Sweden has an aging population and people with dementia often have their needs met in their own homes. The Social Services Act states that eldercare should aim at strengthening older people's ability to live an independent life, in dignity and with well-being. Although few would question the importance of influence and individualized support, research has shown that older people with extensive and complex needs, such as dementia, may be at risk of being excluded from the same opportunities as more privileged groups of eldercare recipients. Care managers and home care staff have to handle both the choice of services and providers as well as the more detailed design of the home care services without developed working methods that support the possibility of informed choices for people with dementia or other cognitive difficulties. The need for some form of decision aid in these situations has therefore been raised. So-called TalkingMats have in previous studies in the United Kingdom (UK) been shown to promote influence in decision-making for people with dementia. Within the framework of this study, our aim is therefore to evaluate the effect of TalkingMats as decision aid in needs assessment and planning home care services for people with mild to moderate dementia. The study is designed as a two-armed RCT study, where the effectiveness and implementation of TalkingMats are evaluated both quantitatively and qualitatively, in collaboration between eldercare in four municipalities in the Sjuhärad region, the Department of Social Work at Göteborg University, the University college of Borås and R & D Sjuhärad Welfare.

The study has a descriptive, analytical and experimental design. It will adopt a pragmatic approach, balancing the criteria of a methodologically sound clinical trial vs the criteria of a clinically useful research study. Thus, participants and context reflect real-world service users and circumstances rather than a highly selected 'clean' sample. The comparative intervention study is designed as a randomized controlled trial (RCT) where the use of TalkingMats (TM) as a decision aid to support influence, choice and control in home care services for people living with mild to moderate dementia will be evaluated through comparison with usual conversation methods (UCM).Two different situations where older people living with dementia have to make decisions about home care services will be studied; first a needs-assessment conversation by care managers working in eldercare concerning the follow up of decisions about home care services as well as choice of provider. The next situation is when home care staff carry out a conversation regarding performance of the decided home care services. The participants will be randomized to one of two study arms. The implementation will be studied and analysed along with the intervention, through video recordings of the clinical encounters and survey to the staff performing the intervention. In addition, explorative qualitative interviews will be performed with staff and study participants in order to gain an understanding of the intervention and its significance as well as of the implementation.

Study population The target group for the study is older people living with mild to moderate dementia, using for home care services at the eldercare authorities in the participating municipalities. Eligibility criteria is age 65+, using home help services at the eldercare authority in Borås, Mark, Tranemo or Ulricehamn municipalities, and living with mild to moderate dementia, defined as scoring between 12-23 at a mini-mental state examination (MMSE).

Procedure The recruitment of participants to the study has started and will continue until the planned number of participants is reached. Older people assessed as meeting the eligibility criteria, contacting the local eldercare authority, will be approached by a care manager and asked if they are willing to participate in the study. If so, they will be given an accessible information sheet about the project. On expression of interest, permission will be sought to forward their contact details to a researcher, who will make initial contact to arrange a convenient time to visit. To make certain that full and informed consent is obtained, the researcher will follow the comprehensive consent procedure undertaken in a previous study ensuring that all documentation is adapted accordingly and explained thoroughly and sensitively. Thereafter, the project assistant will perform a MMSE to assess whether the person meet the eligibility criteria and, if so, collect background data. Randomization and allocation to one of the two study arms will take place immediately after the first visit. To check for comparison group equivalence, we will compare the pre-test dementia scores, functional ability and demographic variables of the two groups.

Participants in the first study arm, interventions group, will use TM as a decision aid for the needs-assessment conversation as well as in conversations planning the provision of social care. Participants in the second study arm, control group, will use UCM in both the needs-assessment conversation and the provision planning conversations.

According to preliminary power calculations, we plan to include 50 participants in each study arm. The calculation was made based on what is needed to detect a moderate difference (.25 effect size) with .8 power, p = .05.

Inclusion Criteria:

• age 65+
• using home care services at the eldercare authority in Borås, Mark, Tranemo or Ulricehamn municipalities
• living with mild to moderate dementia, defined as scoring between 12-23 at a mini-mental state examination (MMSE)

Exclusion Criteria:

-