Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 Nov 1.
Published in final edited form as: Nurs Res. 2010 Nov–Dec;59(6):417–425. doi: 10.1097/NNR.0b013e3181faff52

Effects of a Creative Expression Intervention on Emotions, Communication, and Quality of Life in Persons with Dementia

Lorraine J Phillips 1, Stephanie A Reid-Arndt 2, Youngju Pak 3
PMCID: PMC3010410  NIHMSID: NIHMS252969  PMID: 21048483

Abstract

Background

Effective nonpharmacological interventions are needed to treat neuropsychiatric symptoms and improve quality of life for the 5.3 million Americans affected by dementia.

Objective

To test the effect of a storytelling program, TimeSlips, on communication, neuropsychiatric symptoms, and quality of life in long-term care residents with dementia.

Method

A quasi-experimental, two-group, repeated measures design was used to compare persons with dementia who were assigned to the twice-weekly, 6-week TimeSlips intervention (n = 28) or usual care (n = 28) group at baseline and postintervention at Weeks 7 and 10. Outcome measures included the Cornell Scale for Depression in Dementia, Neuropsychiatric Inventory-Nursing Home Version, Functional Assessment of Communication Skills, Quality of Life–AD, and Observed Emotion Rating Scale (this last measure was collected also at Weeks 3 and 6 during TimeSlips for the treatment group and during mealtime for the control group).

Results

Compared to the control group, the treatment group exhibited significantly higher pleasure at Week 3 (p < .001), Week 6 (p < .001), and Week 7 (p < .05). Small to moderate treatment effects were found for Week 7 Social Communication (d = .49) and Basic Needs Communication (d = .43). A larger effect was found for pleasure at Week 7 (d = .58).

Discussion

As expected given the engaging nature of the TimeSlips creative story-telling intervention, analyses revealed increased positive affect during and at 1-week post-intervention. In addition, perhaps associated with the intervention’s reliance on positive social interactions and verbal communication, participants evidenced improved communication skills. However, more frequent dosing and booster sessions of TimeSlips may be needed to show significant differences between treatment and control groups on long-term effects and other outcomes.

Keywords: art, dementia, emotions, behavioral symptoms


Alzheimer’s disease and other dementias affect an estimated 5.3 million Americans and could affect as many as 7.7 million by 2030(Alzheimer’s Association, 2010). Dementia is prevalent in long-term care patients, affecting up to 70% of nursing home residents (Alzheimer’s Association, 2010). Although antidementia drugs (cholinesterase inhibitors and memantine) show benefits over placebo in attenuating cognitive decline, their effects on the emotional and behavioral changes associated with dementia are inconsistent (Cummings, 2004; Raina et al., 2008; Sink, Holden, & Yaffe, 2005).

The literature on dementia supports the efficacy of nonpharmacological interventions for treating common neuropsychiatric symptoms (e.g., agitation, depression, passivity, aggression) and improving quality of life. Such approaches may offer PWD opportunities for social contact and engagement in meaningful activities, which are care goals endorsed by the American Geriatrics Society and the American Association for Geriatric Psychiatry (2003; Lyketsos et al., 2006). These interventions include behavior therapy, multisensory stimulation, cognitive stimulation, exercise therapy, and recreation therapy (Burgener et al., 2008; Fitzsimmons & Buettner, 2003; Livingston, Johnston, Katona, Paton, & Lyketsos, 2005; Verkaik, van Weert, & Francke, 2005). In addition, the therapeutic use of arts and humanities has been a relatively recent development in dementia care (Gerdner, 2000; Phillips & Conn, 2009). Preliminary research in the arts and other creative expression activities, such as the TimeSlips storytelling program, has shown potential for improving communication, attention, and pleasure as well as the neuropsychiatric symptoms of dementia (Basting, 2003; Gerdner, 2000). Several randomized controlled trials (RCTs) of movement and music therapy found significant improvements in agitation, general neuropsychiatric symptoms, and speech content and fluency (Brotons & Koger, 2000; Raglio et al., 2008; Sung, Chang, Lee, & Lee, 2006).

Progress in creative expression research has been hampered by multiple methodological limitations. These include use of small convenience samples, nonexperimental designs, subjects with unconfirmed diagnosis or dementia staging, researcher-developed instruments with limited psychometrics, inadequately described interventions, and a failure to link outcome variables to the intervention (Gräsel, Wiltfang, & Kornhuber, 2003; Marshall & Hutchinson, 2001). The Dementia Day Camp Research Group (DDC) recently reviewed expressive arts interventions, including writing programs, art and creative activity interventions, and dance therapies, and found minimal supporting evidence for these programs (Richeson et al., 2008). The DDC, however, recommended continued investigation of these programs due to their versatility, benign nature, and potential benefits.

Theoretical Basis of Dementia Care

As a leading proponent of person-centered care for PWD, Kitwood (1997) posited that the primary task of dementia care is to maintain personhood in the face of cognitive decline. This requires that a defined set of psychological needs be met, including comfort, identity, inclusion, occupation, and attachment. Interactions that meet these psychological needs help maintain personhood. Furthermore, Kitwood hypothesized that a deficient social-psychological environment, one that fails to respect PWDs, may actually accelerate neurological degeneration.

Kitwood (1997) identified 12 types of positive interactions - recognition, negotiation, collaboration, play, timalation, celebration, relaxation, validation, holding, facilitation, creation, and giving- and asserted that good dementia care entails varied positive interactions in continuing succession. Group activities, such as TimeSlips, provide an outlet for spontaneity and creativity and an opportunity for caregivers to initiate and sustain positive interactions with multiple residents, thereby meeting psychological needs and affirming personhood.

TimeSlips

TimeSlips is a nationally recognized group storytelling program for PWD that encourages open storytelling in PWD by stimulating imagination rather than relying on factual reminiscence (TimeSlips Creative Storytelling Project, 2010). One of the important principles of TimeSlips is that because it relies on creative abilities and not memory, participants are not frustrated by cognitive deficits. TimeSlips aligns with Kitwood’s (1997) theory of dementia care by promoting self-expression in a safe, accepting environment. Communication occurs naturally as individual contributions are encouraged, recognized, and validated in a collaborative setting. Creative expression interventions like TimeSlips emphasize the uniqueness of persons, enabling them to feel included, supported and valued (Younger & Martin, 2000).

In a typical storytelling session, PWD are seated comfortably in chairs surrounding a facilitator. A funny or staged photograph printed on 8.5 × 11″ copy paper is distributed to each person in the group. Facilitators use open-ended prompts to engage the storytellers, recording verbatim responses on a flip chart or marker board. Facilitators do not correct storytellers, but instead provide whatever is needed (e.g., more time, prompts) to allow the PWD to respond to the image (Basting, 2003). Responses are woven into an inclusive narrative and periodically read back to the storytellers to help them develop the story further or bring it to a close. Based on extant literature, the present study was designed to evaluate the potential benefits of a TimeSlips intervention on functioning and quality of life among residents in skilled nursing and assisted living facilities. Given the intervention’s emphasis on, and unconditional support of, verbal communication, as well as prior research on TimeSlips and other humanities- and art-based interventions, we predicted that participation in TimeSlips would result in improved communication. In addition, we anticipated that the socialization encouraged by the group format, and the creative, collaborative storytelling task of the TimeSlips intervention, would lead to improved psychological well-being as indexed by measures of psychological symptoms and quality of life. To assess the potential benefits of TimeSlips, data were collected pre- and post-intervention, as well as at a mid-point of the intervention. As this was a pilot study, designed with the goal of providing preliminary data regarding treatment effect that would indicate whether longitudinal follow-up was warranted, a relatively brief follow-up interval of 1 week and 4 weeks post-intervention was used.

Subsequent to initiation of the present study, Fritsch, Kwak, Lang, Montgomery, and Basting (2009) published the results of a RCT which compared 10 nursing homes randomly selected for TimeSlips training and implementation with 10 control nursing homes. Time-sampling was used to observe resident engagement and affect across different types of activities, although not specifically during TimeSlips sessions. Researchers found significant improvement in resident general alertness (identified as an affect item in the resulting manuscript), resident engagement, quantity and quality of staff-initiated interactions, and staff attitudes toward PWD in treatment homes compared to control homes. Findings from this RCT suggested that diffusion of the TimeSlips program within facilities may have a multipronged effect on resident care and outcomes. Resident engagement and affect evinced improvements, and staff members who participated in the TimeSlips training reported more positive views of PWD, an effect of which may have promoted more positive staff-resident interactions.

While providing support for some potential benefits of TimeSlips, because Fritsch et al. (2009) targeted facility-level programming, the effect of TimeSlips on individual resident outcomes such as neuropsychiatric symptoms, communication, and quality of life was not evaluated. Therefore, by obtaining estimates of the effect of the TimeSlips program on measures of communication skills, quality of life, neuropsychiatric symptoms, and observed emotions in 50 to 60 long-term care residents with dementia, this pilot study provided unique insights into the potential effects of TimeSlips in the context of person-centered dementia care.

Method

We used a quasi-experimental, two-group, repeated measures design to compare PWD receiving the TimeSlips intervention (n = 28) with persons receiving no intervention (n = 28). Nine long-term care facilities known to the principal investigator (PI) were recruited for the study; between December 2008 and July 2009, four nursing homes and two assisted living facilities sequentially enrolled and completed the study. All facilities were located in two central Missouri metropolitan counties having populations of < 250,000 each. To minimize contamination of the intervention’s effectiveness, individual facilities were assigned to either control or intervention status (rather than having both within each facility). However, at times facilities were not assigned randomly due to concerns about potential imbalance of group sizes.

Participants

The sample initially consisted of 56 participants. Since this was a pilot study, the primary goal was not hypothesis testing, so sample size was not determined by power considerations. Inclusion criteria were age ≥ 60; a dementia diagnosis or current treatment with a cholinesterase inhibitor; a Mini-Mental State Exam (MMSE; Folstein, Folstein, & McHugh, 1975) score of ≥ 11 but < 24; the ability to understand and speak English; hearing adequate to hear the TimeSlips dialogue; and eyesight adequate to view the TimeSlips stimulus picture. We excluded persons: 1) with a Cornell Scale for Depression in Dementia (CSDD) score > 12, indicating severe depression (Alexopoulos, Abrams, Young, & Shamoian, 1988; Watson, Garrett, Sloane, Gruber-Baldini, Zimmerman, 2003), 2) receiving hospice care and, 3) with advanced terminal illness.

Procedures

Per procedures approved by the relevant Institutional Review Board (IRB), recruited facilities provided the PI with names of potential participants and contact information of their family members or legal guardians. A waiver of individual authorization for disclosure of personal health information was obtained so that the facility could share the names of potential study subjects (i.e., persons with a dementia diagnosis). Family members received a letter explaining the study, an opt-out form with return envelope, and a recruitment flier. In accordance with state law, only court-appointed legal guardians received informed consent documents to sign in addition to the other recruitment materials. PWD also received brief letters of invitation.

We telephoned family members to explain the study. In most cases, they provided verbal assent for study personnel to contact and screen potential participants. In the absence of a court-appointed legal guardian, as was the case for the large majority of potential participants, a waiver of consent process was approved and on file with the IRB. No participants in the current study were able to make independent decisions regarding research participation.

One week before starting the intervention at each facility, we screened for eligibility the residents we had permission to contact. From residents’ charts, we collected age, gender, ethnicity, education, marital status, medical diagnoses, and current medications. The PI (first author) or study nurse visited each potential participant to obtain assent for MMSE testing and eventual participation in the study. The assent process was adapted to potential participants’ level of understanding, and any evidence of dissent from the PWD overrode previously obtained responsibility party assent or legal guardian consent (Black, Rabins, Sugarman, & Karlawish, 2010). All assenting residents took the MMSE to determine whether they met inclusion criteria. Nursing staff members involved in the direct care of study participants were invited to participate in the study as staff-informants for specific measures intended for nursing staff or caregivers. Informed consent per IRB guidelines was obtained from each staff participant.

We recruited facilities sequentially to minimize overlap of intervention sessions across facilities. Assignment to the intervention and control conditions was split equally among the facilities, so that two nursing homes and one assisted living facility comprised each group. Data collection for each set of treatment and control homes occurred concurrently. For each data collection visit, study participants received a flower or sugar-free chocolate candy and staff-informants received $20.

Delivery of the Intervention

The TimeSlips intervention was delivered for 6 consecutive weeks to participants residing in the facilities assigned to the treatment condition. We honored participants’ preference to attend or not on any given day. The one-hour sessions were held two mornings a week with groups ranging from 6 to 12 PWD, depending on the facility. In one facility, TimeSlips sessions took place in a private group meeting room; in the other two facilities, sessions were in a shared living space, a condition that resulted in other residents joining the group periodically. The PI and research nurse, both certified TimeSlips facilitators, conducted all sessions together. Each session opened with a personalized greeting and distribution of name tags. Sessions closed with distribution of a flower or sugar-free chocolate candy. During the sessions in which emotion was observed, an additional member of the research team recorded responses to free the PI or study nurse to observe participants. Attendance logs and field notes were recorded at each session.

At the end of data collection in each home, we offered TimeSlips on a limited schedule to the control group. Control group homes also received activity items of their choice to enrich the care of residents with dementia. Both treatment and control participants received a scrapbook of the TimeSlips stories created by their respective groups.

Measures

Unless otherwise noted, we collected the following measures at three time points: baseline (Week 0), 1 week postintervention (Week 7), and 4 weeks postintervention (Week 10). Outcomes were assessed at Week 7 to determine immediate effectiveness and at Week 10 to determine short term retention of benefit. Emotion was observed two additional times: at Weeks 3 and 6 during the TimeSlips intervention for the treatment group and during a mealtime for the control group. At Weeks 0, 7 and 10, emotion was observed during a mealtime for both groups.

The Modified Cumulative Illness Rating Scale (CIRS; Parmelee, Thuras, Katz, & Lawton, 1995) is a standardized instrument for rating the presence and degree of pathology in 13 major organ groups as well as in the category of psychiatric or behavior problems. An Illness Severity Index is represented by the mean of all items, excluding the behavioral or psychiatric item to avoid confounding with cognitive functioning. Items are rated from 1 (None) to 5 (Extremely severe), with higher scores indicating greater illness severity. Although Cronbach’s alpha coefficient was questionable (α = .67), the Illness Severity Index was shown to significantly predict 2-year mortality among frail older institutionalized adults. For this study, CIRS ratings were based on the PI’s review of residents’ charts. Reliability in the present study was poor (α = .45).

The MMSE (Folstein et al., 1975) is an 11-item clinical assessment of global cognitive function. The maximum score is 30, and scores below 24 indicate cognitive impairment (Barrie, 2002). Twenty-eight day test retest reliability of .98 has been documented (Folstein et al., 1975). To limit participant burden in this study, the MMSE was collected only at baseline and Week 10.

The CSDD (Alexopoulos et al., 1988) is a 19-item clinician-administered interview of the patient and a nursing staff member to rate symptoms of depression in individuals with dementia. Items are scored from 0 to 2, with a summed score of > 7 suggestive of depression (Watson et al., 2003). Cronbach’s alpha of .84 has been found (Alexopoulos et al., 1988). Participants’ primary care providers were informed of CSDD > 7 at the time the assessment took place. In keeping with study exclusion criteria, enrolled participants with CSDD scores > 12 at Week 7 were not included in Week 10 data collection. Internal consistency reliability in the present study was acceptable (α = 0.74).

The Neuropsychiatric Inventory-Nursing Home Version (NPI-NH: Wood et al., 2000) is a clinician-administered interview of nursing staff to rate the frequency on a 4-point scale and severity on a 3-point scale of 10 behavioral and 2 neurovegetative items. The domain score for each item is the product of the frequency and severity. The total NPI-NH score is the sum of the domain scores for the first 10 items. A score of 4 or greater in any domain indicates clinically significant symptoms. Statistically significant interrater correlations between staff nurses’ and psychologists’ ratings (r = .28 to .91) have been found in all domains except those of anxiety and appetite change (Wood et al., 2000).

Adapted for use in residential settings, the Quality of Life-Alzheimer’s Disease (QOL-AD: Logsdon, Gibbons, McCurry, & Teri, 2000) resident and caregiver versions are used to assess PWD relationships with friends, family, and staff, as well as physical condition and mood (Edelman, Fulton, Kuhn, & Chang, 2005). The 15-item scale is administered in a structured interview and rated on a 4-point scale (poor, fair, good, or excellent). Scores range from 15 to 60, with higher scores indicating better QOL. Internal consistency reliability of the adapted scale was .92 (Edelman et al., 2005). Alpha coefficients in the present study were .89 for the staff version and .90 for the resident version.

The Functional Assessment of Communication Skills (FACS: Frattali, Thompson, Holland, Wohl, & Ferketic, 1995) is an observational tool for rating communication behaviors in the domains of social communication; daily planning; communication of basic needs; and reading, writing, and number concepts. Interviews were conducted with nursing staff, rather than directly observing residents, based on the recognition that nursing staff would have more opportunities than the research team to observe participants’ communication abilities across a range of circumstances. Only the social communication and communication of basic needs subscales were used in this study. Higher scores indicate better communication ability. Interrater reliability has ranged from .72 to .92 and mean interitem correlation of .82 has been reported (Frattali et al., 1995). Reliability in the present study was excellent for the 21-item social communication subscale (α = 0.92) but poor for the 7-item basic needs subscale (α = 0.52).

The Observed Emotion Rating Scale (OERS: Lawton, Van Haitsma, & Klapper, 1999), previously named the Apparent Affect Rating Scale, is an observational tool for rating two positive emotions (pleasure and general alertness) and three negative emotions (anger, anxiety or fear, and sadness). Over a 10-minute period, the rater chooses one of six possible defined time intervals (e.g., 1 = never; 2 = < 16 seconds; 3 = 16–59 seconds; 4 = 1–5 min; 5 = > 5 min; and 7 = not in view) that a target subject displays each of the five emotions. Higher scores indicate increased duration of expression of that emotion. Interrater reliability has been variable. Fritsch et al. (2009) reported Kappa values ranging from .25 to .81 and Lawton, Van Haitsma, Perkinson, and Ruckdeschel (1999) reported Kappa values ranging from .29 to .66.

All OERS ratings were performed by the PI or study nurse, both of whom received OERS training with a clinical psychologist (also a co-author of this paper). As recommended by Lawton, Van Haitsma, Perkinson, et al. (1999), all three of the aforementioned team members viewed a training video developed by the Philadelphia Geriatric Center (1997). Baseline OERS Kappa values for the first treatment and control homes were .68 for pleasure and .33 for anxiety. Interrater reliability for anger, sadness, and general alertness could not be calculated due to sparse distribution of responses across categories. Periodically Kappa was calculated to check for observer drift. Interrater reliability declined at the midpoint of the study to.64 for pleasure and .21 for anxiety and. Prompt re-training improved Kappa to .86 and .63, respectively.

Data Analysis

Data were entered into a computer database for analysis using SAS v9 software. Data were checked for accuracy and completeness, de-identified to protect privacy, and double-entered to minimize data-entry errors. Descriptive statistics were computed to obtain a demographic profile and to estimate indices of central tendency and pattern of variability for the outcome measures (i.e., CSDD, NPI-NH, FACS, QOL-AD, and OERS). A .05 level of significance was used to evaluate group differences at Weeks 7 and 10 for the variables of CSDD, QOL-AD, and OERS. A more stringent .01 level of significance was used for comparisons of the NPI-NH and FACS (Social Communication and Basic Needs Communication) to reduce Type I error resulting from multiple tests on a sample of this size. The standardized mean difference statistic, d (Cohen, 1988), was calculated in two ways: (a) dividing the difference between the treatment and control groups means at Week 7 by the pooled SD and (b) dividing the difference between the treatment and control groups mean change scores (Week 7 - Week 0) by the pooled SD of the change scores. As the endpoint most proximal to the 6-week intervention, only Week 7 values were used to calculate effect size estimates. Effect sizes are interpreted as small (.20), medium (.50), and large (.80; Cohen, 1988).

A repeated-measures analysis of covariance (RM-ANCOVA) consisted of one between-subjects factor (group) and one within-subjects factor (time), with the baseline value of the outcome measure as the covariate. To account for potential dependencies in the responses from subjects within one facility, the facility (cohort) was treated as a random effect in group comparisons. In addition, MMSE, CIRS, and length of stay were included in the models as covariates if significantly correlated with the outcome variable of interest. Least Squares Means was used to compare the groups at each time period. In the case of model residuals that did not meet the criteria of normal distribution, the Cochran-Mantel-Haenszel test was used to test between-group differences at Weeks 7 and 10, adjusting for the baseline values of the outcome variables and covariates found to be significant in preliminary analyses.

Results

Of the 178 PWD who were invited to participate in the 6-week intervention, 123 family members and legal guardians and one resident with independent decision-making status (who was excluded for MMSE > 23) agreed to follow-up contact. Of the 124 potential participants who received further contact, 26 refused. Of the 98 who agreed to participate in the study, 42 did not meet inclusion criteria (reasons included MMSE > 23 or < 11, acute illness or hospitalization, lack of dementia diagnosis, and discharge from facility). This left 56 participants (28 in the treatment condition and 28 in the control condition) for which baseline data were collected. During Week 2, a treatment participant was removed from the study due to disruptive behavior, and a control participant was withdrawn based on a family member’s objection to depression screening. During Week 10, three of the 54 remaining participants had exited the study: two control participants developed CSDD scores > 12 and one treatment participant was hospitalized.

The sample was largely elderly, White females with moderate cognitive impairment. Sample demographics, baseline background characteristics and study variables are presented in Table 1. The mean MMSE score for the assisted living treatment group was 16.00, whereas mean scores were 13.67 and 13.86 for the nursing home treatment groups. There were no significant differences between groups on baseline background characteristics or study variables, with the exception that the control group had significantly higher years of education. Attendance at TimeSlips sessions averaged 84% overall: 94% and 69% at the first and second nursing home, respectively, and 83% at the assisted living facility.

Table 1.

Sample Demographics and Baseline Background Characteristics and Study Variables (N = 56)

Treatment group (TG) (n = 28) Control group (CG) (n = 28)
Nursing home (TG - 1) Nursing home (TG - 2) Assisted living facility (TG) Nursing home (CG–1) Nursing home (CG - 2) Assisted living facility (CG)
n (%) n (%) n (%) n (%) n (%) n (%)
12 (21) 6 (11) 10 (18) 4 (7) 11 (20) 13 (23)
Demographic Characteristics
Gender
 Female 11 (39) 3 (11) 9 (32) 4 (14) 10 (36) 12 (43)
 Male 1 (4) 3 (11) 1 (4) 0 1 (4) 1 (4)
Ethnicity
 African American 1 (4) 0 0 0 0 0
 Caucasian 11 (39) 6 (21) 10 (36) 4 (14) 11 (39) 13 (46)
Marital Status
 Never married 1 (4) 0 0 1 (4) 1 (4) 0
 Married 3 (11) 3 (11) 0 0 2 (7) 0
 Widowed 6 (21) 2 (7) 8 (29) 1 (4) 8 (29) 12 (43)
2 (7) 1 (4) 2 (7) 2 (7) 0 1 (4)
Separated/Divorced
Education Levela
 No diploma 1 (4) 2 (8) 3 (12) 1 (4) 0 1 (4)
 High school 5 (19) 3 (12) 4 (15) 1 (4) 4 (15) 7 (27)
 College/Technical 4 (15) 0 1 (4) 2 (8) 1 (4) 5 (19)
 Graduate degree 2 (8) 0 1 (4) 0 4 (15) 0
Background Characteristics M (SD) M (SD) M (SD) Treatment Group M (SD) M (SD) M (SD) M (SD) Control Group M (SD)
Age (years) 83.55 (8.79) 80.73 (6.40) 84.89 (8.21) 83.42 (8.0) 81.73 (12.81) 87.59 (4.61) 85.43 (5.96) 85.75 (6.76)
Education (years) 13.75 (3.31) 10.00 (2.19) 11.80 (3.82) 12.25 (3.52)* 13.00 (3.46) 15.70 (3.50) 13.38 (2.22) 14.19 (3.05)*
MMSE 13.67 (4.64) 13.83 (3.71) 16.00 (2.75) 14.54 (3.88) 15.50 (4.20) 16.55 (4.95) 14.92 (3.40) 15.64 (4.09)
CIRS 1.76 (0.24) 1.65 (0.29) 1.65 (0.26) 1.70 (0.26) 1.75 (0.13) 1.69 (0.29) 1.53 (0.16) 1.62 (0.23)
Length of stay (years) 4.75 (3.25) 2.81 (2.02) 1.58 (1.29) 2.15 (1.59) 0.31 (0.16) 4.00 (2.82) 2.51 (2.30) 2.78 (2.61)
Baseline Study Variablesb M (SD) M (SD) M (SD) M (SD) M (SD) M (SD)
CSDD 4.75 (3.25) 5.00 (3.35) 2.50 (1.96) 6.75 (3.77) 4.36 (4.92) 5.77 (3.77)
QOL Resident-rated 40.23 (8.41) 41.17 (8.91) 45.79 (4.86) 43.11 (9.64) 42.18 (9.00) 44.29 (3.56)
QOL Staff-rated 39.75 (6.28) 34.67 (4.13) 44.60 (4.60) 36.00 (5.72) 41.82 (9.52) 43.54 (6.54)
NPI-NH 7.33 (9.00) 7.67 (6.31) 3.10 (3.78) 10.25 (8.18) 9.36 (14.25) 15.92 (12.24)
Social Communication 5.51 (0.92) 6.31 (1.00) 6.20 (0.63) 4.99 (1.65) 5.92 (0.94) 5.51 (0.94)
Basic Needs Communication 5.70 (0.84) 6.31 (0.78) 6.61 (0.38) 5.46 (0.55) 6.00 (0.69) 5.97 (0.89)
OERS pleasure 2.00 (1.04) 1.83 (0.41) 2.00 (1.15) 1.25 (0.50) 2.00 (1.00) 2.50 (1.09)

Note. Percentage totals may not add to 100% due to rounding.

a

Education Level was missing on four participants.

b

See Table 2 for baseline group means.

*

P < .05, Wilcoxon rank sum test (two-sided)

The raw means and SDs for the treatment and control groups on the major study variables as well as effect size estimates for Week 7 are displayed in Table 2. Scores on the CSDD and the NPI-NH improved over time for the control group; therefore, Week 7 effect size estimates on these outcomes reflect beneficial change of the control group and do not represent the intervention effect. However, close to medium effect sizes were found for Social and Basic Needs Communication, reflecting better communication skills for the treatment group at Week 7. A medium effect size was found for OERS pleasure scores.

Table 2.

Mean Scores by Time for Treatment and Control Groups with Standardized Mean Difference Statistics for Week 7

Week 0 Week 3 Week 6 Week 7 Week 10 Effect Size at Week 7 Effect Size of Change Score (Week 7–Week 0)
CSDD
 Control 5.36 ± 4.20 3.63 ± 3.64 3.00 ± 2.47
 Treatment 4.00 ± 2.99 4.30 ± 4.56 4.73 ± 3.27 0.17 0.45
QOL Resident-rated
 Control 43.29 ± 6.85 45.16 ± 5.35 44.20 ± 5.91
 Treatment 42.41 ± 7.62 44.98 ± 6.37 44.12 ± 5.98 0.031 0.14
QOL Staff-rated
 Control 41.79 ± 7.91 41.89 ± 7.28 43.48 ± 7.18
 Treatment 40.39 ± 6.35 39.85 ± 8.07 39.15 ± 8.50 0.27 0.009
NPI-NH
 Control 12.54 ± 12.63 8.00 ± 8.91 5.12 ± 6.25
 Treatment 5.89 ± 7.05 6.30 ± 8.14 10.27 ± 12.20 0.18 0.49
Social Communication
 Control 5.60 ± 1.06 5.35 ± 1.24 5.52 ± 1.22
 Treatment 5.93 ± 0.90 5.87 ± 1.04 5.62 ± 1.14 0.49 0.29
Basic Needs Communication
 Control 5.91 ± 0.77 5.66 ± 1.23 5.76 ± 1.16
 Treatment 6.16 ± 0.79 6.14 ± 0.92 5.96 ± 1.27 0.43 0.29
OERS pleasure
 Control 2.11 ± 1.05 2.04 ± 0.90 2.52 ± 1.22 1.73 ± 0.78 1.84 ± 0.90
 Treatment 1.96 ± 0.96 3.42 ± 0.93 3.48 ± 1.05 2.31 ± 1.12 2.27 ± 0.96 0.58 0.61

Notes. CSDD = Cornell Scale for Depression in Dementia; QOL = Quality of Life-Alzheimer’s Disease; NPI-NH = Neuropsychiatric Inventory-Nursing Home Version; OERS = Observed Emotion Rating Scale

Results of the RM-ANCOVA are presented in Table 3. Significant group-by-time interactions were found for the outcomes of NPI-NH, Social Communication, and OERS. After adjusting for baseline and MMSE, OERS pleasure scores at Weeks 3, 6, and 7 were significantly higher for the treatment than control group. Social Communication scores were higher for the treatment group at Week 7, although this difference did not reach statistical significance. Scores on the NPI-NH favored the control group at Week 10, but this difference was not statistically significant. Although there was no significant interaction effect for the CSDD, the control group had significantly lower CSDD scores than the treatment group at Week 10 after adjusting for baseline and CIRS scores. Due to extreme sparseness in distribution for OERS anxiety, sadness, anger, and general alertness data, further analysis was not considered (e.g., on only three of 261 observations for general alertness were ratings other than 4 or 5 given).

Table 3.

Least Square Means by Time for Treatment and Control Groups based on Repeated-Measures Analysis of Covariance

Week 3 Week 6 Week 7 Week 10 F (df) time effect F (df) group × time effect
CSDD a
 Control 3.38 (0.70) 3.24 (0.57) 0.55 1.21
 Treatment 4.38 (0.70) 5.11 (0.56)** (1, 95) (1, 95)
QOL Resident-Rated
 Control 44.39 (0.95) 43.74 (0.88) 2.06 0.13
 Treatment 44.67 (0.97) 43.59 (0.87) (1, 93) (1, 93)
QOL Staff-Rated b
 Control 41.37 (1.01) 41.79 (1.10) .29 1.65
 Treatment 40.91 (1.01) 39.89 (1.09) (1, 95) (1, 95)
NPI-NH c
 Control 7.48 (2.09) 5.75 (2.24) 1.25 6.87*
 Treatment 6.90 (2.05) 11.20 (2.21) (1, 95) (1, 95)
Social Communication a, c
 Control 5.44 (0.13) 5.55 (0.15) 2.50 14.60***
 Treatment 5.82 (0.13) 5.54 (0.15) (1, 94) (1, 94)
Basic Needs Communication a, c
Control 5.73 (0.17) 5.76 (0.17) 1.05 1.82
Treatment 6.10 (0.17) 5.89 (0.17) (1, 94) (1, 94)
OERS Pleasure c
 Control 2.03 (0.17) 2.41 (0.20) 1.68 (0.19) 1.78 (0.19) 17.24*** 4.0 **
 Treatment 3.44 (0.17)*** 3.55 (0.20)*** 2.34 (0.18)* 2.31 (0.18) (3, 190) (3, 190)

Notes. CSDD = Cornell Scale for Depression in Dementia; QOL = Quality of Life-Alzheimer’s Disease; NPI-NH = Neuropsychiatric Inventory-Nursing Home Version; OERS = Observed Emotion Rating Scale.

Means are adjusted for baseline score and covariates found to be significant in preliminary testing. Standard errors are in parentheses. Significant differences between control and treatment group means at Weeks 3, 6, 7, and 10 are noted by treatment group asterisks. Except for the outcomes of QOL Resident-Rated and QOL Staff-Rated, p-values for between-group comparisons were based on the Cochran-Mantel-Haenszel test.

*

p < .05;

**

p < .01;

***

p < .001

a

Modified Cumulative Illness Rating Scale as covariate;

b

length of stay as covariate;

c

Mini-Mental State Exam as covariate

Discussion

Results of the current study suggest that the 6-week program of TimeSlips was associated with increased expressions of pleasure and better communication for the treatment group compared to the control group. In addition, the treatment group exhibited significantly higher pleasure during the intervention, and this effect persisted at Week 7 follow-up. It is encouraging that social and basic needs communication scores for the treatment group remained stable through Week 7, declining only at Week 10, long after the last TimeSlips session. Spontaneous verbal communication and interpersonal dialogue, both of which are reflected in FACS scores, appear to be upheld by the positive interactions prompted by TimeSlips, specifically recognition, facilitation, and creation. The apparent increase in pleasure manifested during TS may be attributed to two other types of positive interactions, celebration and play. As described by Kitwood (1997), a steady stream of positive interactions is needed to meet the psychological needs of PWD. These findings of benefit, limited to the duration of the TimeSlips intervention, suggest that cognitively impaired persons may need continual reinforcement to sustain gains. Finally, the treatment group, compared to the control group, had significantly poorer scores on the CSDD at Week 10; however, mean depression scores were not in the clinically significant range for either group throughout the study.

This study expanded upon prior research by Fritsch et al. (2009), who did not find a significant difference in observed pleasure between treatment versus control groups, though they did observe a significant group difference in general alertness favoring the treatment group. This discrepancy may be partly explained by differences in study design. Specifically, Fritsch et al. observed residents across a variety of activities, whereas participants in this study were observed only during mealtime and TimeSlips, activities that naturally stimulate general alertness. Fritsch et al. remarked that the nonsignificant difference in pleasure between the control and treatment homes ran contrary to their hypothesis but did not offer further explanation. It may have been the case that activities in control homes were equally pleasurable for residents, albeit not as activating or engaging.

During TimeSlips, PWD have the opportunity to exercise creative potentials and remaining strengths and experience moments of recognition, facilitation, creation, celebration and play. By design, TimeSlips encourages PWD to become active participants, as opposed to passive recipients, of the program. Increasing opportunities for meaningful activities and person-enhancing interactions, as is accomplished by TimeSlips, promotes the positive psychosocial environment foundational to person-centered care.

In considering potential limitations of this pilot study, several features of the sample may be relevant in interpreting study findings. First, in part due to the variety of facilities where recruitment occurred, participants in the study had a range of cognitive impairment. The mean MMSE for the assisted living treatment group was higher than for the two nursing home treatment groups, although this difference could not be statistically analyzed due to sample size. Several interested assisted living residents could not enroll in the study due to MMSE scores greater than 23. The opposite was true in nursing home residents, whose MMSE scores were frequently 10 or less. On the one hand, testing TimeSlips in both nursing home and assisted living residents enhances the external validity of study findings and suggests that this type of activity can be delivered in various settings. On the other hand, given the range in cognitive functioning among study participants, it may be that outcomes important for participants with more advanced dementia, such as neuropsychiatric symptoms and communication skills, were not appropriate for participants with milder disease.

Second, the relatively small group sizes (range = 25–28) may not have provided sufficient power to detect significant differences. Additionally, it is a limitation of the study design that participants were not randomly assigned within the same facility to a treatment or an attention-control group. As such, the active ingredient of TimeSlips could not be isolated from unintended effects secondary to increased attention and socialization in treatment participants, nor from effects of other activity experiences for control participants. Finally, the study sample, though representative of nursing home and assisted living residents in central Missouri, lacked ethnic and racial diversity, a factor that limits the generalizability of results.

Because this was a pilot study, facility staff were not involved in delivering the program and it is not possible to address whether it was feasible for staff to continue TimeSlips after the study ended. Fritsch et al.’s (2009) study of TimeSlips included training of treatment-home staff for the purpose of determining if the interactions between staff and residents during TimeSlips would change the ways in which staff viewed and interacted with residents. They found that staff in treatment homes interacted more frequently with and held more positive views of PWD. This is notable because staff who interact with residents in varied contexts may understand residents’ capacities more fully, which is important for person-centered care.

Despite limitations, this study suggests that, as a program designed for group participation, TimeSlips may be welcomed by residents. Supplies needed to conduct storytelling sessions are inexpensive, making TimeSlips an affordable activity in terms of supply resources. TimeSlips is acceptable to persons with mild to moderate dementia and provides a stimulating alternative to typical activities in long-term care facilities. Activities like TimeSlips, that build social confidence rather than devalue individuality, are fundamental to good dementia care.

Acknowledgments

This research was supported by the John A. Hartford Foundation’s Building Academic Geriatric Nursing Capacity Award Program, the University of Iowa Gerontological Nursing Intervention Research Center, and the University of Missouri Interdisciplinary Center on Aging. Thank you to Richard Madsen, PhD for statistical consultation and Kathleen C. Buckwalter, PhD, RN, FAAN for manuscript review.

Contributor Information

Lorraine J. Phillips, Sinclair School of Nursing, University of Missouri, Columbia, Missouri.

Stephanie A. Reid-Arndt, Department of Health Psychology, School of Health Professions, University of Missouri, Columbia, Missouri.

Youngju Pak, Department of Health Management and Informatics, School of Medicine, University of Missouri, Columbia, Missouri.

References

  1. Alexopoulos GS, Abrams RC, Young RC, Shamoian CA. Cornell Scale for Depression in Dementia. Biological Psychiatry. 1988;23(3):271–284. doi: 10.1016/0006-3223(88)90038-8. [DOI] [PubMed] [Google Scholar]
  2. Alzheimer’s Association. 2010 Alzheimer’s disease facts and figures. Chicago, IL: Author; 2010. [DOI] [PubMed] [Google Scholar]
  3. American Geriatrics Society, & American Association for Geriatric Psychiatry. Consensus statement on improving the quality of mental health care in U.S. nursing homes: Management of depression and behavioral symptoms associated with dementia. Journal of the American Geriatrics Society. 2003;51(9):1287–1298. doi: 10.1046/j.1532-5415.2003.51415.x. [DOI] [PubMed] [Google Scholar]
  4. Barrie MA. Objective screening tools to assess cognitive impairment and depression. Topics in Geriatric Rehabilitation. 2002;18(2):28–46. [Google Scholar]
  5. Basting AD. Reading the story behind the story: Context and content in stories by people with dementia. Generations. 2003 Fall;:25–29. [Google Scholar]
  6. Black BS, Rabins PV, Sugarman J, Karlawish JH. Seeking assent and respecting dissent in dementia research. The American Journal of Geriatric Psychiatry. 2010;18(1):77–85. doi: 10.1097/JGP.0b013e3181bd1de2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Brotons M, Koger SM. The impact of music therapy on language functioning in dementia. Journal of Music Therapy. 2000;37(3):183–195. doi: 10.1093/jmt/37.3.183. [DOI] [PubMed] [Google Scholar]
  8. Burgener SC, Buettner L, Buckwalter KC, Beattie E, Bossen AL, Fick D, et al. Review of exemplar programs for adults with early-stage Alzheimer’s disease. Research in Gerontological Nursing. 2008;1(4):295–304. doi: 10.3928/19404921-20081001-02. [DOI] [PubMed] [Google Scholar]
  9. Cohen J. Statistical power analysis for the behavioral sciences. 2. Hillsdale, NJ: Erlbaum; 1988. [Google Scholar]
  10. Cummings JL. Alzheimer’s disease. The New England Journal of Medicine. 2004;351(1):56–67. doi: 10.1056/NEJMra040223. [DOI] [PubMed] [Google Scholar]
  11. Edelman P, Fulton BR, Kuhn D, Chang CH. A comparison of three methods of measuring dementia-specific quality of life: Perspectives of residents, staff, and observers. The Gerontologist, 45 Spec No. 2005;1(1):27–36. doi: 10.1093/geront/45.suppl_1.27. [DOI] [PubMed] [Google Scholar]
  12. Fitzsimmons S, Buettner LL. Therapeutic recreation interventions for need-driven dementia-compromised behaviors in community-dwelling elders. American Journal of Recreation Therapy. 2003;2(1):17–32. doi: 10.1177/153331750201700603. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state.” A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research. 1975;12(3):189–198. doi: 10.1016/0022-3956(75)90026-6. [DOI] [PubMed] [Google Scholar]
  14. Frattali CM, Thompson CK, Holland AL, Wohl CB, Ferketic MM. The American Speech-Language Hearing Association Functional Assessment of Communication Skills for Adults. Rockville, MD: American Speech-Language Hearing Association; 1995. [Google Scholar]
  15. Fritsch T, Kwak J, Lang J, Montgomery RR, Basting AD. Impact of TimeSlips, a creative expression intervention program, on nursing home residents with dementia and their caregivers. The Gerontologist. 2009;49(1):117–127. doi: 10.1093/geront/gnp008. [DOI] [PubMed] [Google Scholar]
  16. Gerdner LA. Music, art, and recreational therapies in the treatment of behavioral and psychological symptoms of dementia. International Psychogeriatrics. 2000;12(SupplementS1):359–366. [Google Scholar]
  17. Gräsel E, Wiltfang J, Kornhuber J. Non-drug therapies for dementia: An overview of the current situation with regard to proof of effectiveness. Dementia and Geriatric Cognitive Disorders. 2003;15(3):115–125. doi: 10.1159/000068477. [DOI] [PubMed] [Google Scholar]
  18. Kitwood T. Dementia reconsidered: The person comes first. Philadelphia, PA: Open University Press; 1997. [Google Scholar]
  19. Lawton MP, Van Haitsma K, Klapper JA. Observed Emotion Rating Scale. 1999 Retreived March 4, 2008, from www.abramsoncenter.org/PRI (scales page)
  20. Lawton MP, Van Haitsma K, Perkinson M, Ruckdeschel K. Observed affect and quality of life in dementia: Further affirmations and problems. Journal of Mental Health and Aging. 1999;5:69–81. [Google Scholar]
  21. Livingston G, Johnston K, Katona C, Paton J, Lyketsos CG. Systematic review of psychological approaches to the management of neuropsychiatric symptoms of dementia. The American Journal of Psychiatry. 2005;162(11):1996–2021. doi: 10.1176/appi.ajp.162.11.1996. [DOI] [PubMed] [Google Scholar]
  22. Logsdon RG, Gibbons LE, McCurry SM, Teri L. Quality of life in Alzheimer’s disease: Patient and caregiver reports. In: Albert S, Logsdon RG, editors. Assessing quality of life in Alzheimer’s disease. New York: Springer; 2000. pp. 17–30. [Google Scholar]
  23. Lyketsos CG, Colenda CC, Beck C, Blank K, Doraiswamy MP, Kalunian DA, et al. Position statement of the American Association for Geriatric Psychiatry regarding principles of care for patients with dementia resulting from Alzheimer’s disease. The American Journal of Geriatric Psychiatry. 2006;14(7):561–572. doi: 10.1097/01.JGP.0000221334.65330.55. [DOI] [PubMed] [Google Scholar]
  24. Marshall MJ, Hutchinson SA. A critique of research on the use of activities with persons with Alzheimer’s disease: A systematic literature review. Journal of Advanced Nursing. 2001;35(4):488–496. doi: 10.1046/j.1365-2648.2001.01887.x. [DOI] [PubMed] [Google Scholar]
  25. Parmelee PA, Thuras PD, Katz IR, Lawton MP. Validation of the Cumulative Illness Rating Scale in a geriatric residential population. Journal of the American Geriatrics Society. 1995;43(2):130–137. doi: 10.1111/j.1532-5415.1995.tb06377.x. [DOI] [PubMed] [Google Scholar]
  26. Philadelphia Geriatric Center. Recognizing and responding to emotions in persons with dementia [video] Chicago, IL: Terra Nova Films; 1997. [Google Scholar]
  27. Phillips LJ, Conn VS. The relevance of creative expression interventions to person-centered care. Research in Gerontological Nursing. 2009;2(3):151–152. doi: 10.3928/19404921-20090527-03. [DOI] [PubMed] [Google Scholar]
  28. Raglio A, Bellelli G, Traficante D, Gianotti M, Ubezio MC, Villani D, et al. Efficacy of music therapy in the treatment of behavioral and psychiatric symptoms of dementia. Alzheimer’s Disease and Associated Disorders. 2008;22(2):158–162. doi: 10.1097/WAD.0b013e3181630b6f. [DOI] [PubMed] [Google Scholar]
  29. Raina P, Santaguida P, Ismaila A, Patterson C, Cowan D, Levine M, et al. Effectiveness of cholinesterase inhibitors and memantine for treating dementia: Evidence review for a clinical practice guideline. Annals of Internal Medicine. 2008;148(5):379–397. doi: 10.7326/0003-4819-148-5-200803040-00009. [DOI] [PubMed] [Google Scholar]
  30. Richeson NE, Buettner L, McKenzie SE, Burgener SC, Buckwalter KC, Beattie E, et al. Evidence supporting expressive arts interventions for persons in early-stage Alzheimer’s disease. Activities Directors’ Quarterly for Alzheimer’s & Other Dementia Patients. 2008;9:10–27. [Google Scholar]
  31. Sink KM, Holden KF, Yaffe K. Pharmacological treatment of neuropsychiatric symptoms of dementia: A review of the evidence. JAMA. 2005;293(5):596–608. doi: 10.1001/jama.293.5.596. [DOI] [PubMed] [Google Scholar]
  32. Sung HC, Chang SM, Lee WL, Lee MS. The effects of group music with movement intervention on agitated behaviours of institutionalized elders with dementia in Taiwan. Complementary Therapies in Medicine. 2006;14(2):113–119. doi: 10.1016/j.ctim.2006.03.002. [DOI] [PubMed] [Google Scholar]
  33. Time Slips Creative Storytelling Project. University of Wisconsin at Milwaukee, Center on Age & Community; 2010. Retrieved February 12, 2010, from http://www.timeslips.org/ [Google Scholar]
  34. Verkaik R, van Weert JC, Francke AL. The effects of psychosocial methods on depressed, aggressive and apathetic behaviors of people with dementia: A systematic review. International Journal of Geriatric Psychiatry. 2005;20(4):301–314. doi: 10.1002/gps.1279. [DOI] [PubMed] [Google Scholar]
  35. Watson LC, Garrett JM, Sloane PD, Gruber-Baldini AL, Zimmerman S. Depression in assisted living: Results from a four-state study. The American Journal of Geriatric Psychiatry. 2003;11(5):534–542. [PubMed] [Google Scholar]
  36. Wood S, Cummings JL, Hsu MA, Barclay T, Wheatley MV, Yarema KT, et al. The use of the neuropsychiatric inventory in nursing home residents: Characterization and measurement. The American Journal of Geriatric Psychiatry. 2000;8(1):75–83. doi: 10.1097/00019442-200002000-00010. [DOI] [PubMed] [Google Scholar]
  37. Younger D, Martin GW. Dementia care mapping: An approach to quality audit of services for people with dementia in two health districts. Journal of Advanced Nursing. 2000;32:1206–1212. doi: 10.1046/j.1365-2648.2000.01591.x. [DOI] [PubMed] [Google Scholar]

RESOURCES