Article

OBJECTIVE: To validate previously proposed findings and to develop an objective,
feasible and efficient bifactorial (risk factors: gait impairment and balance
disorders) fall risk assessment. DESIGN: Prospective follow-up study Setting:
Nursing homes (Halle/Saale, Germany). SUBJECTS: One hundred and forty-six nursing
home residents (aged 62-101 years) were recruited. METHODS: Gait data were
collected using a mobile inertial sensor-based system (RehaWatch). Postural
regulation data were measured with the Interactive Balance System. Falls were
recorded in standardized protocols over a follow-up period of 12 months. Main
measures: Gait parameters (e.g. spatial-temporal parameters), posturographic
parameters (e.g. postural subsystems), number of falls. RESULTS: Seventeen (12%)
of the participants had more than two falls per year. The predictive validity of
the previously selected posturographic parameters was inadequate (sensitivity:
47%). The new measurement tool defined 67 participants showing an increased risk
of falls. In reality, only 8 participants actually fell more than twice during
the follow-up period (positive predictive value (PPV): 12%). The negative
predictive value (NPV) was 88%. The posturographic frequency range F2-4
(peripheral-vestibular system), stride time and standard deviation of landing
phase were the most powerful parameters for fall prediction. Gait and postural
variability were larger in the high-risk group (e.g. gait speed; confidence
interval (CI)(high): 0.57-0.79 vs. CI(low): 0.72-0.81 m/s). CONCLUSION: RehaWatch
and the Interactive Balance System are able to measure two of the most important
fall risk factors, but their current predictive ability is not satisfactory yet.
The correlation with physiological mechanisms is only shown by the Interactive Balance System.

Langue : ANGLAIS