School-based Interventions for Grief Counselling

Evidence-Based Practice Project

Researchable Question (using the COPES framework)

The question proposed for this evidence-based practice assignment: Are community-based interventions more effective than school-based interventions for helping youth under the age of 18 to decrease their feelings of grief and bereavement, and to help increase their coping skills? Using the PICO question format: the population is youth under the age of 18, the intervention is community-based therapy compared to school-based therapy, and the outcome is decreased feelings of grief and bereavement, as well as enhanced coping skills.

This question was selected because children under the age of 18 who have lost a loved-one are more likely to experience problems related to their mental health, are more vulnerable to substance misuse, and are at an increased risk of death by suicide (Dowdney, 2000). The impact that the early experience of loss may have across the lifespan is thus an important area of practice for clinicians, and correspondingly, a relevant topic for research. For the purposes of this discussion, school-based interventionswill describe any form of grief-counselling and support that is provided to students under the age of 18 in school, whereas community-based interventions describe any other form of bereavement counselling or support provided outside of school. 

Search Plan

The databases searched using the aforementioned COPES/PICO question include: ProQuest, PsycINFO, Social Service Abstracts and Google Scholar. The Randomized Controlled Trial MOLES provided to the class by the University of Toronto Library Services, were used for each of the databases in an effort to narrow the articles found (J. Webb, personal communication, September 18th, 2018).

Additionally, Boolean operators such as “AND” and “OR” were used to combine search terms (J. Webb, personal communication, September 18th, 2018). For example, the Boolean operator “AND” was used to combine each section of the PICO question: bereaved children “AND” therapy “AND” school-based intervention “AND” Cognitive Behavioural Therapy “AND” Emotion-Focused Therapy “AND” Family-therapy “AND” enhanced cop* skills, etc. Additionally, the Boolean operator “OR” was used to combine the following related search terms: therapy “OR” counselling “OR” family therapy “OR” support group “OR” “school-based.” Furthermore, the use of truncation operators was used within the search terms to produce all alternate endings of the keywords such as cop* for coping, cope, coped (J. Webb, personal communication, September 18th, 2018). Quotation marks were also used for search terms in order to keep certain key words together, such as “school-based interventions” and “family therapy” (J. Webb, personal communication, September 18th, 2018).

Inclusion criteria was youth under the age of 18 who experience grief or bereavement from the loss of a loved-one, and exclusion criteria was adults over 18, as well as individuals who had experienced weight-loss [seeing as the word loss often generated results relating to childhood obesity]. Adults were excluded because the COPES question was specific to the experiences of bereaved youth, and although family therapy is mentioned in the search terms, this is only so due to the lack of research present with children-alone considering their status as minors. Additionally, articles included were not limited to North America, and include randomized controlled trial studies conducted in other Nations. According to Rubin and Bellamy (2012), the use of randomized control trials is a way to limit threats to internal validity. As such, randomized control trial studies were chosen to generate greater confidence with respect to evidence-based practice for the purposes of this research question

Results of Evidence Search

The use of the search terms provided above produced limited results that were relevant to the PICO question, and little to no randomized controlled trials.Notwithstanding, a number of the articles found provided information for the literature review used to support the articles chosen for the critical appraisal. Consultation with a University of Toronto librarian resulted in the same lack of findings, and search terms from the PICO question needed to be deconstructed in order to find relevant articles (J. Webb, personal communication, October 22nd, 2018).

As such, searching “interventions for adolescents bereaved by relatives” in Google Scholar yielded one relevant article, and searching “effectiveness counselling in school” in the same database resulted in another article selected for the critical appraisal. Searching within Scholars Portal [accessed through the University of Laurier Library website] with the search line “bereavement randomised control trial adolescents” provided the third critical appraisal article, and the final one was found using the Main Article Database within the University of Toronto libraries search engine, with the search line “bereavement programs for youth.” Please see the appendix for the final copy of the search plan and key words used.

Literature Review

The experience of losing a loved-one can be devastating for many and carries profound impacts for children and adolescents (Edgar-Bailey & Kress, 2010). At this tender age, youth bereavement can bring out a variety of negative clinical symptoms and behaviours, while also causing regression in socio-emotional and physiological development (Hilliard, 2001). Overall, the results uncovered for community and school-based interventions show great potential for effective treatment. Community-based interventions generally showed clinically significant results where family was included in the intervention, and tended to offer more specialized, long-term interventions (Edgar-Bailey & Kress 2010; Harrison & Harrington, 2001; Slyter 2012), whereas school-based interventions showed greater consistency in interventions used, proving manualized treatments can improved students’ grief symptoms, increase adaptability, and while foster feelings of connectedness among peers (Herres et al., 2017; Salloum & Overstreet 2008; Tillman & Prazak, 2018). Although this literature review has been abbreviated, it is important to note that in nearly all studies reviewed, authors highlighted the need for more research regarding youth and bereavement, a finding consistent with the outcome of this critical appraisal.

Critical Appraisal

Pfeffer, Jiang, Kakuma, Hwang, & Metsch (2002). In the Pfeffer et al. (2002) study the population sampled was children aged 6-15 whose parent or sibling had committed suicide. Families were identified from a medical examiners’ lists of suicide victims from January 1996-November 1999; a study that cannot be replicated seeing as the region or nature of medical examiners were not noted in the study. 52 families participated in the study, with 75 children aged 6-15. Of these 75 children, 39 were put in the intervention group, and the other 36 in the control group. It is important to note that the children in the intervention group turned out to be significantly younger than those in the non-intervention group, and that no significant differences were found regarding the demographic variables, or the way in which their loved-one completed the suicide (Pfeffer et al., 2002). The rate of attrition in the control group was very high, at 75% (Pfeffer et al., 2002).

The measures used included the Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present State (K-SADS-IVR); the Beck Depression Inventory (BDI); the Childhood Posttraumatic Stress Reaction Index (CPTSRI); the Children’s Depression Inventory (CDI); the Revised Children’s Manifest Anxiety Scale (RCMAS); and the Social Adjustment Inventory for Children and Adolescents (SAICA) (Pfeffer et al., 2002). Significantly greater reduction of anxiety and depression was generally found for children in the intervention group compared to the control group (Pfeffer et al., 2002). Even so, it is worth mentioning that when the age gap was controlled for, statistically significant findings in the alleviation of anxiety [such as p ≤ .01 for the RCMAS-T], were no longer statistically significant when this variable was accounted for [p ≤ .07 for RCMAS-T]; this trend was consistent across other outcome measures (Pfeffer et al., 2002).

 Enhancing the quality of this study, therapists who facilitated the intervention needed to achieve a specific score on the Therapist Performance scale [≥ 4] in order to participate (Pfeffer et al., 2002). Additionally, those who administered the semi-structured interviews were blind to the treatment or control-group status of the participants in an effort to control confounds associated with participants and researchers (Rubin & Bellamy, 2012). This study is also strengthened through the use of two pilot tests. Moreover, each of the intervention sessions were taped and evaluated by another Psychologist (Pfeffer et al., 2002). Finally, the intervention manual chosen had interrater reliability (coefficient of 0.9) and was measured for internal reliability (alpha coefficient of 0.7) (Pfeffer et al., 2002). 

Sandler et al., 2003. In the Sandler et al. (2003) study, non-probability sampling was used through mail solicitation, personal presentations to agencies that have contact with bereaved children, newspaper articles, and media presentations. Inclusion criteria was that participants must be between the ages of 8 and 16 and had experienced the death of a parent within 4-30 months of beginning the study (Sandler et al., 2003). 244 children and adolescents from 156 families participated in this study with their families; 57.7% of families (135 children) were randomly assigned to the intervention group, and 42.3% of the families (109 children) to the control group. The intervention group used the Family Bereavement Program (FBP) requiring caregivers, children, and adolescents to meet in separate groups for 1-2-hour sessions. Caregiver intervention groups emphasized techniques for improving relationships between the caregiver and child, whereas child and adolescent intervention groups focused on behaviour modification (Sandler et al., 2003). Control group members were given three books and a syllabus each month, outlining important issues in the book as appropriate to their category (Sandler et al., 2003).

 Over 15 different measures were used to assess mental health, routine, stability, coping skills, discipline, and self-esteem of participants, most of which were tested for reliability (Sandler et al., 2003). Even so, the extensive use of outcome measures may have resulted in a Type 1 error where participants appear to experience more mental health issues due to the number of measures (Sandler et al., 2003). Overall, the study findings show that the FBP led to improvements in mental health, risk and protective factors immediately after the intervention, but not to the same extent at 11-month follow up. It is also worth mentioning that attrition rates for this study were very low, and that the findings from this study may be generalizable due to the diverse representation of individuals across socioeconomic status, ethnicity, causes of death, gender of the deceased, and time of death (Sandler et al., 2003).

Thurman, Luckett, Nice, Spyrelis, & Taylor, 2017. In the randomized controlled trial conducted by Thurman et al., (2017) participants were ninth-grade females between the ages of 13 to 17, who spoke the language of Sesotho, and who had reported experiencing a loss, as well as a desire to participate in a support group. Participants were from 11 different schools in towns across the Free State province of South Africa.  The schools were chosen by the provincial Department of Education, in consultation with a local non-profit organization called the Child Welfare Bloemfontein & Childline Free State (CWBFN), based on areas identified as most in need of bereavement-counselling(Thurman et al., 2017).  As such, participants [and schools] were not randomly selected for study inclusion. Even so, after being separated into two groups [one for orphans, and one for non-orphans] respondents were randomly assigned to treatment or control groups using a computer-generated numbering system, thus accounting for several confounds associated with researchers and participants (Thurman et al., 2017).

A total of 453 participants 226 were randomly assigned to the intervention and 227 to the waitlisted control group. 401 completed the baseline survey (200 from the intervention and 201 from the control). There was a high retention rate at post-test [96.5%] (Thurman et al., 2017). Pre, post, and follow-up surveys were translated to Sesotho, translated again into English, and pre-tested for reliability (Thurman et al., 2017). All surveys were administered by trained researchers from an independent company. The intervention was also pretested for cultural applicability, and cognitive interviews were conducted with participants from the pilot study in order to select and modify grief measures (Thurman et al., 2017). The intervention used was a support group where principles of cognitive behavioural therapy were combined with and South African games, stories, and cultural rituals regarding death, and facilitated by specially trained, local practitioners (Thurman et al., 2017).

Measures used include Core Bereavement Items (CBI-G) for normative grief; the Intrusive Grief Thoughts scale (IGTS) and the Inventory of Complicated Grief-Revised for Children  ICG-RC) to measure maladaptive grief; the 20-item Center for Epidemiological Studies-Depression Scale for Children (CES-DC); the 2-Way Social Support Scale (SSS-R) in order to measure understanding of available supports; and the Brief Problem Monitor-Parent Form (BPM) in order to measure behaviour (Thurman et al., 2017, p. 608).  It is also important to note that Chronbach’s alpha was also used to measure internal reliability and was rated above 80% for all outcome measures tested (Thurman et al., 2017). Intervention groups showed significant decrease in two of the grief measures (IGTS p=0·000; and ICG–RC, p=0·015) behavioural problems (BPM-P, p=0·017), and depression (CES–DC); all statistically significant findings (Thurman et al., 2017). Reduction in grief measures was noted, but not statistically significant (CBI-G) (Thurman et al., 2017).

Fox & Butler 2009. In the Fox and Butler (2009) study, participants sampled were between the ages of 11-17, attended one of the 13 National Society for the Prevention of Cruelty to Children (NSPCC) Schools Counselling Teams spread across England, Wales and Northern Ireland, and were assessed from January 2001-July 2003. Counsellors from the NSPCC personally invited youth to whom they provided support to participate in the study, lowering the validity of the study due to what would likely be a high level of selection bias (Fox & Butler, 2009). 219 students (90 males and 124 females, 5 unidentified) completed this pre-intervention questionnaire, and 104 completed the TEEN CORE post-intervention test. Only 29 students completed the questionnaire at 3-month follow-up (Fox & Butler, 2009).

The TEEN CORE questionnaire was used as the primary outcome measure, a tool designed to assess multiple levels of distress for youth aged 11-18 (Fox & Butler, 2009). It is also important to note that internal and external validity of this outcome measure was compromised seeing as at the time of publication, an updated version of this outcome measure had been developed called the YP-CORE. Even so, coefficient generated using Cronbach’s alpha test for reliability was greater that 70%, rendering the tool as reliable for the purposes of this study (Fox & Butler, 2009).

 The difference between TEEN CORE scores before and after counselling were statistically significant, where the mean score after counselling was significantly lower than before (p <0.001), with a large effect size of 1.06 meaning that students reported significantly less distress after participating receiving the intervention [individual grief-counselling] (Fox & Butler, 2009). Additionally, this study did not have a comparison or control group, meaning that it was not a randomized controlled trial; the absence of a control group made it impossible to properly account for confounds associated with time, participants, or researchers that may have led to the positive outcomes some students experienced (Rubin & Bellamy, 2012).

Strength of Research Designs/Discussion. The research design of the community-based intervention study conducted by Pfeffer et al. (2002) was moderately strong, with a clear definition of treatment, an alternating order process to assign families to the treatment or control group, an analysis of baseline measures, the use of a treatment and control group with over twenty individuals, tests for statistically significant findings, and the assessment of measurement tools for validity and reliability. Moreover, those rating the outcomes were blind. Unfortunately, though, the high attrition rate of control group members (75%) decreases the validity of the study findings. On the other hand, the community-based intervention study conducted by Sandler et al. (2003) was also moderately strong using the same criterion listed above. This study also randomly assigned participants to intervention and control groups, and produced statistically significant findings (Sandler et al., 2003). As mentioned previously though, it is important to consider that the number of outcome measures used may have contributed to the exaggeration of certain outcomes, having the potential to significantly alter the results (Sandler et al., 2003).

The research design of the school-based intervention study conducted by Thurman et al., was amongst the stronger articles with its clear definition of treatment, randomization procedures, use of a control and treatment group, measures for statistical significance, and tests for internal validity and reliability. The validity of the school-based intervention study by Fox & Butler (2009) however, is significantly compromised because teachers were able to personally select students to participate in their study, making for a sample that was less representative of the population of students receiving support from bereavement clinicians in UK middle and high schools. 

Except for the article by Thurman et al., these studies did not have an overly strong research design, meaning that one cannot confidently say that community-based interventions are more effective than school-based interventions. It is also worth mentioning that none of the studies were able to blind participants from being in the treatment or control group, seeing as all control groups involved were on a waitlist, or participating in a self-study program. Moreover, due to the nature of the research, none of the participants were randomly selected for study inclusion.

Although not all outcome measures reported from the studies in the critical appraisal were statistically significant, nor were they all appropriately randomized controlled trials as initially expected, one can say that they are clinically significant when considering that each of the studies did report findings of increased coping skills, and decreased feelings of anxiety, depression, or grief for participants. It is also important to consider that the aforementioned lack of evidence for this topic, in conjunction with its importance also contribute to the clinical significance of these findings. Nevertheless, the inconclusive results of this critical appraisal as well as the problems experienced in searching for these articles, speak to the need for more research and randomized controlled trials on the topics of school and community-based interventions for bereaved youth and adolescents.The information from this critical appraisal and the literature review indicate that school-based interventions and community-based interventions are both viable options for children to receive bereavement-counselling.

Evaluation

In order to develop the PICO question, it was first determined that the COPES effectiveness question would be used for research. This was to ensure that the question would yield meaningful, and specific results regarding the population of interest. In the beginning stages, the focus was directed at a single intervention, specifically community-based interventions however, after careful consideration, it was determined that utilizing a comparison intervention may generate a more suitable effectiveness question.

Additionally, after extensive research regarding the topic of youth who have lost a loved one, it was evident that youth experience a great deal of distress that may be best addressed using varying levels of intervention (Edgar-Bailey & Kress, 2010; Hilliard, 2001; Salloum & Overstreet 2008). Even though many different intervention methods were included in the search as well as a number of relevant outcomes (such decreased symptoms of grief and bereavement), the PICO question was unsuccessful in generating appropriate search results. As mentioned earlier, alternative searches such as Google Scholar, and Scholars Portal were used to obtain the articles analyzed in the critical appraisal section. As such, a barrier to applying evidence to working with this population is the lack of evidence for grief-counselling with youth. One might use the Beck Depression Inventory adjusted for children, seeing as depression is often accompanied by grief (Dowdney, 2000) and a few self-report measures created with the help of the client(s) to meet their respective specific needs and track the improvement of their coping skills. 

Improvements & Recommendations

After critically appraising the above articles throughout the evidence-based practice process, it was found that the use of quantitative studies eliminated evidence produced inductively through in-depth qualitative methods such as semi-structured interviews (Rubin & Bellamy, 2012). In future, this process could be improved by critically appraising qualitative studies that would provide an in-depth understanding of the population of interested (bereaved youth). It may also be helpful to include the American spelling of certain terms – a small change in spelling may lead to very different results.

Through this research, it is apparent that this is a topic requiring more randomized controlled trials, and those that have been conducted need replication. Though the critical appraisal did not lead to a confident conclusion that community-based interventions were more effective than school-based interventions, education is universal in Canada, meaning that children and youth will have more access to counselling services offered in school than in the community where barriers such as finance, transportation and safety, may arise. As such, school-based interventions may be an important area for practitioners to target their advocacy efforts, so that students will be able to access bereavement counselling and support groups in a familiar environment.

References

• Dowdney, L. (2000). Annotation: Childhood Bereavement Following Parental Death. Journal of Child Psychology and Psychiatry, 41(7), 819–830. https://doi.org/10.1111/14697610.00670
• Edgar-Bailey, M., & Kress, V. E. (2010). Resolving Child and Adolescent Traumatic Grief: Creative Techniques and Interventions. Journal of Creativity in Mental Health, 5(2), 158–176. https://doi.org/10.1080/15401383.2010.485090
• Fox, C. L., & Butler, I. (2009). Evaluating the effectiveness of a school-based counselling service in the UK. British Journal of Guidance & Counselling, 37(2), 95–106. https://doi.org/10.1080/03069880902728598
• Harrison, L., & Harrington, R. (2001). Adolescents’ bereavement experiences. Prevalence, association with depressive symptoms, and use of services. Journal of Adolescence, 24(2), 159–169. https://doi.org/10.1006/jado.2001.0379
• Herres, J., Williamson, A. A., Kobak, R., Layne, C. M., Kaplow, J. B., Saltzman, W. R., & Pynoos, R. S. (2017). Internalizing and Externalizing Symptoms Moderate Treatment Response to School-Based Trauma and Grief Component Therapy for Adolescents. School Mental Health, 9(2), 184–193. https://doi.org/10.1007/s12310-016-9204-1
• Hilliard, R. E. (2001). The Effects of Music Therapy-Based Bereavement Groups on Mood and Behavior of Grieving Children: A Pilot Study. Journal of Music Therapy, 38(4), 291 306. https://doi.org/10.1093/jmt/38.4.291
• Pfeffer, C. R., Jiang, H., Kakuma, T., Hwang, J., & Metsch, M. (2002). Group Intervention for Children Bereaved by the Suicide of a Relative. Journal of the American Academy of Child & Adolescent Psychiatry, 41(5), 505–513. https://doi.org/10.1097/00004583-200205000-00007
• Rubin, A., & Bellamy, J. (2012). Practitioners guide to using research for evidence-based practice. Hoboken, NJ: Wiley.
• Salloum, A., & Overstreet, S. (2008). Evaluation of Individual and Group Grief and Trauma Interventions for Children Post Disaster. Journal of Clinical Child & Adolescent Psychology, 37(3), 495–507. https://doi.org/10.1080/15374410802148194
• Sandler, I. N., Ayers, T. S., Wolchik, S. A., Tein, J.-Y., Kwok, O.-M., Haine, R. A., Griffin, W. A. (2003). The Family Bereavement Program: Efficacy Evaluation of a Theory-Based Prevention Program for Parentally Bereaved Children and Adolescents. Journal of Consulting and Clinical Psychology, 71(3), 587–600. https://doi.org/10.1037/0022006X.71.3.587
• Slyter, M. (2012). Creative Counseling Interventions for Grieving Adolescents. Journal of Creativity in Mental Health, 7(1), 17–34. https://doi.org/10.1080/15401383.2012.657593
• Thurman, T. R., Luckett, B. G., Nice, J., Spyrelis, A., & Taylor, T. M. (2017). Effect of a bereavement support group on female adolescents’ psychological health: a randomised controlled trial in South Africa. The Lancet Global Health, 5(6), e604–e614. https://doi.org/10.1016/S2214-109X(17)30146-8
• Tillman, K. S., & Prazak, M. (2018). Kids supporting kids: A 10-week small group curriculum for grief and loss in schools. Counselling and Psychotherapy Research, 18(4), 395–401. https://doi.org/10.1002/capr.12190

Appendix

Fox & Butler

Pfeffer et al.

Sandler et al.

Thurman et al.