Développement et validation d'un outil de prédiction du risque de chute chez la personne âgée institutionnalisée

The fall of the elderly is a public health issue because of the medical, social and economic consequences it has. In addition to the serious consequences caused by falls, these accidents can lead to behaviours that reduce daily living activities and promote harmful psychomotor effects often linked to a fear of falling. Repeated falls account for an average of three falls per person, each year, over the age of 65, with this number increasing with age and loss of autonomy. In institutionalized individuals, advanced age, co-morbidities and associated psychomotor disorders, dependency and sometimes very advanced frailty, the risk of falling is all the greater. Health authorities, notably the Haute Autorité de Santé, have reported the lack of effectiveness of methods for identifying people at risk and emphasize the lack of effectiveness of preventive measures to reduce the number of falls. The sensitivity, to detect people at risk when they are indeed at risk, and the specificity of clinical tests used by rehabilitators, to rule out people at low risk of falling, generally make them as useful as a questionnaire on the number of falls suffered in the last twelve months. This raises the question of the relevance of using them to identify and follow up people potentially at risk of falling (especially since they are upstream of repeated falls). At the same time, adapted physical activity could be a useful non-medicinal treatment to reduce falls and their severity in these elderly subjects. With a prevention objective, the implementation of physical exercises could reduce the number of falls per individual. However, the implementation of such programs leads to an optimization of the management so that it is effectively adapted to the participants and the health care institutions. First, we will focus on the problem of the aging population in relation to falls. Then we will address the age-related alterations that impact the deterioration of balance. The third chapter will look at the impact of aging on the markers of static balance. We will see how we have aggregated the results extracted from the literature in a context where the variability of recording protocols had previously confused the comparison of results. The tool for quantifying static balance and gait, called SmartCheck, is presented in order to highlight the possible expectations of this type of solution providing healthcare professionals with an automated analysis of balance during routine tests. This is made possible by signal processing and machine learning algorithms. The recordings made with elderly patients in hospital consultations have made it possible to calculate more than a hundred posturographic variables and then to compare the values obtained between fallers and non-fallers. The relevance of these indicators is discussed in order to provide a set of variables to be calculated for the analysis of the static balance of elderly people. Finally, the fifth chapter will look at fall prevention programs for older adults living in EHPAD. The aim is to set up the Postadychute-AG clinical trial aimed at validating the relevance of posturographic parameters during a longitudinal follow-up associated with adapted physical activity.