Background
Improving the quality of care in hospitals continues to be a key priority within national dementia strategies. Dementia care, and in particular, dementia training are rapidly developing and dynamic fields of study. Different training packages have been developed, and these are often evaluated using varying methods. Understanding the relationships between the inputs and outputs of a programme (dementia training) is critical if we are to develop theory-based, programmatic models for intervention and policies.

Aims and (where relevant) research/evaluation/project questions
The primary aim of the study is to develop a logic model using the principles of programme theory to identify contextual factors, mechanisms, interaction, facilitators and barriers to dementia training in the hospital setting.

Methods
This study uses a two-phase mixed-methods design:
1 – The ‘experiential review and consultation’ stage (n=30, workshops and one-to-one consultations): The primary purpose of this stage was to consult and reflect with researchers, academics, dementia trainers, clinicians involved in development, implementation and evaluation of dementia training packages in hospital settings, and people living with dementia and carers to identify mechanisms, interaction, facilitators and barriers to dementia training in the hospital setting. This was the key stage to the development of the ‘initial’ programme theory for dementia training in hospital settings.
2 – Critical Interpretive Synthesis review of literature (qualitative, quantitative and mixed study), which aims to provide a systematic review of the current evidence base for dementia training in hospitals and the broader range of structures and dementia initiatives within hospitals. A further developmental step to the programme theory (logic model) was undertaken following review to further develop the ‘initial’ programme theory.

Results
There were various barriers to and facilitators of dementia training in hospitals, including both surface and deep-seated factors for patients living with dementia and carers, hospital staff, and organisation as a whole. These include lack of resources, abundance of varying training packages/courses available, and willingness/support to develop and implement training initiatives, existing beliefs and assumptions of trainers and trainees, as well as affiliation or previous experiences of trainers and staff working with people living with dementia and their careers. Cutting across each element of the logic model of dementia training were three drivers, in which barriers and facilitators manifest:
1. Strong and supportive leadership
2. Resources
3. Recognition of dementia training (incentives/rewards for staff)

Discussion and conclusions
The long list of key factors detailed in our logic model is not an exhaustive list, but serves as a programme theory for planning, developing and evaluating educational programmes in hospital settings among practitioners, trainers, and dementia training initiative developers and evaluators. This model will go through a further development round to utilise primary data from our main study (hospital surveys and case-studies collected as part of the Neighbourhoods and Dementia Work Programme 5 study).

Limitations
Our review includes articles published since 1990, which may have excluded older studies that might further contribute to the development and design of dementia training in the hospital setting. We found a limited number of studies that covered dementia training in the hospitals setting. However, to the best of our knowledge, this is the first programme theory (logic model) that focusses specifically on dementia training in hospital settings.