A major challenge in care of persons with dementia is that illness and hospitalizations can lead to significant decreases in functional abilities. Maintaining functional abilities as high as possible is a key goal in dementia care, because persons are able to function optimally despite cognitive problems and engage in more preferred and enjoyable activities. Functional decline can be ameliorated through rehabilitation, building strength, balance and functional competencies. People with dementia, however, may not able to engage fully in rehabilitation due to their cognitive problems. They may not understand the therapist's instructions and may be fearful of what the therapist is trying to do.

The proposed study will test the efficacy of a multifunctional interactive computer system, the IN2L, that uses video and audio to engage people more fully in rehabilitation. Music and videos elicit and sustain responses that contribute to rehabilitation goals such as increasing balance or strength. With the IN2L, audio and video material that matches the preferences and past experience of the person in rehabilitation can be selected. While this approach shows promise with rehabilitation patients generally, it has particular potential for application with persons with dementia. Specifically, when Physical or Occupational Therapists (PT and OT) are not able to engage a person with dementia in rehabilitation tasks using verbal instructions and explanations, they may be able to involve the person using audio and video stimuli that draw upon abilities that have remained intact despite the disease.

To determine the efficacy of IN2L, the study will use a quasi-experimental design that compares rehabilitation patients with dementia seen at two comparable facilities operated by Presbyterian Senior Care Network. In one facility, the Willows, OT and PT currently use IN2L. In the comparison facility, (Southmont), OT and PT do not currently have access to IN2L. This type of design is optimal, because it prevents contamination effects that could occur with random assignment within the same facility. In that case, therapists in the control condition who were working side-by-side with therapists using IN2L would be likely to incorporate IN2L techniques in their therapy sessions if they perceived it to be useful.

The study will use observational and self-report by patients and their therapists measures to determine if persons with dementia in rehabilitation show greater participation in rehabilitation activities, better attainment of rehabilitation goals and functional improvement when therapists use the IN2L compared to usual therapy approaches.

The objective of the study is to evaluate the effectiveness of the use of a multifunctional interactive computer system, the IN2L. that uses video and audio to engage people with dementia more fully in rehabilitation. The study will compare people seen at a facility where physical therapists (PT) and occupational therapists (OT) currently use IN2L as part of the rehabilitation program (the Willows) with a comparable facility (Southmont) in a different locale that does not currently use the IN2L system. Both rehabilitation facilities are operated by the same organization, Presbyterian SeniorCare Network, which provides comprehensive residential, inpatient and outpatient services to older adults in Western Pennsylvania. The two rehabilitation units function in similar ways apart from use of IN2L. The study evaluates procedures already in place, rather than introducing a new experimental protocol. Participants in the research will receive exactly the same treatment at each facility as they would if they did not participate in the research. To evaluate effectiveness, observational measures of participants' engagement in treatment sessions and information provided by therapists about goal attainment over the course of the therapy will be used.

The study will test the following hypotheses:

1. Patients whose therapists use IN2L will show greater increases in engagement during occupational and physical therapy sessions than patients receiving treatment as usual (without IN2L).
2. Patients whose therapists used IN2L in treatment will show greater attainment of therapy goals than patients receiving treatment as usual.
3. Levels and improvement in engagement will account for (mediate) the association of treatment (IN2L use) with greater goal attainment.

The study will also test the following research question:

1. Patients whose therapists used IN2L in treatment will maintain functioning better at 3 months and have lower rates of falls and hospitalizations.

Maintaining functional abilities is a key goal in dementia care, because persons with dementia (PWD) are then able to engage in more preferred and enjoyable activities despite their cognitive problems. Functional decline is often due to inactivity, such as when a person with dementia is restricted to bed during a hospitalization, or otherwise cannot engage in usual activities. This type of functional decline as well as problems caused directly by illness can often be ameliorated through a rehabilitation program that builds strength and balance and other functional competencies. People with dementia, however, are often not able to engage fully in rehabilitation due to their cognitive problems. They may not understand the therapist's instructions and may be fearful of what the therapist is trying to do. Finding more optimal ways of engaging people with dementia in rehabilitation may help them maintain better functioning, which will allow them to engage in more usual and preferred activities. Better functioning may also lead to lower costs of care by helping people with dementia stay at home longer, or by their requiring less assistance in residential care facilities.

The IN2L was developed specifically to engage people with dementia by presenting familiar and lively visual and auditory stimuli. Given the variability in functioning among persons with dementia, the mode and content of the stimuli need to be varied. The IN2L system allows therapists to draw upon extensive visual and auditory materials so that they can present something likely to be familiar and enjoyable to the person, and if that does not work, to try something else. By giving therapists more tools other than just persuasion, they can gain the attention of persons with dementia and use the video and music to engage them in activities that address rehabilitation goals. The IN2L has been in use at several rehabilitation facilities in the country and has been found to be feasible and acceptable to patients and therapists. It has not, however, been evaluated for its impact on rehabilitation outcomes.

The present study will use a quasi-experimental two group design to compare persons with dementia who receive rehabilitation services at a facility (The Willows) that uses the IN2L to assist in treatment and a facility (Southmont) that does not use the IN2L. Both rehabilitation units are operated by the same organization, Presbyterian SeniorCare. The units have similar levels of staffing and use similar procedures. A quasi-experimental design is appropriate for this type of study. If randomization was used within a rehabilitation unit, patients and therapists in the control condition would see other patients and therapists using the IN2L, and could request use of the IN2L or perceive they are receiving a lesser form of treatment. A quasi-experimental design prevents this type of contamination. Furthermore, quasi-experimental designs with equivalent groups have been shown to be as effective as randomized trial in controlling threats to internal validity and yield reliable findings.

A potential limitation of the design is that therapists in the control unit (Southmont) were aware of the use of the IN2L at The Willows, and so could not be blinded to the intervention. During the orientation to the study for therapists at Southmont, the PI emphasized that their use of their professional skills could result in similar outcomes to The Willows. There was no indication in their comments that they believed that the IN2L was a better approach than usual treatment.

The project is recruiting 50 persons with dementia at each facility (total N = 100). All persons will be consented and where an initial validated screening indicates limited comprehension to give consent, the individual's legally appointed representative (LAR) will be contacted to give consent.

Once enrolled, participants will receive the usual treatments at each setting. Research assistants (RA) at each setting will conduct an initial interview, and conduct observations of OT and PT therapy sessions. The RAs and therapists will each complete ratings of engagement in each session, using the Pittsburgh Rehabilitation Participation Scale (PRPS). Initial reliability between the RAs and PI and the RAs and therapists were established. At completion of treatment, the RAs will conduct another interview obtaining qualitative responses to treatment, including responses to the IN2L for participants at The Willows. Following standard procedures, therapists will assess functioning across a standard set of domains at baseline, and identify individualized goals. At the completion of treatment, they will again assess functioning and indicate on a standardized scale if goals had been met.

The RAs will also conduct brief follow up interviews at one and three months after completion of treatment.

Data analysis:

To test hypothesis 1 (Patients whose therapists use IN2L will show greater increases in engagement during occupational and physical therapy sessions than patients receiving treatment as usual), multilevel models with observations nested within persons will be used. These models test for differences in slope (change) and daily variability between the treatment and control groups. Covariates such as cognitive functioning, age, gender, that show associations with the dependent measures will be included in the model.

For hypothesis 2 (Patients whose therapists used IN2L in treatment will show greater attainment of therapy goals than patients receiving treatment as usual), a MANOVA will be used with relevant covariates to test for differences in mean goal attainment between groups.

For hypothesis 3, If findings indicate that treatment affects engagement or goal attainment, a multilevel model will be used to conduct a mediational analysis to test if increases in engagement are associated with better goal attainment.

The additional research question (Patients whose therapists used IN2L in treatment will maintain functioning better at 3 months and have lower rates of falls and hospitalizations) will be tested using data obtained at the one and three month follow up interviews.

Inclusion Criteria:

• Probable dementia indicated by prior diagnosis and/or a score of 23 or less on the Brief Alzheimer Screen.

Exclusion Criteria:

• No evidence of dementia