King's College London
Service provision for older people who are homeless and have memory problems
Alzheimer's disease & other dementias
This study is for 24 months & incorporates three Work Packages (WP). WP1 includes 2 narrative reviews. First is a scoping review which will investigate literature on prevalence of memory problems among older homeless people, by synthesising audits & surveys; explore range of definitions of memory problems used; & explore barriers & facilitators to uptake of services & support. The second will address objective 4 (how homeless people with memory problems perceive their quality of life) & will be an evidence synthesis of quality of life (QoL) instruments used in assessing people with dementia (all types), to assist investigators & service providers to select optimal measures to specifically address issues & life areas that are of importance to homeless older people. A critical discussion with tabulated evidence is required to compare measurement properties of QoL instruments & their appropriateness for assessing QoL of homeless people with dementia. Objectives are to assess domains of QoL measures, conceptual frameworks, sensitivity to different models & settings of care, identify factors that affect reported QoL, including among homeless groups. Key & relevant databases & grey literature will be searched. For both reviews, two independent reviewers will screen all retrieved titles & abstracts; full text articles will be obtained for eligible abstracts; & data extracted from them using a standard data extraction form. Checklists will assess methodological quality of selected studies. Disagreements will be resolved by discussion and consensus. WP2 will examine extent of memory problems among older hostel residents, staffs awareness of problems, & actions taken when such problems are recognised. The study will be carried out at four diverse sites in England. Ethical consent procedures & practical routes of engagement will be considered throughout the study. After ethical approval, WP2 will involve the following steps: 1) Interviews with homeless people (maximum 50): Staff will identify older hostel clients (those over 50), discuss the study with them & seek permission to pass their name to us. We will then explain the study (with an accessible Information Sheet) to them in detail & seek written consent to participate. Short interviews will be conducted with older homeless people to collect background information, including demographic factors; brief histories of homelessness; past & current health (& problems). Then, we will administer Mini-Mental State Examination (MMSE) to each participant to screen for presence of cognitive impairment. Cut-off score of 25 (out of 30) will be used. 2) Interviews with hostel managers: Hostel managers will be interviewed at each site. Using an Interview Schedule, generic information will be collected about what the hostel does / does not do about identifying people with memory problems, daily practices & possible referrals. We will also seek information about resources, finances & policies & explore funding sources for the service. 3) Interviews with hostel workers: Keyworkers will be interviewed for ratings of whether they think participants interviewed have cognitive impairment or not. Other information will be collected, such as details of training, support & professional development. 4) Interviews with consultant: In each site, Dr Timms (psychiatrist) will interview (blinded) a sub-sample of 38 older homeless people. He will review participants hostel records, conduct cognitive testing, confirm presence or absence of cognitive impairment, & identify any confounding diagnoses which might have implications for service use. We aim to compare accuracy of hostel staff perceptions of whether client has memory problems vs. MMSE scores vs. expert judgement. For data analysis, open-ended responses will be coded, & themes identified in a systematic manner, borrowing on framework analysis principles. A coding framework will be developed & interpretations & commonalities will be drawn out. Quantitative analyses will examine MMSE scores by demographic & homelessness factors. Comparisons between staff perceptions, researcher scores & expert judgement will be made. Factors identified by staff which led to them successfully recognising or suspecting memory problems will be identified. WP3 uses a longitudinal case comparative approach to determine health, housing, social & other support services, received by older homeless people with & without cognitive impairment. Service pathways will be charted for both groups & costs of provision will be calculated & compared over 6 months. Two matched samples will be developed: 30 of those in WP2 who had cognitive impairment will be in an observation group (OG). Thirty without cognitive impairment will be in a comparison group (CG). Questionnaires will collect information from OG & CG at baseline, 3 months, & 6 months about problems & needs, current support, referrals & use of services, perceptions of their QoL, & unmet need. The Client Service Receipt Inventory will be customized for collecting service use. Before use, the questionnaire will be piloted with 6 people from St Mungos service user group & revised accordingly. Participants will be followed up over 6 months. Contact will be maintained if participants move; research team has experience of this. Additional service use data will be collected from keyworker, other providers & case records. Analyses will compare changes over 6 months for both groups, in relation to memory problems, support needs, contact with services & help received. Service-delivery factors influencing provision of support will be examined, including hostel staffs capacity & skills to meet this client group’s needs & availability & accessibility of local services. Personal factors influencing provision of support & persons willingness to accept help will also be examined, including age, duration of homelessness, physical health problems, alcohol problems, social contacts, perceptions of QoL & hopes for the future. Health & social care services used by people in OC & CG will be converted to costs using nationally validated data on unit costs & information from local commissioning managers. Regression modelling will explore main cost drivers, including presence (or not) of memory problems, background factors & location-specific influences. A full study report; a review of suitability of QoL instruments for homeless people; practice & resource commissioning guides will be developed.