Grief is a natural reaction to loss. Time really does heal, and as it passes, most bereaved people gradually come to terms with their grief.
However, a significant minority do not and remain very distressed, experiencing abnormally persistent grief over an extended period. For some, this presents as depression and/or post-traumatic stress symptoms, but others (1 in 10: Lundorff et al, 2017) experience a very particular kind of loss-related distress: persistent yearning for and preoccupation with the dead person, intense emotional pain and major problems in successfully getting on with everyday life. This syndrome of ‘Prolonged Grief Disorder’ is now included as a diagnosis in ICD-11 (with a minimum duration of 6 months post loss). Prolonged, ‘pathological’ or ‘complicated’ grief, as many of us know it, is bad both for our longer-term health and quality of life (see this 2015 elf blog by Olivia Maynard: Are changes in routine health behaviours the missing link between bereavement and poor physical and mental health?).
There have been considerably more randomised controlled trials published since previous systematic reviews of psychological interventions for grief (Currier et al, 2008; Wittouck et al, 2011). These newer trials have also used a clearer definition of prolonged grief when deciding both who to include in studies, and how to judge success of outcomes.
A new meta-analysis by Maja Johannsen and her colleagues from Aarhus, Denmark, evaluates the efficacy of psychological interventions for grief in bereaved adults. It also explores possible moderating influences (a range of population, intervention and study characteristics that might explain variation in outcomes).
A systematic literature search was conducted by two reviewers who independently searched electronic databases, reviewed and selected eligible studies, evaluated their methodological quality, carried out a meta-analysis and explored for publication bias.
- Studies of adults ≥ 18 years at time of loss
- Bereavement defined as: ‘loss of live-born person’ excluding pets, jobs, stillbirth
- Included quantitative measure of prolonged grief symptoms
- Reported data from a psychological intervention (cognitive-behavioural, emotional-supportive and or psychoeducation delivered by health professionals)
- Only randomised controlled trials (RCTs).
- Studies of bereaved children and adolescents
- Intervention with main focus on physical aspects (e.g. acupuncture, relaxation, massage)
- Complementary and alternative interventions
- Interventions aimed at professionals only
- Reported only proxy measure of grief as outcome (e.g. other types of complicated bereavement reactions such as post-traumatic stress or depression symptoms)
- Qualitative studies
- Grey literature.
- An adapted version of the JADAD quality assessment tool for RCTs was used
- Scored on 0-9 point scale with 7 items
- randomisation, (0-2 points)
- blinding, (0-2 points)
- dropouts and withdrawals, (0-1 point)
- a priori primary outcome and (0-1 point)
- power calculation, (0-1 point)
- inclusion based on high level of primary outcome (0-1 point)
- used one of 3 valid measures of complicated grief: the Inventory of Complicated Grief (ICG) (Prigerson et al, 1995) the revised version (ICG-R) (Prigerson & Jacobs, 2001) or the Prolonged Grief Instrument (PG-13) (Prigerson et al, 2009) (0-1 point)
- Effect sizes (ESs) were calculated for each study
- An overall pooled ES was calculated for the efficacy of psychological interventions on grief symptoms at post-intervention (first assessment after) and follow-up (longest time point). Hedge’s g (Hedges & Olkin, 1985) was chosen as the standard ES parameter
- ESs of 0.20, 0.50 and 0.80 were considered small, medium and large (Cohen, 1988)
- Moderator analyses were planned to explore possible sources of variation in outcomes between studies.
- 32 studies were identified, including 4,760 bereaved adults of whom 73% were women. 31 randomised controlled trials (RCTs) were included in the meta-analysis (data needed was unavailable for 1)
- 53% measured grief with the ICG-(R) or PG-13, with the others using 10 different measures
- 88% used a grief specific intervention, 59% provided individual therapy and 78% face-to-face with a mean of 10 sessions (range 1-20)
- The mean quality score was 5 (the range was 1-9)
- Statistically significant pooled effects of psychological intervention on grief symptoms were found for both:
- Post-intervention (Hedges’s g = 0.41, p > .001, K = 31)
- Follow-up (g = 0.45, p > .001, K = 18)
- The effect was smaller, but still significant, at post-intervention when adjusting for possible publication bias (g = 0.31)
- Compared with the remaining studies, larger post-intervention effect sizes were found for studies with:
- Individually delivered interventions (Beta = 49, p < .001),
- The ICG-(R)/PG-13 questionnaire as the grief instrument (Beta = 0.46, p < .001),
- Participants who were ≥6 months post-loss (Beta = 0.58, p < .001),
- Participants included based on high baseline symptom levels (Beta = 0.40, p = .002) and
- Higher study quality (Beta = 0.06, p = .013).
The authors conclude that:
Given the recent introduction of Prolonged Grief Disorder in the ICD-11, the results of the present meta-analysis are timely and of clinical relevance.
Based on our results, psychological intervention appears efficacious for alleviating grief symptoms in bereaved adults, with several study characteristics as possible moderators of the effect.
Strengths and limitations
Strengths of the review
- Timely and rigorously conducted update on previous reviews
- Comprehensive investigation of possible moderators to explain variation in outcomes.
- Heterogeneous studies (both in methods and quality). Surprisingly higher quality studies predicted larger effects. In my opinion, study quality is a major limiting factor
- Evidence of publication bias
- Limited clinical value of the results: effect sizes were small, and it isn’t known what a clinically meaningful change would be for any of the self-report grief measures used
- This review tells us little about what is likely to work in terms of a psychological intervention for grief and for who. As the authors state:
results should be interpreted in the light of the considerable diversity within the current field of psychological grief management.
Implications for practice
As someone who has both personally experienced persistent grief after the death of a parent and cared for many people with complicated grief, I recognise it as something which is frequently misunderstood and underestimated. ‘Grief is normal- you’ll get over it,’ just doesn’t work for some. Those who experience normal bereavement don’t need therapy. They need to know grief will pass with time. However, those who are experiencing complicated grief do need it.
As a professional, I found Shear’s ‘Complicated Grief Treatment’, which combines both talking about and coming to terms with the loss as well as re-engaging in every life, a useful clinical approach, and there is good evidence for it (Shear et al, 2005). However, I’d be interested to know how many psychological therapy services offer specific grief therapies (not simply referral to a self-help group)?
This review shows that waiting 6 months before offering specific grief therapy is wise. But what about when a loss is complicated by PTSD and/or has resulted in depression, neither of which were addressed in this review? Although depression can be diagnosed too early and easily (Dowrick & Frances 2013) the rigid delay before even seeing someone after a bereavement, operated by some psychological therapy services, seems unacceptable.
It’s time to pay more attention to the suffering caused by prolonged grief both in research and clinical practice.
Statement of interests
Linda Gask does not have any conflicts of interest other than being a non-executive director of a social enterprise (Six Degrees in Salford, Greater Manchester) that provides psychological therapy via Improving Access to Psychological Therapies.
Johannsen M, Damholdt MF, Zachariae R., Lundorff, M et al (2019) Psychological interventions for grief in adults: a systematic review and meta-analysis of randomized controlled trials. Journal of affective disorders. 253: 69-86
Cohen J (1988) Statistical Power Analysis for the Behavioral Sciences. Lawrence Erlbaum, Hillsdale, N.J.
Currier JM, Neimeyer RA, & Berman JS (2008) The effectiveness of psychotherapeutic interventions for bereaved persons: a comprehensive quantitative review. Psychological bulletin, 134(5), 648.
Dowrick C & Frances A (2013) Medicalising unhappiness: new classification of depression risks more patients being put on drug treatment from which they will not benefit. bmj, 347, f7140.
Hedges LV & Olkin I (2014) Statistical methods for meta-analysis. Academic Press
Lundorff M, Holmgren H, Zachariae R et al. (2017) Prevalence of prolonged grief disorder in adult bereavement: A systematic review and meta-analysis. Journal of Affective Disorders, 212, 138-149.
Prigerson HG, Horowitz MJ et al. (2009). Prolonged grief disorder: Psychometric validation of criteria proposed for DSM-V and ICD-11. PLoS medicine, 6(8), e1000121.
Prigerson HG & Jacobs SC (2001) Diagnostic criteria for traumatic grief: A rationale, consensus criteria, and preliminary empirical test. Handbook of bereavement research: Consequences, coping, and care, 613-645.
Prigerson HG, Maciejewski PK, Reynolds III CF et al. (1995) Inventory of Complicated Grief: a scale to measure maladaptive symptoms of loss. Psychiatry research, 59(1-2), 65-79.
Maynard O. (2015) Are changes in routine health behaviours the missing link between bereavement and poor physical and mental health? The Mental Elf 6 July 2015
Shear K, Frank E, Houck PR et al. (2005) Treatment of complicated grief: a randomized controlled trial. Jama, 293(21), 2601-2608.
Wittouck C, Van Autreve S, De Jaegere E et al. (2011) The prevention and treatment of complicated grief: a meta-analysis. Clinical Psychology Review, 31(1), 69-78.