Beyond Reminiscence

Abstract

Videos and multimedia are increasingly used to stimulate reminiscence in dementia care. However, they are also valuable in eliciting a wide range of language patterns that are not necessarily keyed to reminiscence about self. Low-technology, home-made generic and personalized videos were tested with 2 samples of persons with dementia, to increase engagement and support the retention of identity. Participants showed a slight, though not significant, preference for looking first at personalized videos and produced a wider range of conversational language topics and phrasal patterns in response to the generic videos.

Background

Dementia of any kind, especially in its most common form, Alzheimer’s disease, is more than a constant fear for Americans. The latest Facts and Figures ¹ report from the Alzheimer’s Association reports that dementia is not only the sixth leading cause of death, affecting 1 in 3 seniors, but also that it is responsible for ever-escalating health care costs. ^2,3

Dementia gradually affects memory, language, gait, and the ability to participate in normal activities of daily living, ranging from bathing to conversing and dressing. Quality of life, for both the patient and the caregiver, decreases under such conditions. The burden of care for family as well as for formal caregivers is high. Family caregivers are impacted in terms of socioemotional, economic, physical, and health-related factors ⁴ ; thus, caregiver burden can be seen as multifaceted. ⁵ In addition, formal caregivers of persons with dementia often experience burnout: for example, with little education and low pay, direct care workers have a high turnover rate. ^6,7

Cooper et al ⁸ point out in their systematic review that there is the greatest supporting evidence for the success of tailored activities for persons with dementia combined with training in coping strategies for caregivers. Almost every systematic review of caregiver support interventions points out that multicomponent interventions ⁹ delivered over time are best for caregivers ^{10 –12} so that their combination with tailored activities that could be individualized to the person with dementia would certainly be beneficial to both the person with dementia and the caregiver. For example, Gitlin et al ¹³ studied sensory (ie, videos) and single- and multi-step activities with community-living persons with dementia, while training their family caregivers in how best to enhance engagement and participation.

A number of studies of psychosocial interventions and activities or stimuli report that social interaction with persons having dementia is by far the most beneficial. ^{14 –16} Visual stimuli can provide individual engagement as well as social interaction. Accordingly, we propose that caregivers may well welcome some brief respite provided by a sensory activity such as multimedia tailored to the individual being cared for, which can incorporate stimuli for ongoing and immediate social interaction or can be viewed alone and discussed later. The ultimate goal of such tailored media is to enhance quality of life for both the person with dementia and the caregiver by increasing engagement that promotes talk.

Literature Review

Identity and Person-Centered Dementia Care

All persons, but especially persons with dementia, are dependent upon social interactions for the creation and maintenance of their personhood and sense of self. ^{17 –19} Due to the neurodegenerative effects of Alzheimer’s disease and related dementias, persons with dementia require special assistance in maintaining their personhood and self-identity. ^{20 –24} Research indicates that a relatively new paradigm of dementia care, person-centered care, is effective in rementing (ie, reversing the psychosocial degenerative course of Alzheimer’s disease), retarding, and perhaps preventing destruction of the personhood of persons with Alzheimer’s disease and related dementias. ^23,25,26

Person-centered care is based on the knowledge that the care environment plays a large role in determining the nature of the experience and the course of the disease. The mental powers that fail as a result of dementia are generally those related to thought and memory, whereas feelings and sociability need not be as seriously affected. With help and support, a person with dementia can remain in a state of well-being to a far greater extent than has otherwise been expected and maintain their identity and sense of self. ^{21 –23,27} As Vittoria suggests, caregivers should “assume there are surviving selves in the Alzheimer’s residents and endeavor to preserve, protect, support, and engage those selves as the essential part of their work.” ²⁸ As Tappen et al^29(p124) explain: “With an understanding of personhood as more than cognitive functions measured by mental status scores, we can also search for strategies that will assist caregivers in developing ways of caring that avoid dehumanization, foster the expression of feelings, and encourage therapeutic relationships.” Through a person-centered approach, it becomes the role of others in the environment to help people with dementia preserve their sense of dignity and personhood through social engagement.

Retaining Identity in Dementia

Various approaches have been suggested in the literature for understanding the conception and retention of self and identity for the person with dementia. It is generally accepted that an individual’s identity includes both internal/personal aspects and external/public aspects. Sabat and Harré ²⁴ differentiate between personal selves and social selves. Post identifies the “then self” and “now self” of the person with Alzheimer’s disease. ³⁰ Small et al ³¹ refer to the preservation of self-identity being dependent upon internal (cognitive) and external (social) conditions and their study investigated the integrity of self (internal) and personae (external). Golander and Raz ²⁰ define “personal identity” as one’s awareness of one’s self and “social identity” as the way you are perceived by and interact with those around you. They explain that people with dementia certainly retain a sense of personal identity, but avenues for self-expression are commonly broken. ²⁰ That is, the multiple personae that constitute one’s social identity, and which require the cooperation of others in order to come into being, can be lost as an indirect result of dementia. ¹⁸

Multimedia and Video as Identity Supports

These conceptions of self and identity are seen as important factors in retaining identity, despite chronic illness or cognitive impairment as with Gene Cohen’s ³² use of videotaped snapshots of old family pictures. Cohen ^32,33 developed a video intervention called Therapeutic/Restorative Biographies (TR-Bios). The biographies are created using videotaped snapshots of old photographs that are explained and narrated by family members. Preliminary study showed these TR-Bios are therapeutic for people with dementia. They are also effective in enabling families to feel they are helping their loved ones and for use by formal caregivers in institutional care communities.

Smith et al ³⁴ reported good results from involving families in creating personalized multimedia biographies; see also Guerra et al ³⁵ and Davis et al ³⁶ who involved other residents of an Assisted Living residence, some of whom were spouses. A previous pilot ³⁷ study on residents of a long-term care community helped choose personal photos to represent incidents in a friend’s or spouse’s life and recorded voice-over in the PowerPoint of the photoslides. Slide presentations were made into individualized videos and shown to recipients who lived in the secure care area of the residential community. Pulse rate and the Gueldner Well-Being Picture Scale ³⁸ were measured before and after both the designing and the viewing of the slide presentation. All participants, both with and without cognitive impairment, had a decrease in pulse rate after either making or viewing the PowerPoint presentations. In addition, scores in general on the Well-Being Scale trended upward for both groups after making or viewing the PowerPoint, with variation for individual categories such as feeling “bright” or “sharp” (as opposed to cloudy or dull). Similarly, Alm et al ³⁹ , Kelson, ⁴⁰ and Purves et al ⁴¹ utilized personalized video or interactive computer systems to stimulate conversation between persons with dementia and their caregivers. Alm et al’s Computer Interactive Reminiscence and Conversation Aid (CIRCA) system uses the choices made by the conversation partners with dementia to determine what video is made available in a particular session.

Fels and Astell ⁴² conducted a study utilizing generic photos as prompts for storytelling with people having dementia. They stress that since they did not use personal photographs, there was clearly no sense of right or wrong answers. Although the approach was effective, there were many instances in which the photographs did not elicit any stories and it was unpredictable which images resulted in stories and what the stories were about. We have noticed, however, that in Hamilton’s ⁴³ discussion of language elicited in an art museum tour, discussed subsequently, that viewers with dementia prefer easily seen colors, representations of people, and rooms or landscapes that invite the viewer “into” the painting. Similar to our focus in the present study, their emphasis was on engagement rather than eliciting accurate memories.

Sparking Talk Beyond Reminiscence

Videos and other visual stimuli can spark a range of responses in addition to talking about family members. Hatakeyama et al ⁴⁴ tested both home-made personalized videos and commercial ones that included a range of cultural and life events, finding each kind to be successful in reducing behavioral problems. Both YouTube.com ⁴⁵ and the Library of Congress’s American Memory Project (memory.loc.gov/) provide online access to video clips of events from earlier decades. For example, the American Memory Project has 54 collections for the period 1930 to 1945, which includes links to audio interviews of Pearl Harbor, photos of early baseball players, newsboys in Chicago, and the Great Depression, and Work Projects Administration posters, as seen in Figure 1.

Figure 1.

Work Projects Administration (WPA) posters from the American Memory Project.

Engagement

Engagement is defined as the act of being occupied or involved with an external stimulus. The study of engagement is a necessary foundation for the development of nonpharmacological interventions for persons with dementia:

The analysis of different forms of engagement of persons with dementia is expected to help such persons by reducing boredom and loneliness, and by increasing interest and positive emotions. It is also expected to help staff members by providing them with tools that they can utilize in caring for these persons. (Cohen-Mansfield et al, 2009, p. 2)

Identifying when someone is engaged by an artifact, an activity, a person, or a task is difficult, and measuring the duration and intensity of engagement may be even more so, particularly if that someone has dementia. Over the last 2 decades, Cohen-Mansfield and her team have added much to our understanding of the process of engagement, defined briefly as a person’s being occupied with external stimuli, particularly as it combats apathy. ⁴⁶ Their model, on which the current study is based, identifies 5 engagement outcome measures including (1) rate of refusal to engage with the stimulus, (2) duration of involvement, (3) attention to the stimulus, (4) attitude toward the stimulus, and (5) actions toward the stimulus.

The Observational Measure of Engagement is widely used ^14,47 as a pencil–paper observational tool in which the observer records duration of perceived engagement and intensity of attention using a Likert-type of rating, keyed to her theoretical model, the Comprehensive Process Model of Engagement. ⁴⁸ This measure of engagement was used in the current study. Thus, this video project was designed to measure engagement with persons with dementia that promotes talking. The focus is not on stimulating reminiscences, but rather engagement, that promotes talk: any kind of talk.

Methods

In phase 1 of our ongoing study, 8 graduate student volunteers created and showed both personalized and generic videos to 10 residents with late moderate dementia in 4 memory care units in Charlotte, North Carolina. ⁴⁹ This initial project was designed to investigate whether either or both of the 2 activities—utilizing personalized or generic videos—could be implemented by student volunteers to engage people with dementia without supervision of a staff member over a 6-week period. Their ability to do so was demonstrated clearly and consistently. We report on phase 1 here.

An in-person meeting was held early in the semester to provide training, designate the teams, and prepare the student volunteers to begin the project. Student teams were assigned to the various sites and audio- and video-recording equipment was provided to each team. Background readings and resource materials were placed on a class Web site. The group met for regularly scheduled video-conference calls in order to discuss progress and identify challenges. We met again in person in the middle and again at the end of the semester. Half of the students had received some minimal training in conversing with persons having dementia in a previous class; those who had not taken the class received no specific training, although they were provided with the PowerPointused in the conversation training.

Each student team was instructed to develop a generic video of clips from YouTube and other Internet sources that included music, television shows, commercials, and images from the 1940s and especially the 1950s, which was used with each of their participants as “Activity 2.” They were asked to include 1 news event, 1 clip of cars, 1 of fashions, and 1 of children; videos typically either showed big bands from the 1940s or used background music from the 1940s and 1950s. Each team’s generic videos aimed at being 3½ to 5 minutes long.

Utilizing photographs from the participant’s room and/or provided by the family, students next created a PowerPoint presentation for each participant which was used as “Activity 1.” They were asked to develop the PowerPoint into a personal video of about 4 minutes, using the participant’s photographs and memorabilia. Soft “elevator” music was used in the background.

Each pair of students worked with 2 or 3 individual consented persons having dementia at a particular residential community, for interactions lasting from 5 to 20 minutes each time, depending on resident’s interest in one or both activities. Student researchers were trained to begin with an introduction, greetings, and small talk. One student took the lead and introduced the second student; the same student took the lead with a particular participant each time. Students were reminded that under no circumstances should they ask “Do you remember us”? After the introduction, they were trained to continue with:

“I brought you two things for us to look at, if you’d like. We’ll look at them on this little computer.” [Turn on computer, pull up screen with two icons, one for each PowerPoint]

“Tell me which one you want to look at first.”

Although participants had the opportunity to look at any of the videos more than once during a session, none of the participants chose to do this. Both videos were shown at each session, with roughly 90 seconds between them as the student closed 1 video on the laptop and opened the other one.

Students were reminded during training and during the online conferences and meetings throughout the experiment that the purpose of the video project was not to stimulate reminiscences but rather to identify engagement that promotes talk. Interactions were videorecorded by one of the 2 members of the student team, using a flipcam.

Success of the interactions and 2 activities was measured in several ways. Measures of engagement were assessed using scales adopted and adapted from Cohen-Mansfield et al, 2010. These scales include the following: (1) rate of refusal to look at each activity; (2) duration of engagement with each activity; (3) choice of activity to view first; (4) attitude toward each activity measured on a one-line, 7-point scale, reporting “____’s reaction on ____(date) was: very negative … to … very positive,” and (5) action toward or stimulated by each activity, annotated in field notes of motor or verbal behavior in response to the activity. The student not holding the camera took field notes and recorded responses to the adapted engagement scales.

Feasibility was demonstrated by resident willingness to participate in 5 or more sessions, and engagement was taken as the average level for each of the 3 countable measures described earlier. These 3 countable measures are the rate of refusal to look at each activitity, the duration of engagement with each activitity, and their attitude toward each activitity. Each student team submitted a set of field notes along with these engagement measures and the video transcripts.

We next identify the characteristics of the first cohort and present our findings in using the Observational Measurement of Engagement tool. ⁴⁸ This tool records rate of refusal to engage with the interventional videos, the choices made about which to watch first, the length of time spent watching each type of video, and the attitude toward each video, as observed by the research assistants who video recorded the interactions of the participants with the video. The Functional Assessment Staging (FAST) scale was completed by the faculty principal investigators with a staff caregiver to evaluate the functional level of each participant. ⁵⁰ A higher score indicates less function. Table 1 displays the characteristics of the 10 participants.

Table 1.

Demographics of the First Cohort of Participants With Dementia.

Participant	Age	Gender	Race	FAST
C2	73	Female	White	5
C3	83	Female	White	5
C4	85	Female	White	6
E1	87	Male	White	4
MS3	67	Female	Black	6
MS4	84	Male	Black	2
PE1	89	Female	White	5
PH1	97	Female	White	7
PH2	82	Male	White	7
PH3	96	Female	White	7

Abbreviation: FAST, Functional Assessment Staging.

To analyze language patterns elicited while viewing the videos, we have followed Hamilton ⁴³ (cf Davis et al ³⁶ ) in her analysis of the response of persons with Alzheimer’s disease to art exhibits. Hamilton comments that most people, in contact with a work of art, can describe some part of it, could present some kind of evaluation, interpretation, judgment, visceral reaction, fanciful imagined event, or connection to some other work of art, and could provide personal or general memories.

Persons with dementia maintain multiple kinds of speech acts. In talking about art or paintings, Hamilton found they demonstrated the following:

Perceive

There’s a lot of candles on that cake

Act

I agree with you

Evaluate

Mmm, that looks comfortable

Remember/general

They didn’t wear mini-dresses in those days

Remember/personal

We lived in a big old house

Imagine

There could be bad news in that letter.^43(p.176)

Findings and Discussion

Measures of Engagement: Rate of Refusal

All 10 of the participants agreed to watch both of the videos during each of the 6 visits. Each participant seemed willing to be engaged with interaction with one of the graduate students and with watching a video. There were no refusals.

Measures of Engagement: Duration of Engagement With Each Activity

In general, residents spent almost exactly the same amount of time viewing the generic video during each of the 6 sessions as it was in length. They spent almost the same amount of time viewing the personal video during each session as the length of the personal video. The average amount of time spent watching the personal videos had slightly more variation among the viewers. Interestingly, 2 viewers, PE1 and E1 determined the greater variation between the 2 sets of videos by spending less time viewing the personal video and more time viewing the generic video (Figure 2).

Figure 2.

Average time spent viewing personal photo video compared to video length.

Measures of Engagement: Choice of Activity to View First

Each of the 10 participants agreed to view both videos during each of the 6 visits, and 9 of them made a choice of which video to watch first all 6 times. The final participant indicated a choice of which video to watch first, 5 of the 6 visits, but did not indicate a choice during her final session. One-half of the participants chose to watch each video first during one-half of the visits. Four of the participants chose the personal video more often (ie, 4 or 5 times), while the remaining participants always chose to watch the generic video first.

Measures of Engagement: Attitude Toward Each Activity

The participant’s attitude toward each activity was measured on a 1-line, 7-point, Likert-type scale, reporting “____’s reaction on ____(date) was: (1) very negative … to … 7) very positive.” This assessment was completed by the student researchers at the conclusion of each of the 6 sessions. According to the individual averages, 7 of the participants preferred the personal videos, 1 preferred the generic video, and 2 enjoyed them equally.

The participants’ actions toward or stimulated by each of the 2 activities are being assessed here through the student researchers’ field notes of motor or verbal behavior in response to the activity. In general, the participants’ reactions to the personal videos were positive and included lots of smiles. The verbal reactions to the personal videos tended to be comments about their family members and friends who were pictured. For example, student’s notes about 1 participant’s engagement with their personal video indicated: “Resident comments on seeing family members in photos. Identified all family members. States: beautiful, my grandchild, smiles, very engaged.” Some participants looked away from the screen to locate the photos and items in their room and were curious about how the student researcher got the pictures.

Although differences in response to the videos were not statistically significant, they indicate that the participants as a group slightly preferred the personal videos (5.58 average score compared to the generic videos (5.18 average score). The participants’ reactions to the generic videos varied widely but were consistently qualitatively different from their responses to the personal videos. Although the personal videos elicited comments about specific people and relationships, the generic videos elicited various forms of engagement. Some of the participants commented on and/or moved to the music. Several of them read the captions and advertisements that were pictured and talked about the commercials and television shows. More of them were reported to have talked throughout the generic video as compared to the personal videos.

Adding Pulse Measurement

For phase 2, with a second set of 2 residents and 2 graduate students, we added the measure of pulse with an oximeter. The data suggested that variations in pulse rate implied a heightened level of engagement while watching the videos. The heightened engagement indicated by change in pulse rate correlated with the field notes taken by the graduate student researcher on brief intervals when participant 2 showed some excitement while seeing a video clip (Figure 3). Participant 1 had a more stable pulse rate (ie, no discernible variations); the graduate student reported that participant 1 was either less engaged or fell asleep most of the time and therefore did not have good engagement levels.

Figure 3.

Pulse rates (Kailas A, Personal manuscript, 2013).

Analysis of Elicited Language Patterns

One of the graduate students also served as the project research assistant. She transcribed talk by the residents on the videos taken by the students while the residents were watching the PowerPoint videos; the transcripts were reviewed and edited by the first author. Using the same methods as Hamilton, ⁴³ we tallied (1) participant turns-at-talk by number of minimal responses, single clause responses, and responses of more than one clause; (2) assertions versus questions; and (3) whether they were immediate responses of perception, action or evaluation, or displaced responses keyed to general or personal memory or imagination. Interestingly, with the exception of 1 person who nodded, smiled or frowned, but did not speak, the generic video consistently elicited more diverse language patterns than the personalized video.

We used Hamilton’s coding to examine participant responses, color-coding them for visual effect:

• Perceive = PURPLE

• Evaluate = RED

• Act = GREEN

• Imagine = YELLOW

• Remember-general = ORANGE

• Remember-personal = BLUE (Summarized to protect anonymity)

The color-coding in Figure 4 shows the complexity of language patterns in response to a generic video.

Figure 4.

C2’s language patterns in response to the generic video.

Limitations

Although our findings are suggestive, each phase has included a small sample with no ethnic diversity and skewed gender representation. Phase 2 included 1 male and 1 female, but all members of phase 1 were female and none were minority. This is a limitation from our recruitment strategy: although we recruited persons from 2 different memory care units, each unit had very few males or non-Anglo residents. In addition, all were first-language American English speakers. In future trials of the impact of generic video, we will need to expand our recruiting base.

Conclusions

We report the results of our on-going study to utilize inexpensive, low-level technology to engage people with dementia. Our research employing personal and generic videos to engage people with dementia has generated several key findings. First, we have found that both personal and generic videos can be implemented by minimally trained individuals to engage people with dementia. Second, we have shown that both personal and generic videos are effective in engaging people with dementia and third that different individuals have slight preferences for one or the other type of video. Fourth, generic videos generate more diverse comments on a broader range of topics compared to the personal videos.

The implications of these findings are important both in terms of meeting the needs of people with dementia, along with supporting both family and professional caregivers. Both generic and personal videos can be used by caregivers to increase engagement in the people with dementia for whom they are caring, whether in in-home or group settings. Generic videos offer unique possibilities not only because of our findings about the variety of dialogue they can generate but also because those findings suggest they can be used effectively with groups of residents. Further research should study the use of generic videos to engage people with dementia in small groups. We look forward to additional exploration of these inexpensive, low-technology approaches to engaging people with dementia while at the same time supporting their family and professional caregivers.