Evaluation of grief reactions in children: A retrospective review

Aim: To investigate common grief reactions experienced by children ages 5-17 after the loss of a significant person in their life. Methods: A retrospective review of the records of 83 children for common grief reactions indicated by parent(s)/caregivers of children attending a bereavement camp from July 2009 through July 2011. The review included common grief reactions, gender, age, relationship of the child to the deceased and professional support. Results: The independent t-test analysis indicated that boys reported significantly higher grief reactions than did girls. Total grief scores indicated the age groups did not significantly differ based on their grief scores. Children who had received professional support prior to attending the camp had significantly higher grief scores than those who had not received professional support. Conclusions: Future studies evaluating common grief reactions in children need to be conducted. Research is also indicated after bereavement interventions are utilized with children. Correspondence to: Teresa Howell, DNP, RN, CNE, Associate Degree Nursing Program Coordinator, Professor of Nursing, 316 West Second Street, Morehead State University, 201N Center for Health Education and Research, Morehead, KY 40351, USA; Tel: 606-783-2815; E-mail: t.howell@moreheadstate.edu


Introduction
Camp SMILE (Sharing Memories In a Loving Environment) is a bereavement camp in rural Eastern Kentucky for children ages 5-17 who have experienced the loss of a significant person in their lives. The purpose of this program is to help children develop positive coping skills and behaviors for dealing with the grief associated with the loss of a significant relationship. Participants benefit from gathering with peers and sharing their experience with loss in a supportive environment.
The camp has been held for 3-5 days annually each summer since 2007. The camp which is provided at no cost to the campers is funded by grants, fund raising activities and donations. Children are given "interest" cards for their parents to mail to St. Claire Regional Hospice (SCR) to obtain an application to attend camp. Applications are reviewed by counselors prior to camp. The application asks parents to rate their children according to common reactions [1]. Trained bereavement counselors help children approach their grief through various interactive group activities during each morning of camp. Snacks and lunch are provided. In the afternoon campers can attend dance studio, various gym activities and go swimming. The intent of the camp is to help children learn about death and explore their grief with peers.
Bereavement is believed by many to be one of the most demanding experiences in life [2]. Bereavement not only affects adults but is a significant problem for children in our country. The death of a parent is the reason many children live in single parent households [3]. The manner in which a child grieves is expressed differently depending on the developmental stage of the child. If unresolved, bereavement in children and adolescents can lead to "depression, anxiety, social withdrawal, behavioral disturbances, and secondary school underachievement" [4].
A study by Kennedy et al. [5] established evidence that every bereaved child will require support after the death of a significant person in their life. The nature of support will differ according to the needs and the circumstances surrounding the loss. Adolescents learn how to grieve by watching their parents, grieving properly in front of teens and expressing feelings suggests that death is not a unmentionable subject [6].
Research proposes that support groups allow children to communicate with others who have experienced loss [2]. It is important to assess willingness to interact within the peer group when preparing bereavement interventions for children and adolescents especially in cases of sudden loss. The most important concerns for clinicians are indications of threats to a child's social, emotional or physical well-being since grief occurs over time. Group interaction is believed to provide a therapeutic milieu for learning since children have strong needs for interaction and acceptance. Interaction with a support group provides a safe environment for children to interrelate in a nonjudgmental setting. Non-bereaved children lack the ability to effectively interact with their peers who are bereaved. Through group communication children develop an awareness of caring supportive people who are still in their lives [2].
Even though programs exist for adults who have suffered the loss of a significant person in their life, few programs have been developed for grieving children and adolescents [7]. Unresolved grief in bereaved youth can lead to negative long-term outcomes (depression, withdrawal, social difficulties, and decline in academic performance). Nurses are ideally positioned to contribute to the knowledge base about the grief experience of adolescents, which can serve as a foundation for the development of programs that help adolescents develop positive coping strategies and avoid long term negative outcomes associated with bereavement. Mahon [8] identified a gap in services provided to bereaved children.
Many people believe that children should be sheltered from the death experience failing to see the child is already suffering the loss.

Method
A retrospective analysis was conducted of data collected from 2009-2011 at Camp SMILE, a bereavement program for children who have experienced a loss. The analysis was performed on 83 of 93 deidentified parent(s)/caregiver information forms which provided data related to common grief reactions the children were experiencing prior to attending Camp SMILE. Verification of appropriate data values and identification of missing data were determined by inspecting the frequency distributions of individual variables and when appropriate mean substitution was used for isolated missing data (e.g., for missing individual grief item responses). However, the data from ten participants was not included in the subsequent analyses due to more substantive missing data (e.g., > 30% of grief items).
The data collection instrument (Appendix 1) used by the camp contained 21 common grief reactions and asked parent(s)/caregivers to rate the degree that the child was experiencing each reaction (never, rarely, often or always). De-identified data related to age, relationship to the deceased and professional support were also analyzed. The analysis was focused on the association between the degree of common grief reactions related to gender, age, relationship to the deceased and professional support. The total sample size of the study was 83 youths. This included 32 boys and 51 girls ranging from 4 to 17 years of age with a mean age of 9.15 years and SD 3.05.
Verification of appropriate data values and identification of missing data were determined by inspecting the frequency distributions of individual variables and when appropriate mean substitution was used for isolated missing data (e.g., for missing individual grief item responses). However, the data from ten participants was not included in the subsequent analyses due to more substantive missing data (e.g., > 30% of grief items.

Statistical analysis plan
Prior to main analyses, internal consistency analysis was conducted using Cronbach's coefficient alpha (1951) to evaluate the acceptability of using a total grief score for main analyses that was calculated by summing the scaled responses (Never=0, Rarely=1, Often=2, Always=3) to the twenty-one grief items. As shown in Appendix 2, the internal consistency of the total grief score was sufficiently good (α = .871). Univariate analysis of variance (ANOVA) and planned and Tukey-Kramer post-hoc t-tests were conducted. The underlying assumptions for each analysis were evaluated to determine the acceptability of the analyses and no robust violations were detected. Tukey-Kramer post-hoc t-tests were chosen given differing sample sizes per group and in an attempt to control the family-wise error rate given multiple comparisons were made within each set of analyses.
Univariate analysis of variance (ANOVA) and planned and Tukey-Kramer post-hoc t-tests were conducted to test research hypotheses.
The underlying assumptions for each analysis were evaluated to determine the acceptability of the analyses and no robust violations were detected. Tukey-Kramer post-hoc t-tests were used to control the family-wise error rate given multiple comparisons were made within in set of analyses.

Findings Age and gender analyses
The relationship between age and grief reaction was initially evaluated in 83 participants using a Pearson correlation for age and total grief score. The resulting correlation (r 83 = -0.09, p = .41) was nonsignificant. The relationship of age and grief reactions was subsequently evaluated in relation to how children were placed into grief program groups based on their age. Those groups were 4-8 years old, 9-12 years old, and 13-17 years old. A one-way analysis of variance was conducted to evaluate the total grief scores based on these three age groups (children 4-8, 9-12, and 13-17 years old). The mean grief scores for the three age groups of children did not significantly differ according to the one-way ANOVA, F (2, 80) = 0.727, p = .49, η 2 = .02. Subsequent Tukey-Kramer post-hoc t-test analyses did not yield significant differences for The relationship of gender and grief reactions was subsequently evaluated independent of age as well as its combined influence with age. Independent t-test analysis indicated that boys (M = 20.21, SD = 11.68) reported significantly higher grief reactions than did girls (M = 13.27, SD = 9.38), t (81) = 2.98, p = .004, d = .66. A one-way ANOVA on gendered age groups (six groups: males: 4-8 year old, males: 9-12 years old, males: 13-17 years old, girls: 4-8 years old, girls: 9-12 years old, and girls: 13-17 years old) indicated a significant overall difference among the groups, F (5, 77) = 2.403, p = .04, η 2 = .14. As can be found in Table  1, subsequent analysis of the group means using Tukey-Kramer posthoc t-tests revealed non-significant trends (p = .08) suggestive that boys in 9-12 year old group had higher grief scores than the group of 4-8 year old girls.

Relationship of the deceased and grief reaction analyses
Mean differences in grief scores based on the child's relationship to the deceased (mother, father, sibling and relative) are shown in Table  2. A One-way ANOVA did not yield significant differences between children who lost a mother, father, sibling, or a relative, F (3, 77) = 1.480, p = .23, η 2 = .06. As evident in the Table 2, grief scores associated with the loss of the child's father appear higher relative to the other relationship categories, but were not found to statistically differ based on Tukey-Kramer post-hoc t-test analyses (ps > .05).

Relationship of Receipt of Professional Support and Grief Reaction analyses
Children who received professional support (M = 19.56, SD = 10.96) had higher overall grief scores relative to children whom did not receive professional support (M = 13.64, SD = 10.08), t (80) = 2.525, p = .014, d = .56. Because multiple sources of possible professional support were able to be indicated on the data collection form, subsequent analysis indicated that children receiving two or more sources of professional support (M = 19.21, SD = 12.69) did not have statistically higher grief scores relative to children receiving one source of professional support (M = 18.91, SD = 10.50), or no professional support, F (3, 77) = 1.480, p = .23, η 2 = .06.

Discussion
As described by Kübler-Ross [9] grief is a process consisting of of distinctive stages that mourners traverse. However, the order, degree of intensity and duration of the stages may differ. Children and adolescents may need help to progress in an effective manner through the stages of grief because of the cognitive, emotional and behavioral aspects in each stage [10].
The manifestations of symptoms of grief vary in children and adolescents and may not resemble symptoms seen in adults. Children not only differ in their response to death but also in their needs [11]. Most children do not have the ability to process all the components of death until age 10. This might lead one to presume that children of different age groups require varying degrees of grief interactions or show different reactions to treatment [12]. The association between age and grief reaction for the three age groups of children in this study did not significantly differ.
There is little research in the literature describing the function that gender plays in a child's adjustment after a loss [12]. The relationship of gender and overall grief scores indicated a significant overall difference among the groups in this study. Independent t-test analysis indicated that boys reported significantly higher grief reactions than did girls. In our culture boys are commonly conditioned to be "strong" and hide their emotions, whereas it is more acceptable for girls to express their emotions.
The results of this study indicate that children who had received professional support had higher overall grief scores than did children who did not receive professional support. Perhaps this is an indication of why the remaining parent/caregiver brought the child to the camp. If the child had already received professional support and was still experiencing grief reactions, interaction with peers who have had a similar experience may be indicated. Also, if the parent(s)/caregiver had sought professional support due to the severity of the grief reactions one would expect those children to have higher overall grief reaction scores than children who did not receive professional support depending on the length of time since the loss.
Grief scores associated with the loss of the child's father appear higher relative to the other relationship categories, but were not statistically significant. Children learn from an early age by imitating behaviors of adults. Females in our culture openly mourn for lost loved ones, whereas males tend to hide their emotions. Mothers (as the remaining parent) may grieve more openly causing the child to exhibit more grief reactions. The loss of the father may leave the mother emotionally distressed as well as stressed over financial aspect of the loss, especially if the father was the primary provider. Even if the mother contributed financial support to the family, men in our society tend to earn more money, thereby creating another form of stress if the father dies.
A limitation of the study was the fact that the grief reactions were indicated by the parent(s)/caregiver perceptions and the not the child. Strength of the study is that it provides research retrospectively on common grief reactions indicated by parent(s)/caregivers who voluntarily brought their children to the camp for bereavement support. Rosner et al. [12] found that in the majority of research on youth and family bereavement the subjects were recruited by the researchers. Studies of children and family members who request therapy are uncommon.

Implication for nursing practice
There is limited research available for nurses to establish evidence based practice in the area of children and bereavement. Few studies exist due to the protected nature of the subjects. Nurses need evidence as the foundation for interventions related to children, common grief reactions and bereavement. Nurses need to be trained in bereavement support as do teachers (educators). More resources need to be available for children suffering a loss. Common grief reactions in children need to be identified in the schools. School counselors and school nurses must be proactive in bereavement interventions for children in their care who have suffered a significant loss in their life. Bereavement support should be added to the curriculum of nursing schools.
In addition, a clearer picture of the extent to which children are involved in bereavement support should be developed. Advanced practice nurses working the community setting need to develop and refine skills in bereavement support, the recognition of common grief reactions in children, and foster the development of a therapeutic relationship with bereaved children. There is a need to transform available bereavement care evidence into practice. The care nurses provide to grieving families (including children) and children in the clinical arena is limited. Promotion of positive coping skills are needed to embrace the bereavement needs of children and rather than focusing on negative outcomes.
Evidence in the nursing literature is available relating to grief associated with the loss of a child but relatively few studies are available concerning grief reactions and bereavement support for the surviving child. The grief process for children may be different depending on their age but it is a process that they must navigate successfully for emotional healing. Children display common grief reactions (Lehmann et al., 2001) that advanced practice nurses need to be aware of in order to support the child through the course of their bereavement.
Future research is indicated in relation to this study to evaluate the overall grief scores of the children after completing the camp intervention. Walijarvi et al. [13] indicated the need for research that contributes knowledge to the grief process and assists in the appraisal of the efficacy of grief interventions. Children need to learn about death and explore their grief with peers.

Conclusion
The incidence of bereavement among children is a common occurrence in our society. In fact, the U.S. Bureau of the Census (2001) indicated that 3.4% of children experience the loss of a parent before they are 18. Countless other children experience bereavement through the loss of siblings, grandparents, other relatives, pets and even divorce. Peer support has been indicated as a positive attribute in the adjustment of children and adolescents to loss. In particular, the support of a child's peers may be vital for bereaved children when the remaining members of the child's family are also bereaved and incapable of providing support to the child [14].