Portrait of a Caregiver: Ethnodrama Development and Its Influence on Caregiver Well-Being

Abstract

Background and Objectives

The majority of long-term care needs are placed upon family members who often receive minimal support. In this study, we collaborate with family caregivers to create an ethnodrama about their experience and assess outcomes of participation, including caregiver well-being.

Methods

Participants met over 4 months to discuss their roles as informal caregivers. Discussions were analyzed in a two-phase process and themes were developed into a script. Member checks included script review and revisions, culminating in viewing a professional performance of the play followed by a post-performance discussion and reflection. Data were gathered at six timepoints to assess caregiver well-being and longitudinal analysis was used to assess change during the course of the intervention.

Results

Twenty-two caregivers completed intervention activities. Participant reactions to the process evolved overtime, from an initial hesitance about what individuals had to offer leading to an acknowledgement of feeling heard and a desire to help others. Caregivers had high levels of burden and positive perceptions towards caregiving. Those with high levels of self-rated health had the lowest levels of burden and the most positive perceptions of caregiving.

Discussion and Implications

Developing an ethnodrama in partnership with caregivers is a unique and feasible method of caregiver support, mentoring, reflexivity, and meaning making. Promoting caregiver health early in the caregiving trajectory has the potential of reducing burden and elevating positivity towards caregiving.

The majority of long-term services and supports (LTSS) for persons 65 and older are provided by informal family caregivers. It is estimated that 17.7 million people are providing this care in the United States. By 2030, more than one in five older adults in the United States will be 65 and older (National Academies of Science, Engineering, and Medicine, 2016). People are living longer leading to increased co-morbidity, chronic illness, and complexity of care needs. Adding to this complexity is the decreasing numbers of children and increasing distance between family living proximity. When compared with those not providing care, family caregivers experience higher levels of stress, depression, and anxiety. Additionally, they have lower self-rated physical health, higher rates of chronic disease, and greater financial risk (Bom, Bakx, Schut, & van Doorslaer, 2019; National Academies of Science, Engineering, and Medicine, 2016).

Family caregivers save the United States an estimated $470 billion each year (Reinhard, Feinberg, Choula, & Houser, 2015). Keeping older adults in their homes is necessary, as the current infrastructure will not support the number of people needing LTSS (Castora-Binkley, Noelker, Ejaz, & Rose, 2011; Reinhard et al., 2015). Informal caregivers receive minimal training and are often not integrated into care until the recipient faces a crisis (National Academies of Sciences, Engineering, and Medicine, 2016). Limited supports augment the negative effectiveness of caregiving (Castora-Binkley et al., 2011). Informal caregivers rarely have time or the ability to prepare for this role, which leads to hesitation and fear of mistakes, increasing caregiver stress and suboptimal care (Reinhard, Levine, & Samis, 2012). In addition, self-identifying as a caregiver often does not occur until most are overburdened. Current solutions rely on absorbing large amounts of information and resources that the caregiver does not know how to find, which makes it difficult to solve problems and find answers (National Academies of Sciences, Engineering, and Medicine, 2016). Informal caregivers describe learning by “doing” which does not allow for practice without the fear of making harmful mistakes. There is a need for caregiver interventions which foster confidence, sustain well-being, improve coping skills, and prepare individuals for future challenges (Bom et al., 2018). One approach to practice is a form of high-fidelity simulation through applied theatre which has been used in a variety of health care settings and populations to improve care communication (McIntyre & Cole, 2008), clinical practice (Blignault et al., 2010), and the ability to problem solve (Noice & Noice, 2009).

Ethnodrama is one form of research-based theatre that focuses on retaining the words of the participants while staying true to research findings. The process of development in partnership with research participants has been described and tested (Eaton, 2015, 2017), demonstrating feasibility and positive outcomes in nursing students’ improved understanding of older adults, future patients, and intergenerational collaboration. This model uses transformative learning theory to guide ethnodrama techniques in the process of collaboration and development. The ethnodrama becomes an activating event that triggers transformative learning (Eaton, 2015) with the goal of facilitating self-reflection in order to motivate an individual to move past habitual ways of thinking or habits of the mind (Cranton, 2002; Mezirow, 1997). This process of partnering with research participants to develop and produce an ethnodrama is multisensory, embodied, and incorporates techniques of applied theatre that have the potential of improving approaches to preparation for caregiving.

Ethnodrama serves as a form of community-based research that retains scientific findings, honors the words of the participants, and incorporates participatory collaboration (Eaton, 2015, 2017). This method has the potential of augmenting caregiver support, during ethnodrama development, while also creating a product that is accessible to a broad audience. Thus, the purpose of this mixed method research was to collaborate with early and experienced caregivers to create an ethnodrama targeting early caregiving intervention. The primary qualitative component uses community-based participatory collaboration to develop the ethnodrama, while the secondary quantitative component uses multiple data points to assess caregiver well-being during the process of ethnodrama development and the culminating event of viewing a professional performance of the play. A secondary aim of producing performances for larger audiences as a form of caregiver advocacy will not be presented here and has been reported elsewhere (Eaton & Madsen-Wilkerson, 2017).

Methods

With approval from the University of Utah and the Utah Department of Health and Human Services Institutional Review Boards, recruitment and community outreach occurred through a partnership with the Utah Caregiver Support Program. The research team consisted of the primary investigator with training in theatre and gerontology and the community partner with training in social work, gerontology, and clinical experience leading community-based programs for family caregivers. We sought to recruit 24 adults, in the Mountain West of the United States, providing care for an older adult family member who is 65 or older. A convenience sample was used to recruit 12 early caregivers (with experience in providing care for less than 5 years), and 12 experienced caregivers (with experience providing care for 5 or more years). Excluded from recruitment were professional caregivers, those providing care to non-relatives, and those providing care to adult children. Recruitment activities involved dissemination of flyers in local communities (libraries, senior centers, etc.), social media, and email. In-person presentations were made to community partners, caregiver coalitions, Area Agencies on Aging, and other aging service providers. A snowball approach to recruitment was encouraged in an effort to share the opportunity within caregiving communities. Recruitment continued until 24 eligible participants completed the informed consent process.

Caregiver Meetings

Participating caregivers met a total of four times, each meeting lasting 60–90 min to complete transformative learning activities that led to ethnodrama development and creation. Participants received training in interviewing, the use of digital recorders, and protocols for data management, submission, and confidentiality. Each meeting ended with self-reflexive journaling.

The first three meetings involved caregivers conducting semi-structured interviews with one another regarding their caregiving experiences. Digital recorders were used to document all interactions. Caregivers were given example questions, but also encouraged to approach the interviews as a conversation, asking questions that may be missing and sharing mutual experiences. Topics included: caregiver needs, what participants desired to learn, what they wished they knew at the beginning, barriers and facilitators to caregiving, challenges, and strengths. At the fourth meeting, caregiver dyads were given an early draft of the ethnodrama script, reading it out loud and recording the session to document feedback regarding theme validity and representation. This feedback was then incorporated into the final draft of the script.

Ethnodrama Performance and Discussion

The culminating event involved viewing the first professional performance of the ethnodrama with all participants and any family members they chose to invite. Walk-Ons, Inc., a non-profit performing group of professional actors, partnered on this project to direct and perform the final script. Following the viewing, a post-performance discussion occurred in an effort to evaluate the participant experience of viewing their own caregiving through performance. Participants were asked to provide feedback regarding the experience of participating in this research.

Data Collection

Caregiver meetings and data collection were led by the primary investigator (Eaton). Digital recordings were made to document each meeting and then transcribed in preparation for analysis. A video recording of the final performance and post-performance discussion was gathered to assess physical and verbal reactions to the experience.

The intervention process was documented through investigator journaling, logs of participation, detailed notes on recruitment efforts, accrual, attrition, adherence, challenges, and actions taken to overcome problems. Caregiver meetings involved self-reflexive journaling to identify reactions to the intervention process.

Caregiver well-being was assessed through surveys completed by caregiver participants at six times during the intervention process. Research Electronic Data Capture (REDCap) was used to store data and reduce data entry error (Harris et al., 2009). Demographics were gathered to document age, sex, marital status, education level, race/ethnicity, number of years as a caregiver, and relationship to care recipient. This survey included the Caregiver Burden Inventory (CBI) which examines the impact that caregiving has on social relationships (α = .69), flexibility with time (α =.85), life course development (α =.87), physical health (α =.86), and emotional well-being (α =.81). Scores near or above 36 indicate that a caregiver is at risk for burning out (Caserta, Lund, & Wright, 1996; Novak & Guest, 1989). This scale was chosen as it measures a multidimensional approach to caregiving (Caserta et al., 1996; Novak & Guest, 1989), is used to assess caregiving in a variety of populations, and has established validity and reliability (Greco et al., 2017; Lane et al., 2017; Valer, Aires, Fengler, & Paskulin, 2015). The second survey used in assessing well-being was the Positive Aspects of Caregiving (PAC) scale which measures perceptions of the benefits of caregiving (with reported Cronbach’s alpha of .89) (Tarlow et al., 2004). The addition of this scale provides a holistic view of the caregiving experience, allowing the measurement of both burden and the more positive elements of caregiving.

Analysis

Ethnodrama creation

Qualitative data were organized using NVivo software (QSR International, 2016). Creation of the ethnodrama involved constant comparative analysis of caregiver interview transcripts during the intervention process. Phase 1 coding identified in vivo codes, which retain the words of the participant and assess concepts expressed throughout the interviews. A second phase of coding organized all in vivo codes into patterns, which allowed us to identify which themes were most prominent (Saldaña, 2009). These findings were used to create an outline for the ethnodrama script. Quotes representing each pattern were incorporated into the outline to create the script.

Documentation of the intervention process

Documents were read throughout the data collection and analysis process. All audio and video were transcribed verbatim. In vivo and pattern coding were used to code these data in an effort to identify themes that highlighted problematic and significant features of the intervention as perceived by participants.

Quantitative analysis

Descriptive statistics were used to characterize participants. Longitudinal growth modeling (Singer & Willett, 2003) was used to assess changes in initial status and rate of change in caregiver well-being.

Results

Our initial goal was to recruit 12 early and 12 experienced caregivers to participate in the creation of the ethnodrama. Twenty-six individuals providing care to older adult family members enrolled, however, due to the time commitment, four dropped out before the intervention began. A total of 22 individuals participated in the intervention. Twelve (54%) were new to caregiving, while 10 (45%) were experienced. Descriptive statistics are reported in Table 1.

Table 1.

Demographic Characteristics

Characteristic	Phase 1 (n = 22)
	Mean (SD)
Age	57.59 (12.57)
	No. (%)
Sex
Male	3 (13.6%)
Female	19 (86.4%)
Other	0 (0%)
Race
White	21 (95.5%)
Black/African American	1 (4.5%)
Asian	0 (0%)
Pacific Islander	0 (0%)
Native American	0 (0%)
Other	0 (0%)
Ethnicity
Hispanic	1 (4.8%)
Not Hispanic	20 (95.2%)
Relationship status
Single (never married)	1 (4.5%)
Separated or divorced	2 (9.1%)
Married	14 (63.6%)
Registered domestic partnership or civil union	0 (0%)
Widow or widower	3 (13.6%)
Unknown	2 (9.1%)
Sexuality
Heterosexual/straight	18 (81.8%)
Lesbian or gay	3 (13.6%)
Other	1 (4.5%)
Prefer not to answer	0 (0%)
Education
Did not graduate from high school	1 (4.5%)
High school graduate	7 (31.8%)
Bachelor’s degree	9 (40.9%)
Master’s degree	5 (22.7%)
Doctorate	0 (0%)
Post-graduate degree	0 (0%)

The intervention consisted of four meetings held between December 2015 and March 2016. Each meeting lasted 60 min. Participants met in small groups, interviewing one another on the topics of caregiver needs, facilitators and barriers to caregiving, and important learning events. Activities at meetings correlated with the seven steps of transformative learning (Cranton, 2002). Data from the first three meetings were recorded, transcribed, and analyzed in a two-phase process, involving first in vivo coding and second pattern coding (Saldaña, 2009). This process organized thoughts into major themes, which were then used to outline the script for the ethnodrama. Fifteen patterns were discussed most frequently by participants. These patterns were used to outline the ethnodrama script. Themes and frequency are presented in Table 2.

Table 2.

Patterns From Family Caregiver Interviews

Patterns	Number of sources (interviews)	Number of times referenced
Feelings and learned	19	261
Relationships	14	132
Activities	18	116
Health	12	78
Caregiver needs	14	78
Early stages	13	49
Finances	12	35
Support	11	27
Like a child	10	22
Definition	10	18
Story	12	18
End of life	8	13
Later stages	5	8
Sacrifice	5	6

At the final meeting, participants read the script out loud and provided feedback regarding representation of themes, concerns, strengths, and topics missing from the script.

Fifteen participants (68%) attended all four caregiver meetings. Seven missed one of the four meetings, however, no one missed more than one meeting. Participants were asked to share their experience participating in this process. They used words such as “honored,” and “respect” when describing what it felt like to share their experiences. One person stated, “I felt needed in the group discussions” and another describe it as “healing to be able to talk to others who have the same experiences. I felt like a burden was lifted, if only for a few hours.” The process of sharing confirmed that they were not alone, a theme repeated frequently. One woman stated:

It felt odd at first to be honest. I wasn’t sure that I had a lot to offer. I’m still not sure – but I have gained a lot. I know now I’m not alone. I’m not crazy. The things I am experiencing is normal.

Others described gaining “respect for other caregivers,” and enjoying “the freedom to share hard things.” Participants valued “the experience to listen to others and get to know their stories and situations. I felt a lot of validation listening and talking to people.” One person discussed respect from learning about “their sacrifices on behalf of the ill family member were amazing.” The process promoted a desire to give back, “Being able to understand what others face has helped me to want to be available to help.”

Seventeen participants (77%) attended the first professional performance of the ethnodrama, titled Portrait of a Caregiver. Five missed the performance due to family emergencies (n = 2), work conflicts (n = 1), and unknown reasons (n = 2). Those not in attendance were provided a digital recording of the performance at the end of the study; however, we did not assess whether or not these were viewed. Participants also invited family members to this performance. A total of 30 were in the audience to view and provide feedback, including children, parents, care recipients, caregivers, and friends. Post-performance discussion centered on those participating, with only a few family members commenting. Viewing the play was an unique and emotional experience, “even though we had read through the script beforehand, I was surprised how different it was presented by real actors.” One person stated, “I loved it! I think the style was brilliant—it was ‘jumbled’ a bit which is actually exactly how it feels to be a caregiver. You lose some control.” Another participant described the following, “When I saw my part I cried—it made it real and by the audience response I knew I was appreciated.”

One participant’s spouse shared thoughts from viewing the performance:

As someone married to a caregiver, I would hear from her many times…that she thought she was the only one experiencing this. And when your work is done here, it would allow her to realize that she’s not alone…She’d come back from the meeting and say “oh my gosh, they’re going through the same thing” and that she found somebody who finally understands.

Participants described the experience viewing their own story performed for them as a benefit: “It was as if I was watching my life in front of me and could see more clearly what I was dealing with.” Many wished “all of my family/siblings had seen it.” When asked why, it was explained that “…the performance was able to show a dimension that those who haven’t been caregiving could see and understand.” Participants shared a belief that “the drama will have a positive affect on people” and that it “will help others see why caregiving is crucial to the health of our communities.” Another stated, “It was satisfying hearing some of your own comments in the drama. It felt like perhaps you are helping in some way.”

Caregiver Well-being

The multilevel model for change (Singer & Willett, 2003) was used to assess change in caregiver well-being during the intervention. Several steps occurred to identify a longitudinal model that identifies status and change for participants. This is called the unconditional means model (Level 1) (Singer & Willett, 2003). Three aspects were identified as important to capture and better understand change during the intervention. Intercept represents the initial status of participants, gain represents the overall change from baseline, and trend measures rate of change from the beginning of the intervention. These three elements of change arise for the dependent Y when observed at time j for each study participant i:

$${\displaystyle \begin{array}{l}{\displaystyle Y_{ij}=\mathrm{I}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{c}\mathrm{e}\mathrm{p}{\mathrm{t}}_i+\mathrm{G}\mathrm{a}\mathrm{i}{\mathrm{n}}_i+\mathrm{T}\mathrm{r}\mathrm{e}\mathrm{n}{\mathrm{d}}_i\ast (T_{j1}-1)+{\epsilon }_{ij}}\end{array}}$$

The first post-baseline observation to the last observation, T_{j > 1} − 1, allows assessment of Trend, or true rate of change during the intervention. This first level of modeling includes the average effects for the population on the three components of intercept, gain, and trend. In addition, individual differences are assessed through systematic variance components. In order to retain variance components and improve accuracy of the estimates, a lenient standard of significance was used (p = .20). The second level of modeling tests the influence of covariates on intercept, gain, and trend. These models are summarized in Supplementary Table S1.

The unconditional means model (Model A) establishes a baseline for the overall sample specific to caregiver burden. In this model, the intercept is the sole fixed effect and random effect. The grand mean across all individuals and occasions is the fixed effect of intercept. It demonstrates that caregiver burden was elevated and caregivers were at risk for burnout (CBI parameter = 38.24, p < .001) and positive attitudes of caregiving were high, indicating a higher level of positivity towards caregiving (PAC parameter = 36.42, p < .001). Scores demonstrate that caregiver well-being was on the high side of both instruments (CBI > 36; PAC > 27). The average caregiver varied in initial status over time and burden differed from other caregivers. The within- and between-person variance components demonstrate this as the total variation in burden (CBI = 87%) and positive aspects of caregiving (PAC = 29%) are due to caregiver differences.

Supplementary Table S1 outlines the differences in Models B, C, and D. Model E is the unconditional growth model which facilitates comparisons with the unconditional means model (Model A). Comparing the two allows the identification of true change trajectories. Random effects of Model E include intercept, gain, and trend. The true change trajectory of caregiver burden (CBI) included an intercept of 39.36, gain of −3.12, and trend of 0.85. The null hypothesis was rejected for intercept (p < .001) and trend (p < .10). The non-zero intercept of PAC was 36.59 (p < .001), while the non-zero slope was −0.77 for gain and 0.29 (p < .20) for trend. The within-person residuals of Model A and Model E were used to calculate the pseudo-R² to assess the variation associated with Time, which was 41% for CBI and 30% for PAC.

Level 2 modeling examines the influence of covariates as predictors. A variety of models were assessed, and covariates included gender, education, relationship, age, sexuality, race, and ethnicity. Those models that were not significant are not reported here. Caregiver Self-Rated Health (Caregiver_Health) was identified as a Level 2 predictor of intercept and change. Model F and Table 3 outline the findings from this model. When compared with Model E, this model has a difference of 22.40 (p < .001) for CBI and 15.65 for PAC (p < .001). Caregiver_Health had an overall significance for CBI (F(1, 22) = 7.80, p = .00) and PAC (F(5, 22) = 4.98, p = .003). The average caregiver had an estimated initial CBI of 15.39 (p < .05), with an estimated gain indistinguishable from 0 and an estimated slope rate of change of 0.85 (p < .10). A minor gain was identified (p < .05), accounting for −6% of the variance. The within-person variance was 40%, while the between-person variance accounted for 69% of the difference in burden. Additionally, PAC was significant overall, as the estimated initial score for the average caregiver was 41.05 (p < .001), and estimated gain and trend were indistinguishable from 0. The intercept accounted for 59% of the difference in positive attitudes, while very little of the variance (9%) was due to within person variance. Model F was chosen as the final model. When compared with Model E, this model had a difference of 22.40 (p < .001) for CBI and 15.65 for PAC (p < .001).

Table 3.

Results of Fitting Model F, Using the Multilevel Model for Change, to the Well-being Measurement Data (n = 22)

	Fixed effects magnitude	F(df) for fixed effects	Random effects		Pseudo-R2	CI of the variance components for covariance parameters
			Variance	SD		Lower	Upper
CBI
Repeated measures			36.66***	6.05	.00	25.81	52.08
Intercept	15.39	144.16*** (1, 23)	106.25*	10.31	.69	49.87	226.34
Gain	−3.13	1.19 (1, 26)	121.29*	11.01	−.06	53.95	272.67
Trend	0.85	3.28~ (1, 26)	1.03	1.01	.15	0.06	17.01
Caregiver healtha		7.80*** (5, 22)
	Mean (SE)
Level 1	55.45 (11.75)
Level 2	55.98 (11.74)
Level 3	56.17 (11.74)
Level 4	52.08 (4.44)
Level 5	26.69 (4.44)
Level 6	14.22 (6.92)
PAC
Repeated measures			7.11***	2.67	.09	5.00	10.12
Intercept	41.05	591.25*** (1, 24)	12.04*	3.47	.59	5.04	28.75
Gain	−0.76	0.39 (1, 27)	21.16*	4.60	−.61	9.23	48.52
Trend	0.29	2.01 (1, 24)	0.15	0.39	−.25	0.003	6.43
Caregiver health		4.98** (5, 22)
	Mean (SE)
Level 1	25.11 (4.24)
Level 2	24.98 (4.24)
Level 3	29.84 (4.24)
Level 4	35.56 (1.63)
Level 5	39.33 (1.63)
Level 6	40.90 (2.51)

Note: CBI = Caregiver Burden Inventory; PAC = Positive Aspects of Caregiving scale.

aCaregiver health represents a question that asks participants to rate their health on a 7-point Likert scale, with 1 being poor health and 7 being excellent health. The question states: How would you best describe your general health over the past two months according to the scale below?

~p ≤ .10. *p < .05. **p < .005. ***p < .001.

Self-rated caregiver health was a strong predictor of initial status for both CBI and PAC (Figure 1). Caregivers with more positive levels of self-rated health had lower levels of caregiver burden (CBI). Those with the most excellent health had the lowest burden scores (M = 14.22) well below risk for burnout and close to the low end score of 9, which represents no burden. Those with poor health had much higher burden scores (M = 55.45), well above the mean score for burnout. Only those with the highest general health scores had burden beneath the burnout cutoff, while those with poor through average health were above the cutoff by 16 points and more, demonstrating that even those with general overall self-reported health were at risk for burnout. Caregivers with higher levels of general self-reported health also had higher scores on PAC, representing more positive perceptions of the benefits of caregiving. Those with a self-rating of excellent health also had very high PAC scores (M = 40.90), only 5 points away from the PAC high of 45. Those with lowest general health had very low PAC (M = 25.11), 16 points above the lowest score of 9.

Figure 1.

Influence of self-rated general health on caregiver well-being.

In summary, general caregiver health demonstrated that caregivers with high self-rated health had lower levels of burden and more positive perceptions of caregiving (Figure 1). Caregiver health predicted the initial status of both CBI and PAC. Higher levels of self-rated health were related to lower levels of burden and more positive perceptions of caregiving. Post-intervention overall benefit (Gain) was small, with a decrease in burden (CBI) and positive perceptions (PAC). Gain included data from Time 1, which was baseline information from before the intervention began. Social desirability may have influenced Time 1 scores, causing them to be higher than reality. This may be influencing the minor decrease we see in burden and positivity. Trend was the post-intervention rate of change, which excluded Time 1 data thus only including change that happened during the intervention. This reduces the potential social desirability may have on outcome data. During the intervention process, CBI exhibited the greatest rate of change, with both instruments demonstrating slight increases. This means that both burden and positive perspectives increased slightly over the course of the intervention.

Discussion

Outcomes from this study include clarifying the process of collaborating with community-based older adult caregivers, examining the feasibility of activities, questioning definitions of caregiving experience, reinforcing the importance of caregiver health on caregiver well-being, and the difficulty of applying innovative approaches in caregiving communities.

This project demonstrates the feasibility of collaborating with community-based informal family caregivers to create an ethnodrama. Our research team was advised, frequently, that family caregivers were too busy to participate in a project like this. However, we discovered that while family caregivers are some of the busiest individuals, they also desired to give back by helping those new to caregiving and those promoting the needs of family caregivers. The family caregivers participating in this project were generous with their stories, offering the gift of vulnerable experiences. The meaning inherent within ones’ story and its use to continue promoting the needs of others in similar situations is important for meaning making, advocacy, and the opportunity for caregivers to make a difference (Eaton & Madsen-Wilkerson, 2017). Previous work on meaning making emphasizes that such processes encompass more than verbal communication, but include sensory and embodied forms of communication and perception (McGovern, 2011). Ethnodrama supports a multisensory approach to communicating research findings. It has the potential for being a supportive method to enhance the processes surrounding meaning making. Research in meaning making emphasizes its importance when facing situations of high stress (Park, 2010), it is used as a tool to make sense of such experiences (Proulx & Inzlicht, 2012), and is often connected to well-being (Steger, 2012). In addition, caregiving has been documented as a process of meaning making in and of itself (Ellis & Astell, 2008). This study confirms that caregivers found the process of participating in the development and viewing of an ethnodrama was meaningful.

The definition of early and experienced caregivers is complex and difficult to define solely by years. Our use of years to clearly delineate this definition made recruitment difficult. Those who are new to caregiving may not define themselves as caregivers, while those with more experience may no longer be in an active caregiving role or no longer view themselves as a caregiver. This knowledge allowed us to change our recruitment efforts, removing the actual word “caregiver” from all recruitment flyers and asking, instead, for individuals who provide assistance to an older adult family member. Recruitment efforts improved immensely after this action, allowing us to find individuals in the very early stages of caregiving and those no longer in active caregiving roles who still have much lived experience to share. This has implications for caregiver health, in that previous research has documented the importance of training and support for caregivers to reduce their burden and improve their health (Hughes, Shuman, Wiener, & Gould, 2017; Levine, Halper, Peist, & Gould, 2010). This study supports this previous work by demonstrating that those caregivers with higher self-rated general health had lower levels of burden and more positive perceptions of caregiving (Figure 1). A recent systematic review found evidence that caregiving has a negative effect on the health of the caregiver (Bom et al., 2018). Interventions and activities that promote caregiver health early in the caregiving trajectory may play an important protective role in reducing burden and elevating positivity towards caregiving. Earlier intervention and support may prepare caregivers in a manner that retains their health and ability to provide care for longer periods of time.

The development of an ethnodrama in partnership with community-based older adult family caregivers promotes interaction and support for those attending intervention activities. Script development and revision may not always be clear to participants who are unfamiliar with theatre development, performance, and viewing. Study participants clearly preferred viewing a professional ethnodrama performance to reading the script as a group. They had a difficult time understanding the intent of the script due to the use of direct interview quotes. The narrative, without visual cues and action, was difficult to understand. Participants described the script as “jerky,” “disorienting,” and said that it troubled them. Yet, at the culminating intervention event, viewing the performance, participants expressed that they were moved, felt connected to the overall themes, and fully understood the purpose and vision of the ethnodrama. This demonstrates the importance of aesthetics and multiple steps in a transformative process. Ethnodrama is multisensory, allowing participants to learn through sound, observation, and even smell. In order to have a strong effect on audiences, performances need to be high quality so the audience forgets the format and focuses on words, movement, and sound. Using a script with participant words is challenging and requires a high level of performance skill to translate the results into an effective performance that moves audiences to focus more on the subject than on the form. The individual differences of each participant will influence their understanding of intervention activities, which emphasizes the importance of a variety of tasks that target multiple senses, and various ways of learning and viewing the world.

The advantages of translating qualitative research into a dramatic form is the added benefit of more fully communicating the realities of caregiving. Challenging words and terminology become accessible to a broad audience and research is moved out of the academic realm. In addition, it facilitates a variety of learning styles, pushing viewers to connect in new ways with the goal of facilitating others to step into the world of the research subject.

While the form of ethnodrama is an innovative method to translate research, it also has its challenges. It requires a complexity in coordination between multiple stakeholders, not just community partners, but professional actors, directors, and technicians with the skills to translate challenging scripts. Ethnodrama also requires clearly communicating the purpose of the research and methodology to artists in an effort to make sure performances hold true to research findings while also honoring the words of the participants. In order to access professionals with such skills, money is required to support the artistic team through rehearsal and performances. This team should be included during the initial stages of study design, in order to assess their needs and develop an agreement on the costs required to support their work. Finally, this partnership should incorporate an agreement on the types of performing spaces allowed. In this case, it was important that the ethnodrama could be performed in a wide variety of spaces without the limitations of traditional theatrical environments. Finding spaces that allow for all audience members to adequately view the performance takes additional effort and planning.

Rigor

Multiple data sources were used to create the ethnodrama. These consisted of interviews with both early and experienced family caregivers, as well as member checks via participant approval of the final ethnodrama script. The various steps in the creation of the ethnodrama were documented with notes retained regarding all intervention meetings, logs of attendance and surveys completed, as well as notations on attrition. Participants submitted a total of four reflexive journals which were also used to document participation and any changes that occurred throughout the process. Quantitative data were also collected, providing another form of data which allowed the triangulation of findings.

Limitations

Limitations to this study stem from the design which did not include random assignment to groups, which influences threats to internal validity, such as regression artifacts, selection, maturation, and history. Enrolling participants from a variety of sources until the required sample size was met may mean that those participating had a greater ability and higher motivation to participate. Specific design strategies were put into place to reduce limitations. In order to reduce testing effects, data points were scheduled for 2 or more weeks apart, and the precision of measurement was strengthened through the use of six timepoints (Donaldson, 2008). Using a novel intervention may have elevated outcomes, influencing survey results based on a change to their normal activities. However, the study occurred over time, which may have reduced reactions to novelty. The use of mixed methods strengthened analysis and findings through the use of multiple types of data and different ways of knowing. For example, reading the script with participants served as a step in member checking the outcomes of interview analysis. While the generalizability of these findings are limited to certain settings, times, and individuals, the findings may apply to caregivers in similar situations.

Conclusion

The purpose of this study was to examine the influence of participating in the development of an ethnodrama on caregiver well-being. Family caregivers were dedicated to the process, attending most meetings and committing to participation throughout intervention duration. Professional performances of the ethnodrama continue to find sponsorship within the wider community for caregiver trainings and advocacy efforts. While we plan to continue professional performances, next steps include examining the use of such performances on education and training, as well as developing materials to support its use for professional caregivers, extended family of caregivers, and students. Future studies should consider the addition of post-intervention time points to examine the long-term impact on caregivers. In addition, such outcomes should consider the differences between those new to caregiving in comparison with those who have more experience. This resource has the potential of being a tool to humanize the caregiving experience and spread understanding, discussion, and action to the wider community.

Supplementary Material

Supplementary data are available at The Gerontologist online.

Supplementary Table S1. Comparison of fitting models to the well-being measurement data (n = 22).

Funding

This work was supported by the John A. Hartford Foundation and the Gerontological Society of America Hartford Change AGEnts Action Award.

Conflict of Interest

None reported.