Activity Monitoring in Frail Elderly Patients. A Methods Comparison Study.

Elderly subjects will be recruited from a residential-care complex in Zwolle, the Netherlands. All subjects will be asked to fill in the Groningen Frailty Index. We aim to select patients with varying GFI scores to ensure that we include patients with different degrees of frailty. The participating subjects will be asked to wear the DynaPort MicroMod during six hours. Two research employees will observe all activities of the participating subjects in their home environment. The participant will be instructed to act like it is a normal/usual day. Every 5 minutes the type of activity at that moment and movement intensity of the last 5 minutes (METs scale) is manually entered onto a case report form (CRF) by the two employees. The timing has to be exact in order to compare the data to the data of the DynaPort MicroMod. Type of activity is described as well as categorised into lying, sitting, standing or walking. Movement intensity is categorised according to a MET-value (Metabolic Equivalent of Task).

Sensitivity, specificity, and percentage of overall agreement will be calculated for each type of activity (standing, sitting, lying, walking) and for intensity. The primary analysis will be the comparison between walking and not walking (lying, sitting or standing). An overall agreement/sensitivity/specificity of 90% or higher is considered as sufficient agreement. The Bland and Altman method will also be used as multiple observations are recorded per patient, the limits of agreement will also be calculated with the Bland and Altman method for repeated measures.