*Article* **The Reactions of Adolescents, Parents and Clinicians to Participating in Qualitative Research Interviews Regarding Adolescents Bereaved by Suicide and Other Traumatic Death**

**Karl Andriessen 1,\*, Karolina Krysinska <sup>1</sup> , Debra Rickwood <sup>2</sup> and Jane Pirkis <sup>1</sup>**


**Abstract:** There are concerns that involving adolescents bereaved by suicide and other traumatic death in research may cause distress and harm. However, no study has investigated such bereaved adolescents' research experiences. In addition, no study has looked at the experiences of parents and clinicians as participants in adolescent suicide and traumatic death bereavement research. This study aimed to explore the short-term impact of research participation experienced by adolescents, parents, and clinicians. A total of 61 participants (adolescents, *n* = 17; parents, *n* = 12; clinicians, *n* = 32) filled out a short survey within two weeks of having taken part in a qualitative interview study. Data were analyzed descriptively. Most participants had experienced no distress while participating and no negative effects of participating; rather, participation was experienced as helpful for them and they would highly recommend participating in a study like this to others. A few adolescents and parents reported some distress, related to anxiety about participation and the unpleasantness of grief memories. The study clearly indicates that bereaved adolescents, parents and clinicians can safely participate in research interviews regarding their experiences of grief and help after suicide, generally valuing the opportunity to share their experience. To prevent and mitigate potential distress, training of research staff and implementation of appropriate participant distress protocols are imperative. Future studies could include longitudinal follow-up of participants to assess any longer-term consequences.

**Keywords:** grief; bereavement; suicide; traumatic death; research participation; ethics; research ethics

### **1. Introduction**

### *1.1. Rationale*

Adolescents who are bereaved by the death of someone close to them, such as a family member or a friend, often experience short-term and long-term impacts regarding their grief, mental health and social functioning [1–5]. Experiencing a death is often an unfamiliar, disruptive and stressful event in their lives, leading to acute grief reactions such as crying, and feelings of numbness, sadness and longing for the deceased person [2–5]. Compared to other types of bereavement, adolescents bereaved by suicide and other traumatic death can experience more pronounced feelings of shock, guilt, anger, and abandonment [3,4,6]. They can struggle more with "why" questions, finding meaning in the loss, and experience less social support [1–3,6]. In addition, they have an increased risk of mental health problems, such as depression, anxiety, posttraumatic stress disorder, and long-term increased risk of suicidal behavior compared to other bereaved and non-bereaved adolescents [7–9].

While negative grief reactions are more prominent, there is also emerging evidence of personal or posttraumatic growth in this population [10,11]. The growth is understood as a positive psychological transformation that occurs as the result of a struggle with a traumatic and highly distressing event [12]. Traumatically bereaved adolescents can experience

**Citation:** Andriessen, K.; Krysinska, K.; Rickwood, D.; Pirkis, J. The Reactions of Adolescents, Parents and Clinicians to Participating in Qualitative Research Interviews Regarding Adolescents Bereaved by Suicide and Other Traumatic Death. *Int. J. Environ. Res. Public Health* **2022**, *19*, 452. https://doi.org/10.3390/ ijerph19010452

Academic Editor: Paul B. Tchounwou

Received: 18 November 2021 Accepted: 24 December 2021 Published: 1 January 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**<sup>\*</sup>** Correspondence: karl.andriessen@unimelb.edu.au

personal or posttraumatic growth in various domains, including increased appreciation of life and relationships, increased maturity and self-care, and finding new opportunities, for example, regarding school or professional career paths [3,11,12]. Nonetheless, bereavement by suicide and other traumatic death among adolescents can disconnect them from their friends and rupture the family equilibrium [13–15], which in turn may affect their parents' and guardians' capacity to support them and/or refer them to professional help [16].

Given the potential ramifications of bereavement by suicide and other traumatic death in adolescents, conducting research with this population poses important ethical challenges. The Australian National Statement on Ethical Conduct in Human Research (2007) (Updated 2018) [17] places value on principles of research merit and integrity, justice, beneficence, and respect. Researchers are required to minimize and manage potential risks to participants, and risks are only justified when they are outweighed by potential benefits for participants or the community. The National Statement [17] stipulates that research with bereaved minors/adolescents requires specific attention to participants' capacity to understand the research and consent.

Important concerns regarding participant safety and the potential negative impact of research participation have been voiced in various research fields involving vulnerable populations, including suicide [18–21], trauma and violence [22,23], palliative care [24], psychiatry [25,26], and bereavement [27,28]. In studies with adults bereaved by suicide, research ethics committees have expressed concerns about potential harm to research participants, particularly that talking about grief experiences may traumatize them and increase their suicide risk [29–31]. Further, there are concerns about whether research participants who are negatively impacted will receive enough support [31].

Nonetheless, a recent systematic review found that most participants in suicide bereavement studies experience research participation positively [32]. Positive experiences included gaining insight into their grief experience and providing opportunities for helping others. However, a minority of participants reported unpleasant or negative experiences when participating in studies, such as being reminded of painful experiences, e.g., [27,33]. Despite the importance of these findings, the reviewed studies had several limitations [32]. Most were psychological autopsy studies (in which participants provide information about the deceased person rather than about themselves), and only one study included adolescents in the sample [34]. Hence, little is known about the experiences of adolescents of being involved in research regarding their own grief and help-seeking after a loss by suicide or other traumatic death. In addition, no study has looked at the experiences of parents and clinicians as participants in adolescent suicide and traumatic death bereavement research. This study aimed to address this gap by exploring the short-term impact of research participation experienced by adolescents, parents, and clinicians.

### *1.2. Background: Original Study*

### 1.2.1. Sampling

The original study, about which participants were then asked about their participation experience, was a qualitative study that examined how to best help adolescents bereaved by suicide and other traumatic death. Although details have been published [13,35], we summarize the original study here to provide context for the current study of participant experiences.

The original study adhered to the COREQ criteria [36] and involved a purposive sample of three groups of participants (adolescents, parents and clinicians) recruited in Australia between October 2019 and March 2020. Adolescents could participate if they had lost a family member or friend through suicide or other cause when they were aged between 12 and 18 years, and had experienced the death between six months and 10 years before participating in the study. Parents could participate if they were the parent of an eligible adolescent. Adolescents and parents could participate whether or not their parents or adolescent children participated. Clinicians could participate if they had at least five years of experience with providing professional help to bereaved adolescents.

Study participants (*N* = 72) included a total of 20 adolescents, 18 parents and 34 clinicians. The adolescent participants (16 girls, 4 boys) were aged 14 to 26 years (*M* = 19.50, *SD* = 2.95). They had lost their father (*n* = 9), brother (*n* = 3), sister (*n* = 2), mother (*n* = 2), other family member (*n* = 2), or friend (*n* = 2), by suicide (*n* = 18) or by accident (*n* = 2), on average 4 years previously (*M* = 3.92, *SD* = 2.49, range 1 to 10 years).

The parents (18 mothers) were aged 43 to 60 years (*M* = 53.20, *SD* = 4.35). The deceased person was the child's father (*n* = 10), brother (*n* = 4) or sister (*n* = 4), and the person had died by suicide (*n* = 13), accident (*n* = 2), manslaughter (*n* = 1), illness (*n* = 1) and undetermined (*n* = 1), on average 5 years previously (*M* = 5.31, *SD* = 2.89, range 1.5 to 10 years).

The clinicians (28 females, 6 males) were aged 26 to 71 years (*M* = 48.47, *SD* = 11.35). About one in four clinicians had five to nine years of experience (*n* = 8, 23.5%), 12 others (35.3%) had 10 to 19 years, and 14 (41.2%) had more than 20 years of experience.

### 1.2.2. Data Collection and Analyses

Participants could choose between taking part in an individual interview by telephone or an in-person group interview [37–41]. We conducted 28 individual interviews, and 11 group interviews with 44 participants. Individual interviews lasted, on average, 46 min (range 19–76), and group interviews were an average of 77 min (range 40–102).

The interview guide was adaptable for individual and group interviews. It consisted of open-ended questions allowing for probes and follow-up questions. The lead questions addressed different aspects of the help (for example: "In your opinion, what help should be provided to a bereaved adolescent?", "How should the help be provided?", "How long after the loss?", "What is the role of professional versus peer support?", "What are the characteristics that make help helpful?"). Adopting an inductive approach, we analyzed the interview data through a codebook-based thematic analysis [42–44].

### **2. Materials and Methods**

### *2.1. Survey*

We created a short survey with five questions to assess the participants' experiences of taking part in the individual or group interviews. Table 1 lists the survey questions. Participants were asked to answer the first four questions on a 5-point Likert-type scale (1. Not at all; 5. Absolutely). After each question, participants could write a free text comment. Question five was an open-ended question asking the participant if anything important to them was not discussed during the interview. The survey was anonymous, did not collect sociodemographic data, and could be filled out in less than five minutes.

**Table 1.** Survey questions.


### *2.2. Sampling*

Participants who took part in an in-person group interview received a hard copy of the survey at the end of the interview, and/or by email, as requested. We emailed the survey to those who had taken part in a telephone interview. We asked all participants to return the survey within two weeks, and 85% (61 out of 72) did so. This included 85% (17 out of 20) of the adolescents, 67% (12 out of 18) of the parents, and 94% (32 out of 34) of the clinicians. This amounted to 96% (*n* = 42) of group interview participants and 68% (*n* = 19) of individual interview participants.

Of those who returned the survey, 54% (*n* = 33) provided at least one comment in the free text boxes of the four survey questions and the open-ended question. This included 41% (*n* = 7) of the adolescents, 58% (*n* = 7) of the parents, and 59% (*n* = 19) of the clinicians. There was no difference between the three groups (χ 2 (2) = 1.589, *p* = 0.452). Additionally, an equal proportion of individual (53%, *n* = 10) and group participants (55%, *n* = 23) provided comments (χ 2 (1) = 0.024, *p* = 0.877).

### *2.3. Analyses*

We uploaded all data into SPSS version 26 [45]. We analyzed the quantitative data descriptively and results are presented as frequencies and percentages. Levels 1 and 2 of the 5-point Likert-type scale are considered as low, level 3 is moderate/medium, and levels 4 and 5 are high. We used the Kruskal–Wallis H test to test if there were differences between the three groups of participants. We used Kendall's tau-b correlation coefficient (2-tailed) to investigate the correlations between the data derived from the four Likert-type questions.

As most comments provided in the free text boxes were short (a few words or a short sentence), we opted to summarize the qualitative data allowing for a quantitative and qualitative report of the findings [46]. The summary applied a deductive approach, based on the survey questions. Two researchers (KA and KK) summarized the data independently and compared their report; there were no discrepancies. The research team discussed the progress and results to maximize consistency throughout the study.

### *2.4. Ethical Approval*

The Human Research Ethics Committee of The University of Melbourne approved the study (ID 1955213). All participants provided written informed consent.

### **3. Results**

### *3.1. Quantitative Findings*

3.1.1. At the Time of Participating, Did You Feel Distressed When You Participated in the Focus Group/Interview?

Most participants (75%, *n* = 46) reported that they had no or hardly any distress (18%, *n* = 11) while participating in the interview. Four participants (6.6%, three adolescents, one parent) reported moderate levels, and none reported high levels of distress (Figure 1). The Kruskal–Wallis test indicated a significant difference between the three groups (*H*(2) = 11.772, *p* = 0.003). Pairwise comparisons showed that adolescent participants scored higher than clinicians (*adj. p* = 0.002). *Int. J. Environ. Res. Public Health* **2022**, *19*, x FOR PEER REVIEW 5 of 14 11.772, *p* = 0.003). Pairwise comparisons showed that adolescent participants scored higher than clinicians (*adj. p* = 0.002).

**Figure 1.** Experienced level of distress. **Figure 1.** Experienced level of distress.

three groups (*H*(2) = 0.210, *p* = 0.900).

**Figure 2.** Perceived helpfulness.

0%

10%

20%

30%

40%

50%

4.9% 3.3%

3.1.2. Today, Do You Think That Participating Helped You in Anyway?

3.1.3. Today, Do You Feel That Participating Had Any Negative Effects for You?

16.4%

1. Not at all 2 3 4 5. Absolutely

About 95% (*n* = 58) of participants reported having experienced no or hardly any negative effects of participating (Figure 3); 1.6% (1 adolescent) reported a moderate level, and 3.2% (1 adolescent, 1 parent) a high level of experienced negative effects. The Kruskal– Wallis test found a significant difference between the three groups (*H*(2) = 11.836, *p* =

About 75% (*n* = 46) of participants reported they felt that participation was helpful,

29.5%

45.9%

3.1.2. Today, Do You Think That Participating Helped You in Anyway? 3.1.2. Today, Do You Think That Participating Helped You in Anyway?

18.0%

*Int. J. Environ. Res. Public Health* **2022**, *19*, x FOR PEER REVIEW 5 of 14

higher than clinicians (*adj. p* = 0.002).

75.4%

**Figure 1.** Experienced level of distress.

0%

10%

20%

30%

40%

50%

60%

70%

80%

About 75% (*n* = 46) of participants reported they felt that participation was helpful, 16% (*n* = 10) reported a medium level of perceived helpfulness, and 8% (2 adolescents, 1 parent, 2 clinicians) reported low levels (Figure 2). There was no difference between the three groups (*H*(2) = 0.210, *p* = 0.900). About 75% (*n* = 46) of participants reported they felt that participation was helpful, 16% (*n* = 10) reported a medium level of perceived helpfulness, and 8% (2 adolescents, 1 parent, 2 clinicians) reported low levels (Figure 2). There was no difference between the three groups (*H*(2) = 0.210, *p* = 0.900).

6.6%

1. Not at all 2 3 4 5. Absolutely

0.0% 0.0%

11.772, *p* = 0.003). Pairwise comparisons showed that adolescent participants scored

**Figure 2.** Perceived helpfulness. **Figure 2.** Perceived helpfulness.

3.1.3. Today, Do You Feel That Participating Had Any Negative Effects for You? 3.1.3. Today, Do You Feel That Participating Had Any Negative Effects for You?

About 95% (*n* = 58) of participants reported having experienced no or hardly any negative effects of participating (Figure 3); 1.6% (1 adolescent) reported a moderate level, and 3.2% (1 adolescent, 1 parent) a high level of experienced negative effects. The Kruskal– Wallis test found a significant difference between the three groups (*H*(2) = 11.836, *p* = About 95% (*n* = 58) of participants reported having experienced no or hardly any negative effects of participating (Figure 3); 1.6% (1 adolescent) reported a moderate level, and 3.2% (1 adolescent, 1 parent) a high level of experienced negative effects. The Kruskal– Wallis test found a significant difference between the three groups (*H*(2) = 11.836, *p* = 0.003). Pairwise comparisons showed that adolescent participants scored higher than clinicians (*adj. p* = 0.002). *Int. J. Environ. Res. Public Health* **2022**, *19*, x FOR PEER REVIEW 6 of 14 0.003). Pairwise comparisons showed that adolescent participants scored higher than clinicians (*adj. p* = 0.002).

3.1.4. Would You Recommend Participating in a Study like This to Others?

Almost all participant (97%, *n*= 59) said they would absolutely or close to absolutely recommend participating in a study like this to others, with the other 3.3% (2 parents)

Table 2 presents the correlations, showing that there were significant associations between questions 1 and 3 (*p* = 0.007), and questions 2 and 4 (*p* = 0.026). This reveals that greater feelings of distress during participation were weakly related to more negative experienced effects of participation, and that greater perceived helpfulness of participation

1. Not at all 2 3 4 5. Absolutely

16.4%

80.3%

**Figure 3.** Experienced negative effects. **Figure 3.** Experienced negative effects.

**Figure 4.** Recommending taking part to others.

was weakly related to recommending participation to others.

0.0% 0.0% 3.3%

3.1.5. Correlation Analysis

0%

10% 20% 30% 40% 50% 60% 70% 80% 90% 3.1.4. Would You Recommend Participating in a Study Like This to Others? 3.1.4. Would You Recommend Participating in a Study like This to Others?

8.2%

*Int. J. Environ. Res. Public Health* **2022**, *19*, x FOR PEER REVIEW 6 of 14

nicians (*adj. p* = 0.002).

86.9%

Almost all participant (97%, *n*= 59) said they would absolutely or close to absolutely recommend participating in a study like this to others, with the other 3.3% (2 parents) being somewhat likely to recommend (Figure 4). There was no difference between the three groups (*H*(2) = 0.568, *p* = 0.753). Almost all participant (97%, *n*= 59) said they would absolutely or close to absolutely recommend participating in a study like this to others, with the other 3.3% (2 parents) being somewhat likely to recommend (Figure 4). There was no difference between the three groups (*H*(2) = 0.568, *p* = 0.753).

1. Not at all 2 3 4 5. Absolutely

1.6% 1.6% 1.6%

0.003). Pairwise comparisons showed that adolescent participants scored higher than cli-

**Figure 4.** Recommending taking part to others. **Figure 4.** Recommending taking part to others.

**Figure 3.** Experienced negative effects.

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

#### 3.1.5. Correlation Analysis 3.1.5. Correlation Analysis

Table 2 presents the correlations, showing that there were significant associations between questions 1 and 3 (*p* = 0.007), and questions 2 and 4 (*p* = 0.026). This reveals that greater feelings of distress during participation were weakly related to more negative experienced effects of participation, and that greater perceived helpfulness of participation was weakly related to recommending participation to others. Table 2 presents the correlations, showing that there were significant associations between questions 1 and 3 (*p* = 0.007), and questions 2 and 4 (*p* = 0.026). This reveals that greater feelings of distress during participation were weakly related to more negative experienced effects of participation, and that greater perceived helpfulness of participation was weakly related to recommending participation to others.

**Table 2.** Correlations.


\*. Correlation is significant at the 0.05 level (2-tailed). \*\*. Correlation is significant at the 0.01 level (2-tailed).

### *3.2. Qualitative Findings*

3.2.1. At the Time of Participating, Did You Feel Distressed When You Participated in the Focus Group/Interview?

Six participants provided a comment: 2 adolescents, 1 parent, and 3 clinicians. Two adolescents reported having felt some anxiety but were otherwise fine, as stated by one: "Have anxiety but it was really okay". One parent reported feeling "distress" after hearing stories of other participants, but the participant continued that this was experienced as a "helpful connection and catharsis", which helped to reflect "how far" they "had come in

their own grief". Three clinicians reported having felt safe and supported while participating in the interview, as stated by this clinician: [It was] "Very supportive and informative".

### 3.2.2. Today, Do You Think That Participating Helped You in Anyway?

Twenty-two participants commented on this question: 5 adolescents, 5 parents, and 12 clinicians. Four adolescents and two parents experienced participating as being helpful for themselves, as stated by this adolescent: "It helped me open up and share my experience". One adolescent and three parents referred to being able to use their experiences to help others in similar situations, for example, one parent wrote: "I appreciate being able to share to help those behind me". Eight clinicians commented that participating in the interview was a learning experience, as stated by this clinician: "It's always beneficial to talk about one's practice and reflect on the strategies used in your daily work". Three other clinicians experienced the interview also as a validation and encouragement, as said by this clinician: "It reminds practitioners to continue doing great work". Still, one clinician commented on the interview as a "one-way conversation".

3.2.3. Today, Do You Feel That Participating Had Any Negative Effects for You?

Two participants noted a comment (1 adolescent, 1 parent). The adolescent stated that: "It brought up unpleasant memories but not overwhelming". The parent reported having felt sad and having cried after the interview "particularly after listening to other participants, their loss of their loved ones".

### 3.2.4. Would You Recommend Participating in a Study Like This to Others?

Seven participants provided a comment (2 parents, 5 clinicians). The parents emphasized the importance of listening to others "across cultures and genders" to "identify evidence-based approaches that will work". In addition, it may help us "learn how to best equip and cope with a suicide loss". The clinicians would recommend participating because it "gives support and facilitates time to reflect, focus, and share ideas about a neglected cohort". In addition, it can "increase education and awareness" for clinicians, and "support for young people".

3.2.5. Was There a Topic Very Important to You That You Thought We Should Discuss but Did Not?

Eighteen participants (4 adolescents, 3 parents, 11 clinicians) wrote a comment in the last free text box of the survey, although several comments appeared to be more about what participants thought was important regarding grief and bereavement, rather than about topics that were not addressed in the interviews. Three adolescents emphasized what they saw as crucial for helping bereaved adolescents. One wrote: "trust and knowing that there is good information and discussion out there", and another highlighted the "need to let the younger teens know it's not weak to speak and get help". In addition, one parent elaborated on the "secrecy around suicide" and the tendency of blaming someone for a suicide, which may hinder bereaved adolescents. Subsequently, this parent argued for increased literacy around dealing with grief in society. Two clinicians pointed at cultural sensitivity in providing help, and financial impact of the loss, as important topics for further discussion.

One parent commented on the composition of the group in which she participated. Most participants in her group had lost a husband to suicide and only one participant had lost a child to suicide. Hence, the participant recommended that groups in future studies could be more balanced regarding types of relationship.

Twelve participants (1 adolescent, 2 parents, 9 clinicians) confirmed that everything important had been discussed in their interview, as exemplified by one adolescent who noted: "I believe we covered everything vital", and participants expressed gratitude for having been involved, as stated by this clinician: "I'm sure there is more, but right now it feels good".

### **4. Discussion**

This study was a first to investigate the reactions of adolescents, parents, and clinicians regarding their participation in a study on adolescents bereaved by suicide and other traumatic death. Most participants reported that they experienced no distress while participating and no negative effects of participating. Rather, they found participation helpful for them and said they would definitely recommend participating in a study like this to others. These findings reflect results of bereavement studies with parents [47,48], siblings [49], and people bereaved by suicide [30,32], which reported that research participation was mostly a positive experience, with few negative experiences, and participants perceiving participation as being beneficial for themselves and others [30,32,47]. In addition, the correlation analysis showed that there is no contradiction between reporting distress or negative effects of participation and experiencing participation as helpful and recommending it to others. This finding is corroborated by suicide-related research with adolescents and trauma research with adults suggesting that emotional distress can be understood as an indicator of engagement in a data collection process rather than as an indicator of harm [50–52].

Although the overall levels of distress and negative effects reported in our study were low, adolescents reported the highest scores, and a few mentioned having experienced anxiety or unpleasant memories. This has also been reported with regard to adolescents participating in health-related studies [53]. As suggested above, these emotional reactions may be due to participants' engagement with the research project [51], and a study by Hawton and colleagues [34] suggested that this may lead to them benefitting more from participating than adults. However, it may also point to the presence of emotional problems [53,54]. One study in a systematic review [53] examining children's and adolescents' reactions to participating in biomedical and health-related studies found that the presence of emotional problems in children was the only variable associated with short-lived negative research participation experiences [55]. Objective variables such as age, gender, methods of data collection, and topic/health condition examined in the studies, were not associated with participants' appraisals of study participation [53,55]. Another study on pain also reported that age and level of pain did not affect responses regarding experiences of research participation [56]. However, as in our study, young participants still recommended participation to others [56]. Nonetheless, while only a few participants reported negative experiences, the phenomenon needs further investigation in bereaved adolescents, especially since both short-term deterioration and improvement in mood have been reported in adolescent mental health research [54].

Parents in our study were very positive about research participation. This is corroborated by other research with parents who participated in a study concerning the sudden and traumatic death of their children [27]. All parents in that study reported positive experiences and none regretted participating, despite 73% experiencing the interviews as (a little to very) painful [27]. As in our study, having the opportunity to share experiences and being able to help others contributed to a positive experience [27]. Still, one parent in our study reported being emotionally affected, although this parent still perceived participation as helpful. Follow-up data from adults participating in suicide research revealed that participants can be susceptible to short-term deteriorations in mood [57]. However, any negative effects of participation were confined to the days immediately following the study and this temporary deterioration in mood did not increase risk of suicidal thoughts [57].

As with research with bereaved adolescents, research with suicide bereaved adults indicated that objective factors such as gender of participants, their relationship to the deceased, the method of suicide, and time since loss appeared to have little effect on their experiences of participating in a research interview [58]. Moreover, research regarding mental health in adults showed that the presence of mental health problems did not differentiate participants with negative or positive experiences of research participation [59]. Hence, it may be that the presence of emotional problems has a stronger impact on the

research experience of children and adolescents compared to adults, though further research is needed to ascertain such differences between children/adolescents and adults.

Clinicians in our study found research participation to be a learning experience. This is supported by findings from studies with clinicians who had lost a patient to suicide, who reported their research participation as a learning and therapeutic experience [60,61]. The finding is also in line with literature on substance abuse research participation [62]. Clinical research participants were more willing to use research findings in practice, especially those with favorable attitudes toward evidence-based practices and whose agencies supported professional growth. The combined findings of our study and others from the literature [60,62] suggest that research participation may reinforce clinicians' willingness to use research findings in practice, thus contributing to quality of services.

Although our participants expressed gratitude for being involved in the research, a phenomenon also noted in the literature [63], our data suggest that from the three groups of participants, adolescents potentially experience the most distress and negative effects of participation in research interviews. Thus, researchers must inform potential participants, especially adolescents, about both the potential benefits and distress of research participation, and implement appropriate participant distress management protocols. As Parkes [64] pointed out, if a participant becomes distressed, "the needs of the respondent should take priority over the needs of the research" (p. 174). According to such protocols, researchers must provide support to a distressed interview participant according to the level of distress. This can range from allowing them to pause or withdraw from participation, providing emotional support to participants, providing or arranging referral to an appropriate support service (which may include a parent or guardian), or calling medical emergency services.

The literature indicates the important roles of narrating and sharing grief experiences in the processes of meaning-making and personal growth in those bereaved by suicide, which may also contribute to a positive experience as a study participant [10,58]. Interactions with a skilled and empathetic interviewer can also contribute to positive experiences for bereaved study participants [48,65]. While researchers must be sensitive to potential distress in participants, and provide emotional support if needed, they must also be aware of the potential methodological and ethical challenges when research and therapy blur [50,66]. For example, Biddle and colleagues [50] cautioned that participants may share information that they do not want to be used for the research, and researchers may not have the skills or capacity to deal with the distress or unintended shared information from the participant [50,64,66]. Hence, our findings and the broader literature imply that research interviews must be conducted by experienced and properly trained interviewers who can deal with participants' emotions and make judgements about pausing or continuing an interview, or referring participants to external support [50,64,66]. Further research is needed to clarify the role of the researcher and their approach to participants regarding balancing data collection and being empathic and supportive for the research experience of bereaved participants [48,65,67,68].

To fully understand the study findings, it is important to note that the study involved participants who volunteered to share their experiences, and study participants may not be representative of the population from which they are recruited. In a bereavement study, Akard and colleagues [69] found that those who are motivated and have the capacity to participate tended to respond to the initial researchers' invitations, and sending more than three invitations hardly increased the response rate [69]. Bereaved people who perceive research participation as too difficult may either decline participation or refuse passively by not responding [69,70]. These findings of the literature indicate the soundness of (potential) participants' judgements about research participation [50] and indicate that participants make appropriate cost–benefit appraisals of their participation [52]. Research is needed to confirm these observations in adolescents bereaved by suicide, parents or other family members of bereaved adolescents, and clinicians. Nonetheless, researchers and research ethics committees may consider these when designing and assessing research studies in

this field. In addition, participants in our study had experienced the bereavement at least six months before participating, which may also have contributed to participants reporting little distress.

### *Limitations*

The study involved a modest sample from a qualitative interview study. Despite the high response rate, the findings may not reflect the experiences of those who were invited to participate and chose not to. It is also not known whether findings apply to participants of other interview studies or studies utilizing other methods of data collection. Further, the survey did not include definitions of 'distress' or 'negative effects', and data were collected only at one point in time shortly after research participation. Future studies could include pre- and post-measures and longitudinal follow-up to assess any longer-term consequences.

### **5. Conclusions**

The study clearly indicates that bereaved adolescents, parents and clinicians can safely participate in research interviews regarding their experiences of grief and help after suicide and other traumatic death. Participants reported that they experienced little distress and would recommend participation to others. To prevent and mitigate potential distress, training of research staff and implementation of appropriate participant distress protocols are imperative. Future studies could include longitudinal follow-up of participants.

**Author Contributions:** Conceptualization, K.A., D.R. and J.P.; methodology, K.A.; software, K.A.; data collection, K.A. and K.K.; formal analysis, K.A. and K.K.; writing—original draft preparation, K.A.; writing—review and editing, K.A., K.K., D.R. and J.P.; supervision, D.R. and J.P.; project administration, K.A.; funding acquisition, K.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** K.A. is supported by a National Health and Medical Research Council Early Career Fellowship (GNT1157796) and an Early Career Researcher Grant of The University of Melbourne (ECR1202020). J.P. is funded by a National Health and Medical Research Council Investigator Grant (GNT1173126).

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Human Research Ethics Committee of The University of Melbourne (ID 1955213, 8 October 2019).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** According to the ethics approval, data can only be accessed by the members of the research team.

**Acknowledgments:** We are most grateful to the participants for sharing their experiences, and to the numerous individuals and organizations who helped with recruitment and/or hosted the focus groups, especially: The Compassionate Friends Victoria, Support After Suicide Jesuits Social Services, Standby Support After Suicide, headspace National, Anglicare Suicide Prevention and Grief Support South Australia, Sabrina's Reach 4 Life, and the Australian Centre for Grief and Bereavement.

**Conflicts of Interest:** The authors declare no conflict of interest. The funding bodies had no role in the design of the study, the collection, analysis and interpretation of data, the writing of the manuscript, and the decision to submit it for publication.

### **References**


## *Review* **Continuing Bonds after Loss by Suicide: A Systematic Review**

**Rebecca Goodall <sup>1</sup> , Karolina Krysinska <sup>2</sup> and Karl Andriessen 2,\***


**Abstract:** The concept of continuing bonds as an alternative to detachment from the deceased person has gained traction in grief literature over the years. Those bereaved by suicide are likely to experience various grief reactions and may be at-risk for adverse grief and mental health outcomes. However, it remains unclear how those bereaved by suicide experience continuing bonds. To address this gap, we conducted a systematic review according to PRISMA guidelines. Searches of peer-reviewed literature in Medline, PsycINFO, Embase, Emcare, EBM Reviews, and Scopus identified 15 studies (2 quantitative and 13 qualitative) reporting on 12 samples, published between 2010 and 2021. The study quality of the quantitative studies was poor, but it was fair amongst the qualitative studies. People bereaved by suicide experienced continuing bonds across a variety of domains and reported mostly positive experiences. Factors that tended to have an impact on the expression of continuing bonds included time since bereavement, relationship to the deceased, social expectations, sex of the bereaved, and the ability of the bereaved to make meaning of the death. The review concludes that most participants reported positive experiences with continuing bonds. However, discrepancies between males and females and between those bereaved by suicide and those bereaved by other causes warrants further investigation. In addition, longitudinal community-based research involving representative samples is needed to understand the evolution and experience of continuing bonds over time in those bereaved by suicide and to inform future efforts in supporting them.

**Keywords:** continuing bonds; grief; bereavement; suicide; systematic review

### **1. Introduction**

Suicide claims the lives of over 700,000 people per year globally [1]. A recent metaanalysis of population-based studies indicated that approximately one in twenty people experience a loss by suicide in one given year and one in five do so during their lifetime, although the impact of the death may depend on the closeness of the relationship [2]. Due to the complex nature of grief after suicide, individuals bereaved by suicide have a higher risk of adverse mental health outcomes such as posttraumatic stress disorder, depression, suicidal ideation, and attempted suicide than those bereaved by other causes [3]. People bereaved by suicide also have particular needs for professional help in dealing with their grief (for example, regarding feelings of guilt and struggles with 'why' questions) compared to those bereaved by natural causes [4].

Initially, grief literature suggested that bereaved individuals progress through different stages of grief in order to achieve detachment from the deceased person [5]. However, Klass and colleagues [6] challenged this theory and proposed the concept of continuing bonds as an alternative to detachment. They defined continuing bonds as the presence of an ongoing inner relationship with the deceased person [6]. Still, the concept of continuing bonds has been described in cultural literature and various religions over the centuries [7]. Klass and colleagues drew attention to specific rituals which demonstrated the practice of continuing bonds long before the theory gained traction in grief literature [6,7]. They gave

**Citation:** Goodall, R.; Krysinska, K.; Andriessen, K. Continuing Bonds after Loss by Suicide: A Systematic Review. *Int. J. Environ. Res. Public Health* **2022**, *19*, 2963. https:// doi.org/10.3390/ijerph19052963

Academic Editor: Paul B. Tchounwou

Received: 29 January 2022 Accepted: 1 March 2022 Published: 3 March 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

the example of Japanese Buddhism, in which the deceased become part of the spirit family and are accessible to the bereaved, and in Judaism, in which the deceased are remembered through prayer and the physical lighting of a candle [6,7].

Recent studies have aimed to characterise continuing bonds as positive or negative and to specify the ways in which they manifest [7]. The process of meaning-making emerged as an important tool in the formation of positive bonds [8]. Meaning-making refers to the way in which bereaved individuals process the death in a way that holds meaning for them [9]. The concept of continuing bonds has gained traction over the years, as witnessed through the development of designated instruments [10] and studies in various bereaved populations (such as after the death of a parent or a child) [7]. However, it remains unclear how continuing bonds manifest themselves in those bereaved by suicide and how this population experiences continuing bonds. To-date, no review on this topic has been conducted. Still, a better understanding of this aspect of grief in people bereaved by suicide may provide useful information for future efforts in supporting them.

This review will address this gap by synthesising and analysing the research on continuing bonds in individuals bereaved by suicide. It aims to characterise aspects of the continuing bond including how they manifest, and whether those bereaved experience them as positive or negative.

### **2. Materials and Methods**

### *2.1. Search Strategy*

The review adhered to PRISMA guidelines [11], and the protocol was registered in the PROSPERO database (CRD42021271971). The review involved systematic searches of the following databases: Medline, PsycINFO, Embase, Emcare and EBM Reviews (all accessed via Ovid), and Scopus. The search in Medline comprised MeSH and text words: (Continu\* bond\*.mp OR continu\* relationship\*.mp OR continu\* connection\*.mp OR Ongoing bond\*.mp OR Ongoing relationship\*.mp OR Ongoing connection\*.mp OR meaning making.mp OR Memorial\*.mp OR Ritual\*.mp) AND (Grie\*.mp OR Grief/ OR Mourn\*.mp OR Bereav\*.mp OR Bereavement/ OR Suicide bereave\*.mp OR Bereave\* by suicide.mp OR loss by suicide.mp OR suicide loss survivor\*.mp) AND (suicide.mp OR Suicide/). A similar search string including headings and keywords was used in the other databases.

One researcher (R.G.) conducted the search in August 2021. It was limited to publications in English but not by date of publication. Two researchers (R.G., K.A.) assessed titles and abstracts for eligibility and any discordance was resolved through discussion with the third researcher (K.K.). The same two researchers then assessed the full text of potentially relevant studies using the inclusion/exclusion criteria. Researcher R.G. hand-searched the references of the included studies and conducted a forward-citation search in Google Scholar to identify any further studies. Figure 1 summarises the search strategy.

### *2.2. Inclusion and Exclusion Criteria*

Studies were included if: (1) the study population consisted of people bereaved by suicide; (2) the study provided empirical qualitative and quantitative data on continuing bonds with an individual lost to suicide; (3) the study was published in the English language; and (4) the study was published as a paper in a peer-reviewed journal.

The review excluded: (1) studies not providing data specifically on people bereaved by suicide; (2) studies not providing data on continuing bonds in the context of suicide bereavement; (3) studies based on other methods, such as case studies, reviews and opinion paper; (4) studies not written in English; and (5) studies which were not peer-reviewed. Unclarity on the inclusion of three studies was resolved through discussion with the third researcher (K.K.).

**Figure 1.** PRISMA Flow Diagram. **Figure 1.** PRISMA Flow Diagram.

### *2.3. Data Extraction*

*2.2. Inclusion and Exclusion Criteria*  Studies were included if: (1) the study population consisted of people bereaved by suicide; (2) the study provided empirical qualitative and quantitative data on continuing bonds with an individual lost to suicide; (3) the study was published in the English lan-Two researchers (R.G., K.A.) independently extracted the following data: author, year and location of study, eligibility criteria, sample size, participants' sex, age, time since bereavement and relationship to the deceased, setting, study design, and main findings.

#### guage; and (4) the study was published as a paper in a peer-reviewed journal. *2.4. Quality Assessment*

The review excluded: (1) studies not providing data specifically on people bereaved by suicide; (2) studies not providing data on continuing bonds in the context of suicide bereavement; (3) studies based on other methods, such as case studies, reviews and opinion paper; (4) studies not written in English; and (5) studies which were not peer-reviewed. Unclarity on the inclusion of three studies was resolved through discussion with the third researcher (K.K.). Two researchers (R.G., K.A.) independently conducted the quality assessment and resolved disagreements through discussion. No eligible study was excluded based on its quality. Quantitative studies were assessed using the Newcastle-Ottawa Quality Assessment Form for Cohort Studies [12], comprising eight items across three domains: (1) selection (four items), (2) comparability (one item), and (3) outcome (three items). Scores in each domain were totalled to determine study quality as good, fair, or poor. The interrater reliability was high (κ = 0.82).

*2.3. Data Extraction*  Two researchers (R.G., K.A.) independently extracted the following data: author, year and location of study, eligibility criteria, sample size, participants' sex, age, time since bereavement and relationship to the deceased, setting, study design, and main findings. The qualitative studies were assessed using the Consolidated Criteria for Reporting Qualitative Research (COREQ) [13] consisting of thirty-two items across three domains: (1) research team and reflexivity (eight items), (2) study design (fifteen items), and (3) analysis and findings (nine items). For each study, the number and percentage of items satisfied

within each domain and across all domains was calculated. The interrater agreement was high (κ = 0.92).

### **3. Results**

### *3.1. Study Characteristics*

The searches identified fifteen studies reporting on twelve samples. Five studies were conducted in Australia [14–18], three in the USA [9,19,20], three in the UK [21–23], two in Israel [24,25], one in Canada [26], and one in Switzerland [27]. There were two quantitative and thirteen qualitative studies. The quantitative studies collected data via online questionnaires [19,24]. One also utilised the Two-Track Bereavement Questionnaire (TTBQ) [10] to measure the level of functioning of bereaved individuals and their relationship with the deceased person [24]. Eleven of the thirteen qualitative studies involved semi-structured or in-depth interviews either in-person or via telephone, one utilised a survey [20] and one collected data through a series of workshops over twelve weeks [18].

The quantitative studies had sample sizes ranging from *N* = 159 [24] to *N* = 1301 [19], whilst qualitative studies had sample sizes ranging from *N* = 7 [14] to *N* = 50 [27], with the exception of Jahn [19,20] who used the same data set for both a quantitative and qualitative study. Tables 1 and 2 summarise the quantitative and qualitative studies, respectively.

A large percentage (67–89%) of participants in most studies (*n* = 9) were female [16, 19–22,24–27]. Three studies had equal numbers of both sexes [15,17,23], two had slightly more male (57–60%) than female participants [9,14], and one study did not report the sex of participants [18]. Across studies, the age range was between 8 and 85 years [19,23].

Whilst most studies included a broad range of relationships between the bereaved individuals and the deceased, six focused on a specific relationship, including bereaved parents (*n* = 3) [15–17], siblings (*n* = 2) [14,25], and children (*n* = 1) [23].

Time since death ranged between two months and over forty-five years [18,24]. Two studies focused on the grief reactions from five months onwards [21,22], whilst others looked at continuing bonds and grief reactions on a longer-term from five years onward [9] or within a ten-year period [14]. Eight studies did not limit participants by time since loss [16,19,20,23–27], and three included longitudinal data [15,17,18].

### *3.2. Quality Assessment*

Appendix A presents the methodological quality of the quantitative studies, and the two included studies received a rating of 'poor' quality. One study [24] scored well in the 'comparability' domain, but both studies [19,24] tended to score poorly in the 'selection' and 'outcome' domains by using selected samples and relying on self-reported outcomes. Appendix B outlines the quality assessment of the 13 qualitative studies. The studies satisfied between 34% [22] and 75% [15,17] of the COREQ criteria [13]. Most studies reported only few items across the 'research team and reflexivity' domain (on average 35% of items were reported). However, most items (on average 63%) in both the 'study design' and 'analysis and findings' domains were reported.

### *3.3. Study Findings*

### 3.3.1. In What Way Can Continuing Bonds Manifest?

Findings from the included studies showed that continuing bonds manifested or were experienced in various ways. Quantitative data showed that 63% of participants reported having spiritual experiences [19]. The most common experiences included dreams about the deceased (73.4%), feeling the presence of the deceased (53%), and profound coincidences (40.3%) [19]. In qualitative studies, participants reported the use of rituals and memorials (both physical and online) [9,16,21,23], spiritual and religious beliefs [9,26,27], metaphysical experiences [20,26], physical objects [9], and suicide-related thoughts and behaviours, such as a desire to join the deceased and/or understand what they went through [9].



Jahn et al. (2014) USA [19]

Survivors of

M = 135 F = 1155 U = 11

40–85

<1 y (25.7%; *n* = 334), 4 y–10 y (22.0%; *n* = 286), 11 y–20 y (12.2%; *n* = 159), and >20 y (8.2%; *n* = 107).

(*n* = 204), child (*n* = 180), friend (*n* = 112), niece/nephew (*n* = 21), grandchild (*n* = 9), another relationship (*n* = 119).

one demographics and one about experience of suicide. Both containing both qualitative and quantitative info

characteristics of loss and chi squared to examine demographical differences between responses

(ii) spiritual experiences of survivors of suicide, including questions about type and frequency of spiritual experiences

regarded as positive with 74.5% of participants (*n* = 610) finding them 'helpful' and only 4.8% (*n* = 39) viewing them as 'harmful' (19.9%, *n* = 163 saw them as neither helpful nor harmful). Being female and having a closer pre death relationship (family member as opposed to friend or client) was positively associated with spiritual experiences.

suicide loss *N* = 1301


**Table 2.**

Summary

 of qualitative

 studies.




**Table 2.** *Cont.*

Suicide bereaved individuals tended to experience continuing bonds across multiple domains both concurrently and over time [15,25]. It was common to experience continuing bonds across a continuum, for example, engaging in both public and private expressions of continuing bonds [25]. A public manifestation may include participation in ceremonies or rituals such as a funeral whilst a private manifestation may include writing to or talking to the deceased [25]. Although the process of developing continuing bonds was never described as linear, time since bereavement did appear to affect how the bonds were experienced. Entilli and colleagues [15] and Ross and colleagues [17] described experiences with continuing bonds as they appeared at 6, 12, and 24 months, observing that memorialisation and intrusive thoughts were present early on whereas paranormal beliefs and the maintenance of the relationship with the deceased were new themes at 24 months. Similarly, Sands and colleagues [18] reported themes as they appeared during workshops spanning over 12 weeks. Participants experienced thoughts around 'why' the person had died by suicide before reconstructing the death story and repositioning the relationship in order to establish positive ongoing bonds [18]. Suicidal ideation with a desire to join the deceased and/or to understand what they went through was another common initial reaction that typically subsided over time [9,14].

Qualitative studies found that most participants wanted to experience continuing bonds and actively pursued them [14,26]. They achieved this through the creation of online memorials [21,22], visiting places that were frequented by the deceased [16], keeping objects that belonged to the deceased [26], actively reminiscing about the deceased [26], and writing to the deceased [15]. However, some participants found that the social expectations of grief and the stigma associated with suicide limited their public expression of continuing bonds, forcing them to express their grief privately [16,21,22].

There were differences in the types of bonds experienced according to sex and type of relationship to the deceased person [15,20,21]. A qualitative study found that female participants and those who had a close pre-death relationship with the deceased family member or partner had more spiritual experiences with the deceased [20]. Females also created more online memorials than males [21]. Conversely, males appeared to engage more in avoidant coping strategies and would express their grief less openly [15]. Though children may experience continuing bonds in similar domains as adults, research suggests they tend to interpret their experiences differently and this varies depending on age [23].

### 3.3.2. Continuing Bonds as a Positive or Negative Experience

Participants generally perceived continuing bonds as positive experiences [16,17,19–21,27], which they associated with comfort and hope [26]. Although less common, negative experiences were those that were beyond the control of the bereaved individuals and included subconscious thoughts or feelings [25], spiritual or metaphysical experiences [26], or the unexpected discovery of facts relating to the deceased that were not in-line with the perceptions of the bereaved [21]. A quantitative study [19] demonstrated that 74.5% of participants interpreted spiritual experiences with the deceased as helpful whilst only 4.8% viewed them as harmful [19]. However, another quantitative study comparing outcomes of those bereaved by suicide with those bereaved by sudden or expected death found that those bereaved by suicide had lower levels of close positive relationships with the deceased both pre- and post-death, including less expression of continuing bonds [24].

### **4. Discussion**

This review aimed to synthesise the research literature concerning continuing bonds in people bereaved by suicide, including how they manifest and whether the bereaved experienced them as positive or negative. The review identified 15 studies reporting on 12 samples. Factors that affected the way in which continuing bonds manifested included time since death, type of relationship, societal expectations, sex of the bereaved person, and their ability to make meaning of the death. In all but one study [22], participants reported mostly positive thoughts, feelings and experiences regarding continuing bonds.

### *4.1. Factors That Affect the Manifestation of Continuing Bonds*

Aside from two longitudinal studies [15,17], there appears to be little research on the effect of time since bereavement on the manifestation of continuing bonds in those bereaved by suicide. Participants from one study [9] reported that the intensity of the bonds had not lessened up to 10 years following the suicide. This is supported by a quantitative study [24] which found that when compared to those bereaved by sudden or expected deaths, those bereaved by suicide tended to maintain more intense bonds for a longer period of time and consequently took longer to engage in 'meaning-making'. These findings are important due to the strong association between intense grief and less positive personal transformation [24], highlighting the need for more longitudinal, comparative data.

Societal beliefs and expectations may impact on the manifestation of continuing bonds in several ways [16,21,22,27]. Some participants felt they were forced to express their ongoing relationship more privately due to the stigma surrounding suicide [9,21], and the expectation to resolve grief [16]. Our review revealed that suicidal ideation [9,14] in suicide bereaved individuals can be a way to reconnect with the deceased or to understand what they went through, suggesting that suicidal ideation may be a manifestation of continuing bonds in those bereaved by suicide. Interestingly, participants claimed that this process aided their grief work by helping them to make sense of the suicide; however, reported they felt discouraged from sharing these thoughts with others [9]. These findings illustrate the importance of addressing suicidal thoughts in a constructive manner as opposed to inadvertently reinforcing suicide stigma by silencing them [9].

The stigma associated with suicide in the context of religions may be particularly strong, with some bereaved individuals choosing to hold private ceremonies following a suicide death [27,28]. This is supported by a qualitative systematic review that found that religious individuals often isolate themselves due to feelings of shame and suicide-related stigmatisation emanating from religious doctrines [29]. Interestingly, taking part in religious ceremonies was common practice even in those participants who identified as agnostic [27]. A possible explanation for this may be that societal and cultural expectations dictate how we grieve, insofar as rituals such as funerals are expected in western culture [30]. Future studies conducted in different cultures may shed light on the expression of continuing bonds with regards to spiritual and religious themes and the possible stigma in this context.

Referring to the 'Dual Process Model of Coping with Bereavement' [31], studies reported differences between males and females with females engaging in more 'lossoriented' behaviours, including continuing bonds, and males engaging in more 'restorationoriented' behaviours, including learning new skills [15]. Whilst it may be the case that males experience continuing bonds to a lesser extent than females and/or in different domains [32], literature suggests that societal perceptions of gender and masculinity may influence the expression of grief, and grief after suicide, in males [33], resulting in males engaging in restoration-oriented activities as a distraction from or expression of their grief [33]. This view is supported by Entilli and colleagues [15] who found that fathers avoided discussing their feelings regarding the loss by suicide. Future studies may further compare the experience of continuing bonds between males and females and inform services to be directed accordingly.

The process of 'meaning-making' was mentioned throughout the reviewed studies. It refers to the ability of the bereaved person to make sense of or find meaning in the death [8,34] and has been widely recognised in grief literature [8,34]. Milman and colleagues described meaning-making as a method of alleviating the cognitive dissonance caused by the death through purposeful reflection [34]. They highlight the importance of differentiating this process from rumination as the latter fails to negotiate this discrepancy and is characterised by passive, repetitive and negative thoughts [34]. Our review indicated that those who were able to make meaning of the suicide experienced more positive continuing bonds whilst those who were unable to make meaning experienced more negative continuing bonds. In addition, high resilience characteristics and a strong social support network tended to predict less intense continuing bonds and higher levels of posttraumatic growth [24].

### *4.2. Continuing Bonds as a Positive or a Negative Experience*

Most participants in the included studies reported positive feelings experienced with continuing bonds [16,18,23], a finding that is mirrored by those of general bereavement studies [35,36]. However, a quantitative study comparing different forms of bereavement indicated that those bereaved by suicide reporting lower levels of close and positive relationships with the deceased both pre-and post-death [24]. The incorporation of multiple factors, including close and positive, and pre-and post-death relationships, make it hard to accurately compare these results with other studies. As demonstrated by Leichtentritt and colleagues [25], a 'close' relationship is not necessarily synonymous with a 'positive' one. Likewise, as suggested throughout this review, the process of meaning-making transforms the relationship so that the pre- and post-death relationships may not be equivalent. Further studies comparing experiences of continuing bonds between those bereaved by different causes may further clarify how to understand these experiences.

Negative experiences occurred when continuing bonds in the form of memories or objects faded or were lost [14,21], when participants were not able to make meaning of the loss [15], or when aspects of continuing bonds were beyond the control of the bereaved [22,25]. Participants experienced feelings of distress at fading memories of the deceased [14], or due to the sudden and unexpected disappearance of online memorials [21], which was described as a 'double loss' [22]. Some expressed apprehension and guilt as contributing to the continuing bonds as they did not want to abandon the deceased person [10]. Additionally, participants experienced cognitive dissonance when their religious beliefs were in opposition to the actions of the deceased [27], or upon the discovery of new information about the deceased that was contradictory to their own internal representation of the deceased person [21,22]. These psychological discrepancies were found to hinder the ability of the bereaved person to make meaning of the death and enhanced negative feelings associated with continuing bonds [9]. Further research may clarify in which circumstances it may be more beneficial for the bereaved person to relinquish than to maintain the bond [35,36].

Participants utilised several tools and behaviours to create or maintain positive continuing bonds. Many studies observed the tendency of participants to omit certain aspects of the deceased personality or behaviour when recalling memories of them [23], or to hold more symbolic or fictitious representations of the deceased [25,27]. Some bereaved individuals favoured online memorials due to them being more interactive than traditional mourning objects [21]. This is supported by literature reporting that this allows the bereaved to explore the relationship in more depth and to discover new things about the deceased, thus enabling the relationship to evolve [37]. Nonetheless, a potential negative aspect of online grief-related activities is the ease with which those bereaved can become preoccupied with the relationship, hindering the meaning-making process [21,38].

Positive and negative experiences with continuing bonds were often intertwined. Wood [23] found that negative thoughts about the deceased's personality or behaviour initially evoked distress but with reflection and perspective, understanding and personal growth could be achieved. This was echoed by Maple [16], who found that parents could initially be distressed at the ongoing 'presence' of their deceased child but in time came to enjoy these experiences. Nonetheless, participants reported a bitter 'aftertaste' associated with positive thoughts about the deceased as they tended to trigger difficult emotions and memories [23]. Remembering the more holistic picture (both good and bad) was beneficial in making sense of the suicide and relieving responsibility and guilt in those bereaved by suicide [23,39].

### *4.3. Limitations*

The studies included in the review and the review itself entailed a few limitations. Most studies were qualitative, or cross-sectional (of relatively poor quality) with primarily female participants from western countries, making the impact of culture, sex and time since bereavement difficult to discern. This highlights the need for international research involving representative samples. Also, further studies involving control groups or adopting

community-based longitudinal designs may enable researchers to capture the experiences of continuing bonds over time. While the review involved searches in six databases, future reviews can broaden the scope by including more databases as well as grey literature, which may increase, for example, the likelihood of finding negative experiences in the context of continuing bonds.

### *4.4. Implications*

A better understanding of continuing bonds in suicide bereavement may inform future interventions and enable service providers to deliver more accurate and targeted support. In particular, the review highlights the substantial impact that societal expectations and stigma continue to have on the experience of grief in suicide bereaved individuals [40]. The findings of this review may direct future studies and help to characterise the experiences of continuing bonds in people bereaved by suicide with greater clarity.

### **5. Conclusions**

People bereaved by suicide commonly experience continuing bonds and generally interpret these as positive experiences. Factors such as time since bereavement, social and cultural expectations, sex of the bereaved person, and the ability of the bereaved to make meaning of the death may influence how continuing bonds manifest and whether they are experienced as positive or negative (although further research is needed). Postvention efforts should consider the process of meaning-making in creating the basis for positive continuing bonds and should attempt to address the stigma and societal expectations surrounding suicide bereavement. Future studies should involve representative samples, compare with continuing bonds after other causes of death, and investigate continuing bonds in suicide bereavement longitudinally.

**Author Contributions:** Conceptualization, K.A., K.K. and R.G.; methodology, R.G., K.K. and K.A.; software, R.G.; validation, K.K. and K.A.; formal analysis, R.G. and K.A.; investigation, R.G.; data curation, R.G. and K.A.; writing—original draft preparation, R.G.; writing—review and editing, R.G., K.K. and K.A.; supervision, K.K. and K.A.; project administration, K.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** K.A. was supported by a National Health and Medical Research Council Early Career Fellowship (GNT1157796). The funder had no role in the design of the study, in the collection, analyses, or interpretation of data, in the writing of the manuscript, or in the decision to publish the results.

**Institutional Review Board Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **Appendix A**

**Table A1.** Quality assessment <sup>1</sup> of quantitative studies.


#### **Table A1.** *Cont.*


one star for each numbered item within the Selection and Outcome categories. A maximum of two stars can be given for Comparability. Thresholds for converting the Newcastle-Ottawa scales to AHRQ standards (good, fair, and poor): Good quality: 3 or 4 stars in selection domain AND 1 or 2 stars in comparability domain AND 2 or 3 stars in outcome/exposure domain. Fair quality: 2 stars in selection domain AND 1 or 2 stars in comparability domain AND 2 or 3 stars in outcome/exposure domain. Poor quality: 0 or 1 star in selection domain OR 0 stars in comparability domain OR 0 or 1 stars in outcome/exposure domain.




1

2

3

4

5

6

7 8

9 10

11

12

13

14 Setting of data collection

15 16 Description

17

18

19 Audio/visual

 recording

 p. 325

 p. 75

 p. 378

 p. 4

 p. 2

 p. 102

 p. 337

 p. 1106

 p. 59

 p. 2

 p. 103

 p. 879

 Repeat interviews

 Interview guide

 p. 325

 p. 74–75

 p. 378

 p. 3–4

 p. 2

 p. 2

 p. 102

 p. 102

 p. 337

 p. 5–6

 p. 1106

 p. 59

 p. 2

 p. 2

 p. 103–104

 p. 878–879

 of sample

 p. 325

 p. 74

*Data collection*

 p. 378

 p. 5

 p. 2–3

 p. 101

 p. 336

 p. 5

 p. 1105

 p. 59

 p. 2–3

 p. 103

 p. 877–878

Presence of non-participants

 p. 325

 p. 75

 p. 4

 p. 2

 p. 336

 p. 5

 p. 59

p. 59

 p. 2

 p. 103

 p. 878


*Int. J. Environ. Res. Public Health* **2022**, *19*, 2963

**Table A2.** *Cont.*

### **References**

