An information letter was sent to nursing homes in the Lille region (France) to present the objectives of the study. The nursing homes that agreed to participate received additional information for implementing the study. Randomisation was carried out in clusters, i.e. the sites were randomised to one of the three experimental conditions. The randomisation was carried out using Excel (Microsoft Corporation, version 18.2110.13110.0). For each site, a random value was generated and then a rank was assigned for each value. The rank value was divided by three, corresponding to the number of experimental conditions. These values were then rounded up. Each site therefore has a value of "1", "2" or "3" corresponding to an experimental condition: "1" = arm #1; "2" = arm #2; and "3" = arm #3.
The investigators of the 9 participating nursing homes screened the residents and family caregivers according to inclusion, non-inclusion and exclusion criteria. The study was then presented to residents and family caregivers who met the criteria by giving them an information letter so that they could take time to consider whether to participate in the study, with a minimum two-week cooling-off period between the information and the written consent.
The socio-demographic data collected will concern for all participants: age; gender; level of education; relationship in the dyad; previous practice of Tai Chi; usual sports and leisure activities. For residents, the Mini Mental State Examination (MMSE) score will also be collected from the medical record of each included resident and must be less than six months old. If the MMSE score is older than six months, the nursing homes will be asked to carry out the MMSE.
Two assessments of the residents and family caregivers will be carried out before the intervention and two weeks after the intervention. The assessments will be carried out by psychologists who have been previously informed and/or trained in the use of the questionnaires and scales. The information and/or training will be done by telephone. Evaluations will be conducted blind, i.e. psychologists will not be aware of the experimental condition that will be assigned to each participant. Each evaluation will last between 30 minutes and 1 hour.
Data collection in the observation books will be carried out via a code assigned to each participant in a non-identifying manner. A table of correspondence between the participants' first and last names and the codes will be kept securely by the principal investigator at each site. Once the data collection is finalised, these tables will be deleted and the codes will not appear on the computer files used for data processing and statistical analysis.
Data will be stored in each nursing home for the duration of the study. Only authorised principal investigators will have access to the consent forms and observation books for the duration of the study, which will be anonymised. The external evaluators will have access to the observation books via by the principal investigators of each sites only during the evaluations.
The coordinating investigator will make monitoring visits to ensure the proper implementation of the study and the accuracy of the consent and data collection. The coordinating investigator will also be in regular contact by email and telephone with the principal investigator at each site.
The results of the tests and scales will be processed descriptively and analysed statistically with comparisons of means between the three experimental conditions. The results will also be analysed taking into account signs of anxiety, depression and the quality of the caregiver-caregiver relationship before the start of the study. The qualitative data will be analysed for content.
