Abstract
For patients who underwent hip hemiarthroplasty, usually due to a femoral neck fracture from a fall, subsequent falls can have life-threatening consequences. This study identified predictors for falling in patients from a clinical trial comparing the direct lateral and posterolateral approach. Patients were included if they participated in the APOLLO trial. This secondary analysis divided patients into 'fallers' during 6 months follow-up, 'non-fallers,' and 'unknown.' Two logistic regression analyses were conducted: one using variables collected pre-surgery and one using variables available at discharge. The outcome variable 'fallen within 6 months after hemiarthroplasty' was dichotomous. Among the patients with known fall status, 219 (48%) fell one or more times and 240 (52%) were non-fallers. Fall status was unknown for 384 (46%) of the 843 APOLLO trial participants. Falling was associated with five pre-surgery variables: impaired mobility, ADL difficulties, neurological and pulmonary comorbidities, and dementia. At discharge, falling was associated with impaired mobility, ADL difficulties, and in-hospital complications. The pre-surgery model had an R(2) of 0.217 and an Area Under the Curve (AUC) of 0.730 (95% CI 0.7-0.8), while the discharge model had an R(2) of 0.216 and an AUC of 0.709 (95% CI 0.7-0.8). Both models showed moderate predictive ability for falling. This study highlights the substantial incidence of falls after hemiarthroplasty for femoral neck fractures and identified predictors for falling within six months. The multidisciplinary team should prioritize fall prevention, and future research should validate prediction models and assess whether fall prevention strategies can effectively reduce falls.