Complicated grief.
While a clinical grief diagnosis was not included in DSM-V revisions, a diagnosis of prolonged grief disorder (PGD) is proposed for ICD-11, due to be launched in 2017 (Maercker et al., 2013). This disorder is characterised by persistent and pervasive yearning and longing for the deceased and a persistent (> 6 months) preoccupation with the deceased that exceeds cultural norms and interferes significantly with the person’s functioning (Prigerson et al., 2009). Those who qualify for such a diagnosis exhibit both increased mental and physical morbidity and impairment of quality of life (Boelen & Prigerson, 2007; Prigerson et al., 1997).

A prolonged grief reaction is one of several complications that can follow a child’s death. Because the child’s death often occurs suddenly and unexpectedly, it is common for parents to experience trauma, with posttraumatic stress reactions occurring as a result. Both avoidance and intrusive memories are cardinal symptoms of PTSD. Intrusive memories are characterised by unwelcome, intrusive images and thoughts as well as distressing dreams. Avoidance is characterised by eschewal of thoughts, places and situations reminiscent of the deceased. Increased arousal is also an integral part of the traumatic effects.

Interventions for the bereaved.
In recent years there has been an increase in grief intervention studies, so much so that there have been several so-called meta-analytic studies of these interventions. In these studies it has been concluded that grief interventions are not particularly effective for normal grief (Currier, Neimeyer, & Berman, 2008; Jordan & Neimeyer, 2003; Kato & Mann, 1999; Schut, Stroebe, van den Bout, & Terheggen, 2001; Wittouck, Van Autreve, De Jaegere, Portzky, & van Heeringen, 2011). However, interventions for complicated grief show more promise and Schut and co-workers (2001), for example, divided interventions into primary interventions (for all), secondary interventions (for those considered to be in a risk group) and tertiary interventions (for those who struggle with complicated grief). For the first group, interventions were regarded as of little value, except for those aimed at children. For the second group, the results were somewhat more positive. For those who had experienced traumatic deaths, such as the loss of children, the results were promising. For the third group, interventions were more strongly supported.

In the aforementioned meta-analytic studies (e.g., Jordan & Neimeyer, 2003), grief groups were not particularly effective, and in some cases it was suggested that they could worsen the situation. In many Norwegian municipalities, as in other parts of the Western world, various grief groups have emerged to support the bereaved. These groups vary in the types of deaths they are formed for, as well as in terms of structure, leadership and duration. In other studies, various aspects of these groups from the perspective of both group leaders and group participants are outlined (Dyregrov, Dyregrov, & Johnsen, 2013, 2013–2014). The findings for these studies have shown that those with complicated grief reactions perceive less benefit from such groups than other bereaved (Johnsen, Dyregrov, & Dyregrov, 2012). Such groups are usually not formed to accommodate the needs of those who evidence complicated grief reactions.

While randomised controlled trials (RCT) studies are carefully controlled, use random allocation to groups, adhere to treatment manuals and usually exclude clients with comorbid conditions, effectiveness studies are often more flexible regarding the use of interventions and have fewer exclusion criteria. RCTs are often considered to have little relevance for clinical practice (Westen, Novotny, & Thompson–Brenner, 2004). In the present study, we compared the pre-test and post-test means for intervention groups, following the recommendation of van Ingen, Freiheit and Vye (2009) for a so-called efficiency study. This method involves more flexible treatment delivery and fewer exclusionary criteria. It demands less costly research resources, allow studying change in a small sample, and is a convenient way of testing a new group approach to complicated grief. While randomized controlled studies obviously has a stronger methodology, the approach used here is more applicable when RCT designs are not optimal and user organizations are not comfortable with withholding intervention for a control group. However, results must be interpreted and discussed with great care as the lack of a control group allow for competing explanations for any change observed.

Based on a small previous study (Dyregrov, Mortensen, & Dyregrov, 2003) and clinical experience with self-help methods such as systematic writing, we developed a program for parents who had lost children where we used therapeutic weekend gatherings as the arena for interventions. Systematic writing followed a protocol based on the Pennebaker writing (Yule, Dyregrov, Neuner, Pennebaker, Raundalen, & van Emmerik, 2005). This writing manual has been tried out successfully with bereaved adolescents (and young adults) (Kalantari, Yule, Dyregrov, Neshatdoost, & Ahmadi, 2012). They are asked to write about their innermost feelings about their loss. The instructions change subtly to direct them first to describe their psychological reactions and latterly to what they would advise someone who had had a similar loss, and what they have learned that was most helpful.

Intervention

Purpose and target group.
The program entailed the testing of a therapeutic grief processing program developed at the Centre for Crisis Psychology. While grief groups usually emphasise conversations based on topics that participants bring to bear, these weekend meetings were based on a firmer structure with a scheduled program within a problem-oriented change perspective. The goal of the intervention was to reduce the incidence and consequences of traumatic/complicated grief among participants. The target group was parents who had lost children through different types of sudden deaths. Members of national associations for parents following (unexpected) child deaths were given the opportunity to participate. The study investigates the program’s effect on traumatic/complicated grief and general psychic distress.

Procedure.
The intervention had three steps. For step 1, 93 parents signed up for a daylong psychoeducational introductory presentation/seminar on grief led by the first author (delivered in three different locations in 2006 and 2007). The seminar lasted eight hours. There were lectures on topics such as grief over time, complicated grief, difficulties in the family and siblings’ situations. Advice was given about how to best cope with the loss, including instructions in self-help methods. There were good opportunities for questions and comments. Following this seminar, the 37 participants who scored above the cut-off levels (for levels see method section) on the Impact of Event Scale (Horowitz, Alvarez and Wilner, 1979) and/or the Prigerson Inventory of Complicated Grief (ICG) (Prigerson et al., 1995) were offered the opportunity to participate in step 2: therapeutic weekend gatherings. Twenty-one of them accepted the offer. Two groups were formed, and the gatherings were held in Norway’s two largest cities. Two psychologists from the Center for Crisis Psychology led three weekend gatherings held a month apart. The first meeting lasted for two days; the next two covered a day each. The aim was to start the processing of complicated grief and trauma reactions.

The gatherings had a fixed structure, and each of the three weekend meetings began with a ritual: a candle was lit for each child, and these were placed alongside pictures of the respective deceased children, which the parents had brought along. The sessions consisted of discussions, the presentation of themes for the weekend and instructions in self-help methods that they practiced during the session. Between weekends, homework was given for participants to complete before the next gathering.

During the first meeting, the participants introduced themselves, and the group leaders presented the rationale behind the weekends. This created a common focus for the gatherings. The participants were asked to think about and present what they felt they struggled most with in their daily lives and what they felt they needed help with during the sessions. Feedback from the participants partly formed the basis for the next meeting. In addition, issues related to complicated grief reactions were addressed.

In the first session, we also examined how the child died and the worst experiences related to this. Topics that were discussed during the first session included the course of grief, social network reactions, complicated grief and traumatic aspects of grief. During both this session and the next two weekend sessions, the participants wrote about the loss, for a total of six half-hours. Specific written instructions were handed out before each writing session, with the writing tasks adapted from a manual prepared for the Children and War Foundation (Yule et al., 2005). During the first gathering, the participants wrote about the loss and the innermost thoughts and feelings associated with it.

Several of the participants reported having difficulty with experiences directly related to their loss. At this gathering, they were taught different imagery tasks to take better control of their memories (for a description, see Dyregrov, 2010). Many of the group conversations were about intrusive memories and thoughts, guilt, anger and strong bitterness. They were given specific advice on how these could be handled, often exemplified in conversations around the topics. In addition, they were instructed in thought management techniques and relaxation exercises.

In the second gathering, we first examined how the participants had been doing recently. Topics for the second gathering included mastery of separation distress, with a focus on longing and yearning that did not decline. The participants were challenged to confront and change habits and rituals that could prolong their grief. Such rituals could include daily, almost obsessional visits to the grave or keeping the child’s room and personal effects unchanged (i.e., clothing, toys, bed, stroller, etc.). At the same time, more functional ways of remembering their children were discussed. The use of therapeutic rituals to end aspects of their grief was also described, and they were taught distraction techniques and attention control. The participants were advised to set aside a regular time for grief. The instruction was modelled on the so-called «postponed worry exercise» (Wells & Sembi, 2004). If thoughts of the child should appear outside of the allotted time for grieving, the participants were advised to record the thoughts that came, to remind themselves that these belonged in their «go near» time and to notice that the thought could fade.

The writing tasks in the second gathering emphasised writing about trauma reminders in different sensory channels and any feelings of guilt. They were also asked to organise their stories with a beginning, a middle and an end to create more structure for their narrative and to organise the timeline.

In the third gathering, the emphasis was on the way ahead, meaning-making and how to best cope with their everyday lives. The writing assignments aimed to achieve more reflection and perspective, with an emphasis on what they had learned from their loss experience and how they would advise other parents who have lost a child.

In conversations, the participants’ common reactions and coping measures were spontaneously recognised by the others or brought to the attention of the group by the group leaders. The leaders also presented specific and general advice to the participants when difficult topics were raised.

Individual psychotherapy, which comprised step 3, was offered to 11 people (7 who scored above cut-off levels at the last assessment and 4 who asked for individual help), and eight accepted the offer. Therapists with specific knowledge of grief/trauma therapy techniques were recruited to assist in this step where possible, limited by accessibility in different parts of Norway. This help was offered shortly after step 2, without unnecessary delay. Six sessions were covered by the project, and therapists were chosen based on their background from work with trauma and loss. No information on treatment length or content beyond the offered sessions is available. In Figure 1 the flow of the study participants are outlined.

FIGURE 1: Flow of study participants.

Method

Participants.
In all, 83% (77) of the participants answered questionnaires before step 1. Of these, 51 were women and 26 were men. Their mean age was 36.5 years (SD = 6.6), with the youngest participant 22 years old and the oldest 57 years old. Thirty-seven people had scores indicating complicated grief reactions and were offered a place at weekend gatherings. Of these, 21 accepted the offer of the weekend gatherings. Eighteen were women and three were men. Mean age for this group was 35.4 years (SD = 8.86) with the youngest 22 years old and the oldest 57 years old. The mean time since they lost their child was 2 years and 7 months. The cause of death varied and consisted mostly of death due to illness (65%) including heart disease, stillbirth and genetic disease, accidental deaths (25%), deaths from accidents (25%), and deaths from SIDS (Sudden Infant Death Syndrome) (10%).

The response rate among the participants who attended the weekend sessions was very high (90% for weekend 2 and 95% for weekend 3). For those who attended the weekend sessions, an average of 3 years and three months had elapsed since the death (range 6 to 84 months) when the first weekend seminar started. Around 30 percent (28.6%) had experienced the death within the previous year, nearly 40 percent (38.1%) between 1 and 3 years before and about 1/3 (33.3%) over 3 years before. Participation in the survey was based on written informed consent.

Questionnaire.
A self-made questionnaire was used to map the demographic factors and for assessment of the course content. The Impact of Event Scale (IES) (Horowitz et al., 1979) was used to obtain a measure of the incidence of traumatic reactions. It provides a general measure of the degree of traumatic reactions. Two subscales measure the degree of intrusive memories and avoidance. The scale consists of 15 questions with four possible answers scored ‘not at all’ (0), ‘rarely’ (1), ‘sometimes’ (3) and 5 (‘often’). Individuals who score above 35 (cut-off score) on the IES are considered to struggle with a high degree of traumatic reactions at a level consistent with post-traumatic stress disorder (PTSD) (Neal et al., 1994).

The Prigerson Inventory of Complicated Grief (ICG) (Prigerson et al., 1995) was used to measure the extent of complicated grief, with an emphasis on maladaptive and debilitating symptoms. The questionnaire captures complicated grief beyond normal and natural grief. The form has 19 items with five possible answers scored ‘never’ (0), ‘rarely’ (1), ‘sometimes’ (2), ‘often’ (3) or ‘always’ (4). Individuals who score at 25 or above (cut-off) on this questionnaire have been shown to be significantly affected in their social and mental functioning and in their physical health (Prigerson et al., 1995; Prigerson, Varderwerker, & Maciejewski, 2008).

The General Health Questionnaire (GHQ) (Goldberg, 1978) was used as a measure of general psychic distress and captures problems over a wider range than the previous questionnaires. The 28-item version was used with four answer categories and a Likert score (the two first answer options were scored 0 and the two remaining scored 1). Different values have been used for the determination of cut-off scores, usually 6 or more (Goldberg, 1978). In this article, we use a conservative cut-off score of 7 or more. People who score higher than the cut-off score are considered to be at risk for mental disorders.

Procedure.
The project was organised by the Norwegian SIDS and Stillbirth Society, which, in addition to its own members, also arranged for the distribution of recruitment information to another organisation for bereaved parents (Vi som har et barn for lite). Participants were invited to sign up for the grief seminar through the organisations’ membership magazines and websites.

Data collection was performed at three times: those who signed up to participate were sent a questionnaire prior to the first weekend gathering. This represents the baseline measurement. Analyses of these data formed the basis for the selection of who was invited to participate in stage 2.

Additional data collection was conducted about a month after attendance at the weekend gatherings, i.e., at the end of step 2. The reason for the delay is the observation at previous seminars that immediately after a weekend where the loss is focussed there will be days (sometimes weeks) before loss and trauma activated emotions and cognitions are processed. Questions for the participants’ evaluation of stage 1 and 2 were included. Analyses of data from the IES, ICG and GHQ were used to determine who was to be offered individual psychological help (step 3).

The third data collection took place approximately nine months after the completion of step 2 to examine the long-term effects of stage 1 and 2. The measurement points were consistent with the previously conducted study (Dyregrov, Mortensen, & Dyregrov, 2003) to allow for a comparison of the results.

Statistics.
To examine changes between the different times of measurement, we used a general linear model (GLM) with repeated measures. Only statistically significant findings (p < .05) are discussed in this study. The eta squared is a measure of strength. In addition, the effect size was calculated as the difference between mean values divided by the standard deviation at the starting point (van Ingen, Freiheit, & Vye, 2009), with sizes .2 indicating a small effect, .5 a medium effect, and .8 a large effect. The data’s reliability was assessed as good, with all Cronbach’s alpha > .88. Statistica and SPSS were used for data processing. On the main variables, ICG, IES and GHQ, data were missing for a few people. Missing data were imputed with the (expected maximisation) EM procedure. To analyse data not only on the group level but also on an individual level, latent growth modelling was used (Bollen & Curran, 2006). This allows for the analysis of non-linear changes. With such a small sample, the model fit may be unstable.

Results

All results are calculated for the participants who filled out the questionnaires at times 1, 2 and 3.

Traumatic grief.
The Inventory of Complicated Grief was previously described as a measure of the extent of complicated grief focusing on maladaptive and debilitating symptoms. Figure 2 depicts developments over time on this measure.

FIGURE 2: Development over time on the complicated grief measure (ICG) (N = 21).

The reduction is significant over the three measurement points (F(2, 40) = 13.99, p < .001), and it is significant from point 1 to 2, but not from point 2 to 3. The partial η² was .41. The effect size from T1 to T2 was 1.00 (large effect), and it was .19 from T2 to T3 (not reaching a small effect). Despite a sharp decline in the mean scores of the ICG over time, they were above 25 at all times (cut-off score), which means that, despite significant improvements, the participants still had a lot of grief over their lost children.

The observed scores indicated a nonlinear change over time. A latent growth model was therefore analysed with the initial level set to 0 and the final level set to 1 in the change factor. The model showed that 77% of the change took place in the period from T1 to T2. The model fit was good (χ² = .67, df = 2, p = .72, RMSEA = .000, RMSEAc.i. = .00 to –.32, RMSEAclose fit = .73). A statistically significant variance was found in both the baseline level and the change factor. Thus, over time the individuals differ in their changes around the mean change. We found no correlation between the baseline level and change, which indicates that the degree of change is not related to the baseline level. This harmonises well with the observed scores.

Impact of Event Scale.
The average scores on the IES (see Figure 3) showed a clear significant decline over the three measurement periods (F(2, 40) = 18.68, p < .001). The decline from point 1 to 2 and from point 2 to 3 was significant. The partial η² was .48. The effect size from T1 to T2 was .60 (medium effect), and it was .26 from T2 to T3 (small effect). While the average score prior to the start of the educational day was above the cut-off score, it was down to this level at the end of the weekend gatherings and continued to drop up to the 9-month measurement point.

FIGURE 3: Development over time in traumatic after effects (IES) (N = 21).

Not shown in a separate figure, the decline in the subscales IES­-Intrusion and IES-Avoidance was statistically significant for the intrusion at all times (p < .05), while the decline for Avoidance was only significant from time 1 to 2. Latent growth modelling showed that 62% of the change occurred from T1 to T2. Thus, more of the total change here came later compared to what was seen for the grief measure, although most of the reduction was still found from T1 to T2. The model fit the data well (χ2 = 2.35, df = 3, p = .50, RMSEA = .000, RMSEAc.i. = .00 to -.34, RMSEAclose fit = .52). There was statistically significant variance in the scores at baseline and in individual variation in change over time.

Psychic distress.
For GHQ, there was a significant decline in scores over the three time points (F(2, 40) = 12.98, p < .001) (see Figure 4). The partial η² was .39. The effect size from T1 to T2 was .61 (medium effect), and from T2 to T3, it was .24 (small effect). The decrease between point 1 and 2 was statistically significant. Before the start of the day seminar, the average score was considerably higher than the cut-off score. After the weekend gatherings, the average score dropped to just above the level of the cut-off score, while at follow-up, the average score was below the level set for the cut-off. Given the conservative level set, the level must still be considered somewhat elevated.

FIGURE 4: Development over time in General Health (GHQ) (N = 21).

For this variable, no latent growth model showed good fit.

Discussion

Complicated grief is frequent among parents who lose children and represent a long-lasting threat to parents’ health (Dijkstra, 2000; Dyregrov, 2000, 2003; Dyregrov & Dyregrov, 1999; Murphy, Johnson, Wu, Fan, & Lohan, 2003). This study aimed at evaluating the use of weekend gatherings, to reduce the negative effects of complicated grief on parents. The results of the questionnaires illustrate various aspects of the extent to which the goal of reducing the incidence and impact of complicated grief and to improve the psychosocial health of the participants was obtained.

Inventory of Complicated Grief.
Participants showed a clear decrease in the extent of complicated grief measured just after participation in stage 2 compared to baseline and a further decline nine months after the intervention. The difference is statistically significant from point 1 to 2. Although parental grief diminished, the parents were still bothered by their grief. The loss of a child is likely to be relatively strong in parents’ memory for the rest of their lives (Dyregrov & Dyregrov, 1999), and it may very well be that the cut-off score we use for ICG, which is essentially established from studies of bereaved populations other than parents who lose children, is set too low and that the level of the parents evidenced here may be “normal” after such a loss. In a survey conducted in a total population of parents who lost children through suicide, accidents and sudden infant death (Dyregrov, 2003) all groups (SIDS, accidents and suicide) evidenced ICG levels comparable to what we observed after these weekend gatherings. After the gatherings, the average score for ICG was 33.6 (all 21 responded to this point), while the average score for the SIDS group in the previous study was 32.0 (35.3 for suicides and 38.0 for accidents). As the majority of the participants in the intervention program had lost a child several years previously, it is unlikely that a spontaneous recovery as a result of time alone would explain the results. While in a previous intervention (Dyregrov, Mortensen, & Dyregrov, 2003), the decline in scores on the various measures first was observed at the nine-month measuring point, the decline in the scores in the present study was mainly accounted for at the second measurement point. This suggests that it was the group sessions, more than the individual psychological help that was offered after step 2 that led to the effect of the intervention. Data can be interpreted to mean that the intervention had a liberating effect on entrenched grief.

Impact of Event Scale.
The average scores on the IES showed a statistically significant decrease from measurement point 1 to 2. The decline is considered large, dropping from a high start level to a level significantly below the cut-off score. The mean score is 29.0, which is close to the level (27.9) of a national SIDS sample in a previous study (Dyregrov, Nordanger, & Dyregrov, 2003). The other two groups in the 2003 survey showed, respectively, 35.8 following suicide and 36.3 following accidents). Thus, it seems that the intervention had a particularly good effect on the incidence of posttraumatic reactions, which were also targets for many of the methods that were utilised during the weekend gatherings. The further decline occurring after the end of the weekend gatherings may indicate that parents continued to use methods they had learned, but it may also reflect the fact that they had received individual psychological help. However, the decline was greatest before this help was initiated.

General Health Questionnaire.
A statistically significant decrease from time 1 to 2 was found for scores on the GHQ, indicating that the general health of the participants was improved through the intervention. This change was much more pronounced than in the previous intervention (Dyregrov, Mortensen, & Dyregrov, 2003), suggesting that the program had a broader effect this time.

General remarks.
The participants in this study evidenced strong and continuing reactions following the loss of a child which, in most cases, was several years previously. Following therapeutic weekend gatherings, the effects of symptom reduction continued in the three areas measured: complicated grief reactions, posttraumatic reactions and general psychic distress. In line with van Ingen, Freiheit and Vye (2009), the effect sizes were calculated and showed a significant effect of the intervention on all three measures, with the greatest effect appearing between time 1 and 2 and the highest effect size for complicated grief (ICG). These power measurements reflect only the force of change, not how much of this change came as a direct result of the content of the intervention or other factors related to the process.

The group that participated in the program had complicated grief reactions. The weekend gatherings started, on average, almost two years following their losses, and it must therefore be assumed that the parents’ grief had become entrenched or set. This further suggests that the observed decline over the measurement points does not come as a function of time alone but as a consequence of the intervention. It is natural to assume that the active processes undertaken and implemented by the parents during and between the gatherings have made it possible for them to utilise the new strategies they had learned to achieve a change in their grief. The combination of group and individual assistance in the intervention program may have contributed to the further decline in scores following the gatherings.

The approach used included methods used in other bereavement interventions found helpful for people with complicated grief reactions (i.e., Pearlman, Wortman, Feuer, Farber, & Rand, 2014; Rosner, Pfoh, & Kotoučov, 2011; Shear, Frank, Houck, & Reynolds, 2005). These methods are usually eclectic, drawing on different approaches, although methods from cognitive behavioral approaches are dominant.

Limitations and sources of error.
It is obvious that a randomised study with a control group or control intervention would have improved this study, as it would have allowed us to see whether the changes observed were reserved for participants in the weekend gatherings or not. The bereavement organizations involved would not allow randomizing of parents to an intervention and a control group. Though this would have been highly desirable from a research point of view to know if the change was instigated by the program, it is unlikely that the observed changes are a result of a natural decline in reactions over time. The time between loss and intervention, which was almost two years, on average, challenges the assumption that the reaction declined naturally. In meetings, participants reported that their reactions had continued from their loss and onwards. Previous bereavement research has found little evidence of delayed grief (Bonanno & Field, 2001), studies have shown that complicated grief is persistent (Lannen et al., 2008), and research has further shown that changes in waiting list controls and other control groups are minimal without treatment (i.e., Nathan & Gorman, 2002; Weisz, Doss, & Hawley, 2006). All parents were beyond the six-month limit suggested for prolonged grief disorder (Prigerson et al., 2008). When the weekend gatherings ended, feedback from participants in the summaries at the end of the training program were clear and mirrored the impression that it had been very important to mitigate grief to a more “normal” level. Nevertheless, our design allows us to make no fixed conclusions about the effectiveness of the intervention. The study is also limited by its exclusive reliance on self-report inventories without a structured assessment of grief.

Conclusion

Many grief groups are conducted without any attempt at systematic evaluation or learning. This study indicates that a systematic program with weekend gatherings aimed at helping the bereaved with complicated grief can have a beneficial effect. The results show that the participants reported good effects on several levels, influencing complicated grief, posttraumatic reactions and general psychic distress. However, the study requires replication with a larger group of parents and a more rigorous methodological design.

References

Boelen, P. A., & Prigerson, H. G. (2007). The influence of symptoms of prolonged grief disorder, depression, and anxiety on quality of life among bereaved adults. European Archives of Psychiatry and Clinical Neuroscience, 257, 444–452. doi: 10.1007/s00406-007-0744-0.

Bollen K. A., & Curran P. J. (2006) Latent curve models: a structural equation perspective. Hoboken, N.J.: Wiley-Interscience.

Bonanno, G. A., & Field, N. P. (2001). Examining the delayed grief hypothesis across 5 years of bereavement. American Behavioral Scientist, 44, 798–816. doi: 10.1177/0002764201044005007.

Currier, J. M., Neymeier, R. A., & Berman, J. S. (2008). The effectiveness of psychotherapeutic interventions for bereaved persons: a comprehensive quantitative review. Psychological Bulletin, 134, 648–661. doi: 10.1037/0033-2909.134.5.648.

Dijkstra, I. (2000). Living with loss. Parents grieving for the death of their child. Utrecht: CIP-Gegevens Koninklijke Biblioteek, Den Haag.

Dyregrov, A. (1990). Parental reactions to the loss of an infant child: A review. Scandinavian Journal of Psychology, 31, 266–280. doi: 10.1111/j.1467-9450.1990.tb00839.x.

Dyregrov, A. (2010). Supporting Traumatized Children and Teenagers. London: Jessica Kingsley Publishers.

Dyregrov, A., & Dyregrov, K. (1999). Long-term impact of sudden infant death: A twelve to fifteen year follow-up. Death Studies, 23, 635–661. doi: 10.1080/074811899200812.

Dyregrov, A., Mortensen, Ø., & Dyregrov, K. (2003). Evaluering av et sorgbearbeidingsprogram for etterlatte ved brå spedbarnsdød [Evaluation of a grief program for bereaved following sudden death in infancy]. Tidsskrift for Norsk Psykologforening, 40, 839–847.

Dyregrov, K. (2003). The loss of child by suicide, SIDS, and accidents: Consequences, needs and provisions of help. Doctoral dissertation (dr. philos). HEMIL, Psykologisk fakultet. University of Bergen. ISBN 82-7669-099-8.

Dyregrov, K., Dyregrov, A., & Johnsen, I. (2013). Participants’ recommendations for the ideal grief group. A qualitative study. Omega: Journal of Death and Dying, 67(4), 363–377. doi: 10.2190/OM.67.4.b.

Dyregrov, K., Dyregrov, A., & Johnsen, I. (2013–2014). Positive and negative experiences of grief group participation. A qualitative study. Omega: Journal of Death and Dying, 68(1), 45–62. doi: 10.2190/OM.68.1.c.

Dyregrov, K., Nordanger, D., & Dyregrov, A. (2003). Predictors of psychosocial distress after suicide, SIDS and accidents. Death Studies, 27, 143–165. doi: 10.1080/07481180302892.

Espinosa, J., & Evans, W. N. (2013). Maternal bereavement: The heightened mortality of mothers after the death of a child. Economics & Human Biology, 11(3), 371–81. doi: 10.1016/j.ehb.2012.06.002.

Goldberg, D. (1978). Manual of the general health questionnaire. Windsor: NFER.

Harper M, O’Connor R. E, & O’Carroll R. C. (2011). Increased mortality in parents bereaved in the first year of their child’s life. BMJ Supportive & Palliative Care, 1, 306–309. doi: 10.1136/bmjspcare-2011-000025.

Horowitz, M. J., Wilner, N., & Alvarez, W. (1979). Impact of event scale: A measure of subjective stress. Psychosomatic Medicine, 41, 209–218. doi: 10.1097/00006842-197905000-00004.

Johnsen, I., Dyregrov, K., & Dyregrov, A. (2012). Participants with prolonged grief – how do they benefit from grief group participation? OMEGA: Journal of Death and Dying, 65(2), 87–105. doi: 10.2190/OM.65.2.a.

Jordan, J. R., & Neimeyer, R. A. (2003). Does grief counseling work? Death Studies, 27, 765–786. doi: 10.1080/713842360.

Kalantari, M., Yule, W., Dyregrov, A., Neshatdoost, H., & Ahmadi, S. J. (2012) Efficacy of Writing for Recovery on Traumatic Grief Symptoms of Afghan Refugee Bereaved Adolescents. OMEGA — Journal of Death and Dying, 65(2), 139–150. doi: 10.2190/OM.65.2.d.

Kato, P. M., & Mann, T. (1999). A synthesis of psychological interventions for the bereaved. Clinical Psychology Review, 19, 275–296. doi: 10.1016/S0272-7358(98)00064-6.

Lannen, P. K., Wolfe, J., Prigerson, H, G., Onelov, E., & Kreicbergs, U. C. (2008). Unresolved grief in a national sample of bereaved parents: Impaired mental and physical health 4 to 9 years later. Journal of Clinical Oncology, 36, 5870–5876. doi: 10.1200/JCO.2007.14.6738.

Li, J., Precht, D. H., Mortensen, P. B., & Olsen, J. (2003). Mortality in parents after death of a child in Denmark: a nationwide follow–up study. The Lancet, 361(9355), 363–367. doi: 10.1016/s0140-6736(03)12387-2.

Lyngstad, T. H. (2007). Four essays on marital dissolution. Doctoral dissertation. Faculty of Social Sciences, University of Oslo.

Maercker, A., Brewin, C. R., Bryant, R., Cloitre, M., van Ommeren, M., Jones, L. M., et al. (2013). Diagnosis and classification of disorders specifically associated with stress: proposals for ICD-11. World Psychiatry, 12(3), 198–206. doi: 10.1002/wps.20057.

Murphy, S. A. (2008). The loss of a child: Sudden death and extended illness perspectives. In M. S. Stroebe, R. O. Hansson, & H. Schut, & W. Stroebe. (2008). (Eds.), Handbook of bereavement research and practice (pp. 375–395). Washington: American Psychological Association.

Murphy, S. A., Johnson, C., Wu, L., Fan, J. J., & Lohan, J. (2003). Bereaved parents’ outcomes 4 to 60 months after their children‘s death by accident, suicide, or homicide: A comparative study demonstrating differences. Death Studies, 27, 39–61. doi: 10.1080/07481180302871.

Nathan, P. E., & Gorman, J. M. (2002). A guide to treatments that work. New York: Oxford University press.

Neal, L. A., Busuttil, W., Rollins, J., Herepath, R., Strike, P., & Turnbull, G. (1994). Convergent validity of measures of post–traumatic stress disorder in a mixed military and civilian population. Journal of Traumatic Stress, 7, 447–455. doi: 10.1002/jts.2490070310.

Pearlman, L. A., Wortman, C. B., Feuer, C. A., Farber, C. H., & Rando, T. A. (2014). Treating traumatic bereavement. A practitioner’s guide. New York: The Guilford Press.

Prigerson, H. G., Bierhals, A. J., Kasl, S. V., Reynolds, C. F., Shear, K., Day, N., et al. (1997). Traumatic grief as a risk factor for mental and physical morbidity. American Journal of Psychiatry, 154, 616–623. doi: 10.1176/ajp.154.5.616.

Prigerson, H., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K., et. al. (2009). Prolonged grief disorder: Psychometric validation of criteria proposed for DSM-V and ICD-11. PLoS Med, 6(8), e1000121. doi: 10.1371/journal.pmed.1000121.

Prigerson, H. G., Maciejewski, P. K., Reynolds, C. F., Bierhals, A. J., Newsom, J. T., Fasiczka, A., et al. (1995). Inventory of Complicated Grief: A scale to measure maladaptive symptoms of loss. Psychiatry Research, 59, 65–79. doi: 10.1016/0165-1781(95)02757-2.

Prigerson, H: G., Vanderwerker, L. C., & Maciejewski, P. K. (2008). Complicated grief as a mental disorder: Inclusion in DSM. In M. S. R. O. Stroebe, Hansson, & H. Schut, & W. Stroebe. (2008). (Eds.), Handbook of bereavement research and practice (pp. 165–186). Washington, DC: American Psychological Association.

Rosner, R., Pfoh, G., & Kotoučová, M. (2011). Treatment of complicated grief. European Journal of Psychotraumatology, 2, 1–10. doi: 10.3402/ejpt.v2i0.7995.

Rostila, M., Saarela, J., & Kawachi, I. (2012). Mortality in parents following the death of a child: a nationwide follow-up study from Sweden. The Journal of Epidemiology and Community Health, 66, 927–33. doi: 10.1136/jech-2011-200339.

Schut, H., Stroebe, M., van den Bout, J., & Terheggen, M. (2001). The efficacy of bereavement interventions: Determining who benefits. In M. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of Bereavement Research: Consequences, Coping, and Care (pp. 705–737). Washington, DC: American Psychological Association Press. doi: 10.1037/10436-029.

Shear, K., Frank, E., Houck, P. R., & Reynolds, C. F. (2005). Treatment of complicated grief. Journal of the American Medical Association, 293, 2601–2608. doi: 10.1001/jama.293.21.2601.

van Ingen, D. J., Freiheit, S. R., & Vye, C. S. (2009). From the lab to the clinic: effectiveness of cognitive-behavioral treatments for anxiety orders. Professional Psychology: Research and Practice, 40, 69–74. doi: 10.1037/a0013318.

Weisz, J. R., Doss, A., & Hawley, K. M. (2006). Evidence-based youth psychotherapies versus usual clinical care: A meta-analysis of direct comparisons. American Psychologist, 61, 671–689. doi: 10.1037/0003-066X.61.7.671.

Wells, A., & Sembi, S. (2004). Metacognitive therapy for PTSD: a core treatment manual. Cognitive and Behavioral Practice, 11, 365–377. doi: 10.1016/S1077-7229(04)80053-1.

Westen, D., Novotny, C. M., & Thompson-Brenner, H. (2004). The empirical status of empirically supported therapies: Assumptions, findings, and reporting in controlled trials. Psychological Bulletin, 130, 631–663. doi: 10.1037/0033-2909.130.4.631.

Wilcox, H. C., Mittendorfer-Rutz, E., Kjeldgård, L., Alexanderson, K., & Runeson, B. (2014). Functional impairment due to bereavement after the death of adolescent or young adult offspring in a national population study of 1,051,515 parents. Social Psychiatry and Psychiatric Epidemiology. doi: 10.1007/s00127-014-0997-7.

Wittouck, C., Van Autreve, S., De Jaegere, E., Portzky, G., & van Heeringen, K. (2011). The prevention and treatment of complicated grief: A meta-analysis. Clinical Psychology Review, 31, 69–78. doi: 10.1016/j.cpr.2010.09.005.

Yule, W., Dyregrov, A., Neuner, F., Pennebaker, J., Raundalen, M., & van Emmerik, A. (2005). Writing for recovery. A manual for structured writing after disaster and war. Bergen: Children and War Foundation.