Evaluating the Benefits of the TimeSlips Creative Storytelling Program for Persons With Varying Degrees of Dementia Severity

Abstract

Objectives:

To evaluate the benefits of TimeSlips, a group creative storytelling intervention used in residential care settings, on quality of life (QOL), interactions with caregivers, and Mini-Mental State Examination (MMSE) scores for persons with varying degrees of dementia severity.

Design:

A mixed-methods approach was used weekly over a 6-month period to measure QOL and resident–caregiver relationships.

Setting:

A dementia care unit in Pennsylvania.

Participants:

Twenty-two residents with mild-to-severe dementia.

Measurements:

Dementia severity and QOL were assessed using the MMSE and Greater Cincinnati Chapter Well-Being Observational Tool. Resident–caregiver interactions were analyzed using a modified version of the Quality of Interactions Schedule. Semistructured interviews were conducted upon the study’s conclusion.

Results:

Quantitatively, participants initially classified with mild–moderate dementia were significantly more likely to experience positive benefits compared to those initially classified with severe dementia. There were no significant changes in dementia severity over time. There was also no change in QOL or resident–caregiver relationships for those with mild–moderate dementia over time, although there was a decrease in certain measures of QOL and resident–caregiver relationships for those with severe dementia. Qualitative analysis identified consistent benefits for residents with both mild–moderate and severe dementia over time.

Conclusions:

Mixed-methods analyses helped identify benefits of TimeSlips for persons at all levels of dementia severity, but particularly for those with milder dementia. Such an observation helps demonstrate how arts-based programs like TimeSlips can uniquely benefit people with advanced memory impairments and thereby support QOL.

Introduction

Over 46 million people live with dementia worldwide, and it is estimated that this number will increase to 70 million in 2030 and 131.5 million by 2050. ¹ At present, there is neither a cure for dementia nor an effective treatment that slows or stops its progression, and the estimated global cost for the condition is US$818 billion. ^2,3 Given the limited efficacy of pharmacological interventions in managing symptoms, as well as the need to move away from overuse of antipsychotic medications in residential care settings, ⁴ evidence-based nonpharmacological treatments have emerged in recent years to improve care and decrease problems experienced by people with dementia, such as behavioral disturbances, depression, and caregiver burden. ⁵ One such nonpharmacological intervention is TimeSlips, a group creative storytelling initiative developed in the 1990s that is now used across the world. ⁶

Creative expression programs such as TimeSlips are grounded within a “person-centered” care framework. Unlike traditional reminiscence therapies that use biographical details to capture a sense of who a person with dementia (PWD) was in the past, TimeSlips encourages participants to produce something novel that has meaning to them and to others. ⁷ The activity involves facilitators using surreal, staged visual images to encourage participants to respond in a spontaneous, improvisational manner as part of a storytelling group. By engaging PWDs in a creative, failure-free activity in which there are no “right” or “wrong” answers and all contributions are considered valuable and significant, TimeSlips aims to support remaining strengths and capacities while creating a meaningful social role for residents in an institutional setting. Although evidence of the program’s efficacy is limited, preliminary studies have shown that participation positively benefits PWD, improving engagement and alertness while fostering more frequent staff–resident interactions and more positive staff views of the residents, ^8,9 engendering pleasure and positive affect in participants, ¹⁰ increasing creativity, reducing agitation and anxiety, ¹¹ improving quality of life (QOL), and positively altering behavior while further producing benefits for the morale of staff members and nursing home community as a whole. ^8,12

There are, however, no known studies that have analyzed the effects of TimeSlips beyond a 10-week period or brought mixed methods to bear in carrying out such an analysis. Moreover, while TimeSlips is intended for persons with mid- to late-stage dementia, few studies have stratified participants by their “stage” of dementia (eg, mild, moderate, and severe) to determine which group(s) of PWD experience the greatest relative effects of the activity. This study sought to address these gaps in the literature by using both quantitative and qualitative methods to evaluate over a 6-month period which patient population experienced the greatest relative benefits of TimeSlips as measured by QOL and interactions with caregivers.

Methods

Study Site

This study was conducted in the dementia unit at “CM”—a retirement community in a rural town in south central Pennsylvania. CM is home to over 250 residents in independent, personal care and skilled nursing settings.

Study Sample

Study participants were selected via convenience sampling. All 32 residents living in the secured dementia unit at the time of the study were eligible to participate. The principal investigator (PI) first met with each potential participant in person and assessed their mental status via the Mini-Mental State Examination (MMSE), a 30-point questionnaire used in clinical and research settings to measure cognitive deficiency. ¹³ If participants scored ≥24, the PI (A.A.V.) discussed the study with the potential participants directly and provided the consent form if potential participants were interested. Consent was obtained from legally authorized representatives of eligible participants who were incapable of providing consent themselves, that is, scored <24 on the MMSE. For example, for those who scored <24 on the MMSE, the PI contacted their legally authorized representatives via mail, providing a cover letter with an explanation of the study, a consent form, an opt-out card, and a self-addressed stamped return envelope. Instructions were given to either (1) mail back the opt-out card or (2) wait for the PI to call. The PI then called the legally authorized representative if she did not receive an opt-out card, explained the study, and potentially received consent to participate. Signed consent forms were then mailed back after a discussion took place.

Twenty-two of those residents who met the inclusion criteria (residence in CM, diagnosis of dementia) consented to participate in the study. Eligible residents declined to participate for the following reasons: (1) legally authorized representatives felt their family member’s health was deteriorating too rapidly or the legally authorized representatives were not interested in having their family member participate (n = 5), (2) eligible residents relocated from CM prior to the start of the study (n = 2), (3) legally authorized representatives could not be reached after multiple attempts (n = 1), (4) eligible residents did not participate in TimeSlips prior to the initiation of the study, so they were unlikely to participate during the study period (n = 1), and (5) legally authorized representatives chose not to read through the consent form (n = 1). Table 1 describes the demographic information of each participant and their mental status throughout the study as determined by the MMSE. Given the small sample size, MMSE categories of “questionably significant dementia” (scores of 25-30), “mild dementia” (scores of 20-25), “moderate dementia” (scores of 10-20), and “severe dementia” (scores of 0-10) were combined into 2 groups: mild–moderate dementia (scores of 10-30) and severe dementia (scores of 0-10). Of note, all participants were categorized as “Caucasian” and were between the ages of 80 and 97. On average, each participant had spent approximately 2 years in the dementia care unit. Several participants did not have complete MMSE scores as they either died or left the dementia unit during the study.

Table 1.

Study Sample Demographics at Time of Study.

	June MMSE Scores		August MMSE Scores		October MMSE Scores		December MMSE Scores
	Mild-Moda	Severe	Mild-Mod	Severe	Mild-Mod	Severe	Mild-Mod	Severe
Total Participants (N = 22)b	11	11	10	11	8	12	8	6
Males (n = 6)	4	2	4	2	3	2	4	0
Females (n = 16)	7	9	6	9	5	10	4	6

^a Mild-Mod indicates mild–moderate dementia.

^b The total number of participants decreased over the course of the study, as several participants died or left the dementia unit.

Data Collection

Data collection was conducted by one member of the study team, the PI (A.A.V.), who also attended all TimeSlips sessions. Although both participants and nonparticipants participated in TimeSlips each week, the PI only recorded data for the participants.

This study began in June 2016 and ended in January 2017. During that time, 22 participants were observed over the course of 21 sessions as they participated in the weekly, hour-long TimeSlips program at CM. Dementia severity was assessed prior to the start of the study and subsequently 3 times during the study at 2-month intervals using the MMSE. Assessment scales and observation tools were utilized to measure dementia severity, QOL, and resident–caregiver relationships during each weekly session. Brief field notes were written during each TimeSlips session by A.A.V. and were expanded upon 20 minutes after the conclusion of each session. The data collected were deidentified and were documented in an excel spreadsheet.

Measures

With regard to QOL, there is generally a lack of consensus on how to measure this concept with PWD. ¹⁴ However, one QOL tool that has been validated is the Greater Cincinnati Chapter Well-Being Observation Tool, an instrument that evaluates 7 domains of well-being: interest, sustained attention, pleasure, negative affect, sadness, self-esteem, and normalcy. ^15,16 This tool has specific indicators for each of the domains of well-being, giving objective criteria for each category. For example, in the context of this study, a person who was angry during TimeSlips or appeared agitated during the activity would be categorized as having a “negative affect.” Each of the 7 domains are then rated on a scale of 0 to 4 in order to express the amount of time in which the participant experienced that domain during TimeSlips. For example, a score of “0” indicates that a resident “never” spent time experiencing a particular domain. A score of “1” indicates a resident “rarely” spent time in that domain, or approximately one-fourth of the program. A score of “2” indicates that a resident spent “some” of the time in a domain, or roughly one-half of the program. A score of “3” indicates a resident spent “most” of the time, or three-fourth of the program in a domain, while a score of “4” indicates a resident spent the “full duration” of the program in a particular domain. In this study, A.A.V. observed participants and documented their actions as they engaged in TimeSlips throughout the duration of the 6-month study and analyzed QOL based on the tool’s designated criteria.

In order to measure resident–caregiver interactions, several approaches were utilized. As has been established as best practice in previous studies, a single member of the team—A.A.V.—observed and documented resident–caregiver interactions and analyzed them using a modified version of the Quality of Interactions Schedule (QUIS). ^9,17 The QUIS stratifies interactions into the following categories: positive social, positive care, negative protective, and negative restrictive and lists examples of each type of interaction. Participants’ actions were also categorized into the following using the QUIS: disengaged, nonsocial engagement, social engagement, and challenging behavior. Similar to the scoring of the Greater Cincinnati Well Being Observation Tool, A.A.V. recorded whether each participant’s interaction met a specific QUIS interaction category and then rated the frequency of interactions on a scale of 0 to 4. A score of “0” indicated that a particular interaction did not occur, with “1” indicating that of all the interactions during the session, a particular interaction occurred 25% of the time, “2” representing that a particular interaction occurred 50% of the time, “3” representing 75% of the time, and “4” signifying that the particular interaction was present 100% of the time (ie, the exclusive mode of interaction for a participant).

Table 2 describes the corresponding definitions to the following terms utilized in the Greater Cincinnati Well Being Observation Tool and the QUIS: interest, sustained attention, pleasure, negative affect, sadness, self-esteem, normalcy, positive social, positive care, negative protective, negative restrictive, disengaged, nonsocial engagement, social engagement, and challenging behavior.

Table 2.

Corresponding Terms and Definitions.

Term	Definition
Interest	Participant makes eye contact, eyes following object or person; attempts to socialize by extending hand, pat on shoulder; turning toward or moving body toward person; chats with others (does not have to have sustained conversation or even intelligible conversation); smiles; offers and receives support from others during session
Sustained attention	Participant is able to attend to project or activity for 10 minutes; once started on the activity, participant requires rare prompts to complete simple steps to complete the task at hand, or to return to the activity; if verbal prompts or cues required, participant returns to activity and works or pays attention until distracted; participant may engage in conversation with facilitator during the activity but returns to activity when chatting is finished.
Pleasure	Verbal expression of pleasure while participating in the actual activity; eyes crinkled, smiles, laughter, relaxed facial expression; nods positively, relaxed body language
Negative affect	Closed body language, frown on face, angry verbal outbursts, facial grimacing or brows furrowed, psychomotor agitation (hand tapping, moving in chair, leg jiggling, wincing); rapid breathing, eyes wide, frightened
Sadness	Flat affect or weeping quietly; verbalization of feeling sad over situation; eyes drooping; sighing, head in hand, eyes/head turned downward and face expressionless
Self-esteem	Verbal expression of pride and satisfaction; nonverbal expression of pride, pleasure, and satisfaction of having completed an activity (clapping, smiling, tearfulness, nodding head; expressing an internal sense of pride verbally through reminiscence in response to piece of art created by the individual)
Normalcy	Verbal expression of experiencing a sense of normalcy, which arises from belonging to a group and deriving enjoyment from participant in a group activity
Positive social	Interaction principally involving good, constructive, beneficial conversation, and companionship
Positive care	Interactions during the appropriate delivery of physical care, for example, (1) interactions in which caregivers physically helped residents (eg, a caregiver helping a resident find a seat in the group) and (2) interactions in which caregivers encouraged general participation during TimeSlips (eg, a caregiver asking a resident his thoughts about a picture). Unlike “positive social” interactions, “positive care” interactions do not involve in-depth conversations or deeper connections between caregivers and residents
Negative protective	Providing care, keeping safe, or removing from danger but in a restrictive manner, without explanation or reassurance
Negative restrictive	Interactions that oppose or resist residents’ “freedom of action” without good reason, or which ignore participants as a person
Disengaged	Resident is inactive, sitting passively or sleeping, or involved in unpurposeful activity such as fiddling with clothes, smoking, or aimless walking
Nonsocial engagement	Resident is engaged in purposeful activities that do not involve social interaction with others; for example, combing hair, reading, watching television, actively listening to music, knitting
Social engagement	Resident is engaged in some form of communication with others where there is a state of reciprocity with at least one other person or the resident is initiating contact with another person
Challenging behavior	Resident is engaged in solitary repetitive, nonfunctional motor activity, self-injury, aggression to others, damage to property, or other inappropriate behavior

Finally, semistructured interviews were conducted at the completion of the study by A.A.V. with 2 of the participating caregivers and 2 of the participating residents to gain a richer understanding of their perceptions of TimeSlips. Caregiver interview questions underwent multiple iterations prior to use and were provided prior to the interview. Caregiver 1 and caregiver 2 were selected for the interviews since they facilitated the majority of the TimeSlips sessions. Participating residents were selected to be interviewed based on their convenient accessibility and willingness to be interviewed. Given the potential burdens that can be imposed on research participants affected by cognitive frailty, participant interviews consisted of 3 brief questions that also went through multiple iterations. These questions were previously piloted in order to better understand participants’ capacity to respond. Interviews were conducted in a private room by one researcher (A.A.V.). Only one set of interviews was conducted, and these conversations were digitally recorded and transcribed. Transcripts were not returned to participants for comment; however, A.A.V. asked clarifying questions during each interview as needed.

Ethical Approval

Approval of the study was obtained from the Penn State Hershey institutional review board (study ID: 00004985). Written consent from participants or their legally authorized representative was required.

Intervention

Dr Anne Basting, a full professor of Theatre at University of Wisconsin–Milwaukee who specializes in strategies for incorporating the arts and humanities into the long-term care of people with cognitive disabilities, created TimeSlips in 1998. TimeSlips is now a therapeutic modality used around the world, and the organization became an independent nonprofit in 2013. Currently, there are active facilitators in approximately 44 states and 15 countries.

Although TimeSlips was initially designed to be used with long-term care residents affected by mid- and late-stage dementia, it also can be used with those not residing in long-term care facilities. Ultimately, it encourages participants to be performative in the moment rather than relying on long- or short-term memory recall. During the program, participants form into a storytelling circle and receive a staged picture (eg, a child stealing a cookie from a jar); a facilitator then prompts the group with a series of open-ended questions in reference to each picture that seeks to elicit imaginative response. Participants are encouraged to vocalize anything that comes to mind, and all verbalizations—including noises, gestures, and songs—are incorporated into the writing of one composite story that is ultimately read back to the group. By fostering a “failure-free” environment and including all narrative contributions from participants, TimeSlips aims to stimulate community interaction, creativity, and belonging while affirming each person’s contribution to the collective construction of stories. In this study, each TimeSlips session at CM lasted approximately 1 hour. One member of the research team (D.R.G.) is a trained TimeSlips facilitator and had previously introduced the intervention to CM in 2010. All staff who facilitate TimeSlips sessions have received formal training.

Caregiver Involvement

In general, 2 CM caregivers, current CM staff members with formal TimeSlips training, worked in the dementia unit and facilitated the weekly TimeSlips sessions on a rotating basis in alternating weeks. If neither of the 2 primary caregivers were available, 2 additional CM caregivers would facilitate the weekly sessions. All 4 of these caregivers were eligible to participate in this study, and each elected to participate and consented to the study. All 4 caregivers were female.

Statistical Analysis

Our main study objectives were to use mixed methods to (1) determine the effects of TimeSlips on participant QOL; (2) determine the effects of TimeSlips on resident–caregiver interactions; (3) determine which resident population, that is, those with mild–moderate dementia versus those with severe dementia, experienced the greatest relative benefits of TimeSlips over a 6-month period; (4) determine whether scores, such as the MMSE, changed over time; and (5) identify potential qualitative themes and trends not immediately evident in the quantitative analysis.

We used linear and generalized linear mixed-effects models for the outcome variables, which are appropriate for longitudinal data analysis. ^18,19 All participants were engaged in TimeSlips, so there were no randomized groups for comparison. For analyzing the MMSE total score, we used a linear mixed-effects model to determine whether there was a statistically significant time trend. In this statistical model, we only included a fixed-effect regression parameter for the elapsed time from the initial visit and a random effect for the study participant. For analyzing the QOL outcomes, the caregiver outcomes, and the resident engagement outcomes, we used a generalized linear model that consisted of ordinal logistic regression with fixed-effect parameters for the participant’s initial MMSE total score and the session number and a random effect for the study participant. We then estimated the odds ratio for a change in the outcome variable with respect to a 1-unit change in the session number (elapsed time).

Finally, we partitioned the participants into 2 subgroups of 11 participants each—namely, those with and without a “severe” initial MMSE score, and we conducted the analyses described above within each subgroup. Because there were 10 outcomes of interest, we applied Hochberg’s step-down procedure to adjust the P values for determining statistical significance within each subgroup. ²⁰

Qualitative Analysis

Additionally, A.A.V. conducted a preliminary qualitative review at the conclusion of the study using a thematic analysis approach. This approach was chosen because of its flexibility and ability to identify patterns and make interpretations from qualitative data. ²¹ First, A.A.V. reviewed original documentation of each TimeSlips session, which contained detailed information describing each participant’s behaviors and interactions during each of the 21 TimeSlips sessions; relevant direct quotes were also documented. Session date, session number, caregiver involvement, and dementia severity relating to each participant entry were noted. Data were then reorganized based on similar content and A.A.V. began to generate initial codes such as “memory,” “laughter,” “sharing,” and “unengaged.” For example, field notes that mentioned similar phases such as “participant remembered,” “participant recalled childhood stories,” or “participant reminisced” were grouped together. Quotes were also utilized in this process. For example, several direct quotes began with “I remember…” These were grouped with the other notes mentioned above. Through this reorganization and coding of the data, A.A.V. noted recurrent characteristics and trends over the 6-month intervention. Through this process, 3 major themes emerged from the data analysis.

Results

Our mixed-methods approach demonstrated overall positive benefits of TimeSlips for both groups of participants, those with mild–moderate dementia and severe dementia. These quantitative analyses demonstrated a statistically significant benefit in QOL for those with mild–moderate dementia, while qualitative analysis revealed 3 main themes (bonding and interactions, memory reminiscence, and laughter) that potentially explain the overall QOL benefits. Additionally, qualitative analysis further describes potential QOL benefits for those with severe dementia, a nuance that was not present in the quantitative analyses alone.

Statistical Results

Our statistical analyses demonstrated that those initially classified as having mild–moderate dementia were more likely to experience increased interest, pleasure, self-esteem, normalcy, and sustained attention during TimeSlips compared to those with severe dementia. Additionally, those with mild–moderate dementia were more likely to exhibit an increase in positive social interactions and an increase in positive care interactions with their caregivers during TimeSlips than those with severe dementia. Participants with mild–moderate dementia were also less likely to be disengaged than those with severe dementia. Additionally, participants with mild–moderate dementia exhibited greater nonsocial engagement and social engagement than those with severe dementia. These results are summarized in Table 3.

Table 3.

Odds Ratios and Corresponding Confidence Intervals for a 1-Unit Change in (1) the TimeSlips Session Number and (2) the Initial MMSE Total Score.

	Initial MMSE, Total Score
Quality of life: Interest	1.20a (1.11-1.30)
Quality of life: Sustained attention	1.24a (1.12-1.38)
Quality of life: Pleasure	1.23a (1.12-1.34)
Quality of life: Negative effect	0.98 (0.89-1.07)
Quality of life: Self-esteem	1.24a (1.10-1.40)
Quality of life: Normalcy	1.27a (1.16-1.40)
Caregiver: positive social	1.12a (1.07-1.17)
Caregiver: Positive care	1.14a (1.06-1.23)
Patient disengaged	0.84a (0.78-0.92)
Patient nonsocial engagement	1.19a (1.00-1.42)
Patient social engagement	1.22a (1.11-1.34)

^a Results indicate statistical significance, with P < .05. Of note, the odds ratio (OR) represents the odds of a higher outcome score for a session X + 1, relative to the odds of a higher outcome for session X. In other words, it depicts if later sessions display higher outcome scores than earlier sessions. For example, when looking at “Quality of life: Interest”, the estimated odds ratio is 1.20. This indicates that the odds are 1.2 times higher for an individual to have a higher value of the “Quality of life: Interest” score at visit 12 than the odds at visit 11.

When looking at changes over time, those with severe dementia experienced a decline in interest during TimeSlips over the 6-month course of the study. These participants also demonstrated less social engagement and had less positive care interactions with caregivers over the course of the study. In comparison, those with mild–moderate dementia experienced no change in interest over the 6-month study and did not demonstrate a change in social engagement or a significant decline in positive care interactions over the course of the study. These results are summarized in Table 4.

Table 4.

Odds Ratios and Corresponding Confidence Intervals for a 1-Unit Change in TimeSlips Session Within initial MMSE Nonsevere and Initial MMSE Severe Categories.

	TimeSlips Session Numbera (MMSE Nonsevere)	TimeSlips Session Number (MMSE Severe)
Quality of life: Interest	1.00 (0.93-1.07)	0.86b (0.78-0.95)
Quality of life: Sustained attention	1.02 (0.95-1.10)	0.91 (0.82-1.01)
Quality of life: Pleasure	0.94 (0.91-1.04)	0.88 (0.79-0.97)
Quality of life: Negative effect	0.90 (0.78-1.03)	1.02 (0.88-1.17)
Quality of life: Self-esteem	1.02 (0.94-1.11)	0.95 (0.83-1.09)
Quality of life: Normalcy	0.97 (0.90-1.04)	0.88 (0.79-0.98)
Caregiver: Positive social	1.06 (0.99-1.14)	0.94 (0.85-1.04)
Caregiver: Positive care	0.92 (0.86-0.99)	0.87b (0.79-0.96)
Patient disengaged	1.00 (0.93-1.07)	1.11 (1.01-1.23)
Patient nonsocial engagement	–	–
Patient social engagement	1.00 (0.93-1.08)	0.86b (0.78-0.96)

^a “TimeSlips session number” refers to the fact that the study was performed during the latter 6 months of 2016. Sessions were enumerated from 1 (July 5, 2016) through 21 (December 13, 2016).

^b Results indicate statistical significance after application of the Hochberg step-down procedure, with P < .05. For example, when looking at “Quality of life: Interest,” the estimated odds ratio in the MMSE severe category is 0.86. This indicates that the odds of an individual to experience a greater interest at visit 12 versus the odds at visit 11 are decreased by 0.86 times.

As mentioned above, participants’ MMSE scores were determined 4 times over the course of the study. Although many participants experienced changes in their MMSE scores over time, only 2 participants experienced changes in dementia severity categories over the course of the study. Both of these participants started in the “mild–moderate” dementia severity group and ended in the “severe” dementia severity group. We performed a sensitivity analysis in which we eliminated those 2 participants from the data analysis, and the resultant odds ratios were nearly identical to those of the original analyses, suggesting that these changes do not significantly alter our results or conclusions.

Although statistically it appears those with severe dementia do not experience the same relative positive benefits as those with mild–moderate dementia, qualitative analysis identified possible benefits of TimeSlips for these participants.

Qualitative Results

Theme 1: Bonding and interactions

Increased connections and interactions between participants was a prominent theme of the TimeSlips sessions that appeared to enhance participant QOL. During every session, at least 2 residents interacted with each other by having their own side conversations, sharing pictures, or helping each other in some way (ie, assisting in sitting down, moving to create space for a new participant, picking up personal belongings that were dropped, and returning them to the owner). While several residents sat next to and spoke with the same people each week, new conversations between different participants were continuously occurring. When asked the question “What do you like most about TimeSlips?” Participant 7, a resident with mild–moderate dementia stated, “It’s interesting to find different things every time.” Of note, participant 7 often sat next to participant 1, another resident with mild–moderate dementia. Together, the 2 participants would collaboratively discuss the picture.

At the end of the intervention, several of the caregivers spoke directly to these interactions and relationships between participants. Caregiver 1 commented, “I think it’s a wonderful bonding experience because it’s different than me sitting there teaching them to do a craft or trivia. It’s just a new way for us to interact once a week, to joke around…” Caregiver 2 commented,

The thing that surprises me is…how different people can see different things and it’s just always neat…to see the interactions between the residents too because you know, you have residents who are more coherent and will say ‘oh, that’s not a little boy or that’s not a little girl’…it’s interesting to see them interact.

Bonding and interactions between participants were not limited to just those with mild–moderate dementia but occurred across all levels of severity. For example, one participant with severe dementia would smile and “babble” in response to her neighbor’s comments. Other participants with severe dementia would share their TimeSlips picture with their neighbors. One participant with severe dementia even acknowledged her neighbor and said, “Thank you!” when her neighbor retrieved the picture that she dropped. While these interactions were more limited than those of participants with mild–moderate dementia, they were still present.

Theme 2: Memory reminiscence

TimeSlips provided participants with the opportunity to spontaneously reminisce about their lives and childhoods. In over half the sessions, the pictures used in the TimeSlips sessions triggered participants to talk about their personal histories. For example, participants told stories about bicycles they used to own, activities they participated in at county fairs, their children, and so on. When looking at a picture of a baseball player during one session, several participants started to sing the song “Take Me Out to the Ballgame.” Other pictures prompted residents to talk about the perms they used to get at the hairdresser or the cars they used to drive. When looking at a picture of leaves falling in autumn, one resident remembered “bobbing for apples” and “dancing” at a fall festival. The caregivers also noticed that TimeSlips informally elicited memories. When discussing her perception of TimeSlips, caregiver 1 stated, “I think it’s a great program…a great way for our residents to open up and really use their imagination. Like the other day when [participant] used the name ‘Joanna’—that was a friend he used to have here.” Although caregiver 2 explained that “TimeSlips doesn’t always [foster] a cohesive group” she echoed caregiver 1’s sentiment, stating “…it’s beneficial for all of them actually, because…you get different perceptions from all of them from different levels.”

While the majority of participants who reminisced about their lives and recalled memories during TimeSlips had mild–moderate dementia, TimeSlips also evoked memories for several participants with severe dementia. For example, when prompted by the caregiver, a participant with severe dementia talked about the milkman who used to come to his home. Caregiver 1 commented, “Well, I’ve even seen some residents who never talk…they’ll make noises or they’ll say something [in reference to a picture]. It’s exciting for me to watch.”

Theme 3: Laughter

Finally, laughter was a prominent theme throughout each TimeSlips session, being observed in both in-person sessions and in the caregiver interviews. Laughter was also perceived as improving QOL and was observed in participants with both mild–moderate and severe dementia. At least 3 residents laughed during each of the 21 TimeSlips sessions. In reflecting on her participation in the sessions, participant 1, a resident with mild–moderate dementia, affirmed, “It’s unique. We are always laughing!”

Conclusions

This study sought to understand which groups of PWD experience the greatest relative benefits of TimeSlips over a 6-month period using both quantitative and qualitative methods. Overall, our longitudinal evaluation of TimeSlips using this mixed-method approach demonstrated positive benefits for both people with mild–moderate and severe dementia. Quantitatively, participants with mild–moderate dementia were significantly more likely to experience more positive resident–caregiver interactions and improved QOL during TimeSlips, compared to those with severe dementia. Qualitatively, TimeSlips demonstrated a range of benefits for residents with both mild–moderate and severe dementia each week.

Using validated instruments, we found that participation in the program confers greater relative benefits on QOL and resident–caregiver interactions for participants with mild–moderate dementia. Given the degenerative nature of the illness, as well as TimeSlips’ emphasis on verbal/improvisational performance, this finding is perhaps unsurprising. Through the use of preliminary qualitative thematic analysis, it was also notable that benefits of TimeSlips such as encouragement of participant bonding, memory reminiscence, and elicitation of laughter appeared to span all levels of dementia severity. Although neither group experienced a statistical change in pleasure over time (ie, smiles, laughter, relaxed facial expression, relaxed body language, etc), it is noteworthy that participants with severe dementia were still experiencing these benefits alongside those with mild–moderate dementia. Such an observation helps demonstrate how the arts can uniquely reach people with advanced memory loss and thereby support QOL. This observation also demonstrates the importance of utilizing a multi-method analytical approach. Although these interpersonal elements (eg, sharing stories, laughing, creative expression) are often difficult to measure quantitatively, qualitative analysis can perhaps capture or unpack their richness and significance.

There were several limitations to this study. First, we focused only on the experiences of 22 participants and 4 caregivers and lacked a control group. There was also an imbalance between the gender of participants that could have affected outcomes, and the study population—owing to the rural setting of the residential care home—was also racially homogeneous. Moreover, although having only one member of the research team (A.A.V.) undertake data collection on dementia severity and outcomes for each TimeSlips session eliminates possible interrater reliability challenges, it is possible that interactions and behaviors were missed during the sessions or that bias may have entered into the longitudinal evaluations. Additionally, A.A.V. did not ask for participant feedback with regard to the findings, which could have been valuable in further identifying or clarifying themes. It would also have been helpful to elicit caregiver feedback as to how TimeSlips could be improved or augmented in residential care settings. For instance, caregiver 1 voluntarily offered that she felt participants experienced increased benefits when new TimeSlips facilitators were introduced, and this would seem to be a question deserving of future inquiry.

Indeed, future studies can use rigorous mixed-methods designs to replicate the results of this study and further assess if creative, arts-based programs such as TimeSlips can improve QOL and caregiver–resident relationships for those with less advanced forms of the illness, lower the use of psychoactive drugs like antipsychotics that may mitigate QOL, or improve biometric indicators/risk factors such as blood pressure or cortisol. Mixed-methods studies might be particularly helpful when studying the aging population, as they have the ability and flexibility to capture significantly rich information throughout a constantly evolving time period. Subsequently, these results could be useful in developing personalized care plans for residents with varying degrees of dementia. If, for example, a PWD used to play sports, creative arts-based programs could be specifically tailored to that patient (eg, using images of athletic events) to evoke more personalized memories and initiate deeper and more meaningful reflection. Additionally, if TimeSlips continued to demonstrate positive benefits for PWDs, the use of creative, arts-based programs could become a standard practice in dementia care units. It would also be useful for future research to seek out the perspective of PWD to better understand the experiential benefits and limitations of activities such as TimeSlips on QOL, as there is presently a dearth of studies that elicit the subjective experiences of cognitively frail participants.