fracture; home care; nursing homes; osteoporosis; registers.

Figure 1

Transitions between care states over the course of 24 months, by time period (T), among hip fracture patients, age- and sex-matched controls in the general population, and propensity-score–matched controls (n = 61,719), Sweden, 2014–2017. Left-hand panels show care transitions among hip fracture patients (A), age- and sex-matched controls (C), and health-matched controls (E). Right-hand panels show the care distributions among survivors for hip fracture patients (B), age- and sex-matched controls (D), and health-matched controls (F). T1, baseline; T2, 3 months after baseline; T3, 12 months after baseline; T4, 24 months after baseline. Blue shading: no formal elder care; light red shading: home care for <40 hours/month; dark red shading: home care for ≥40 hours/month; yellow shading: residence in a nursing home; gray shading: death.

Figure 2

Structure of the multistate model of care-status transitions after experiencing a hip fracture and number of transitions observed within the first 3 months after hip fracture (n = 20,573), Sweden, 2014–2017.

Figure 3

Hazard ratios (HRs) and 95% confidence intervals (CIs; bars) for care-status transitions during the 3 months after hip fracture among 20,573 individuals with hip fracture, Sweden, 2014–2017. Results are shown for transitions from no care to at-home care (A), no care to residence in a nursing home (B), no care to death (C), at-home care to residence in a nursing home (D), at-home care to death (E), and residence in a nursing home to death (F). Frailty was based on the Gilbert frailty index (30). Scores of 4.9 or less indicate low frailty risk; scores of 5.0–15.0 indicate intermediate frailty risk; and scores over 15.0 indicate high frailty risk. Sec., secondary.