Introduction
The loss of a loved one can be one of the most painful life experiences. Although many can adjust to this loss (referred to as ‘normal grief’), others may experience complicated grief (CG). CG is a condition in which emotional reactions after the death of a loved one are prolonged and accompanied by complicating thoughts, behaviours, and dysfunctional emotional regulation (Shear & Shair, Reference Shear and Shair2005). Unlike normal grief, CG can lead to adverse health outcomes, such as high blood pressure (Prigerson et al. Reference Prigerson, Bierhals, Kasl, Reynolds, Shear, Day, Beery, Newsom and Jacobs1997) and sleep disturbances (Germain et al. Reference Germain, Caroff, Buysse and Shear2005; Maytal et al. Reference Maytal, Zalta, Thompson, Chow, Perlman, Ostacher, Pollack, Shear and Simon2007). The development of an appropriate intervention method for reducing adverse effects of CG requires the identification of protective factors, such as social support (Bonanno et al. Reference Bonanno, Wortman, Lehman, Tweed, Haring, Sonnega, Carr and Nesse2002; Vanderwerker & Prigerson, Reference Vanderwerker and Prigerson2004; Burke et al. Reference Burke, Neimeyer and McDevitt-Murphy2010).
Social support can help mitigate adverse psychological responses to bereavement (Bonanno et al. Reference Bonanno, Wortman, Lehman, Tweed, Haring, Sonnega, Carr and Nesse2002; Vanderwerker & Prigerson, Reference Vanderwerker and Prigerson2004; Burke et al. Reference Burke, Neimeyer and McDevitt-Murphy2010), whereas unhelpful support may produce negative outcomes (Krause & Markides, Reference Krause and Markides1990; Litwin & Landau, Reference Litwin and Landau2000; Burke et al. Reference Burke, Neimeyer and McDevitt-Murphy2010; Nam & Hyun, Reference Nam and Hyun2014). Unhelpful support, as reported by bereaved individuals, includes encouraging recovery, wrong advice, speaking in a callous manner, and fighting during stressful events (Dyregrov, Reference Dyregrov2004; Wilsey & Shear, Reference Wilsey and Shear2007). Although there is a lack of quantitative research on the issue, such forms of unhelpful support are associated with negative psychological outcomes in bereaved individuals, and individuals surrounded by people who make them angry or upset are more likely to exhibit CG (Burke et al. Reference Burke, Neimeyer and McDevitt-Murphy2010).
Unhelpful support may be the result of inexperienced individuals providing support during grief and unexpected loss, anxiety regarding continuous interactions with individuals experiencing loss, or insensitivity regarding the emotional pain experienced by the bereaved (Dyregrov, Reference Dyregrov2004; Wilsey & Shear, Reference Wilsey and Shear2007; Burke et al. Reference Burke, Neimeyer and McDevitt-Murphy2010). Theoretically, acute emotional pain after the death of a loved one may be exacerbated by significant others’ insensitivity to this pain because the bereaved individual may feel insulted by this insensitive support and believe that only the return of the deceased would assuage feelings of loss (Bowlby, Reference Bowlby1980; Wilsey & Shear, Reference Wilsey and Shear2007).
The above observations suggest that potential supporters of bereaved individuals may benefit from psychoeducation and particularly appreciate the importance of providing helpful support and avoiding unhelpful support. However, no previous studies have developed or tested an intervention programme designed to teach supportive ways of helping bereaved individuals. Previous studies have shown that psychoeducation benefits bereaved persons by providing access to people who understand and aid discovery of the universal aspects of experiencing a death (Zisook & Shuchter, Reference Zisook and Shuchter2001) and have demonstrated that psychoeducation for CG is a critical component of treatment, as mentioned in Complicated Grief Treatment and Cognitive Behavioral Treatment for CG (Shear et al. Reference Shear, Frank, Houck and Reynolds2005; Boelen et al. Reference Boelen, de Keijser, van den Bout and van den Bout2007). The present study assesses and describes a preliminary psychoeducation intervention method directed at those providing support to the bereaved that integrates psychoeducation for CG and provides information on helpful support. In this regard, I sought to answer the following questions:
-
(1) Are levels of CG and depressive symptoms diminished by access to appropriately trained supportive individuals?
-
(2) Does perceived helpfulness of supporters mediate the effects of the intervention on CG and depressive symptoms?
Method
Participants
Study participants were recruited at two hospice service centres and one community service centre for older adults in two South Korean cities (Seoul and Incheon). These hospice and community service centres identified older adults who sought bereavement support and referred them to the researchers; flyers containing project information were also distributed. The inclusion criteria were age >18 years and recent loss of a loved one. Individuals with CG identified by community service workers [trained by the author using the Brief Dimensional Complicated Grief Assessment (BDCGA; Shear, Reference Shear2011)] and with a total BDCGA score of >5 were included in the initial contact list. Of the 56 individuals initially included, 45 were recruited (11 discontinued the screening process due to life issues, such as difficulty arranging work schedules), and of the remaining 45, three dropped out after the initial assessment refusing to participate because they feared public disclosure of the cause of death (i.e. a stigma). After completing the recruitment process, the 42 participants were randomly assigned to an experimental or a control group as determined using a computer program. The 21 participants assigned to the experimental group were encouraged to bring someone they perceived as an individual who continuously provided substantial support, such as a close friend or relative. All members of the experimental group brought such a supporter. The final sample included 42 participants and 21 supporters of the 21 individuals in the experimental group. All 42 participants received psychoeducation on CG, and 21 participants of the experimental group and their supporters received psychoeducation on providing helpful support.
Of the 42 participants, 24 (57.1%) were female, and age ranged from 21 to 73 years (mean = 45.57; s.d. = 13.04 years). Regarding mode of death, 76.2% had lost a loved one due to natural causes (e.g. illness) and 21.43% due to an accident. In terms of relationships with the deceased, 73.8% had lost a spouse and 26.2% a friend. The average time since the loss was 8.4 months.
Of the 21 members of the experimental group, 13 (61.90%) were female, average age was 47.05 (s.d. = 13.50) years, 71.43% had experienced the loss of a loved one due to natural causes and 28.57% due to an accident, and 76.19% had lost a spouse and 23.81% a friend. The average time since the loss was 218 (s.d. = 82.43) days.
Of the 21 individuals in the control group, 57.14% were female, the average age was 45.43 (s.d. = 13.44) years, 80.95% had experienced loss due to natural causes and 19.05% due to an accident, and 19.05% had lost a spouse and 90.95% a friend. The average time since the loss was 191 (s.d. = 79.21) days. Demographic and bereavement variables were not significantly different between the two study groups.
Intervention
The intervention in the experimental and control groups was conducted in two phases. The first phase involved two psychoeducational sessions on CG and was attended by all 42 participants. The second phase involved psychoeducation on helpful social support, which was provided to only 21 bereaved participants and 21 supporters of the experimental group. The two phases of psychoeducation each consisted of two 60-min weekly sessions, which were conducted on consecutive weeks (total 4 weeks); sessions were held on consecutive weeks.
At the first session, an information booklet defining CG and its symptoms, causes, and risk factors was developed and provided to all participants. The booklet was written in a simple manner to enable participants to understand regardless of their literacy and education levels. Participants received 60 min of one-to-one instruction from a mental health professional with a doctoral degree, who explained what CG is and answered questions about the booklet. It should be noted the professional was aware of the group assignments. At the end of the first session, the members of the experimental group were told to ask their supporters to participate in the supporter intervention programme, and all supporters agreed to do so.
At the second session, all participants received psychoeducation regarding supportive relationships after bereavement. At this time, supporters received separate psychoeducation in the form of a 60-min lecture. At the lecture, supporters were instructed that the presence of social support was not the same as the absence of unhelpful support, and supporters were also educated on ways of offering helpful support and of not providing unhelpful support. For example, they were instructed to speak in a warm manner, to behave in ways that honoured the deceased, and to provide affection with physical contact as required. These programmes were guided by an interventionist with a master's degree in psychology and social work.
At the third and fourth sessions, which were held in the third and fourth weeks, bereaved participants of the experimental group attended a face-to-face interview session with the interventionist or the author. These two sessions consisted of conversations on interpersonal relationships they had with anyone including supporters who participated in the intervention. After these reviews, the members of the experimental group received education regarding supportive relationships in bereavement. At the same time as the session was being held, the interventionist or the author had an interview session with supporters and discussed the support that supporters provided and provided psychoeducation regarding ways of helping in bereavement. Control groups received no intervention in these third and fourth sessions.
During the psychoeducational sessions, participants and supporters were told that if they had questions regarding CG to contact the author in person, by email, or phone. All 42 participants completed the intervention programme.
Outcomes
All participants underwent a pre-intervention (immediately before the intervention programme) and two post-intervention assessments (at 4 and 8 weeks). Two independent evaluators with master's degrees, who were not involved in the intervention and were trained to use the rating instruments, assessed psychological symptoms pre- and post-intervention. The rating instruments used were the Inventory of Complicated Grief (ICG; Prigerson et al. Reference Prigerson, Maciejewski, Reynolds III, Bierhals, Newsom, Fasiczka, Frank, Doman and Miller1995) and the Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, Reference Radloff1977), which are widely used to assess the mental health of older adults residing in community service centres who have experienced bereavement (Schulz et al. Reference Schulz, Boerner, Shear, Zhang and Gitlin2006; Holland & Neimeyer, Reference Holland and Neimeyer2011; Kersting et al. Reference Kersting, Brahler, Glaesmer and Wagner2011). Three mental health experts with doctoral degrees, fluent in English and Korean, and independent of this study used the standard ‘forward-backward’ translation method to prepare Korean versions of the ICG and CES-D.
The ICG is a 19-item tool designed to assess CG, which requires responses to statements, such as ‘I think about the person so much that it's hard for me to do the things I normally do’ and ‘Memories of the person that died upset me’. Participants were asked to rate the extent to which they agreed with items using a five-point Likert-type scale ranging from ‘never’ (0) to ‘always’ (4). The total score of the 19-item ICG ranges from 0 to 76, in which higher scores indicate more severe symptoms. In the present study, the internal consistency of the ICG was high (Cronbach's α = 0.85).
The CES-D is a 20-item instrument used to determine the severity of depressive symptoms. Subjects are asked to comment on statements, such as ‘I am bothered by things that didn't usually bother me’ and ‘I do not feel like eating; my appetite is poor’. The participants were asked to indicate the extent to which they agreed with each item as applied to the previous 7 days using a four-point Likert-type scale ranging from ‘rarely or never’ (0) to ‘always’ (3). The total score of the CES-D ranges from 0 to 60, and higher scores indicate more severe depressive symptoms. In the present study, the internal consistency of the CES-D was excellent (Cronbach's α = 0.91).
Additionally, at the post-intervention assessments, participants were asked ‘Do you find the support you received helpful?’ The responses to this question were analysed to determine the relationship between the provision of helpful support and psychological outcomes.
Data analysis
This analysis was designed to address whether psychoeducation for CG in bereaved individuals produces better outcomes than introductory psychoeducation on CG and thus aid the provision of helpful social support. The analysis focused on differences between the ICG and CES-D scores of the experimental and control groups pre- and post-intervention.
Data were first descriptively analysed, and the scores of the experimental and control groups were compared. A repeated-measures ANCOVA was conducted to compare the ICG and CES-D scores of these two groups pre-intervention and at 2 and 8 weeks’ post-intervention with respect to demographic variables, death-related variables (mode of death and relationship with the deceased), and time elapsed since loss. The indirect effects of the perceived support from supporters on the relationship between the intervention programme and outcome variables were investigated.
Results
Descriptive analyses of outcome variables
Table 1 shows the overall means and standard deviations of outcome variables. No significant difference was observed between the pre-intervention ICG scores of the experimental and control groups (t = −0.25, p = 0.81). However, significant differences were observed between post-intervention (2 weeks after intervention) ICG scores (t = −6.03, p < 0.001) and follow-up (8 weeks after intervention) ICG scores (t = −5.41, p < 0.001). No significant between-group difference was observed between the pre-intervention CES-D scores (t = −0.07, p = 0.53), but a significant difference was found between scores at 2 weeks (t = 2.05, p = 0.02) and at 8 weeks (t = 2.13, p = 0.02) post-intervention.
Table 1. Means and standard deviations of ICG and CES-D scores
ICG, Inventory of Complicated Grief; CES-D, Center for Epidemiologic Studies Depression Scale.
a 2 weeks’ post-intervention.
b 8 weeks’ post-intervention.
ANCOVA
No sphericity issue of the ICG scores was found in the analysis period (Mauchly's W = 0.940, p = 0.349). The intervention programme had significant effects of time (F = 63.41, p < 0.001; Table 2) with a large effect size (η 2 = 0.64); specifically, members of the experimental group showed larger reductions in ICG scores post-intervention than controls (Fig. 1).
Fig. 1. Improvements in Inventory of Complicated Grief (ICG) scores during the course of bereavement in the experimental group and control group.
Table 2. ANCOVA results of ICG scores
ICG, Inventory of Complicated Grief; MS, mean squares.
A sphericity issue was found for CES-D scores over the analysis period (Mauchly's W = 0.470, p < 0.001). Nevertheless, statistics were interpreted based on the assumption of no sphericity. The intervention programme had significant effects of time (F = 21.58, p < 0.001; Table 3) with a large effect size (η 2 = 0.38), in which participants in the experimental group showed larger decreases in CES-D scores post-intervention than controls (Fig. 2).
Fig. 2. Improvements in depressive symptoms during the course of bereavement in the experimental group and control group. CES-D, Center for Epidemiologic Studies Depression Scale.
Table 3. ANCOVA results of CES-D scores
CES-D, Center for Epidemiologic Studies Depression Scale; MS, mean squares.
Indirect effect analyses
Given the significant positive observed effects of the intervention programme on ICG and CES-D scores, the indirect effects of the intervention programme on ICG and CES-D scores according to perceived helpfulness of the support received were examined. It was found that the intervention programme had significant indirect effects on follow-up ICG and CES-D scores, but these effects depended on whether the support received was considered to be helpful (Table 4).
Table 4. Indirect effects of intervention on ICG and CES-D scores through perceived helpfulness of support
ICG, Inventory of Complicated Grief; CES-D, Center for Epidemiologic Studies Depression Scale.
If the confidence interval does not contain zero, then the indirect effect is significant.
Discussion
CG is a distinct type of distress experienced by individuals who have lost a loved one. Although social support is considered a promising protective factor in CG, little is known about whether helpful support can be provided through an intervention programme. Using a randomized controlled design, the analysis conducted in the present study focused on testing the effectiveness of the proposed psychoeducation intervention programme at various levels of helpful support. The results provide support for the hypothesis that psychoeducation incorporating helpful support provides a protective buffer against CG symptoms. This result is consistent with the findings of previous studies, which reported that only helpful support can reduce the severity of psychological outcomes during bereavement (Wilsey & Shear, Reference Wilsey and Shear2007; Somhlaba & Wait, Reference Somhlaba and Wait2008; Burke et al. Reference Burke, Neimeyer and McDevitt-Murphy2010). It is noteworthy that CG was symptomatic in both study groups at the pre-intervention assessment but that members of the experimental group showed steep decreases in CG and depressive symptoms after the intervention programme, whereas controls did not. This finding suggests the effectiveness of implementing a psychoeducation intervention programme on helpful social support after individuals have experienced the loss of a loved one. Moreover, the effect remained even 8 weeks after completing the intervention programme, indicating that the programme successfully provided protracted protection against CG and depressive symptoms. Furthermore, the indirect effects of the intervention programme on outcomes were found to depend on whether the participant perceived the support as helpful. This preliminary examination provides a better understanding of the mechanism underlying the benefits of psychoeducation on the provision of helpful support for CG and depressive symptoms in bereavement, but future research is required to provide a deeper analysis of the mechanism involved to enable a solid theoretical base for a robust model of psychoeducation to be constructed.
This study is the first to show that psychoeducation of the bereaved and supporters of the bereaved can provide a means of mitigating CG severity. In particular, our results suggest that the proposed psychoeducation programme can benefit individuals with CG and indicates that the provision of appropriate support should be integrated into any successful intervention programme. Previous studies based on cognition-based psychoeducation programmes have discussed the key roles played by relatives and friends with respect to helping the bereaved confront difficult memories of the deceased or realities surrounding death (Shear et al. Reference Shear, Frank, Foa, Cherry, Reynolds, Bilt and Masters2001, Reference Shear, Frank, Houck and Reynolds2005). This study extends the evidence on psychoeducation-based intervention methods by demonstrating the effectiveness of educating helpers to provide useful support.
Although the results of this study confirm the beneficial effects of psychoeducation regarding the provision of helpful support, our results should be interpreted with caution. Although the proposed programme educates supporters based on previous research, data on helpful support during bereavement remain scarce. In this regard, future research should be conducted to investigate additional factors relevant to support during bereavement, including when a bereaved individual perceives support to be helpful and whether the receipt of support is affected by situational factors (e.g. mode of death). Furthermore, in the present study, the mediating effect of perceived helpfulness from supporters on the relationship between intervention and psychological outcomes was examined, and a significant indirect effect of perceived helpfulness from supporters was observed. However, this result is limited because it was based on responses to only one question. In future studies, an assessment tool such as the Multidimensional Scale of Perceived Social Support (Zimet et al. Reference Zimet, Dahlem, Zimet and Farley1988) should be employed to comprehensively measure the perceived helpfulness of support from close social networks and to examine the mechanism underlying the relationships between psychoeducation on the provision of helpful support and psychological outcomes in bereavement. Additionally, the analysis did not consider factors that might influence the supporters’ ability to provide helpful support, and future research should address this limitation to enhance the quality of the proposed intervention method. Finally, the analysis employed a single-blind approach, and intervention assignments should have been made independently based on a double-blind design to better examine the effect of psychoeducation in bereavement.
In summary, the study proposes a concise intervention programme for psychoeducation that can protect against CG during bereavement. Although future research should adopt larger and more diverse samples, the results of the present study provide support for the potential benefit of psychoeducation on CG. In this regard, future research should further elucidate the effects of helpful support on CG in various situations to improve social support for bereaved individuals experiencing a difficult time in their lives.
Declaration of Interest
None.
