Fall-related hospitalization in people with Parkinson's disease

Background and purpose: Falls are common in people with Parkinson's disease (PD) but few data exist on fall-related hospitalizations in this group. This population-based study compared fall-related hospital admissions, injury rates and consequences in people with and without PD, and determined whether PD was an independent predictor of fall-related hospital length of stay. Methods: This was a retrospective study using probabilistic linkage of hospital data in people aged 65 years hospitalized for a fall between 1 July 2005 and 31 December 2013 in New South Wales, Australia. Rates of hospital admissions and injuries per person admitted over the study period were compared between people with and without PD using Poisson or negative binomial regression. Multilevel linear modelling was used to analyse length of stay by clustering individuals and adjusting for possible confounders. Results: There were 342 265 fall-related hospital admissions in people aged 65 years during the study period, of which 8487 (2.5%) were for people with PD. Sixty-seven per cent of fall-related PD admissions were associated with injury and 35% were associated with fracture. People with PD had higher rate ratios for fall admissions (1.63, 95% confidence interval 1.59-1.67) and injury (1.47, 95% confidence interval 1.43-1.51) and longer median length of stay [9 (interquartile range 1-27) vs. 6 (interquartile range 1-20) days in people without PD; P < 0.001]. PD remained associated with increased length of stay after controlling for comorbidity, age, sex and injury (P < 0.001). Conclusions: This study provides important benchmark data for hospitalizations for falls and fall injuries for older people with PD, which may be used to monitor the effect of fall prevention programmes.