|
Affiliation
|
Asia-Pacific Centre for Neuromodulation, Centre for Clinical Research, The University of Queensland, Brisbane, Queensland, AUSTRALIA. |
|
Abstract
|
OBJECTIVE: This prospective study evaluated whether the Activities-specific Balance Confidence (ABC-16) scale and short-form ABC-6 could predict future recurrent falls in Parkinson's disease (PD) and evaluated the robustness of their predictive capacities.
DESIGN: 12-month prospective cohort study SETTING: General community PARTICIPANTS: 313 PD patients were contacted via neurology clinics and pre-existing databases and invited to participate. Of those contacted, 188 did not respond, 22 were not interested and 4 were deceased. Following screening, a further 18 were excluded; 13 had deep brain stimulation surgery; 3 used walking aids; and 2 had poor cognition. The remaining 81 completed all assessments, but two withdrew before completing the follow-up. INTERVENTIONS: Clinical tests were completed to establish symptom severity, balance confidence and medical history. Over the subsequent 12 months, participants recorded any falls on daily falls calendars, which they returned monthly via reply-paid post. MAIN OUTCOMES: Logistic regression and receiver operating characteristic analyses estimated the sensitivities and specificities of the ABC-16 and ABC-6 for predicting future recurrent falls in this cohort and 'leave-one-out' validations were used to assess their robustness.
RESULTS: Of the 79 who completed the follow-up, 28 fell more than once during the 12-month period. Both the ABC-16 and ABC-6 were significant predictors of future recurrent falls and moderate sensitivities (ABC-16: 75.0%; ABC-6: 71.4%) and specificities (ABC-16: 76.5%; ABC-6: 74.5%) were reported for each tool for cut-off scores of 77.5 and 65.8, respectively.
CONCLUSIONS: The results have significant implications and demonstrate that the ABC-16 and ABC-6 independently identify patients with Parkinson's disease at risk of future recurrent falls.
Language: en |