|
Abstract
|
BACKGROUND: Hospital length of stay (LoS) is believed to be associated with higher mortality in hip fracture patients, however, previous research has shown conflicting results. We aimed to explore the association between LoS and 4-month mortality in different groups of hip fracture patients.
METHODS: The study population in this Swedish register-based cohort study was 47,811 patients ≥65 years old with a first hip fracture during 2012-2016, followed for 4 months after discharge. LoS was categorized by cubic splines and the association between LoS and mortality was analyzed with Cox regression models, adjusted for sociodemographic- and health related factors.
RESULTS: Mean LoS was 11.2±5.9 days and 12.3% of the patients died within 4 months. Both a shorter and a longer LoS, compared to the reference 9-12 days, was associated with higher mortality (HR [95% CI]); 2-4 days 2.15 (1.98-2.34), 5-8 days 1.58 (1.47-1.69) and 24+ days 1.29 (1.13-1.46). However, in fully-adjusted models, only the association with a long LoS remained; 13-23 days 1.08 (1.00-1.17) and 24+ days 1.42 (1.25-1.61). Stratifying by living arrangement revealed that the increased risk for a short LoS was driven by the group living in care homes. For patients living at home, a short LoS was associated with a lower risk, HR 0.65 (0.47-0.91) and 0.85 (0.74-0.98) for 2-4 and 5-8 days, respectively.
CONCLUSIONS: A long LoS after a hip fracture is associated with increased 4-month mortality risk even after considering patient characteristics. The association between mortality and a short LoS, however, is explained by individuals coming from care homes (with higher mortality risk), being discharged early.
Language: en |