In-home physical frailty monitoring: relevance with respect to clinical tests.

BACKGROUND: Frailty detection and remote monitoring are of major importance for slowing down, and/or even stopping the frailty process in home-dwelling older people. Taking the Fried's criteria as a reference, this work aims to compare the results produced by a technological set (ARPEGE Pack) with those obtained by usual clinical tests, as well as to discuss the ability of the Pack to be used for long-run frailty remote monitoring. METHODS: 194 participants were given a number of geriatric tests and asked to make use of the ARPEGE technological tools as well as reference clinical tools to feed Fried's indicators. Spearman or Pearson's correlation coefficients were used to compare the ARPEGE results to the reference ones, depending on data statistical characteristics. RESULTS: Good correlations were obtained for measurements of weight (0.99), grip strength (0.89) and walking speed (0.79). Results are much less satisfactory for evaluation of physical activity and exhaustion (Spearman correlation coefficients 0.25 and 0.41, respectively). CONCLUSION: Correlations regarding weight, grip strength and walking speed confirm the validity of the data produced by the ARPEGE Pack to feed Fried's criteria. Assessing activity level and exhaustion from an abbreviated questionnaire is still questionable. However, for long-run monitoring other methods of evaluation can be explored. Beyond the quantitative results, the ARPEGE Pack has been proved to be acceptable and motivating in such a long-term frailty monitoring.

Identification of the period of stability in a balance test after stepping up using a simplified cumulative sum.

Falls are a major cause of death in older people. One method used to predict falls is analysis of Centre of Pressure (CoP) displacement, which provides a measure of balance quality. The Balance Quality Tester (BQT) is a device based on a commercial bathroom scale that calculates instantaneous values of vertical ground reaction force (Fz) as well as the CoP in both anteroposterior (AP) and mediolateral (ML) directions. The entire testing process needs to take no longer than 12 s to ensure subject compliance, making it vital that calculations related to balance are only calculated for the period when the subject is static. In the present study, a method is presented to detect the stabilization period after a subject has stepped onto the BQT. Four different phases of the test are identified (stepping-on, stabilization, balancing, stepping-off), ensuring that subjects are static when parameters from the balancing phase are calculated. The method, based on a simplified cumulative sum (CUSUM) algorithm, could detect the change between unstable and stable stance. The time taken to stabilize significantly affected the static balance variables of surface area and trajectory velocity, and was also related to Timed-up-and-Go performance. Such a finding suggests that the time to stabilize could be a worthwhile parameter to explore as a potential indicator of balance problems and fall risk in older people.