If you have not read any of the previous blogs yet, it would be a good idea to go back and read our first blog post which talks about the steps of this project, including the site visits to existing services, and to also read Suzanne Timmons blog post describing why people with dementia need palliative care.

The current stage of the Model for Dementia Palliative Care Project involves conducting site visits to existing “excellent” model services across Ireland and the UK to identify what works well for these services, and how they are run, in order to provide good palliative care to people with dementia. Our previous literature scoping review revealed a lack of high-quality evidence to direct the choice of core components for a model for dementia palliative care. Thus we will now visit services to see ‘what makes a good service’.

What do the researchers do when they conduct a site visit?

The research team has already conducted one site visit which was to an Irish hospice. This involved two members of the research team, Dr Siobhan Fox and Ms Niamh O’Connor, spending two days at the service conducting data collection. This involved interviewing a range of staff at the hospice, from managers and consultants, to social workers and healthcare assistants. A focus group with the community team was also conducted which provided a valuable insight into the care of people with dementia who live in the community and who also wish to die at home in their community, and into what the team does to ensure this can happen.

As well as interviews and focus groups, the researchers also distributed anonymous questionnaires to all staff members which asked them about their experiences of working in the service and how the service supports the palliative care of people with dementia. Routinely collected data was also provided to the researchers by the service. This data included information such as the overall number of people who used the service in a given year, the number of people with dementia who used the service in that year, the number of people with dementia who died in the inpatient hospice, the number of people with dementia who were supported to die at home, etc. Of note, this data didn’t include any details of a particular person, just the service users in general.

When doing a site visit the researchers are completely flexible and do not go to the site visit with a rigid time table, but instead work with the service in what suits them best i.e. being very flexible with interviews/focus group times to suit the staff. The research team are currently planning their first site visit to the UK – a service for people with advanced dementia in London. We aim to conduct a further 5-6 site visits across Ireland and the UK.

What will the researchers do with all the data they collect from the site visits?

The data collected from each service will first be described using a “complex logic model”. Logic models are typically used in program theory to describe programmes. A logic model is a visual illustration of a program’s resources, activities and expected outcomes. It is a tool used to simplify complex relationships between various components and is often used during program planning, implementation and evaluation (Public Health Ontario, 2016). Figure 1 shows the outline of the logic model which will be filled with the appropriate information and will be used to describe each service visited by the researchers. Basically, the resources, context, core aim(s) of the service, activities of the service and short and intermediate outcomes will be mapped onto the model. Service characteristics such as duration, size and scale, and complexity will be included, as well as the theory underpinning the model e.g person centred care, palliative care approach. This will allow each service to be individually described as a model, which is important given the expected dissimilarity of the services.

The services will also be evaluated and compared using the RE-AIM framework (Reach, Effectiveness, Adoption, Implementation, Maintenance). This framework will be explained more in a future blog post! Using this dual method will help the researchers to compare across the various services, and to highlight certain components that are present in all services, or only feature in some.

Figure 1: Logic model outline for site visits

What’s next?

The research team is in the process of planning the remaining site visits. Once all these visits are completed, described and evaluated as outlined above, a draft model of palliative care for people with dementia in the community will be produced. By using the learning from the evaluation of existing services though the site visits and the evidence from the scoping review and also stakeholder input (described in blog 1 ), a draft model of palliative care for people with dementia will be produced. A logic model will also be used to illustrate this draft model so that the individual components and their relationships are clearly displayed. This initial version will be very inclusive, to display all possible options, and will be refined through expert consultation and feedback from service users including people with dementia and their carers. The findings of these stages will follow in later blog posts once the researchers have completed their analysis. If you are interested in the use of logic models, the Centre for Disease Control and Prevention provide useful explanation on the different uses of logic models: Link

To keep up to date with the project and to be informed of new blogs posts follow @PallC4Dementia


Centre for Disease Control and Prevention. Programme Evaluation Framework Checklist. Available at: https://www.cdc.gov/eval/steps/step2/index.htm. Programme Performance and Evaluation Office; 2018.

Ontario Agency for Health Protection and Promotion (Public Health Ontario), Abdi S, Mensah G. Focus On: Logic models-a planning and evaluation tool. Toronto, ON: Queen’s Printer for Ontario; 2016.

5. Using Logic Models to evaluate services to develop a model for dementia palliative care

Leave a Comment

Your email address will not be published. Required fields are marked *