Table 1.

Summary of included studies.

Study reference, location	Study design, assessment	Level of evidence (NHMRC grade)	Sample intervention/control N = … Age: M (SD) or range Sex: F/M: n/n (%/%)	Intervention, setting	Outcome measures	Main outcomes	Limitations
Cha et al. (2018) Korea	Prospective cohort study Assessment: -Baseline: 1 week after suicide -Follow-up at 5 months	III-3	N = 956 F/M: 506/450 (53%/47%) Trauma-symptom group (CROPS ≥ 19) N = 83 Age: M = 17.1 (SD 0.8) F/M: 57/26 (69%/31%) Non-trauma group (CROPS <19) N = 873 Age: M = 16.9 (SD 0.8) F/M: 449/424 (51%/49%)	A school-based crisis intervention program conducted 1 week after a peer suicide. Program included screenings, educational sessions, and further interview with psychiatric specialists for the trauma-symptom group. Setting: School	Posttraumatic stress symptoms: Child Report of Posttraumatic Symptoms (CROPS) The University of California at Los Angeles posttraumatic stress disorder (PTSD) reaction index (UCLA-PTSD-RI) Anxiety symptoms: Korean-Beck Anxiety Inventory (K-BAI) Depressive symptoms: Korean-Beck Depression Inventory-II (K-BDI-II) Complicated grief: Inventory of Complicated Grief (ICG)	Significant differences in CROPS, UCLA-PTSD-RI, K-BAI, K-BDI-II, and ICG scores between baseline and follow-up in both groups. Scores of the “trauma” group dropped more compared to the non-trauma group. At follow-up 2.9% of students were in the “trauma” group vs. 8.6% at baseline. A higher proportion of female students showed posttraumatic stress symptoms than male students.	Timing of follow-up determined by school circumstances Various psychosocial factors not examined, such as level of psychological closeness between the deceased and the students, social support, family functioning, or pre-existing psychopathology No unexposed control group
Kramer et al. (2015) Belgium, and The Netherlands	Pre-/post study Mixed methods: self-reported measures and interviews Assessment: -Baseline -Follow-up at 6 and 12 months -Interviews with selected sample after 12 months	IV	N = 270 Age: M = 42.9 (SD 12.4) F/M: 238/32 (87%/13%) Interview subgroup: n = 29 Age: M = 45.3 (SD 10.8) F/M: 26/3 (90%/10%)	Two government-funded web-based peer support forums for the bereaved by suicide. Site visitors can read and/or post messages about a specific topic. The two forums were similar in terms of layout, structure, and most of the predefined sub-forums. Setting: Online	Well-being: WHO-Five Well-being Index (WHO-5) Symptoms of depression: Center for Epidemiological Studies Depression Scale (CES-D) Complicated grief: Inventory of Traumatic Grief (ITG) Suicide risk: subscale of the MINI-International Neuropsychiatric Interview (MINI-Plus) Semi-structured interview about experiences with forum	Significant improvement in well-being and depressive symptoms (both p < 0.001). Small to medium pre-post effect sizes for well-being (6 months: d = 0.24, 12 months: d = 0.36), and small for depressive symptoms (6 months: d = 0.18, 12 months: d = 0.28). No change in grief symptoms (p = 0.08, 6 months: d = 0.05, 12 months: d = 0.12). No change in suicide risk (baseline: 20.8%. 12 months: 17.2%) Main reasons for visiting online fora: sharing with peers, finding recognition	Sample: online help-seeking, self-selected, mostly female Self-report measures subject to recollection bias High drop-out rate (43%) Dutch forum was launched 1 month before recruitment started, was not yet at its full capacity No control group
Peters et al. (2015) Australia	Retrospective study Mixed-methods: self-reported measures (online or hard copy) and interviews Assessment: shortly after intervention	IV	N = 82 Age: 75% over age 45 F/M: 75/7 (91%/9%) Interview subgroup: n = 30	The Lifekeeper Memory Quilt Project, implemented by the Suicide Prevention-Bereavement Support Services of the Salvation Army in 2008 to provide support for the bereaved by suicide and to create greater public awareness of suicide. Setting: Community-based	Participants' Evaluation of Quilt (PEQ-16): 16-item scale developed for the study to measure participant satisfaction Semi-structured interview about participants' experiences with project	High participant satisfaction (M 69.6; SD 9.1) According to 48%, 1 year after the loss was the best time for participating Approx. 92% rated the Quilt project as helpful or extremely helpful Qualitative analysis of the interviews found four themes: healing, creating opportunity for dialogue, reclaiming the real person, and raising public awareness.	Sample: mostly female, self-selected (55% response rate) People who participate in Quilt projects not necessarily representative Grief was not assessed Descriptive study No control group
Scocco et al. (2019) Italy	Pre-/post study Assessment: -Baseline: 4–6 days before intervention -Post: 4–6 days after	IV	N = 61 Age: M = 49.5 (SD 11.0) F/M: 49/12 (80%/20%)	A support program of mindfulness-based residential weekend retreats, including emotion- and grief-oriented exercises Setting: Residential, group	Mindfulness experiences: Five-Facet Mindfulness Questionnaire (FFMQ) Self-Compassion Scale (SCS) Dimensions of affect: Profile of Mood States (POMS)	Significant improvement over time in almost all dimensions of the POMS (mood states). No change in the dimensions of the SCS and FFMQ Compared with first-time participants, the multiple-participation group showed significant improvements over time on the Self-kindness subscale of the SCS and Non-judging subscale of the FFMQ	Sample: mostly female, help-seeking, self-selected participants Preferable, participants had attended self-help group/ counseling Unclear if observed effects were related to intervention or group effects Grief was not assessed No follow-up data No control group
Supiano et al. (2017) USA	Prospective, observational study Analysis of the process of individual participant change in three complicated grief therapy groups	IV	N = 21 Age: M = 53 (range 34–73) F/M: 15/6 (71%/29%)	Complicated grief group therapy (CGGT): a multimodal, manualized group psychotherapy, with 2-h sessions over 16 weeks Setting: Clinical, group	Meaning reconstruction in grief: Meaning of Loss Codebook (MLC) Grief and Meaning Reconstruction Inventory (GMRI)	Therapy facilitated resolution of complicated grief symptoms and integrated memory of the deceased The MLC codes captured most of the statements of participants, helped articulating the therapeutic process, and showed that CGGT facilitated grief. Some participants continued to experience physical distress, depression or anxiety, even with improved self-care.	Sample: small and mostly female Sample limited to people bereaved by suicide with complicated grief Findings may only be generalizable to persons seeking intensive psychotherapy No control group
Visser et al. (2014) Australia	Retrospective cross-sectional study Assessment: after intervention (unspecified)	III-3	Intervention: N = 90 Age: M = 45.7 (SD 15.8) F/M: 73/17 (82%/18%) Control: N = 360 Age: M = 40.1 (SD 13.4) F/M: 311/49 (88%/11%)	Face-to-face outreach and telephone support provided by a professional crisis response team. The service then develops a customized plan, referring clients to other community services matched to their needs. The service is provided only to people who request it. Setting: Community-based	Quality of life: EQ-5DTM ICECAP index of capability Psychological distress: Kessler Psychological Distress Scale (K6) Suicidality: Suicidal Behaviors Questionnaire-Revised (SBQ-R) Work performance: World Health Organization Health and Work Performance Questionnaire (HPQ) Health care usage questions	Standby clients scored better on levels of suicidality (p = 0.006) No significant differences on other scales or health care usage	Sample: self-selected, mostly female Low response rate of clients (23%) Significant sociodemographic differences between the two groups Grief was not assessed Observational design, no control of confounding variables such as age of bereaved, time since death, and other treatments sought by participants
Wittouck et al. (2014) Belgium	Cluster RCT Assessment: Baseline 8-months after study entrance	II	Intervention: N = 47 Age: M = 49.3 (SD 13.8) F/M: 38/9 (81%/19%) Control/No treatment: N = 36 Age: M = 47.6 (SD 12.8) F/M: 25/11 (69%/31%)	Cognitive-behavioral therapy-based psychoeducational intervention, facilitated by clinical psychologists at participants' home 2 h sessions, 4 sessions, frequency not reported Setting: Clinical, group/family	Complicated grief: Inventory of Traumatic Grief, Dutch version (ITG) Depressive symptoms: Beck Depression Inventory (BDI-II-NL) Hopelessness: Beck Hopelessness Scale (BHS) Secondary outcomes: -Grief Cognitions Questionnaire (CGQ) -Utrecht Coping List (UCL)	No significant effect on the development of complicated grief reactions, depression, and suicide risk factors Secondary outcomes: Decrease in intensity of grief, depression, passive coping style, social support seeking and behavioral expression of negative feelings in intervention group only (all p < 0.05)	Sample: small, mostly female sample, possibly subject to selection bias Findings may only be generalizable to bereaved persons at-risk of complicated grief and/or seeking psychotherapy
Zisook et al. (2018) USA	RCT Assessment: -Baseline -Monthly -At week 20	II	Total: N = 395 -Suicide bereaved (SB): n = 58 -Accident/homicide (AH): n = 74 -Natural causes (NC): n = 263 Randomized in 4 groups: medication, placebo, CGT + medication, CGT + placebo Age: SB: M = 47.2 (SD 14.1) AH: M = 51.6 (SD 14.8) NC: M = 54.6 (SD 14.2) F/M: SB: 48/10 (82%/17%) AH: 56/18 (76%/24%) NC: 204/59 (78%/22%)	Manual-based structured Complicated Grief Therapy (CGT), facilitated by social workers, psychiatrists, psychologists Antidepressant medication (citalopram) with individual follow-up CGT: 16 sessions over 20 weeks Medication: 12-week with 2–4 weekly visits until week 20 Setting: Clinical, individual	Psychiatric symptoms: Structured Clinical Interview for DSM-IV-TR Axis 1 (SCID-1) Complicated grief: Complicated Grief Clinical Global Impressions Scale-Improvement (CG-CGI-I) Inventory of Complicated Grief (ICG) Structured Clinical Interview for Complicated Grief (SCI-CG) Grief-Related Avoidance Questionnaire (GRAQ) Suicidality: Columbia Suicide Severity Rating Scale-Revisited (C-SSRS-R) Impaired relationships: Work and Social Adjustment Scale (WSAS) Cognitions: Typical Beliefs Questionnaire (TBQ)	CGT was effective in all bereaved groups regarding CG symptom severity, suicidal ideation, grief-related functional impairment, avoidance and maladaptive beliefs. Lower improvement on clinician-rated CG-CGI-I in SB vs. AH and NC groups (p < 0.5) CGT seemed acceptable treatment in all groups Low acceptability of medication-only treatment	Sample: Underpowered to examine cause of death as a moderator and other possible interactions, for example related to socio-demographic variables High dropout rate in medication only subgroup Heterogeneity within cause of death subgroups No no-treatment control group