Playful Multimodal Intervention, Monitoring and Decision Support for Activation of People With Alzheimer's Dementia

Background. The prevalence of Alzheimer's disease is increasing due to an aging society. Parallel, care and nursing needs and costs rise with disease progression. Promoting and stimulating patients to maintain patients abilities and the independent life is therefore essential to prolong the time at home and reduce direct and indirect costs. Computer-based technologies, such as tablet-based trainings, might have the potential to support the patient itself and the caregiver by reducing the (nursing) care-dependency and stabilizing or slowing the disease progression. Prior studies revealed that patients suffering from dementia in general desire technologies supporting cognitive abilities, communication skills, social integration and activities of daily living (e.g. mobility) (Lauriks et al .2007, Wand et al., 2016). Moreover, a multimodal approach has been demonstrated to be superior compared to unimodal approaches, aiming to foster only one domain or aspect. In this study, a multimodal tablet-based training, focusing on cognitive, psychological, physical and aspects of activities of daily living, for patients suffering from mild Alzheimer's disease was implemented.

The primary aim of this study is to examine the efficacy of the multi-domain approach which as implemented in the tablet-based training in mild AD patients. Secondary aims are to measures structural and functional cerebral changes, quality of life, mobility, motivation, life-style factors, stress, care-dependency, (instrumental) activities of daily living, changes in medical treatments, problems in behavior, burden due to nursing, activity level, arm strength, depressive symptoms and the usability of the data raised for the development of a decision support system.

Methods. This study is implemented as a randomized controlled trial with embedded design. Two-hundred twenty subjects with mild AD (dementia diagnoses according to NINCDS-ADRDA criteria) and the primary caregiver will be included in the study. One hundred ten subjects will receive a guided tablet-training for 1.5 years. One hundred ten subjects are in the control group, receiving "normal" medical treatment and support. Additionally, five nursing persons and 5 M.A.S. (Morbus Alzheimer syndrome) trainers conduct single and group trainings. All subjects and the caregivers will be tested and interviewed, respectively, at study entry, at six, twelve and eighteen months. The study team consist of Joanneum Research (JR), the Sozialverein Deutschlandsberg (SVDL), the Austrian Red Cross (ÖRK), and the Medical University of Graz, Department of Neurology and the Institute of Nursing Science as well as Health System \& Intelligence (HS\&I).

Recruiting. The recruiting process is carried by telephone, flyer, home page, personal meetings, and social media. The approach is convenience sampling.

Randomisation. In advance of the study, a randomisation concept determines the participant's random assignment to intervention or control group. The first half of the 220 participants are 70% likely to be assigned to the intervention group and 30% to the control group. For the second half of the sample the likelihood for group assignment reverses. This step should guarantee that enough participants receive the intervention in case of challenging recruitment. The intervention is favoured over an equal distribution to the groups. At the University Clinic of Neurology, where the examination takes place, only a study nurse knows about the group assignment of the patients, forwarding this information to the personnel of the SVDL and the ÖRK.

Blinding. In the context of the study, blinding of the employees of the SVDL and ÖRK is not possible since the intervention is apparent to the personnel. The neurologists and neuropsychologists are not informed at any time during the study of the patient's group assignment.

Intervention. "Multimodal Activation" (MMA) is an app developed for tablets that contains a serious game and has been previously tested in patients with dementia. It stimulates a multimodal training by games that involve cognitive and physical exercises. These exercises can be individually adapted to content, level of difficulty, sequence and time. The training starts with physical exercises accompanied by music. Afterwards there are cognitive exercises such as a quiz, puzzle, gap-filling, simple arithmetic tasks or listening comprehensions. The patients can use MMA at home as well as in a care facility.

Intervention group (IG). The IG undertakes a 1.5-year tablet PC-based training of cognitive and physical skills using DaheimAktiv, which is carried out every 14 days by a MAS trainer, medical nurse or pedagogically trained person from the SVDL or the ÖRK together with the patient. The tablet remains with the patient at home and the participants are encouraged to continue training without the presence of a trainer (with and without a relative). All training sessions are recorded. DaheimAktiv is a multimodal intervention procedure with cognitive and physical exercises. In addition to the fourteen-day training units with the trainers and independent training, "training cafés" will take place in small groups to promote motivation and increase social integration, i.e. several patients train together.

Control group (CG). In the control group, the medical interventions initiated are continued (standard intervention). Apart from that, these patients are subjected to the same study conditions as the patients of the intervention group. After completion of the evaluation phase, patients of the CG will receive a free, one-year license for DaheimAktiv.

Study outline. First, a pre-screening of potential participants takes place to check for inclusion and exclusion criteria. The patients that fit the criteria are invited to the University Clinic of Neurology for a neurological examination, blood sampling (also for potential genetic analysis) and cognitive assessments (performance tests and interviews). If a diagnosis of Alzheimer's disease is confirmed on a clinical basis, patients are invited to a structural and functional magnetic resonance imaging (MRI) in order to exclude other causes of cognitive deficits and aggregate imaging biomarkers of disease-associated changes such as hippocampal volume. Six and twelve months after study inclusion short cognitive testing and interviews at home will be carried out. After 1.5 years, the patients from the control and the intervention group again are invited to the University Clinic of Neurology performing the same procedures as during baseline.

Hotline. For the duration of the study, the SVDL provides a hotline specific for questions and problems that participants might face.

Statistics. The quantitative data will be analysed using IBM SPSS (version 25.0, 2017). The chosen level of significance is determined at α = .05. Mean and standard deviation or median and quartiles will be part of the descriptive statistics. Absolute and relative frequencies will depict categorical data. In order to test for the hypothesis regarding the positive effect of tablet training on cognition, primary outcome parameters (such as global neuropsychological scores) as well as secondary variables (such as neuropsychological subscores) will be compared between intervention and control group using ANCOVA adjusted for multiple testing. Furthermore, total brain volume, local volume of brain areas such as the hippocampus, cerebral microstructure and functional connectivity will be measured and compared between groups. Other secondary outcome parameters will be analysed using ANCOVA for between group comparisons. Pre-post comparisons (by ANOVA for repeated measures) shall deepen the understanding of how these changes occur over time. The analyses will be adjusted for age, sex, comorbidity and education. Qualitative data from the interviews will be organized and coded via the software MAXQDA. Afterwards a qualitative content analysis will be applied.

Ethics. Patients (or the person with power of attorney) have to give informed consent in order to participate in the study. At any given time, the participants can refuse or withdraw the consent without giving specific reasons. The genetic analysis merely serves a scientific purpose and requires written consent. All samples are analysed and stored using pseudonyms. Only at the University Clinic of Neurology it is possible to connect the sample with the name of the donor. If the results of the genetic analysis will be used for multi-centric research, it is ensured that the identity of the donor remains anonymous for cooperating centres. All aspects of the examination are non-invasive (except blood sampling), state-of-the-art procedures that are applied according to the law. The ethics committee of the Medical University of Graz approved the study on February 5th 2020 (31-556 ex 18/19).

Data protection Every participant receives an individual, pseudonymous code at the University Clinic of Neurology that is shared with the SVDL and ÖRK for patient-code assignment. The other partners of the study team only receive the patient's code. The communication between the project partners and all data transfer occurs using this code to ensure the anonymity of the patients. The SVDL submits questionnaires and interviews to the Medical University of Graz (MUG) via a secured server (protected by password). All personal information about the participants is confidential. Interviews are pseudonymised during transcription. The project partners have access to the data for the purpose of data entry and analysis.

For the tablet-based training with the MMA app, a registration with an email-address and a password is necessary. This information is stored in a coded form in a database. The serial number of the tablet has to be stored as well. Further personal information is optional and hence not required to use the training.

The app also stores information about every completed unit in the programme like speed and precision in the training intervention for the sake of future improvement. For further analysis of the intervention group, the data will be pseudonymous.

The usage of fitness trackers only works in combination with an account that the users have to create in the first place. However, there is no data transfer of the log data to the central database of the project. Only the users can access that data.

Trainings. In order to minimize the inter-rater variability of thr raters for questionnaires and interviews, the data-collecting persons of the SVDL and the ÖRK are trained beforehand on the questionnaires and interviews (individual interviews, focus groups) by employees of the MUG Neurology and nursing. For the tablet training, the SVDL and the ÖRK will train the dementia trainers to use the intervention. The duration will be about 4 hours. All participants/patients receive training in the application of the intervention by the SVDL/ÖRK.

Hotline, contact persons. During the study, the SVDL will set up a hotline for questions and problems of the participants during the study period. Regular times are given for this.