Caregivers’ Loss of the Dyadic Experience after Their Care Partners’ Death

Abstract

Little is known about the experience of family caregivers when their care partner dies and their dyadic relationship comes to an end. This study qualitatively examined and characterized the loss of the dyadic experience for the caregiver after the death of their care partner. Data was accrued as part of a randomized clinical trial in 29 older hospice caregivers. Iterative thematic analysis focused on dyadic processes before, during and post death. Using two relational parameters from Relational Turbulence Theory resulted in a preliminary characterization of a new concept - dyadic dissolution as a cognitive and affective process whereby a remaining member of a dyad experiences relational uncertainty and partner interference while adapting (or not) to the death of their care partner. Findings suggest that asking several open-ended questions about the dyadic relationship will enable assessment for any continuing impact of relational uncertainty and partner interference on bereaved caregivers.

Understanding the important role family caregivers play when caring for adults with chronic illnesses is gaining depth and breadth in a series of important dyadic (adult with chronic illness/family caregiver) theoretical, empiric, and systematic review papers (Buck, Hupcey, et al., 2018; Buck et al., 2019; Buck, Stromberg, et al., 2018; Lee et al., 2015; Lyons & Lee, 2018). This dyadic chronic illness care may result in positive caregiver outcomes such as improved relationship quality (Hooker et al., 2018; Sebern & Woda, 2012) and quality of life (Vellone et al., 2014) but also concerning outcomes such as caregiver strain (Lyons et al., 2019; Strömberg & Luttik, 2015), anxiety and depression (Pucciarelli et al., 2018). Bereavement research supports that the death of a significant other is one of life’s major stressors with psychological, physiological, and behavioral effects (Buckley et al., 2012). However, grief and bereavement are generally studied at the individual (Doering & Eisma, 2016) or system level (Hooghe et al., 2013). Less is known about how the bereaved person’s loss of the dyadic relationship (i.e. the shift from being a dyad to being single) influences their ability to move through the bereavement process in adaptive ways. While there is a body of literature on the continuing bonds between care partners after death (Klass et al., 2014; Neimeyer et al., 2006; Root & Exline, 2014), this loss has not been examined from a dyadic perspective to ascertain what effect the loss of the dyadic experience, itself, may have on the caregiver.

Purpose

The purpose of the study was to qualitatively examine and characterize the loss of the dyadic experience for the caregiver after the death of their care partner. To accomplish this, we used theoretical framing from the Relational Turbulence Theory (RTT). RTT provides a novel way to understand relationships in transition (Droser, 2020; Solomon & Brisini, 2019; Solomon & Knobloch, 2004; Solomon et al., 2016). The theory defines transition as a time when changing relationship status forces individuals to acclimate to new circumstances and roles (Solomon et al., 2016) such as occurs with the death of one partner. RTT is an inherently dyadic theory and so allows an examination of processes at the dyadic rather than individual level as has been done in past bereavement studies (Johannsen et al., 2019; Klass et al., 2014).

Methods

Study Design

This is an analysis of qualitative data accrued as part of a randomized clinical trial in hospice caregivers (R21AG056584–01) (Buck et al., 2020). The interviews occurred at the final data collection point. Caregivers were asked to describe their dyadic relationship (Buck et al., 2013) with the care partner during the interview. Qualitative descriptive methodology (Sandelowski, 2010) was used to achieve the study purpose.

Setting and Sample

Caregivers were included in the larger study if they were over the age of 60, had experienced the death of a care partner at least 12 months prior to enrollment, screened positive for complicated grief, and denied any suicidal ideation or intent. However, by the time we interviewed them their complicated grief was resolved (i.e. below established diagnostic cut points) (Buck et al., 2020). Exclusion criteria were cognitive impairment, psychotic behavior, or current psychotherapy. Recruitment occurred at a local full-service hospice. A university IRB provided ethical oversight (#PRO00032358). Caregivers were recruited between 2017–2019.

This sample was considered especially salient for characterizing the loss of the dyadic experience for several reasons. First, these caregivers had experienced a supported, hospice death yet needed complicated grief treatment suggesting unresolved dyadic issues. Complicated grief is intense, prolonged (acute symptoms last > 6 months) and functionally impairing response to the loss of any close relationship (Shear, 2015). Secondly, these caregivers were at least a year out from the death providing them with some perspective on the experience. Additionally, the sample contained different types of dyadic relationships – spousal, parent/child, and other family relationships. This variability is acceptable given conceptual underpinnings from Interdependence Theory which examines interactions between two individuals irrespective of the type of relationship (Van Lange & Rusbult, 2011). Finally, while it may appear problematic to examine the dyadic experience from the perspective of one informant, meaningful information can be accrued from an individual about dyadic relationships (Knafl & Van Riper, 2017; Thompson & Walker, 1982; Uphold & Strickland, 1989). Taken together these factors suggest that varied informants with recently resolved complicated grief would provide a strong signal when examining and characterizing the loss of the dyadic experience.

Procedures

Following approval of the study protocol, the research assistants were trained by the principal investigator in qualitative interviewing techniques (Speziale & Carpenter, 2007). The semi-structured interview guide began with an exploratory question about the caregiver’s experience in the study which was then followed by more focused probes about dyadic experiences both prior to, during, and after the death of their care partner. Interviews were digitally recorded, and field notes completed for each interview. Digital recordings were transcribed verbatim by a second team member but checked and cleaned by the original interviewer (Sandelowski, 2000). The team reviewed early transcripts as the interview guide probes were tested. All subsequent participants were interviewed using the agreed upon probes. An audit trail was kept which included analytic meeting minutes, decision points, and rationales for any changes.

Data Analysis

Relational Turbulence Theory provided the interpretive lens for this analysis. RTT posits that two relationship parameters, relational uncertainty and partner interference, contribute to biased cognitive appraisals and intensified emotions (Solomon et al., 2016). Relational uncertainty occurs when the individual is unsure of the status of the relationship; this impacts appraisals by removing a clear reference point (Knobloch et al., 2019). Partner interference, while generally understood as active hindrance by a present partner, may occur after death when the caregiver attempts to maintain interdependent patterns despite the death (Knobloch et al., 2019; Rusbult & Van Lange, 2003). While RTT originally focused on the establishment of romantic relationships (Solomon & Knobloch, 2004), it has more recently been applied broadly to transitions into and out of parenting, chronic illness care, and adult caregiving relationships (Knobloch et al., 2018, 2019). A recent examination of marital transitions using RTT found that death of one partner was one of the most frequently reported transitions (Brisini et al., 2018) suggesting that the theory might be useful in examining and characterizing the loss of the dyadic experience. In addition, given the life-course dimensions of the dyadic care relationship (Buck, Hupcey, et al., 2018; Buck et al., 2013) our group hypothesized that many of the processes active in dyadic establishment might be mirrored in its disestablishment.

Data analysis was concurrent and iterative (Sandelowski, 2000). Data were coded according to established approaches and analyzed using iterative thematic analytic techniques (Speziale & Carpenter, 2007). We examined verb tenses (present/past) used when discussing the care partner first within and then across cases. When the team agreed upon the coding structure, a code book was developed and two independent coders progressed with coding all transcripts using Atlas.ti version 6 which allowed examination of the intensity of specific codes (counts) (Sandelowski, 2001). A third team member then checked all coding; any discrepancies were discussed in the team meetings until consensus was reached. During the coding aggregation phase, open codes were synthesized into sub-themes and then categorized into the two a priori relational parameter categories (relational uncertainty and partner interference) by mapping the code definitions to the relational parameter definitions during team meetings. Analytic notes were kept on any constraints imposed by the theoretical framing.

Results

Demographic Characteristics

The 29 caregivers in this study were primarily female, white, and on average 67.4 (± 7.1) years old (Table 1). While the preponderance of informants had been in spousal dyads (n=18), sufficient non-spousal dyads (n=11) were represented to examine dyadic type variations. Their care partners, in those who provided this information (not required), were generally male (n=20) and died of cancer (n=9), dementia or stroke (n=8), or end organ failure (n=6). The interviews (average length = 23.7 lines of text; range = 2–81) resulted 727 lines of text for analysis. Open codes were synthesized into five subthemes -still present, gone but not forgotten, and social isolation subthemes were categorized under relational uncertainty, whereas role change/role reversal and ambivalence towards the partner were categorized under partner interference.

Table 1.

Demographic Characteristics (n = 29)

Characteristic	f
Gender (n = female)	24
Dyadic relationship
Spouse	18
Child	6
Parent	3
Grandchild	1
Sibling	1
Race/Ethnicity
White	28
Black	0
Asian	0
Other	1
Education
Less than high school/high school	8
Some college/technical	7
Associate degree	3
Bachelors’ degree	6
Graduate degree	5
Annual household income
Less than $25,000/year	12
$25,000-$49,000/year	10
$50,000-$74,000/year	4
More than $75,000/year	2
No Response	1

Relational uncertainty

Relational uncertainty or being unsure of the status of the relationship as an interpretive lens uncovered how present the dead partner still was to the living partner. The subthemes of still present, gone but not forgotten, and social isolation set up a continuum on which the caregiver acts as if the care partner is still here or gone (Figure 1). Verb tenses showed that approximately 8% (n=16) statements in this theme were in the present tense suggesting that the care partner was still very present to the informant as the continuing bonds research supports (Root & Exline, 2014). Examples include one adult daughter who stated that she kept several pictures of her father in her dressing room and that, I see him and I’m able to smile and say –“Hey daddy” [ID 1035]. A wife, describing her household responsibilities, used the past tense but then suddenly shifted to the present tense for her husband saying, I took care of the household responsibilities, I’m living and loving an old-world Italian man so, uh, there were a lot of things that were my realm. [ID 1057]. In the subtheme gone but not forgotten informants acknowledge the death but express guilt with adapting to the loss so they try to keep their memories, if not their care partner, alive. A mother who lost an adult daughter described this as:

Figure 1.

The Axes of Dyadic Dissoluation

You know, some people are just not willing to let go of the grief. Sometimes I think, as a mother, at least that’s the way I feel. You almost feel guilty if you don’t feel the grief. Like that person up in heaven, she’s gonna think that I don’t love her anymore and I’ve forgotten that her death, you know, she’s just gone. Oh well, “gone but never forgotten”

[ID 1046].

In this narrative the informant continues to define herself as mother indicating continuing bonds (Neimeyer et al., 2006) or a potentially ambiguous loss (Boss, 2016) where the adult daughter is physically absent but psychologically present. The death of her daughter causes her to be uncertain as to her longtime identity as mother and perhaps delays her re-definition of herself and her role after the loss of the dyadic experience.

In the third subtheme, social isolation, the informants talked about the cost of letting go of the care partner which results in a profound decrease in human contact. One mother of an adult son described this as, “he is my only family, I don’t have any other family members that I’m close to” [ID 1049]. The verb tenses (is) suggest that she remains uncertain about the relational status (subtheme still present) because the cost is isolation due to the lost human connection. A second informant confirmed this with, “And my teammate is gone, you know, so it’s kinda hard” [ID 1046]. In these cases of relational uncertainty, despite very different relationships (adult child/older parent; wife/husband; mother/adult child) uncertainty over the status of the relationship may hamper the caregiver from adapting to the lost dyadic experience.

Partner interference

Partner interference or the partners’ influence on goal attainment as an interpretive lens uncovered two ways in which the care partner continued to actively influence the caregiver: role change/role reversal and ambivalence towards the partner. If relational uncertainty uncovered how present the partner remains, partner interference describes how the partner may continue to interfere with the surviving caregiver’s ability to move forward after losing the dyadic experience. The caregiver vacillates between whether to hang on or let them go on this axis. This negative interference can be seen in practical (role change/role reversal) or emotional (ambivalence towards the partner) responses. Examples of ways in which role change/role reversal interfered with the informant’s current life were both global and specific, but in all cases, there was a sense in which the caregiver did not feel adequately prepared for the care partner’s death. For example, a wife described how before her husband’s death she never called a mechanic or plumber – he always took care of this. She followed this up with a second example which described the economic impact when she stated, “everything is so expensive and I’ve never had to pay big bills before, because he did everything” [ID1039]. Four other informants spoke similarly of the financial impact of the death on their current ability to pay bills or taxes, suggesting again that the partner’s death was potentially interfering with goal attainment by lack of skills or economic constraints. These narratives emphasized the re-definition of roles that the death brings as many dyads share goals and distribute tasks (Hoppmann & Gerstorf, 2009). Similarly, informants expressed ongoing ambivalence towards the partner signifying emotional interference. This subtheme had the strongest signal (35% of codes, n=30/86). Voicing ambivalence is interesting in that it violates the cultural norm of “speaking well of the dead” (De mortuis nil nisi bonum)(Collingwood, 2020). Ambivalence towards the partner had life-course and family system dimensions, when a wife reported, “He was a very lazy person. Even my son mentioned this to me. He said, “Did dad ever do anything around the house?” [ID1022]. In a second example, a wife stated that her husband had always had a Type A personality and never consulted her earlier in their marriage, so she responded that, “I always said I’m no Florence Nightingale and that year he was sick, he took care of everything. I consider myself a caregiver, but he took care of himself and he took care of everything.” [ID 1041]. In these examples care partners continue to influence the caregiver’s relationships with others and conception of themselves as caregivers as they either hang on or struggle to let go of their care partner.

The Axes of Dyadic Dissolution

These data suggested the need for development and characterization of a new process-oriented concept called “dyadic dissolution” to explain the loss of the dyadic experience for the caregiver after the death of the care partner. The term dissolution was selected as it implies an act or process which results in separation into component parts (Merriam-Webster, n.d.) or the breaking of bonds, ties, or unions (Dictionary.com, n.d.). As shown in Figure 1, using the two theoretical parameters of relational uncertainty and partner interference resulted in development of a schematic in which dyadic dissolution is represented as a cognitive and affective process which begins at the end of the dyadic care relationship (death) and concludes with adaptation to the death during bereavement. During the dissolution process the surviving caregiver has to decide if their care partner is here or gone (relational uncertainty) while also determining if they should hang on or let them go (partner interference) so that they can move forward into a future without them. As a process, surviving caregivers can move vertically and horizontally at the same time but the directionality determines where they are in the process (undissolved, upper left quadrant; dissolving, upper right, lower left quadrants; dissolved, lower right quadrant). Dyadic dissolution is a dynamic process which becomes more fixed with time. Failure of the dissolution process to progress to adaptation may indicate a pathologic state (Shear, 2015) but does not have to (Root & Exline, 2014). Based on the qualitative data and theoretically informed schematic, a preliminary definition of dyadic dissolution is that it is a cognitive and affective process whereby a remaining member of a dyad experiences relational uncertainty and partner interference while adapting (or not) to the death of their care partner.

Discussion

The purpose of this study was to qualitatively examine and characterize the loss of the dyadic experience for the caregiver after the death of the care partner. Using Relational Turbulence Theory (RTT), a dyadic theory, we examined processes at the dyadic rather than individual level as previously (Root & Exline, 2014; Waller et al., 2016). This examination resulted in the development and characterization a new concept, dyadic dissolution.

In keeping with earlier work (Brisini et al., 2018; Solomon & Brisini, 2019) transitional periods, like death, are times of increased relational uncertainty and interference. Our caregivers continued to use dyadic strategies, such as seeking interaction (sub-theme still present), feeling connected to each other (sub-theme gone but not forgotten) that they may have successfully used in the past to navigate this new transition (Brisini et al., 2018). However, with the care partner’s death these strategies have potential to become maladaptive (Klass et al., 2014). This study addresses a gap in the continuing bonds literature (Root & Exline, 2014) by providing a better understanding of the post-death experience from a dyadic perspective. In the following we will discuss interesting features of the study in-depth and then suggest some implications.

First, dyadic dissolution may be a key but unexamined risk factor for poorer bereaved caregiver outcomes (Allen et al., 2013; Miles et al., 2016) and reduced likelihood of future caregiving (DiGiacomo et al., 2016). This sample of caregivers recently successfully completed complicated grief treatment, however, factors which predict complicated grief, such as female gender, older age, and reduced income (Kersting et al., 2009) remained. These factors are also the usual characteristics of the caregiving population (International Alliance of Carer Organizations, 2014) as a whole. However, further work is needed to confirm these findings in caregivers who experience normal bereavement. What we can state at this time is that what happens after death disrupts dyadic processes and needs further theoretical and empirical examination.

Second, using RTT in an innovative way accomplished several things. RTT does not clearly assign valence labels to its parameters, because they can be destructive (typically) or constructive (sometimes). For example, relational uncertainty usually corresponds with relational problems, but it can also add a frisson of excitement to a stagnant relationship. Similarly, partner interference typically predicts relational problems, but it also demonstrates the deep and abiding daily interconnections of the partners, i.e. if dyads were not intertwined, they would not interfere with each other’s goals. In the case of the death of one partner, preoccupation with and longing for the other partner may engender an uncertain cognitive appraisal of whether the dyadic relationship remains or not (Boss, 2016). Likewise, the bereaved caregiver may experience difficulty engaging in new meaningful activities and relationships as a single person after the death suggesting that while the partner may no longer be physically present, they still interfere, perhaps through rumination (Watkins & Roberts, 2020) or guilt (Li et al., 2018). These valence free parameters allowed us to examine human relationships in all their complexity. Additionally, using RTT as an interpretive lens links bereavement science with interdependence theory (Van Lange & Rusbult, 2011), existing dyadic chronic illness theories such as the well-known Developmental-Contextual Model of Couples Coping with Chronic Illness (Berg & Upchurch, 2007) and the more recent Theory of Dyadic Illness Management (Lyons & Lee, 2018), and other potential types of dyadic dissolution such is found in divorce (Mikucki-Enyart et al., 2017), failed businesses (Ashraf et al., 2017) and athletic partnerships (Wachsmuth et al., 2017).

Finally, although the past two decades have seen a growing emphasis on the dyadic science of illness with several dyadic frameworks developed (Berg & Upchurch, 2007; Lyons & Lee, 2018; Regan et al., 2015; Trivedi et al., 2016), dyadic understanding of how the care partner’s death impacts the health and well-being of the bereaved caregiver has not been examined in-depth. It is known that uncertainty states, such as dyadic dissolution, are dynamic, highly individualized, and comprised of shifting perceptions of confidence and control (Penrod, 2007). Similarly, family caregivers have been shown to seek a steady state or “normal” in the midst of end of life uncertainty (Penrod et al., 2011) resulting in finally reinventing their definition of normal after the death of the care partner (Penrod et al., 2012). While this earlier theoretical work provides further conceptual support for our findings, the current paper is just the beginning of a line of inquiry that may result in potential interventions to aid in this reinvention process. What is known is that at the individual level how well-prepared someone is for the death (Fujisawa et al., 2010), the quality of the pre-death relationship (Bottomley et al., 2019), and whether the care partner suffered (Allen et al., 2019) influence bereavement. Future dyadic research is needed in comparing how type of dyadic relationship (spouse, child etc.), expected versus unexpected death, and the off-time nature of the death in the life-course (e.g. death of a young adult or child vs. older adult) (Neugarten, 1979) influence bereavement.

There are equally important implications from this study for nurse clinicians and educators. Whether we are providing hands-on bedside care, engaging in health system management or preparing the future workforce, all nurses touch the lives of patients and families at their most vulnerable moments, such as the death of a care partner. Understanding that the individual in front of you might be experiencing a disconnect between the physical and psychological presence of their family member and assessing for where they are in process is person centered care. While there are specific diagnostic criteria for normal vs. complicated grief (Shear, 2015) these do not begin to explicate the myriad ways in which humans grieve. There are negative health outcomes for the surviving caregiver, therefore it is imperative that nurses assess for complicated grief. What this paper suggests is that if you ask several open-ended questions about the nature of their dyadic relationship, you will be able to assess for any continuing impact of relational uncertainty and partner interference that they may be experiencing.

Certain limitation should be kept in mind when reviewing this study. Our sample was limited by the sampling frame, hospice caregivers, which is well known to include a greater number of white patients (Cohen, 2008; Forst et al., 2017). Similar questions in a more racially diverse group may have resulted in different information. As acknowledged earlier, these informants had screened in as positive for complicated grief when recruited into the larger study, however their complicated grief was resolved by the time that they were interviewed. Finally, using a theoretical framework as an interpretive lens shaped our analysis, however, what was gained in using a well-tested dyadic theory far outweighed any potential limitations. However, it is more than likely that using two theoretical concepts, relational uncertainty and partner interference, resulted in a two-dimensional understanding of what is more than likely a multi-dimensional process. Figure 1 should be understood as merely a starting point for understanding the complex interplay between the cognitive and affective tasks that must be accomplished to develop a new normal during dyadic dissolution. The addition of other relational concepts from this theory or other factors will result in a more fully developed model of the experience.