Improving the efficacy of intervention for bereaved individuals : toward a process-focused psychotherapeutic perspective "

Professional psychotherapeutic intervention programs for bereaved individuals are not highly effective. We explore the range of possible reasons for this, based on arguments made in the literature and on further key considerations that have so far been neglected. Limited efficacy may be due to inclusion of bereaved individuals who do not need help; inappropriate types of intervention; inadequate quantity (too much or too little); or incorrect timing of the intervention (too soon or too late). Accordingly, we propose that effective intervention will require (1) motivated bereaved persons, (2) identifying and working on processes that cause or maintain the difficulties presented by the bereaved person, and (3) flexibility on the part of the therapist. We describe a number of processes underlying the difficulties encountered by bereaved people, which need consideration in therapy. In addition to these process-focused interventions, working on the therapeutic relationship will be a funda... Document type : Article de périodique (Journal article) Référence bibliographique Zech, Emmanuelle ; Ryckebosch, Anne-Sophie ; Delespaux, Emily. Improving the efficacy of intervention for bereaved individuals : toward a process-focused psychotherapeutic perspective. In: Psychologica Belgica, Vol. 50, no. 1-2, p. 103-124 (2010) DOI : 10.5334/pb-50-1-2-103 Psychologica Belgica 2010, 50-1&2, 103-124.


IMPROVING THE EFFICACY OF INTERVENTION FOR BEREAVED INDIVIDUALS: TOWARD A PROCESS-FOCUSED PSYCHOTHERAPEUTIC PERSPECTIVE
Emmanuelle ZECH (1) , Anne-Sophie RYCKEBOSCH-DAYEZ (1, 2) , & Emily DELESPAUX (1)  (1) Université catholique de Louvain & (2) Fonds de la Recherche Scientifique -FNRS Professional psychotherapeutic intervention programs for bereaved individuals are not highly effective.We explore the range of possible reasons for this, based on arguments made in the literature and on further key considerations that have so far been neglected.Limited efficacy may be due to inclusion of bereaved individuals who do not need help; inappropriate types of intervention; inadequate quantity (too much or too little); or incorrect timing of the intervention (too soon or too late).Accordingly, we propose that effective intervention will require (1) motivated bereaved persons, (2) identifying and working on processes that cause or maintain the difficulties presented by the bereaved person, and (3) flexibility on the part of the therapist.We describe a number of processes underlying the difficulties encountered by bereaved people, which need consideration in therapy.In addition to these process-focused interventions, working on the therapeutic relationship will be a fundamental factor in helping bereaved individuals cope with their difficult grief.
Several qualitative reviews have been published on the efficacy of grief interventions (e.g., Jacobs & Prigerson, 2000;Jordan & Neimeyer, 2003;Raphael, Minkov, & Dobson, 2001;Schut & Stroebe, 2005;Schut, Stroebe, van den Bout, & Terheggen, 2001).In addition, three recent meta-analyses have been conducted (Currier, Neimeyer, & Berman, 2008;Kato & Mann, 1999;Litterer-Allumbaugh & Hoyt, 1999).Consistently, the reviews have concluded that primary care interventions for bereaved individuals, that is, DOI: http://dx.doi.org/10.5334/pb-50-1-2-103 interventions open to all persons who become bereaved, in order to prevent the occurrence of mental health or grieving problems, are not efficacious (e.g., Schut et al., 2001).These results indicated that most bereaved individuals do not need or will not benefit from a psychological intervention beyond the natural coping processes that are already taking place.With regard to secondary care interventions -interventions directed toward bereaved individuals at risk of developing complicated grief reactions or mental health problems -, the reviews have indicated that studies have produced mixed results, mainly modest in size and rather temporary.The results have also indicated that the selection of specific participants increased the probability to find significant results.In particular, the results have revealed that, when the type of intervention was appropriate, taking into account the bereaved individual's characteristics (e.g., such as their gender), the interventions tended to be more efficacious.Finally, and most relevant to the current article, with regard to psychotherapeutic interventions for the bereaved, which are interventions for individuals selected on the basis of their mental health or grieving difficulties, the reviews have indicated that the interventions, although yielding generally positive results, and more positive than for the primary or secondary types, seemed less efficacious than could be expected.Indeed, effect sizes on grief or depressive symptoms varied from d = .052(Kato & Mann, 1999, on 11 controlled studies) or d = .13(Currier et al., 2008, on 23 studies) to d = .39(Currier et al., 2008), if participants were selected for their risk of presenting pathological grief reactions, and d = .43for shortterm effects in 35 pre-post test studies (Litterer-Allumbaugh & Hoyt, 1999).These effect sizes are pretty consistent and indicate that psychotherapeutic interventions for bereaved individuals are moderate at best.
Explaining the moderate efficacy of tertiary intervention programs for bereaved individuals who suffer complications in their grief processes How can these results be explained?Why are interventions directed to bereaved individuals with grief complications not more efficacious?Next, we examine four hypotheses that could explain this.Moderate efficacy may be due to (1) the inclusion of bereaved individuals who do not need help; (2) the use of inappropriate types of intervention; (3) an inadequate quantity (too much or too little) of intervention, or (4) an incorrect timing of intervention (too soon or too late).These hypotheses, which will be explored in more detail in the following section, are related to a main point we would like to make in this article: bereaved people need to get individualised intervention for their own specific problems, to the right extent, and at the right time.
We propose that the standardisation implied by current gold standards for testing the efficacy of therapeutic interventions makes it less likely to find significant therapeutic effects for bereaved individuals in the context of scientific evaluations of efficacy.Indeed, basic to scientific standards is the requirement of replicability of results.As a consequence, gold standards involve the use of a randomized blind design, manualized therapies with a precise and structured protocol of intervention, a fixed number of sessions, and trials in which patients typically meet the criteria for a single diagnosis (Reed, McLaughlin, & Newman, 2002;Seligman, 1995).For example, the use of protocols implies that therapists will strictly apply and behave in such a way that both validity and reliability are warranted.As a consequence, the space for flexibility and adaptation to a specific client's characteristics is left beyond consideration.We propose that this is precisely why interventions for bereaved individuals who suffer complications in their grief processes might not prove to be very efficacious.The main reason is that bereaved individuals, even when they are presenting difficulties and are looking for professional help, do not present the same symptoms, and even when they do so, the processes that underlie the symptoms (e.g., repetitive thoughts) may actually be different.These will be examined in further detail in the second section of this manuscript and specific intervention directed toward these processes will be outlined.

Inclusion of bereaved individuals who do not need or want help
A first hypothesis as to why intervention for bereaved individuals is only moderately effective is that many bereaved individuals do not need professional interventions in addition to the help or resources that they already naturally developed after bereavement (e.g., Schut & Stroebe, 2005;Schut et al., 2001).There are two things to consider in relationship to this claim.First, only a small proportion of individuals will develop chronic problems and difficulties in the range of pathological reactions that might not resolve without professional help (Forstmeier & Maercker, 2006;Prigerson, Ahmed, Silverman, Saxena, Maciejewski, Jacobs et al., 2002).The prevalence of persons who would be categorised as developing pathological or complicated grief reactions is still not well-established.Estimations have, however, yielded prevalence rates of complicated, traumatic, or prolonged grief ranging from 4.6% (Forstmeier & Maercker, 2006) to 34% (Prigerson et al., 2002).It has been suggested that about 10 to 15% of people suffer pathological grief reactions after 6 months of bereavement (Prigerson, Frank, Kasl, Reynolds, Anderson, Zubenko et al., 1995).Thus, if interventions are given to bereaved people, simply on the basis that they have suffered a loss, this would not be likely to lead to a reduction in difficulties, since it would address 85 to 90% of people who do not suffer to an extent that could be considered pathological.In addition, it is important to remember that suffering is part of normal grieving, which cannot be impacted on by intervention.The effect size would then be considerably diminished.Consequently, interventions need to be targeted toward bereaved individuals who are in need of intervention.
The second point relates to the motivation to search for help.Some bereaved individuals might not ask for professional help when they need it, while others might ask for help although they do not reach diagnostic criteria (e.g., severe functional impairment).In fact, several studies have shown that most bereaved individuals not only do not need professional help, as indicated above, but that they also do not ask for it (e.g., Lund, 2007).In sum, these results suggest that intervention studies might have included bereaved individuals who actually did not need or want help.In intervention and in efficacy studies, it is thus important to select bereaved participants who will receive and want to receive therapeutic intervention because of their grieving difficulties.

Inappropriate intervention
A second hypothesis relates to the type of intervention that pathological or complicated grievers are receiving: they might not get what they need.Until the late '90s, most psychotherapeutic programmes were based on theories that were in line with the "grief work" hypothesis (e.g., Stroebe, 1992;Stroebe & Stroebe, 1991).This hypothesis postulates that bereaved individuals have to do their grief work in order to cope with bereavement.Stroebe (1992, p. 19-20) defined grief work as "a cognitive process of confronting the reality of loss, of going over events that occurred before and at the time of the death, and of focusing on memories and working towards a detachment from the deceased".Following this, traditional interventions largely or even exclusively focused on promoting confrontation strategies and the relocation or relinquishing of the bond to the deceased (e.g., Ramsey's (1977Ramsey's ( , 1979) "flooding technique", Gauthier & Pye's (1979) "systematic desensitisation" or Lieberman's (1978) "forced mourning").More recently, following inconsistencies in research findings concerning the efficacy of grief work, Stroebe and Schut (1999) developed a theoretical model, the Dual-Process Model of Coping with Bereavement (DPM), which not only included confrontation with loss (so-called "loss-orientation") but also avoidant coping strategies as necessary mediators of health and wellbeing.Most importantly, the authors hypothesised a fundamental oscillation process between these various coping strategies.In fact, the diversity of tasks and problems that individuals encounter ("restoration-orientation") in addition to grief over the loss of their loved one would require the use of multiple cop-ing strategies.Research on the efficacy of coping strategies after stressful or major life events confirms that the use of multiple and flexible coping strategies is most efficacious (e.g., Folkman, Lazarus, Gruen, & Delongis, 1986;Suls & Fletcher, 1985).Recent studies that investigated the efficacy of oscillation strategies during bereavement have begun to provide some evidence of the importance of the DPM parameters for positive outcomes (Caserta & Lund, 2007;Richardson & Balaswamy, 2001;Wijngaards, Stroebe, Stroebe, Schut, van den Bout, van der Heijden et al., 2008).
Following the DPM, intervention that promotes grief work alone would not be efficacious for those who suffer complications in their grieving.The DPM postulates that exclusive use of either loss-, or restoration-oriented strategies would lead to pathological grief reactions such as chronic grief in the first case and absent or inhibited grief in the second.2Flexibility (oscillation) is of utmost importance to cope with bereavement: bereaved individuals receiving psychotherapeutic interventions based on the grief work hypothesis alone might actually not get what they need.If the DPM predictions are correct, chronic grievers are already focusing on confrontation strategies related to grief work and they thus focus on thoughts, images of, and bonds to the deceased while they should also focus on avoidant and/or restoration-oriented aspects.Thus, psychotherapeutic interventions should incorporate oscillation, and pathologically-bereaved individuals should accordingly be encouraged to use other strategies than the ones they are ineffectively using.A recent intervention programme developed by Shear, Frank, Houck, and Reynolds (2005) for bereaved individuals presenting complicated -i.e., chronic -grief reactions (Prigerson et al., 1995) was created to address processes postulated by the DPM, including restoration-oriented tasks.In support of the DPM, results revealed that this treatment was more effective than a control treatment (i.e., interpersonal psychotherapy) which focused essentially on loss-focused tasks (i.e., on grief, the relationship between symptoms, grief and interpersonal problems, and a more realistic assessment of the relationship to the deceased).To conclude, in line with other authors (Stroebe & Schut, 1999), we propose that bereaved individuals should not get the same intervention but an intervention that addresses the specific process that is at hand and explains their difficulty.

Inadequate quantity of intervention
A third reason why therapeutic interventions for bereaved individuals might not be that efficacious is that, with the use of structured intervention programmes, bereaved individuals typically not only get a fairly identical -----programme, but also, in principle, the same number of therapy sessions.However, as suggested above, bereaved persons are likely to have specific, individual needs.A complete intervention programme may take between 10 (de Keijser & Schut, 1991) to 16 sessions (Shear et al., 2005).It is possible that the time spent with one specific technique (e.g., for 4 sessions) is not enough to resolve a particular dysfunction, and that the "dose" or quantity offered is thus insufficient.On the other hand, it is possible that other bereaved individuals would need less than or none of what is proposed, in which case the "dose" might be too much.This might lead to either boring or overwhelming the client.
An example of this is provided by the Shear et al. ( 2005) study reported earlier.Half of the clients who left the DPM programme (called "complicated grief therapy" by the authors) did so because they "considered the treatment too difficult and/or did not believe that telling the highly painful story of the death could help them" (Shear et al., 2005(Shear et al., , p. 2606)).In addition, about the same number of patients refused part of the proposed programme which involved an imaginal exposure exercise in which they had to talk to the deceased person "because they considered it too difficult" (Shear et al., 2005(Shear et al., , p. 2606).In the other condition, more than half of those who discontinued treatment did so "because of perceived lack of effectiveness" (Shear et al., 2005(Shear et al., , p. 2606)).These findings suggest that a significant minority of participants in such programmes (15 to 22% in that study) refuse (at times) to comply either because the programme or task was too difficult for them, or because they viewed it as ineffective.Thus, the use of protocols, if they are not sufficiently tailored to the specific needs and difficulties of the bereaved individuals (e.g., also addressing their ambivalence over change), might prove to be less efficacious.

Incorrect timing of intervention
Finally, a fourth explanation of the insufficient efficacy of psychotherapeutic interventions is that the timing of the intervention might not be optimal.It is indeed possible that the intervention is provided at a wrong moment in the grieving process, and with respect to the rhythm with which the bereaved person is addressing his or her own particular concerns.It is still not clear as to whether the timing of the intervention in relationship to the duration of bereavement (i.e., the time elapsed since bereavement) is important for efficacy.First, it is possible that greater efficacy could be found once problems or difficulties have developed and are maintained over time, and thus not early after bereavement, since it is later on that these bereaved individuals will need the intervention most.In a way, this would be consistent with the results found for primary and secondary care interventions for bereaved persons, which have been shown to be less effective than psychotherapeutic interventions that are provided to bereaved individuals who suffer complicated grief reactions and which are typically given later in the grieving process than preventive interventions.Alternative reasoning would suggest that an early psychotherapeutic intervention would actually be more effective since the pathogenic processes might not yet be strongly entrenched in the bereaved person's functioning.With regard to the empirical literature, it seems that no study to date has directly tested whether the timing of the psychotherapeutic intervention made a difference in efficacy.Neimeyer and Currier's (2009) meta-analysis addressed this question by examining whether the timing of the intervention after bereavement moderated the efficacy on outcome.Their results indicated that this was not the case.This should further be confirmed by empirical investigation.For now, there is no consensus with regard to when or whether an intervention should be provided at a specific time after the death.
The literature on the efficacy of psychotherapy in general may actually provide an answer to this "when the intervention should be given" question: findings have indicated that it will be most efficacious when people are ready to engage in therapeutic sessions and motivated for change (see Lambert & Bergin, 1994;Miller & Rollnick, 2002).This suggests that the most important point might not be when after bereavement professional help is given but when the bereaved person is ready to get help.This implies that the person needs to realise that he or she is in need for help and is prepared to accept it.The person might then, in a second step, ask for help.Taking into account motivation and motives for help-seeking will include not only whether the person asks for help, whether the request comes from the person him-or herself versus any other person in the surroundings (or even if help is offered or advertised as in many intervention programmes), but also the reason why the bereaved person asks for help.

Conclusions
A short conclusion of this section could be "one suit does not fit all the bereaved".We have indeed suggested that, to be efficacious, therapeutic interventions for complications in grieving processes need to be individualised in "what is given, to whom, to what extent, and when".This implies taking into account whether a bereaved person really needs professional psychotherapeutic help, as well as his/her motivation and reasons for help seeking.Second, the therapist should provide the psychotherapeutic intervention at the right moment and with the right dose, according to what is best for the client at that moment in his or her grieving process.Third, a psychotherapeutic intervention needs to be flexible, not only with respect to timing, but also the content of the treatment programme.And finally, psychotherapeutic intervention should address the specific difficulties or dysfunctions that are presented by the bereaved individual (we expand on this below).The propositions that are made next go further toward identifying what precisely is needed to address the specificities in the complications that appear in the grieving person.
Reactions and what lies beneath: processes underlying complications in bereaved individuals' grief reactions Next, we examine difficulties or complications in grief processes that are frequently presented by bereaved individuals.Examples of reactions that occur during bereavement (e.g., Rubin, 1999;Ryckebosch-Dayez & Zech, 2010;Stroebe & Stroebe, 1987) are provided and we examine when these reactions might become problematic.More importantly for the message of the present article, we then propose several processes that might be underlying these grieving complications.In doing so, we do not join the debate relating to whether reactions presented by the bereaved individual are (or not) pathological in the psychiatric sense, that is, whether they reach the diagnostic criteria of a disorder (for a discussion about this debate, see Wagner & Maercker, this issue).In line with a person-centred approach (Elliott, Greenberg, & Lietaer, 2004;Rogers, 1957Rogers, /2007)), the psychotherapeutic point of view taken is rather to examine whether the reaction or symptom is problematic from the bereaved person's perspective, whether it is maintained over time, and whether the bereaved person is unable to deal with or get rid of it by him-or herself.This then becomes the definition of a complication or difficulty in the grieving process that will need psychotherapeutic intervention.This way to deal with "complications" in grieving implies that we do not use a category or a norm as a reference and we do not consider that some specific reactions are per se pathological.As a consequence, we propose that psychotherapeutic intervention should not focus on symptoms or reactions that are presented, but rather on the underlying process that maintains the reaction or symptom.This is the reason why we call such strategies "process-focused psychotherapeutic intervention".
The problems and difficulties that bereaved people report and the potential underlying processes that cause or maintain them are presented separately for ease and clarity of presentation.They are, however, linked to one another.The first most obvious reactions after the death of a loved one are emotional reactions (see Emotional reactions).Emotional reactions arise from the evaluation and significance of the eliciting event, in this case the death of a person (e.g., Frijda, 1986).The evaluation is based on the person's frame of reference which includes his or her goals, values, beliefs, self-and other-perceptions.The occurrence and evaluation of the event may also lead to shattering of this frame of reference (shattered beliefs and existential dilemmas) because the information included in the event comes into cognitive conflict with the person's frame of reference.This discrepancy also leads to recurring thoughts that intrude in the person's mind (see Ruminations versus intrusive thoughts).Finally, the attachment Working Models of Self and Other (e.g., Bartholomew & Horowitz, 1991) may also be challenged, since bereavement can imply the loss of an attachment figure (see Attachment: bonds to the deceased person).In the present article, we address only intrapersonal or intrapsychic processes that are initiated by bereavement and will not consider problems that are related either to physical health, or to interpersonal relationships.

Emotional reactions
Why are emotional reactions sometimes a problem?In general, emotions should not per se be considered as problematic.Indeed, emotions signal desires and needs, are initially adaptive (e.g., Greenberg, 2002;Greenberg & Paivio, 1997), and are usually self-regulated (Frijda, 1986).The problem in coping with emotional reactions may come from different sources.First, several emotions may be experienced at the same time and this makes the regulation of emotions complex, since the person may be overwhelmed or confused by various emotions (feelings of shock, depressive affect and grief, helplessness and hopelessness, anxiety, guilt and regrets, anger, anhedonia, loneliness, relief, see e.g., Glick, Weiss, & Parkes, 1974;Shuchter & Zisook, 1993).The complexity in emotional regulation may also come from the fact that these emotions may be contradictory or mixed (e.g., feeling sadness and joy at the same time), complex or subtle (e.g., envy, jealousy, nostalgia), or a first emotion may induce a secondary emotion (e.g., feeling guilty because one experienced joy or relief that the person died or feeling angry at oneself because of being unable to stop feeling sad, see Greenberg, 2002).A second problem is that people may be or become unaware of their emotional reactions.This can occur because these emotions are too painful and the person may repress, suppress, deny, over-control, or avoid them to such an extent that they become unconscious.Finally, although emotions can be fully experienced, people may be encouraged by those around them to avoid these states and this may induce conflict (e.g., bereaved persons may continue to dwell on their suffering while others admonish them to react differently).In these instances, bereaved persons can also get confused, not knowing who to listen to, themselves or others.
How can complex, repressed, or conflicting emotional reactions be regulated?Regulating emotions implies accessing the emotions and concerns underlying one's emotions, acknowledging, distinguishing, and experiencing the various emotions so as to heighten their awareness and their value, on the one hand, but also to accept them, on the other hand.Empathic reflections on the experiences of the bereaved person enable access to these feelings so that they become more conscious and distinct (Greenberg, 2002;Norcross, 2002;Watson, 2007).Acknowledgment of the bereaved person's experiences in a non-judgmental way allows emotions to be accepted or changed by the client (Greenberg, 2002;Norcross, 2002;Watson, 2007).
In addition to these general person-centred principles, we propose that different emotions should not be dealt with through the same strategies.In particular, it is important to distinguish emotions relating to fear or anxiety from other negative emotions, such as sadness, anger, or guilt.There are a number of reasons for this.First, fear is known to habituate with mere exposure while this does not seem to be the case with emotions like sadness or anger, which tend to perpetuate or even increase with repeated exposure.For example, it has been shown that the expression of anger, as when beating a cushion "to let one's anger out", does not decrease the anger but rather increases it, since it may actually lower the threshold at which people would act their anger out (Littrell, 1998).Dalgleish (2004, p. 251) also noted an essential distinction between fear and anxiety, on the one hand, and other negative emotions, on the other hand, describing this as follows: "Fear is prospective.That is, it is about something negative that might happen in the future.Repeated exposure to a traumatic memory involving fear will quickly reveal an absence of future threat, thus allowing the fear to dissipate.In contrast, emotions such as anger, shame, and guilt are retrospective.That is, they are about something negative that has already happened.Repeated exposure to a traumatic memory involving these emotions is merely likely to accentuate what was guilt-, shame-or angerinducing about the original experience".This suggests that fear-related emotions should be dealt with through exposure or confrontation strategies, while other-than-fear-related emotions should be dealt with through recognition, understanding, and acceptance strategies (see above).

Shattered beliefs and existential dilemmas
Another consequence of experiencing bereavement is related to a secondary, more subtle impact of the event, which involves the disconfirmation of expectations and models of the world and the self, the loss of meaning (Rimé, 2009), and the uncovering of existential dilemmas (Yalom, 1980(Yalom, /2008)).Indeed, the death of a dear person involves information to be processed that will come in cognitive conflict with the bereaved person's pre-existing framework, assumptions, values, or beliefs (Neimeyer, 2001;Parkes, 1972).Bereavement may shatter beliefs about justice (e.g., people get what they deserve and de-serve what they get), the predictability of events (e.g., bad things won't happen to me or my beloved ones), their controllability (e.g., bad things can be dealt with or avoided), their own vulnerability (e.g., life will not end suddenly, one can suffer), their self-worth or -esteem (e.g., I am a good or valuable person), self-confidence (e.g., being powerful), and the benevolence of other people (e.g., people are good, they won't hurt me) (Janoff-Bulman, 1992).When experiencing a significant event, fissures arise in this symbolic universe (Rimé, 2009) and people do not understand why this all happened.People facing adversity are often questioning why it happened to them: they search for the meaning of it and try to make sense of the events (Neimeyer, 2001).
In addition, the death of a significant person may induce intense existential questioning related to four existential challenges that humans face (Yalom, 1980(Yalom, /2008)).First, facing the death of someone may bring back the reality, uncontrollability, and inevitability of our own death and limited existence.Second, the loss of a significant other usually brings feelings of loneliness, utter isolation from others, and one's fundamental existential isolation.Third, bereavement may lead one to realise one's existential freedom.This implies psychological freedom to choose the directions in one's life and brings with it responsibility concerning who we were, are, and will become (i.e., responsibility in our values and actions).This means, for example, that the deceased person cannot be regarded anymore as responsible for the directions of the bereaved person's life.Finally, as already indicated above, the loss of a dear person may highlight the existential dilemma that life is per se meaningless.Humans can search and find meaning in their lives, but meaning is not intrinsically given.
As such, these fractures in the bereaved person's pre-existing framework may be non-existent for some, small for others, but intense and prolonged for even other bereaved individuals (Neimeyer, 2001).In this latter case, which represents complications in the grieving process, bereaved people may report that life has become meaningless and not worth living, bringing feelings of bitterness or fundamental injustice in life, absolute loneliness, and uncontrollability.How can this type of experience be dealt with?It should be noted that existential challenges have no ultimate answer and thus no given solution (Yalom, 1980(Yalom, /2008)).This may cause anxiety or terror.This is why people develop various behavioural or symbolic strategies to suppress or rationalise these thoughts (Rimé, 2009).Basic beliefs are part of the symbolic strategies that people develop to fight against existential anxiety and terror (see also Terror Management Theory, Pyszczynski, Greenberg, & Solomon, 1999).Fissures in these representational constructions can be dealt with by finding new behavioural or symbolic strategies that could make sense for the bereaved individual and that could restore their beliefs and values.It is important to recognise that, although people function and behave accord-ing to their own beliefs and values, they are usually not aware of them, or of their associated dilemmas.Being guided toward awareness of them and accepting responsibility renders the person powerful and able to control and make choices in his/her life (Yalom, 1980(Yalom, /2008)).The manner whereby this is reached is important: the therapist needs to respect and recognise the value and usefulness of the client's defences -the client would otherwise protect him-or herself even more (see Miller & Rollnick, 2002).Similarly, Acceptance and Commitment Therapy (ACT, Hayes, Strosahl, & Wilson, 1999) proposes to work on and access the values that people struggle for in their lives (e.g., being a good person).This will enable people to find realistic and alternative ways to direct their life, in line with their values.Thus, psychotherapeutic intervention should first access the bereaved person's values and make them more conscious, in order to be able, in a second step, to work on meaningful ways to reach these values.

Ruminations versus intrusive thoughts
"Negative emotions initiate a state of cognitive dissonance and are thus at the beginning of important cognitive efforts toward dissonance reduction" (Rimé, 2009, p. 64).One of the most frequent cognitive reactions after bereavement has to do with thoughts related to the lost person and the deceased, including the circumstances that led to the death (e.g., Lehman, Wortman, & Williams, 1987;Schoenberg, Carr, Peretz, Kutscher, & Cherico, 1975).To our knowledge, there is no consensus with regard to the definition or proposed processes that define when a repetitive thought might be considered as involving dysfunctional processes.Several propositions of dysfunctional mechanisms have been made in the literature (e.g., Watkins, 2008) but it is beyond the scope of this article to discuss these.We propose that recurring thoughts can take different forms that are underlined by different processes and lead to different emotional experiences.It is thus important to differentiate between them, since the effective intervention will accordingly be different.
A first type of thought is mental rumination (for a review, see Nolen-Hoeksema, 2001).People ruminating are constantly focusing in a passive and repetitive way on their own negative emotions or symptoms of distress and on the possible causes and consequences of these emotions and symptoms.Rumination, even extreme rumination early on after loss, is a normal part of grieving.However, in some more complicated cases, mental rumination can actually be maintained over time in a vicious circle.This type of thought has then consistently been found to be related to periods of depression and anxiety (Nolen-Hoeksema, 2001).Investigations have shown that the vicious circle maintaining rumination can be described as follows.On the one hand, the more people ruminate, the more their distress will be amplified.On the other hand, the more people are distressed and depressed, the more likely they will be to ruminate about themselves and their lives.This self-perpetuating process can be explained in terms of the following processes.First, people who ruminate often say that they do so to try to understand their emotions and solve their problems.They are thus tempted to ruminate more and more.However, mental ruminations have been found to interfere with good problem-solving (Lyubomirsky & Nolen-Hoeksema, 1993).This is so because people ruminating are thinking more negatively about themselves (e.g., they blame themselves more for their problems, are less self-confident) and about their lives (their past and the future) (e.g., Lyubomirsky & Nolen-Hoeksema, 1995;Nolen-Hoeksema, 1991).They thus become less able to find innovative solutions to their problems.Second, mental ruminations have been associated with impaired instrumental behaviours.People who ruminate are less motivated to engage and are actually engaging less in everyday activities, including social interactions, that could increase their sense of control and lift their mood (Lyubomirsky & Nolen-Hoeksema, 1993;Nolen-Hoeksema, 2001).Thus, people who ruminate are less likely to distract themselves or get social support (e.g., Nolen-Hoeksema & Davis, 1999).
For bereaved people who ruminate, one of the ways to stop these vicious circles is to engage in distraction from these ruminative thoughts, either by doing something else (e.g., going out) or thinking about something else than their grief and problems (e.g., Nolen-Hoeksema & Morrow, 1993).In this case, implementing distraction from the ruminative thoughts so as to be replaced either by neutral thoughts (giving relief from negative moods), or by positive ones -even temporarily -may bring relief, which could in turn have long-term positive consequences on the person's well-being and behaviours (Nolen-Hoeksema, 2001).
In contrast to the ruminations which are related to a prolonged and rather voluntary confrontation with bereavement-related thoughts, intrusive thoughts are thoughts that intrude in the person's mind without his or her will.They are typically involuntary and underlined by an involuntary process of access to the thoughts in working memory.They are said "to pop up" into the person's mind and bring up anxiety, horror, or helplessness feelings.Intrusive thoughts are related to anxiety and, in the most extreme cases, to anxious disorders such as post-traumatic stress disorder (Diagnostic and Statistical Manual of Mental Disorders, 4 th ed.; DSM-IV; American Psychiatric Association, 1994).The frequency and intrusive characteristics of these thoughts have been explained by a different mechanism than the one proposed above for mental ruminations.In this case, the thoughts provoke anxiety feelings that are so overwhelming that the person tries to avoid them.Because the person avoids thinking about the event, the event cannot be integrated and thoughts reappear (e.g., Ehlers & Clark, 2000;Horowitz, Bonanno, & Holen, 1993).Intrusive thoughts will thus come back until the cognitive dissonance is resolved (see above).The efficacious strategy to stop the vicious circle responsible for repetitive intrusive thoughts will be confronting them, since it is their prolonged avoidance that leads to their maintenance.To counter experiential avoidance, the therapeutic principle is thus to allow and to accept the experience of (negative) feelings and thoughts (e.g., Hayes et al., 1999).
To conclude, we propose that bereaved people may encounter two types of repetitive thoughts that are underlined by different processes and that should be dealt with by different intervention strategies.On the one hand, if bereaved people show extreme and prolonged mental rumination about their loss and negative emotions, they should be guided toward switching away from an ineffective confrontation strategy (ruminating) to an avoidant strategy (distract from or suppress their thoughts).On the contrary, if they present anxiety-related intrusive thoughts, the effective strategy would be to switch from an ineffective avoidant strategy to a confrontation strategy.In fact, although the underlying processes are different for mental ruminations and intrusive thoughts, the intervention would be in both cases to reinstate the oscillation between different sorts of coping strategies.Bringing back flexibility in one's coping would be the effective process.

Attachment: bonds to the deceased person
The loss of a significant other is an event likely to activate the attachment behavioural system (Bowlby, 1980;Hazan & Shaver, 1987), a motivational system that regulates the proximity to attachment figures, that is, persons who provide protection, support, and care.In bereavement, this system is activated precisely because of the loss of an attachment figure, a person who would normally function as a security base against distress.When does the bond to the deceased person become a problem for the bereaved individual?In theory, there are essentially two cases that are related to insecure attachment dimensions (e.g., Bartholomew & Horowitz, 1991) and that will represent problematic bonding after the death: people scoring extremely high on attachment-related anxiety could develop chronic grief reactions, while people scoring extremely high on attachment-related avoidance would rather present absent or inhibited grief reactions (Stroebe, Schut, & Stroebe, 2005).In the first case, the bond would be too strong and the bereaved would be too dependent on and cling too much to the tie to be able to recover from grief: he/she would be unable to loosen or relinquish it.In the second case, the bond would be too loose or denied and the bereaved would avoid suffering and behave as if nothing had happened.In line with the DPM, Stroebe et al. (2005) proposed that these two types of extremely non-securely attached persons should be guided to oscillate between loss-and restoration-oriented strate-gies.Indeed, people scoring extremely high on attachment-related anxiety could benefit from loosening their ties to the deceased, while people scoring extremely high on attachment-related avoidance could benefit from doing more loss-oriented tasks such as going over memories about the deceased.
In practice, absent grievers or extremely avoidantly-attached persons are less likely to show up in consultation (Lopez, Melendez, Sauer, Berger, & Wyssmann, 1998;Mallinckrodt, Gantt, & Coble, 1995) because their strategy is precisely to avoid or suppress the idea or fact that there might be a problem and because they have a negative view of others, which may make them disinclined to seek help from another person (Bartholomew & Horowitz, 1991).On the contrary, chronic grievers or extremely anxiously-attached persons might consult because of their intense suffering and need to express their emotions and grief (Stroebe et al., 2005).The question then becomes whether, to what extent, and how extremely anxiously-attached persons have to relinquish or loosen the bond related to the deceased person in order to adapt to the loss.An alternative way of looking at this question is to determine the conditions under which the bond could be maintained in a healthy way.Rando (1993) has proposed two conditions for the healthy continuation of bonds: first, the bereaved needs to fully acknowledge that the person is dead as well as the implications of the loss and second, the continued bonds must not interfere with moving forward into a new life.Building on this proposition, Field (2008) suggested that internalisation makes it possible to establish proximity to the deceased at the mental representational level, while at the same time, it is fully accepted that the person will never return in reality.
But how does this take place in concrete terms?We propose that the therapist needs to work on the significance of the bonds or behaviours for bereaved clients rather than on the behaviours themselves (e.g., not being able to touch or to give away a deceased's belongings).This therapeutic strategy was suggested following the results of a study conducted on the relationship between, on the one hand, the attitudes of 113 bereaved adult children toward the inheritance of the belongings of their deceased parent, in particular, the linking objectsobjects that are related to the deceased -and, on the other hand, the intensity of their grief reactions (Coulange, 2005).The results indicated that expressing the need to possess the deceased's belongings was related to more intense grief reactions, while having the belongings or not was unrelated to the intensity of grief reactions.This thus suggested that the actual attachment behaviours were less important for the bereaved person's well-being than the significance they had put on having or not having these linking objects.
In therapy, it is also important to note that most bereaved individuals will report ambivalence over the relinquishment or continuation of their bond to the deceased.Whether the person will actually change his or her behaviours is likely to be determined by three aspects described in Motivational Inter-viewing (Miller & Rollnick, 2002): whether he or she feels (1) it is important to change and thus wants to change, (2) it is possible to change (being able to change which is related to self-efficacy and self-esteem), and (3) ready to change, which is a question of priority (e.g., the good timing).It has been well established that forcing a change will heighten the person's defences and be counterproductive, even leading to a deterioration of the mental health of the client (Goldfried, 2007).In sum, this means that the therapeutic strategy with extremely anxiously-attached persons will be to work on the meanings related to keeping the bonds to the deceased as well as those related to loosening them.
Beyond the focus on specific processes: on the development of a therapeutic relationship We assume, following the general literature, that features to do with the therapeutic relationship will affect the outcome of bereavement intervention too.In fact, we even speculate that the quality of the therapeutic relationship should be even more important for efficacy in the case of grief therapy because the bereaved person has to deal with the loss of an affective relationship.In the general psychotherapy literature, it is now well established that empathy and warmth, that is, one aspect of positive unconditional regard, are substantially and consistently associated with positive therapeutic outcomes (for a review, see Norcross, 2002).As a consequence, the Task Force on empirically validated relationships of Division 29 in Psychotherapy of the American Psychological Association recommended, on the one hand, empathy and warmth as essential for therapeutic efficacy and, on the other hand, congruence and acceptance as being probably effective (Watson, 2007).It has indeed been demonstrated that an empathic, accepting, and congruent response facilitates the regulation of affects through several means: it (1) stimulates the consciousness of emotional responses; (2) helps clients to put their inner experiences into words and to symbolise them; (3) allows the client to modulate his or her emotional responses by using words; and (4) cultivates the client's reflexive ability by emphasising that the client's perspectives are subjective and that other ways of representing the world exist (Watson, 2007;Zech, 2008).The empathetic attitude could also lead to positive outcomes because (1) it increases therapeutic satisfaction and, as a consequence, compliance with the intervention; (2) it creates a corrective emotional experience and makes the client feel valued, respected, and understood; (3) it encourages exploration of feelings and can facilitate emotional processing and (4) it can mobilise the client's efforts toward change (Farber, 2007;Norcross, 2002).A positive unconditional attitude also (1) allows clients to become more responsible for their choices (they decide, they gain psychological freedom, they are in control, they gain self-confidence and esteem), (2) increases the possibility that the client's defences against distress and negative states will not be reinforced, and (3) the client's potential is trusted and growth can occur.By trying to maintain positive unconditional regard, the therapist stays flexible and adapts to his or her client's characteristics and state-of-mind moment for moment.This is made possible if the therapist makes the intervention his/her own, which means that the therapist should be as authentic as possible, that is, able to be self-aware of his or her own attitudes, emotions, and values, and to communicate them when it is relevant (Rogers, 1957(Rogers, /2007)).In sum, the psychotherapeutic intervention should be provided by an authentic, empathic, and warm therapist.

Conclusions
In the first part of this article, we proposed that intervention programmes developed for bereaved individuals might be less effective than they could be, partly because each person gets the same standardised intervention even if their grieving difficulties, reactions, impairments are different and occur at different time points.It was proposed that psychotherapeutic grief interventions need to be flexible and adapted to each bereaved individual according to "what is given, to whom, to what extent, and when".
In the second part of this article, we further argued that interventions for bereaved individuals who suffer from complications in their grieving processes will be most effective if the therapeutic strategy used specifically addresses the processes that underlie the development and maintenance of the problematic reactions or symptoms.The general principle of the intervention should be to make "stuck", static, or vicious circles of reactions become more flexible.Specific propositions were made with regard to pathogenic processes and intervention strategies for handling emotional reactions, shattered beliefs and existential dilemmas, mental ruminations and intrusive thoughts, and insecure attachments to the deceased person.
Finally, in the third part of this article, in line with an experiential and person-centred perspective, we have proposed that the therapist should try to endorse the therapeutic attitudes of congruence, empathy, and positive unconditional regard (Rogers, 1957(Rogers, /2007)).These attitudes are known to improve the quality of the therapeutic relationship, which is an essential explanatory factor with respect to therapeutic efficacy (e.g., Norcross, 2002).In sum, in the process-focused psychotherapeutic perspective described above, we have integrated recent cognitive-behavioural, humanistic, client-centred, and existential views in suggesting guidelines for psychotherapeutic interventions with bereaved individuals who suffer from complications in their grieving processes.
Many of these propositions still need empirical investigation and support.Putting this process model to test will require not only examining whether psychotherapeutic interventions that are adapted to the bereaved individual's pathogenic processes are more efficacious than standardised or manualized interventions for "complicated grievers", but also conducting laboratory studies examining subtle cognitive processes, such as how memories related to the death intrude in the bereaved person's mind and how these thoughts are controlled or avoided by complicated grievers in comparison to normal grievers.