Evaluating the Community Areas of Sustainable Care and Dementia Excellence Model of Care (CASCADE)

The CASCADE ways of working draw on the person's strengths to optimize safety and independent living in their community. The ways of working incorporate learning and development for all stakeholders to enhance interactions with people living with dementia; use of technology to improve access to specialist care as well as safety in the community; and holistic care tailored to the needs of the person to improve and maintain well being. These elements work in tandem to enable people living with dementia to engage in meaningful leisure and activities of daily living.

The evaluation study aims to explore the benefits of the CASCADE model of care (ways of working) on service users (people with dementia and their caregivers), staff and service providers. The evaluation comprises elements of outcome and process evaluation. The outcome evaluation involves quantitative measures of health, well being and cost benefits while the process evaluation explores intervention implementation procedures to find out what worked and how it worked. The CASCADE model of care assumes that using the person's strengths to guide holistic personalized dementia care is a sustainable approach to safe meaningful independent living and improving public perceptions.

Investigators will use a quasi-experiment to assess quantifiable benefits (health and well being outcomes and cost benefits) of the CASCADE model for people with dementia. This means, investigators will compare two groups of people with dementia including one that will receive care under the CASCADE model (intervention group) and the second group will not. The anticipation is that comparing the two groups will make benefits of people receiving care under the CASCADE ways of working more observable. Participants will not be allocated randomly to the study arms because service use will be dependent on assessed need for referrals and need and affordability for self-referred users. This implies there is no opportunity to randomize participants to either the intervention or control groups. Time point measures will include pre-intervention (T0), and two follow up points (T1 and T2).

It is anticipated that 160 people living with dementia will take part in the study (respite intervention group n= 50; respite control group n=50; long term care intervention group n= 30; long term care control group n= 30). Other participants will include family or informal carers (n= 20) and staff delivering care (n=70).

The residential mode of the intervention will minimize cases of missing data. However, in the event of missing data, investigators will use Multiple Imputations Methods (MIM) if missing data are less than twenty percent (<20%). It is unlikely that data will be Missing Completely at Random (MCAR), but data Missing at Random will be assessed using appropriate statistical procedures such as analysis via the patterns command to determine patterns of missingness and their frequency and to examine variables predictive of missing data. Investigators will eliminate dropouts or noncompliance from analyses, if missing data are more than 20%. Similarly, if data are not Missing at Random, investigators will use last observation carried forward (LOCF) method particularly for participants in long term care to maintain the sample size and minimize confounding effects of noncompliance or attrition.