Grief following homicidal loss

Grief following homicidal loss Mariëtte Y. van Denderen

Mariëtte Y. van Denderen

Grief following homicidal loss

Mariëtte Y. van Denderen

© Mariëtte van Denderen, 2017 The research described in this dissertation is funded by Victim Support Fund (in Dutch: Fonds Slachtofferhulp). ISBN:

978-90-367-9386-5

Cover drawing: Windvlaag, by Reinder Homan Typesetting:

Kees-Jan Glashouwer (Glashouwer Design)

Printing:

Ridderprint BV, the Netherlands

Grief following homicidal loss Proefschrift

ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen op gezag van de rector magnificus prof. dr. E. Sterken en volgens besluit van het College voor Promoties. De openbare verdediging zal plaatsvinden op donderdag 12 januari 2017 om 12.45 uur

door

Mariëtte Yvonne van Denderen geboren op 17 mei 1985 te Achtkarspelen

Promotoren

Prof. dr. A. de Keijser Prof. dr. P. A. Boelen Beoordelingscommissie

Prof. dr. R. Kleber Prof. dr. A. Pemberton Prof. dr. R. Sanderman

This dissertation is dedicated to all people who have lost a loved one to violence.

Om ver te geraken moet je beginnen met een kleine stap. En dit eerste stapje is de belangrijkste op je reis. – Krishnamurti



Contents

Preface

Je verliest je kind door moord en dan?

1

Chapter 1

General introduction

3

Chapter 2

Psychopathology Among Homicidally Bereaved Individuals:  A Systematic Review

Chapter 3

Prevalence and Correlates of Self-Rated Post-Traumatic Stress 47 Disorder and Complicated Grief in a Community-Based Sample of Homicidally Bereaved Individuals

Chapter 4

Revenge and Psychological Adjustment after Homicidal Loss

67

Chapter 5

Prolonged Grief, Posttraumatic Stress, Anger, and Revenge Phenomena Following Homicidal Loss: The Role of Negative Cognitions and Avoidance Behaviors

83

25

Chapter 6 Treating Symptoms of Complicated Grief and Posttraumatic Stress 103 Disorder in Homicidally Bereaved Individuals with Cognitive Behavioral Therapy and EMDR: a Randomized Controlled Trial Chapter 7

Summary and general discussion

133



Nederlandse samenvatting

161



Dankwoord

165



Curriculum vitae

169



Publications

171

Je verliest je kind door moord en dan? Het was zaterdagochtend 25 juni 2005 toen mijn vrouw en ik ’s ochtends vroeg een telefoontje kregen van mijn schoonmoeder. Zij had een akelig nieuwsbericht op Radio Rijnmond gehoord. Er was sprake van de vondst van een levenloos lichaam van een jonge vrouw in een huis in Barendrecht. Zij was door geweld om het leven gebracht. Bij aankomst bij het huis trof de politie twee kleine kinderen aan die ongedeerd waren en een man die zich verwond had. Omdat onze dochter in Barendrecht woonde en twee kleine kinderen had, vroeg mijn schoonmoeder mijn vrouw of alles goed was met onze dochter omdat ze zo’n raar voorgevoel had bij het radiobericht. Daarop probeerde mijn vrouw onze dochter telefonisch te bereiken en toen dat niet lukte, is zij direct met de auto naar het woonhuis gereden. Onze dochter woonde nog geen 10 minuten rijden bij ons vandaan en aangekomen bij het woonhuis zag ze de rood-witte linten van de politieafzetting. Op de vraag aan een postende politieman of er iets ergs was gebeurd met de jonge vrouw die in het huis woonde, kreeg ze een bevestigend antwoord. Mijn vrouw werd tegengehouden en mocht niet naar binnen. Politiemensen brachten haar met een auto naar ons huis waar ik in spanning zat te wachten op nieuws van mijn vrouw. Toen mijn vrouw huilend naar binnen werd geleid, was het mij onmiddellijk duidelijk dat er iets vreselijks was gebeurd. Het rare was dat ik niet direct kon huilen. Ik was murw geslagen door het afschuwelijke nieuws. Ik kon het moeilijk bevatten dat mijn dochter er niet meer was. De tranen kwamen later. Dezelfde dag werden de kinderen van onze dochter, onze kleinzoon van een half jaar en onze kleindochter van twee en een half jaar oud, na een korte crisisopvang in een pleeggezin, bij ons gebracht. Vanaf het eerste moment was het voor mijn vrouw en mij duidelijk dat de twee kinderen bij ons moesten komen. Van alles moest geregeld worden. De uitvaart van onze dochter maar ook praktische zaken voor de kinderen als babymelk, luiers, kleding, bedjes en dekentjes. Gelukkig kregen wij onmiddellijk veel hulp van familie, buren en vrienden. De eerste maanden leef je in een roes. Het verdriet, de pijn van het verlies, de zorg voor de kleine kinderen, de rechtszaken, zowel strafrechtelijk als familierechtelijk, de afspraken met de Kinderbescherming, Jeugdzorg en Pleegzorg. De kinderen voelden zich helemaal thuis bij oma en opa. Toch bleek keer op keer dat onze kleindochter veel had meegekregen van het moment van de moord op haar moeder. Hoewel het niet te bewijzen is, bestaat het sterke vermoeden dat zij dingen gehoord of gezien heeft. Lange tijd was zij bang voor messen, voor donker gekleurde mannen (haar biologische vader/dader is Antilliaans) en een rood geblokt dekbedovertrek. Zij was onafscheidelijk van oma. Waar oma ging, ging zij ook. Er was duidelijk sprake van verlatingsangst.

1

Omdat de biologische vader naast langdurige detentie ook veroordeeld was tot dwangverpleging in een tbs-kliniek vanwege persoonlijkheidsstoornissen als asperger, narcisme en theatraal gedrag, maakten wij ons zorgen of onze kleindochter daar iets van meegekregen had. Wij hebben een kinderpsycholoog benaderd om dat te onderzoeken. Tot onze opluchting bleek dat onze kleindochter echter niet deze persoonlijkheidsstoornissen had. De aanhoudende problematiek van onze kleindochter, als bedplassen en vrijwel dagelijkse nachtmerries als gevolg van haar traumatische ervaring, was voor ons aanleiding om na verloop van tijd een psychiater van Pleegzorg te consulteren. Deze psychiater verwees ons door naar een kinderpsycholoog van de GGZ. Na een intake gesprek werd voorgesteld om EMDR op onze kleindochter toe te passen. Tot onze blijdschap begonnen de klachten na een paar sessies snel te verdwijnen en tot op vandaag heeft onze kleindochter geen psychische klachten meer gehad. In 2008 hebben mijn vrouw en ik ons als lid aangemeld bij de lotgenotenorganisatie Vereniging Ouders van een Vermoord Kind (VOVK). Hoewel de gang naar de bijeenkomsten vooral in het begin zwaar was, hadden wij er veel baat bij. Het ontmoeten van lotgenoten die ook een kind door moord hadden verloren, werkte helend. Het besef dat jij niet de enige bent die zoiets vreselijks is overkomen, hielp bij het verwerken van het verdriet en omgaan met de pijn. Inmiddels zit ik alweer een aantal jaren in besturen van lotgenotenorganisaties en houd mij bezig met het organiseren van bijeenkomsten en het behartigen van belangen van nabestaanden van geweldslachtoffers bij diverse overheidsinstanties. De moord op onze dochter heeft een enorme impact op onze levens gehad. Met de kinderen, mijn vrouw en met mij gaat het nu gelukkig goed. We genieten van de kinderen die ons veel positieve energie hebben gegeven en nog steeds geven. De kinderen, inmiddels 12 en 14 jaar oud, staan positief in het leven en hebben zich sociaal prima ontwikkeld. Jan van Kleeff Vader van Simone van Kleeff, overleden 25 juni 2005

Chapter 1 General introduction

Chapter 1 “To my mind, my father is not deceased, he was murdered. If someone asks me whether my father is deceased, I always say ‘no, my father was killed.’ He is dead, but not deceased.” 1 (Ms S.) This is a citation from a daughter whose father was murdered. She becomes angry when people ask her whether her father is deceased because for her, it does not capture the essence of the death cause. Her father is not deceased, he is murdered. This dissertation is about people who lost a loved one due to first degree murder, defined as the deliberate and premeditated killing of another human being, or due to second degree murder, defined as intentional killing, but without premeditation (hereafter called homicide, to capture both types of murder). These types of homicides differs from manslaughter, which is killing without intent, such as a drunken driver who causes an accident in which a passenger dies. He is legally responsible for the death because of his drunken behavior, but it is not his intention to kill someone. Crucial in homicidal loss is that the death was deliberately (and premeditatedly) caused by another person. In the Netherlands, on average between 140 and 150 individuals are murdered each year (Centraal Bureau voor de Statistiek, 2016; Leistra, 2015). While the victim experiences the direct consequences of the violence by losing his or her life, the bereaved parents, children, siblings, partner and friends of the victim are facing the loss on a daily basis. Based on the assumption that every victim leaves on average four close bereaved individuals behind, in the Netherlands between 560 and 600 individuals become confronted with the consequences of homicidal loss each year. While this number is relatively low when compared to other (violent) death causes, the consequences in terms of psychological adjustment may potentially be quite great. In homicidal loss, bereaved individuals are not only confronted with the loss, but also with the traumatic circumstances surrounding the loss, such as waiting for the death confirmation, and an absent or violated body of the victim (Kristensen, Weisaeth, & Heir, 2012). Also, bereaved individuals have to deal with investigation by the police, the criminal justice system, media attention, and the search for the perpetrator (Amick-McMullan, Kilpatrick, Veronen, & Smith, 1989; Kaltman & Bonanno, 2003; Parkes, 1993; Riches & Dawson, 1998; Rynearson, 1994). Homicidal loss differs from traumatic experiences without the loss of a person, such as rape, and from loss without traumatic circumstances, such as loss due to illness2. Other types of losses in which bereavement and trauma are intertwined are war related loss, mass murder and school shootings. Situations in which trauma and loss coexist are a risk factor for delayed recovery (Rynearson & McCreery, 1993). Violent loss, defined as accidents, suicides and homicides (Kristensen et al., 2012), is generally associated with more mental health problems than non-violent loss, such as Complicated Grief (CG), depression, and Posttraumatic Stress-Disorder (PTSD) (Boelen, De Keijser, & Smid, 2015; Breslau et al., 1998; Burke & Neimeyer, 2013; Kristensen et al., 2012; Van Ameringen, Mancini, Patterson, & Boyle, 2008). Before The citations used in this dissertation came from participants of the study, which gave oral and written consent for usage. The distinction made here is simplified. A traumatic experience such as rape may lead to other types of losses, such as loss of innocence or loss of faith in other people. Also, natural deaths may be felt as traumatic by the bereaved individual.

1

2

4

General Introduction

presenting a schematic model of variables central to this dissertation, the following sections will first focus on two types of mental health problems, namely CG and PTSD. Only these two problems are elaborated about because these are the two main outcome measures in the studies included in this dissertation.

1. Psychopathology 1.1 Complicated Grief Most individuals will experience the loss of a loved one once or multiple times during his or her life. Adults, children, or mentally impaired people are all basically able to grief. The majority of bereaved individuals, about 80-90%, are able to deal with the loss, with or without social support from family members and friends. They have the capacity to trust others, engage in social activities, maintain a sense that life has meaning for them and also focus on other things than their loss (Prigerson, 2004). Bonanno refers to these bereaved individuals as resilient and being able to maintain a relatively stable and healthy level of psychological and physical functioning after a potentially disruptive event, such as the death of a loved one or a violent or lifethreatening situation (Bonanno, 2004). For a small percentage of people, about 10% of the bereaved population, there seems to be no natural limitation to grief (Shear et al., 2011; Zisook & Shear, 2009). These people show ongoing intense yearning and searching for the deceased, report constant disbelief about the death, and cannot accept the loss (Prigerson et al., 1995). They might experience intrusive thoughts or recurrent images related to the death, and want to avoid painful reminders of the loss (Zisook & Shear, 2009). Alternatively, they might become overly occupied with the deceased, for example by nourishing personal objects, or over-involved in activities which remind them about the deceased, such as daily visits to the cemetery (Smid et al., 2015; Zisook & Shear, 2009). When these symptoms are accompanied with impairment in work, health and social functioning, they are referred to as complicated grief (CG) (Currier, Holland, & Neimeyer, 2006; Kersting, Brähler, Glaesmer, & Wagner, 2011; Zisook et al., 2010). CG is also termed prolonged grief disorder (Boelen, Van de Schoot, Van den Hout, De Keijser, & Van den Bout, 2010), complicated grief disorder (Maercker, & Znoj, 2010), pathological grief (Jacobs, 1993), traumatic grief (Jacobs, Mazure, & Prigerson, 2000), and Persistent Complex Bereavement Disorder (PCBD, American Psychiatric Association, 2013). The difference between uncomplicated grief and complicated grief lies not so much in different symptomatology, but in impairment in daily functioning and in the duration of symptoms, namely longer than would be expected according to social norms (Shear, 2015). There is much debate about the duration of symptoms before one can speak of CG, and the inclusion of CG in the Diagnostic and Statistical Manual of Mental Disorders (DSM) has been a topic of a wide discussion. In the DSM-5, where CG is termed Persistent Complex Bereavement Disorder (PCBD), the condition is included in section 3 of the manual, as a condition which needs further study (APA, 2013). In the mid-1990s, a panel of experts in the

5

Chapter 1 field of trauma and bereavement was formed to discuss the criteria for CG. The panel proposed that the presence of certain symptoms of grief for at least six months was appropriate to distinguish complicated from non-complicated grief (Boelen, 2005). Later, they proposed that two month was sufficient (Boelen, 2005). Now, the duration of proposed criteria for PCBD in the DMS-5 of 12 months has been established (APA, 2013). As can be seen, the terminology, as well as the duration of symptoms, is still under debate. In the studies included in this dissertation, the Inventory of Complicated Grief (Prigerson et al., 1995) is used to assess CG symptoms, although it is noticed that the symptoms for CG and of PCBD are not totally identical. Prevalence rates of CG vary greatly according to the death cause and time since loss, such as 10-20% following natural loss (Prigerson, 2004), 30% following violence during military operations (2.5 years post-loss) (Ginzburg, Geron, & Solomon, 2002), 38% following war-related violence in Kosovo (seven years post-loss) (Morina, Rudari, Bleichhardt, & Prigerson, 2010), and 12% following the Tsunami in SouthEast Asia (six years post-loss) (Kristensen, Weisaeth, Hussain, & Heir, 2015). There are various risk factors for CG, such as female gender, age of the deceased, previous loss experiences, anticipation of the death, pre-existing mental health difficulties, close kinship, a history of difficult early relationships, alcohol or drug abuse, and lack of social support (Shear, 2015; Zisook & Shear, 2009, see for a review Kristensen et al., 2012 and Lobb et al., 2010). CG is more likely after a sudden death by violent means than after non-violent loss (Currier et al, 2006; Shear, 2015; Simon, 2013; Parkes, 1993). It is therefore particular that CG following homicidal loss is subject of few studies, and empirical, well-designed studies examining CG following homicidal loss are currently lacking (Rynearson, Schut, & Stroebe, 2013). Only one study with prevalence rates of CG following homicidal loss was found (55%) (McDevitt-Murphy, Neimeyer, Burke, Williams, & Lawson, 2012). As further elaborated in Chapter 2, generalizability of the findings was limited because of the overrepresentation of women, ethnicity of the participants and limited time since loss. Because prevalence rates following homicidal loss are largely unknown and there is to date no systematic review of studies about CG and PTSD following homicidal loss, a systematic review was conducted about psychopathology following homicidal loss (Chapter 2). Further, prevalence rates of CG and PTSD in a large sample of homicidally bereaved individuals were examined to gain more insight in the prevalence of psychopathology following homicidal loss (Chapter 3). 1.2 PTSD Another response possibly elicited by homicidal loss is posttraumatic stress reaction (Burke, Neimeyer, & McDevitt-Murphy, 2010; Freedy, Resnick, Kilpatrick, Dansky, & Tidwell, 1994; Murphy, Johnson, Wu, Fan, & Lohan, 2003). Symptoms of Posttraumatic Stress-Disorder (PTSD) may occur after a traumatic event, such as rape, an accident and violence, and include four diagnostic clusters, namely intrusions, avoidance, negative alterations in cognitions and mood, and alternations in arousal and reactivity (APA, 2013). The estimated lifetime prevalence of PTSD in the general population of the US is 7.8% (Kessler, Sonnega, Bromet, Hughes, & Nelson, 1995).

6

General Introduction

Prevalence rates of PTSD following a traumatic event vary widely according to the type of event, such as 20-38% following a natural disaster, 7-72% following terrorism, 4-39% following accidents, 14-28% following injury/disease, and 4-45% following military combat (See for a review Utzon-Frank et al., 2014). In order to experience symptoms of PTSD, the traumatic event does not have to be experienced by the individual himself; also witnessing or hearing about the traumatic event, such as the violent death of a family member or friend, is also classified as potentially traumatic (APA, 2013). In the DSM-III-R and DSM-IV, it was already included that PTSD may follow an event experienced by others that was heard of, thus indirectly instead of directly (APA, 1987; Cougle, Kilpatrick, & Resnick, 2012). In general (thus not restricted to homicidal loss), it has been found that the probability of PTSD by means of direct exposure is higher than by means of indirect exposure (11.1 % vs. 7.3%, Breslau & Kessler, 2001; Cougle et al., 2012). However, the death of a family member due to homicidal loss was found to be associated with risk of PTSD and levels of distress similar to the prevalence of PTSD following direct exposure to an assault or other traumatic event (Cougle et al., 2012). It is with the inclusion of the ‘indirect exposure’ qualifier that homicidal loss which is indirectly learned about was defined as possibly leading to PTSD symptoms (Cougle et al., 2012). Rynearson and McCreery (1993) described for example homicidally bereaved individuals who experienced disturbing intrusive images of the death scene without actually witnessing the homicide. While the relation between homicidal loss and PTSD symptoms has been found in some studies, the amount of studies examining this relation is scarce. The few empirical studies which have been conducted are firm in their conclusions about the traumatic impact of homicide: “there is ample evidence that loss due to homicide is associated with post-traumatic stress phenomena” (Freeman, Shaffer, & Smith, 1996; p. 337). As it is further elaborated in Chapter 2, the firmness of these conclusions seems somewhat premature, given the fact that PTSD prevalence following homicidal loss differs substantially between studies, and studies differ in methodology, time since loss and sample types (Amick-McMullan, Kilpatrick, & Resnick, 1991; Freedy et al., 1994; McDevitt-Murphy et al., 2012; Zinzow, Rheingold, Byczkiewicz, Sauders, & Kilpatrick, 2011). Also, a number of studies combines homicide, suicide and accidents (Murphy et al., 1999; Murphy et al., 2003) and the generalizability of the prevalence rates is limited due to differences between studies (as reported in Chapter 2). Because of these reasons, the prevalence rate of PTSD in a large sample of homicidally bereaved individuals is reported in Chapter 3.

2. How can the difference in psychopathology following violent and non-violent loss be explained? Homicide is a particular type of violent loss. As noted before, research has shown that violent loss leads to more severe emotional distress in bereaved individuals than non-violent loss (Boelen et al., 2015; Boelen, 2015; Kristensen et al., 2012; Pearlman,

7

Chapter 1 Wortman, Feuer, Farber, & Rando, 2014). Although intuitively this may not seem surprising, it may be questioned how this difference can be explained. Violent loss, such as accidents, suicides or homicides, but also disaster and war-related deaths, can be characterized by their violent nature and, most often, by their suddenness and unexpectedness (Kristensen et al., 2012). Studies have indicated that the violent nature of the loss (Kaltman & Bonanno, 2003), and the suddenness of the loss (Boelen, 2015; Valdimarsdottir, Helgason, Fürst, Adolfsson, & Steineck, 2004) is predictive of PTSD symptoms. Following disasters and war-related losses, bereaved individuals may have to wait a significant period of time until the death can be confirmed. This may also be true for relatives of people who are missing and in some cases for homicidally bereaved individuals. A lack of official confirmation of the death may lead to uncertainty about the death, fantasies about what has happened, and denial of the death or a delay of a grieving process (Kristensen et al., 2012). However, research has yet to establish whether lack of confirmation of the death is associated with elevated levels of bereavement-related psychopathology. Another stressor after violent loss may be the interplay between grief and posttraumatic stress reactions. Following suicide, accidents or homicide, bereaved individuals may re-experience finding the body, envisage pictures of the crime scene, or of the violated body of the victim. As described previously in the paragraph about PTSD, such symptoms may also be present if the bereaved individual learned about the loss indirectly (Kristensen et al., 2012). Another explanation for the difference in psychopathology between violent and non-violent loss has been found in making sense of the death. Violent loss is more difficult to make sense of than natural loss, because violent loss may undermine and change fundamental beliefs and assumptions people have about themselves, others and the world around them (Currier et al., 2006; Janoff-Bulman, 1992). This is also known as ‘Theory of shattered assumptions’(Janoff-Bulman, 1992). Rynearson (1988) refers to this by saying: “the manner of dying determines the meaning of death” (p. 214). The inability to find a reasonable sense of understanding of the loss over time is frequently accompanied by grief complications (Currier, et al., 2006; Janoff-Bulman, 1989; Mancini, Prati, & Bonnano, 2011), greater distress (Davis, Wortman, Lehman & Silver, 2000), and traumatic symptoms (Janoff-Bulman, 1992). Homicidal loss may be difficult to make sense of because of its intentional and violent nature, which may leave bereaved individuals with a sense of injustice and assumptions of unfairness about people and the world in general (Currier et al., 2006; Mancini et al., 2011).

3. A cognitive behavioral model of CG: Factors associated with problematic coping The psychological mechanisms following (violent) loss may also be viewed from the perspective of a cognitive behavioral model of CG (Boelen, Van den Hout, & Van den Bout, 2006). This model is not primarily developed to explain the difference

8

General Introduction

in psychopathology between violent and non-violent loss, but to examine the psychological mechanisms which are present in individuals with CG following violent or non-violent loss. Based on this model, it is thought that three processes are crucial in the development and maintenance of CG: 1) problems with the integration of the event in autobiographical memory, 2) maladaptive thoughts, and 3) avoidance behavior. This model is related to cognitive-behavioral models for PTSD (e.g., Bower & Sivers, 1998; Brewin, Dalgleish, & Joseph, 1996; Ehlers & Clark, 2000; Ehlers & Steil, 1995; Foa & Kozak, 1986; Foa & Rothbaum, 1998; Horowitz, 1997; Janoff-Bulman, 1992). The first process which is thought to be important in the development and maintenance of CG regards problems with the integration of the event in autobiographical memory. In individuals with CG, numerous stimuli unintentionally trigger memories about the deceased person, making everything a possible reminder of the loss. Rather than making the loss feel more ‘real’, bereaved individuals continue to be shocked by the loss (Boelen et al., 2006). Individuals bereaved by homicidal loss are very likely to be unwillingly and unexpectedly confronted with reminders of the loss. Fictional movies, documentaries, news bulletins about homicidal cases on television, and games about violence are popular among the general public, but may be a reminder of their violent loss to homicidally bereaved individuals. Through police investigations and ongoing lawsuits, individuals keep being confronted with cause of death, and with the perpetrator who took their loved one. Such lawsuits may last for several years. Thereby, the separation of the deceased one may be more difficult to accept. Also, homicidal loss is rare when compared to natural losses, making the loss possibly feel unreal to the bereaved individual (‘it feels like a movie’), and more difficult to integrate within their existing knowledge about the world and the trustworthiness of other people (‘If I cannot trust my father since he killed my mother, who can I trust then?’). The second process which is thought to be important in the development and maintenance of CG concerns negative cognitions and misbeliefs people might have about their own grief reactions. In general, people have expectations and ideas about their role in life and their purpose. When people adjust to their loss, they are able to adjust their beliefs to the new situation, in which their loved one is dead. For some people, adapting their beliefs to a new situation remains difficult. They may experience negative beliefs, such as ‘the future is worthless’, and ‘I do not have confidence in the future’. When these cognitions become dominant, people may retreat from activities which foster functioning, such as social meetings, going to work and setting new goals (Boelen et al., 2006). Since homicidal loss is culpable, bereaved individuals are likely to blame the perpetrator. They may also blame bystanders who were present during the homicide, or they may blame themselves (‘This would not have happened if I have warned my daughter for her boyfriend’). Another type of maladaptive thoughts that might be more frequent after homicidal loss than after other death causes, are thoughts of revenge. Individuals might become preoccupied with vengeful thoughts about the offender (‘I want him to feel the same amount of pain he caused us), and think they only can ‘move forward’ when the perpetrator

9

Chapter 1 has been punished. These types of thoughts might be especially difficult to cope with in cases where punishment is out of the question, or when victim and perpetrator are part of the same family. Because of the violent nature of the loss, homicidally bereaved individuals may be more inclined to develop negative cognitions than bereaved individuals whose loved one died by a non-violent death cause: negative cognitions about the world (‘the world is no safe place anymore’) and others (‘I can trust nobody’) are easily triggered when the victim died by the intentional and violent act of another human being. Third, people with CG are inclined to use avoidance strategies to avoid the pain connected to the loss. They assume to lose control or go mad when they are confronted with feelings of mourning. Therefore, people, places and objects who remind the bereaved individual of the victim are avoided (Boelen et al., 2006). This type of avoidance is referred to as anxious avoidance, avoiding the reality of the loss. People who are inclined to avoid the reality of the loss may also do seemingly the opposite: they constantly think, talk or ruminate about the deceased or the death cause, and cherish or cultivate personal objects. While it might seem that these bereaved individuals confront rather than avoid the loss, this may function as a way to maintain a strong connection with the deceased and avoid having to admit to the loss and the fact that the deceased is gone (Boelen et al., 2006). Following homicidal loss, bereaved individuals may ruminate more about issues related to the death cause, about questions such as: who is the perpetrator? Why did he murder our loved one? Will he be arrested and punished? Did the victim feel any pain? And could I have warned the victim, or could I have done something else to prevent his/her death? Following homicidal loss, these types of questions are generally unanswered, which might lead to rumination in the bereaved individual. Another type of avoidance is referred to as depressive avoidance, which means refraining from social activities that could foster adjustment (Boelen & Van den Bout, 2010). Bereaved individuals may withdraw from social activities, because they belief that these activities are pointless when encountered without the deceased. Thereby, they miss the opportunity to engage in new social contacts, and they do not challenge negative thoughts and cognitions (Boelen & Van den Bout, 2010). Depressive and anxious avoidance were both associated with symptom-levels of CG and PTSD and were, together with negative cognitions, found to be unique mediators of the association between violent loss and elevated levels of PTSD (Boelen et al., 2015; Boelen & Eisma, 2015). The tendency to depressive avoidance among homicidally bereaved individuals might also be related to the context in which the loss happened. Bereavement following non-violent loss is in most cases a personal and private experience. Homicidal loss on the other hand may often take place in a public context. Television and newspapers often report about homicidal cases, and the search for and punishment of the perpetrator is often a subject of news reports. As a consequence, the grief experience is no longer a private event of the family and friends who are directly related to the victim, but has turned into a public event. Every social encounter, even with strangers, might lead to questions or conversations about the

10

General Introduction

loss, making homicidally bereaved individuals inclined (possibly more than following natural loss) to avoid social encounters. Homicidally bereaved individuals might also be inclined to avoid social situations because of (fear for) negative or stigmatizing social attitudes from other individuals. About 13% of the homicides in the Netherlands occur in a criminal environment. In these types of homicides, the perpetrator or the victim (or both) are involved in criminal activities, such as drug trafficking or liquidations (Nieuwbeerta & Leistra, 2007). These types of losses might lead to shame in bereaved individuals or to negative or stigmatizing social attitudes from other individuals (Armour, 2006). The fear of such (perceived) stigmatizing reactions may make homicidally bereaved individuals more inclined to avoid social situations and encounters with other people, thereby limiting involvement in activities which could foster adjustment. To conclude, in terms of the cognitive behavioral model put forward by Boelen et al. (2006), three processes are crucial in the development and maintenance of CG, namely insufficient integration of the loss, negative cognitions, and avoidance behavior. Research has found that negative cognitions and avoidance are associated with CG and PTSD (Boelen et al., 2015; Boelen, De Keijser, Van den Hout, & Van den Bout, 2011). However, it is to date unknown if the assumptions of the model of Boelen et al., (2006) are also true following homicidal loss. Therefore, it is tested whether negative cognitions and avoidance are associated with CG and PTSD symptoms in homicidally bereaved individuals. Measures of anger and revenge are added, because those are likely to be elicited following homicidal loss and it was deemed relevant to enhance understanding of cognitive-behavioral variables possibly contributing to these latter phenomena (Chapter 5). Before turning to a treatment for homicidally bereaved individuals (reported in Chapter 6), the role of revenge following homicidal loss is further elaborated about.

4. Revenge Vengeful thoughts are likely to occur following homicidal loss. Revenge can be defined as ‘an aggressive, often violent, response to intentional harm that has been inflicted on the avengers and their families’ (Stuckless, 1996, p. 21). Revenge may be associated with negative mental health. In victims of severe interpersonal violence, such as sexual violence and physical assault, revenge was found to be associated with more intense rumination, less life satisfaction, and PTSD (Kunst, 2011; Orth, Montada, & Maercker, 2006). A more forgiving response to the person who has wronged you on the other hand, has been found to be associated with increased mental and physical health (Schultz, Tallman, & Altmaier, 2010). As further elaborated in Chapter 4, revenge may block grief processing following homicidal loss, by maintaining an external, ruminative focus on why and how the loss occurred, that interferes with the elaboration and processing of the reality of the loss and the feelings associated with it. The preoccupation with revenge could be associated with avoidance of acceptance of the death and with CG (Rynearson, 1984) and PTSD (Ehlers & Clark, 2000).

11

Chapter 1 One of the key components of revenge is the attribution of responsibility to the offender (Cota-Mckinley, Woody, & Bell, 2001; Orth, 2004). This is relevant in homicidal loss, since the perpetrator killed the victim premeditatedly and intentionally, making him responsible for the death. Studies found that bereaved individuals experienced more revenge following homicidal loss than following loss by suicide or accidents (Baddeley et al., 2015) and in situations someone could be blamed for the death (Stuckless, 1996; Weinberg, 1994). While these studies seem to give some indication that revenge is possibly elicited by homicidal loss, scientific literature regarding revenge and coping following homicidal loss is limited. The few studies described previously have methodological limitations, such as small numbers of homicidally bereaved individuals and a lack of a measure of CG. Since the association between revenge and coping following homicidal loss is largely unclear, two studies about revenge were conducted (Chapter 4 and 5). In Chapter 4, the association between revenge and CG and PTSD was examined. Two types of revenge were distinguished, namely dispositional revenge, defined as referring to someone’s general attitude toward revenge, and situational revenge, referring to levels of revenge experienced following a specific incident and directed at a specific perpetrator. The two types of revenge were regarded as independent variables and as a possible predictor of psychopathology. In Chapter 5, the association between negative cognitions and avoidance on the one hand, and CG, PTSD, anger and revenge on the other hand was examined. The relation between avoidance behavior and symptom-levels of CG and PTSD was found prospectively and longitudinally in several studies, but primarily in samples of bereaved individuals following natural loss, and not following homicidal loss (Boelen, Van den Hout, & Van den Bout, 2013). In the study presented in Chapter 5, only situational revenge was included, and not dispositional revenge. In this study, revenge was used as an outcome variable. In sum, the association between revenge and psychopathology was first examined (Chapter 4, revenge as independent variable), followed by a further examination of the association between revenge and negative cognitions and avoidance (Chapter 5, revenge as dependent variable). This is also depicted in Figure 1, which is presented after the next paragraph. In Chapter 4 and 5, the general term thoughts and feelings of revenge was used, although it is noticed that revenge has also been referred to as a cognition, emotion, response or behavior (it goes too far to provide a complete conceptualisation of revenge, see for further reading Grobbink, Derksen, and Van Marle, 2015 who wrote a review about revenge). When the difference between dispositional and situational revenge is relevant, these terms were used.

5. Treatment with CBT and EMDR for homicidally bereaved individuals As it is reported in more detail in Chapter 2 and 3, a significant number of homicidally bereaved individuals might experience symptoms of CG and PTSD.

12

General Introduction

In Chapter 6, a treatment study with Cognitive Behavioral Therapy (CBT) and Eye Movement Desensitization  and Reprocessing (EMDR) to reduce self-rated CG and PTSD symptoms is described. The rationale of an intervention with CBT for bereaved individuals was based on the cognitive behavioral model of CG as described previously. Various interventions have been developed for individuals who experience a distorted grieving process, whereby CBT was found to be the most effective treatment for reducing psychological complaints in bereaved individuals (Currier, Holland, & Neimeyer, 2010). The effectiveness of CBT to reduce CG and PTSD following non-violent and violent loss has been reported in several studies (Asukai, Tsuruta, & Saito, 2011; Boelen, De Keijser, Van den Hout, & Van den Bout, 2007; Bryant et al., 2014; Maccallum & Bryant, 2011; Rosner, Pfoh, Kotoucová, & Hagl, 2014; Wagner, Knaevelsrud, & Maercker, 2006). However, in all these studies, different modes of violent death were grouped together, and only a small number of participants were homicidally bereaved individuals. As reported in more detail in Chapter 6, the few studies that have been conducted in samples of homicidally bereaved individuals were not controlled studies and were merely observational in nature (Parkes, 1993; Rynearson, 1994), making the effectiveness of CBT following homicidal loss largely unclear. The treatment reported in Chapter 6 did not only include CBT, but also EMDR. The latter was found to be effective for emotional distress following different types of trauma (Bisson & Andrew, 2007), and is also suggested to be effective in situations where trauma and loss are intertwined (Solomon & Rando, 2007; Solomon & Shapiro, 1997). In traumatic loss, such as homicide, there may be obstacles which can interfere with recovery and adjustment. Examples are memories related to the traumatic death cause and feelings of responsibility for the event, a lack of control, and personal vulnerability and safety (Solomon & Shapiro, 1997). Applying EMDR is hypothesized to be helpful in alleviating CG symptoms by desensitizing traumatic memories and the associated distressing thoughts and painful feelings. Thereby, processing of information held and dysfunctionally stored is accelerated. By reprocessing the mental representations of traumatic material with the use of EMDR, neutral or more pleasant memories associated with the loved one may appear, which may foster positive functioning (Solomon & Rando, 2007; Solomon & Shapiro, 1997). As it can be read in Chapter 6, the effectiveness of EMDR as a treatment for CG has only been demonstrated in case examples (Solomon & Rando, 2007) and in a sample of 50 bereaved individuals following loss due to different violent death causes (Sprang, 2001). This gives some preliminary indication that EMDR could be helpful treating CG although the effectiveness of EMDR after grief following homicidal loss has yet to be established. In Chapter 6, a Randomized Controlled Trial (RCT) was presented, in which the effectiveness of a combined treatment of CBT and EMDR is examined to reduce self-rated CG and self-rated PTSD symptoms. Due to the comorbidity of CG and PTSD in homicidally bereaved individuals (McDevitt-Murphy et al., 2012; Shear, Frank, Houck, & Reynolds, 2005), a combination of CBT and EMDR seems to have added value over only one of the two treatments, because then both trauma and loss-symptoms are addressed (Shahani & Trish, 2006).

13

Chapter 1 In Figure 1, a schematic model of the variables likely to be important in adjustment to homicidal loss is presented. In sum, negative cognitions and avoidance behavior could hinder adaption to the loss. Revenge (dispositional and situational revenge) was added to this model, because thoughts and feelings of revenge are likely to be elicited by homicidal loss, and may complicate the grieving process. By means of the different studies included in this dissertation, the role of negative cognitions, avoidance behavior, and revenge, as well as the treatment effect of EMDR and CBT on CG and PTSD following homicidal loss was examined.

Cognitions

Avoidance

Revenge

CG CG

Homicide

EMDR+CBT PTSD

PTSD

Cognitions

Avoidance

Revenge

Figure 1. Schematic model of variables central to this dissertation.

14

General Introduction

6. The importance of examining the impact of homicidal loss Examining psychopathology following homicidal loss is deemed important because of several reasons. Studies about this topic are needed because the psychological impact of homicidal loss is often assumed by scientists, the bereaved community, and the society, but has not yet been established by empirically sound studies. For bereaved individuals, scientists and clinicians working with homicidally bereaved individuals, it may be important to know whether such claims are based upon empirical evidence. If these claims are not based upon such evidence, they may serve as a reinforcement for pathological behavior: repeatedly getting the message that adjustment after homicidal loss is difficult or perhaps impossible could strengthen non adaptive cognitions and even block adaptive behavioral patterns. In theory, this may increase the subjective psychological pain or pathological behavior and could prolong the time in which the bereaved individual suffer from psychological complaints. If these claims do indeed reflect objective reality, this information can serve as a base for the development of clinical interventions for this population, such as more individualized or specialized treatments. This latter is consistent with sounds of a support group for parents of murdered children in the Netherlands, which called for more specialist treatments to cope with their experience of loss. Clinically, insight in the variables which are likely to play a role in symptomatology may provide input for the development of a treatment for this population. On an individual level, knowledge about the psychological mechanisms following homicidal loss may be used to inform homicidally bereaved individuals by means of psycho-education what they may expect following this type of loss. Further, if the intervention is proven effective in reducing symptoms of CG and PTSD, help-seeking homicidally bereaved individuals can make an informed choice about the type of therapy they can use.

7. Research questions and chapter outline Based on the previous outline, this dissertation includes the following research questions: 1. What is the nature and prevalence of emotional symptoms following homicidal loss (Chapter 2)? 2. What is the prevalence of CG and PTSD in homicidally bereaved individuals, and which socio-demographic and perpetrator-related factors are correlated with CG and PTSD (Chapter 3)? 3. What is the association between dispositional and situational revenge on the one hand, and CG, PTSD and positive functioning in homicidally bereaved individuals on the other hand, and which socio-demographic and perpetrator related factors are associated with dispositional and situational revenge (Chapter 4)? 4. What is the association between negative cognitions and avoidance behavior on

15

Chapter 1 the one hand and CG, PTSD, anger and revenge on the other hand in homicidally bereaved individuals (Chapter 5)? 5. What is the effectiveness of a short intervention with CBT and EMDR on self-rated CG and self-rated PTSD symptoms in homicidally bereaved individuals, and which variables moderate the treatment (Chapter 6)? The research questions will be answered in the different chapters included in this dissertation. In Chapter 1, a general introduction is provided about the variables central to this dissertation. In Chapter 2, a systematic review of the literature on the nature and prevalence of emotional symptoms following homicidal bereavement is provided. A systematic review may give insight in the type of psychopathology experienced following homicidal loss, their prevalence, and the relative importance of the different disorders, i.e., which disorders are more commonly experienced following homicidal loss or are more studied than other disorders. In Chapter 3, the prevalence of CG and PTSD in a sample of 312 partners, family members and friends of homicide victims is examined. Also, socio demographic and perpetrator related correlates of CG and PTSD will be described. Socio-demographic correlates that are studied are sex and age of the participant, time since loss and kinship between the bereaved individual and the victim. These variables are often studied following non-violent loss or in studies where different violent death causes were grouped together, but not frequently following homicidal loss. Perpetratorrelated variables, such as type of relation between the perpetrator and the bereaved individual and the legal status of the perpetrator also have not been examined previously following homicidal loss. In Chapter 4, a cross-sectional study is described in which two types of revenge (namely general revenge and situational revenge) and PTSD, CG and positive functioning among 331 spouses and family members of homicide victims are examined. The study provides insight into the association between revenge and longterm adjustment to homicidal loss. While the association between revenge and PTSD has been found in victims of severe violence other than homicidal loss (Kunst, 2011; Orth et al., 2006), and the association between revenge and CG has been theorized about (Salloum & Rynearson, 2006), it has not yet been examined empirically in a large sample of homicidally bereaved individuals. In Chapter 5, a cross-sectional study is presented about the associations between negative cognitions and avoidance behavior on the one hand, and CG, PTSD, anger and revenge on the other hand. The study was conducted among 331 spouses and family members of homicide victims. Anger and revenge may be important consequences of homicidal loss. The relation between avoidance behavior and symptom levels of CG and PTSD was found prospectively and longitudinally in several studies, but primarily in samples of bereaved individuals following natural loss (Boelen et al., 2013). The studies included only small samples following violent loss, which were also grouped together (Boelen et al., 2015; Boelen, & Eisma, 2015; Boelen, Reijntjes, Djelantik, & Smid, 2016; Boelen, & Van

16

General Introduction

den Bout, 2010). The relative importance of negative cognitions and avoidance behavior with CG and PTSD has not yet been examined following homicidal loss. In Chapter 6, the results of a RCT are presented, in which a combined intervention of EMDR and CBT was used to reduce self-rated CG and PTSD in individuals bereaved by homicide. The treatment effect was expected to be moderated by sex of the participant, time since loss and the recruitment style (i.e., divided into support organisations for homicidally bereaved individuals, by the governmental organization Victim help or via the internet). Examination of these moderator variables was deemed important because the results may tell us for which subgroup of people the treatment is more or less effective and for whom the treatment may require adjustment. In Chapter 7, a general discussion is provided. The main findings from the preceding chapters are put in a broader context, methodological difficulties are discussed, and clinical implications and recommendations for future research are provided. It should be noted that this dissertation is about psychological adjustment following homicidal loss, but that individuals whose loved one died due to manslaughter, such as a drunk driver (in Dutch dood door schuld) were not included. Bereaved individuals of homicides committed in a collective context, such as wars and genocide, were also not included. When referring to the offender, the term ‘he’ is used, where also ‘she’ could be read. Further, the term Complicated Grief was used through trough this dissertation because the term Persistent Complex Bereavement Disorder (PCBD) had not been used when the studies in this dissertation began, and the term is consistent with the scale that was used (the Inventory of Complicated Grief, Prigerson et al., 1995). Although PCBD is now widely used, the term Complicated Grief was used for the sake of consistency.

17

Chapter 1

References American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Washington, DC: Author. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing. Amick-McMullan, A., Kilpatrick, D. G., & Resnick, H. S. (1991). Homicide as a risk factor for PTSD among surviving family members.  Behavior Modification,  15(4), 545-559. doi:10.1177/01454455910154005 Amick-McMullan, A., Kilpatrick, D. G., Veronen, L. J., & Smith, S. (1989). Family survivors of homicide victims: Theoretical perspectives and an exploratory study. Journal of Traumatic Stress, 2(1), 21-35. doi:10.1002/jts.2490020104 Armour, M. (2006). Violent death: Understanding the context of traumatic and stigmatized grief.  Journal of Human Behavior in the Social Environment,  14(4), 53-90. doi:10.1300/ J137v14n04_04 Asukai, N., Tsuruta, N., & Saito, A. (2011). Pilot study on traumatic grief treatment program for Japanese women bereaved by violent death. Journal of Traumatic Stress, 24(4), 470-473. doi:10.1002/jts.20662 Baddeley, J. L., Williams, J. L., Rynearson, T., Correa, F., Saindon, C., & Rheingold, A. A. (2015). Death thoughts and images in treatment-seekers after violent loss. Death Studies, 39(2), 84-91. doi:10.1080/07481187.2014.893274 Bisson, J., & Andrew, M. (2007). Psychological treatment of post-traumatic stress disorder (PTSD). Cochrane Database of Systematic Reviews 2007, Issue 3: CD003388. Boelen, P. A. (2005). Complicated Grief. Assessment, theory, and treatment. Enschede/ Amsterdam, Nederland: Print Partners Ipskamp. Boelen, P. A. (2015). Peritraumatic distress and dissociation in prolonged grief and posttraumatic stress following violent and unexpected deaths. Journal of Trauma & Dissociation, 16(5), 541-550. doi:10.1080/15299732.2015.1027841 Boelen, P. A., & van den Bout, J. (2010). Anxious and depressive avoidance and symptoms of prolonged grief, depression, and post-traumatic stress disorder. Psychologica Belgica, 50(12), 49-67. doi:10.5334/pb-50-1-2-49 Boelen, P. A., & Eisma, M. C. (2015). Anxious and depressive avoidance behavior in postloss psychopathology: A longitudinal study.  Anxiety, Stress & Coping: An International Journal, 28(5), 587-600. doi:10.1080/10615806.2015.1004054 Boelen, P. A., van den Hout, M.A., & van den Bout, J. (2013). Prolonged grief disorder: Cognitive-behavioral theory and therapy. In M. Stroebe, H. Schut, & J. van den Bout (Eds.), Complicated grief. Scientific foundations for health care professionals (pp. 221-234). New York, United States: Routledge/Taylor & Francis Group. Boelen, P. A., van den Hout, M.A., & van den Bout, J. (2006). A cognitive-behavioral conceptualization of complicated grief.  Clinical Psychology: Science and Practice,  13(2), 109-128. doi:10.1111/j.1468-2850.2006.00013.x Boelen, P. A., de Keijser, J., & Smid, G. (2015). Cognitive–behavioral variables mediate the impact of violent loss on post-loss psychopathology.  Psychological Trauma: Theory, Research, Practice, and Policy, 7(4), 382-390. doi:10.1037/tra0000018 Boelen, P. A., de Keijser, J., van den Hout, M.A., & van den Bout, J. (2007). Treatment of complicated grief: A comparison between cognitive-behavioral therapy and supportive counseling. Journal of Consulting and Clinical Psychology, 75(2), 277-284. doi:10.1037/0022006X.75.2.277

18

General Introduction

Boelen, P. A., de Keijser, J., van den Hout, M.A., & van den Bout, J. (2011). Factors associated with outcome of cognitive–behavioral therapy for complicated grief: A preliminary study. Clinical Psychology & Psychotherapy, 18(4), 284-291. doi:10.1002/cpp.720 Boelen, P. A., Reijntjes, A., Djelantik, A. A. A. M., & Smid, G. E. (2016). Prolonged grief and depression after unnatural loss: Latent class analyses and cognitive correlates. Psychiatry Research, 240, 358-363. doi:10.1016/j.psychres.2016.04.012 Boelen, P. A., van de Schoot, R., van den Hout, M.A.., de Keijser, J., & van den Bout, J. (2010). Prolonged grief disorder, depression, and posttraumatic stress disorder are distinguishable syndromes. Journal of Affective Disorders, 125(1-3), 374-378. doi:10.1016/j.jad.2010.01.076 Bonanno, G. A. (2004). Loss, trauma, and human resilience: Have we underestimated the human capacity to thrive after extremely aversive events?  American Psychologist, 59(1), 20-28. doi:10.1037/0003-066X.59.1.20 Bower, G. H., & Sivers, H. (1998). Cognitive impact of traumatic events.  Development and Psychopathology, 10(4), 625-653. doi:10.1017/S0954579498001795 Breslau, N., & Kessler, R. C. (2001). The stressor criterion in DSM-IV posttraumatic stress disorder: An empirical investigation.  Biological Psychiatry,  50(9), 699-704. doi:10.1016/ S0006-3223(01)01167-2 Breslau, N., Kessler, R. C., Chilcoat, H. D., Schultz, L. R., Davis, G. C., & Andreski, P. (1998). Trauma and posttraumatic stress disorder in the community: The 1996 detroit area survey of trauma. Archives of General Psychiatry, 55(7), 626-632. doi:10.1001/archpsyc.55.7.626 Brewin, C. R., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of posttraumatic stress disorder. Psychological Review, 103(4), 670-686. doi:10.1037/0033-295X.103.4.670 Bryant, R. A., Kenny, L., Joscelyne, A., Rawson, N., Maccallum, F., Cahill, C., . . . Nickerson, A. (2014). Treating prolonged grief disorder: A randomized clinical trial.  JAMA Psychiatry, 71(12), 1332-1339. doi:10.1001/jamapsychiatry.2014.1600 Burke, L. A., & Neimeyer, R. A. (2013). Prospective risk factors for complicated grief: A review of the empirical literature. In M. Stroebe, H. Schut, & J. van den Bout (Eds), Complicated grief. Scientific foundations for health care professionals (pp. 145-161). New York, United States: Routledge/Taylor & Francis Group. Burke, L. A., Neimeyer, R. A., & McDevitt-Murphy, M. (2010). African American homicide bereavement: Aspects of social support that predict complicated grief, PTSD, and depression. Omega: Journal of Death and Dying, 61(1), 1-24. doi:10.2190/OM.61.1.a Centraal Bureau voor de Statistiek. (2016). Overledenen; moord en doodslag; pleeglocatie Nederland. Geraadpleegd op http://statline.cbs.nl/Statweb/publication/?DM=SLNL&PA=8 1453NED&D1=0,2-10&D2=0&D3=a&D4=14-18&HDR=G1,T&STB=G2,G3&VW=T Cota-McKinley, A., Woody, W. D., & Bell, P. A. (2001). Vengeance: Effects of gender, age, and religious background. Aggressive Behavior, 27(5), 343-350. doi:10.1002/ab.1019 Cougle, J. R., Kilpatrick, D. G., & Resnick, H. (2012). Defining traumatic events: Research findings and controversies. In J. G. Beck, D. M. Sloan, J. G. (. Beck & D. M. (. Sloan (Eds.), (pp. 11-27). New York, NY, US: Oxford University Press. doi:10.1093/oxfordhb/9780195399066.013.0002 Currier, J. M., Holland, J. M., & Neimeyer, R. A. (2006). Sense-making, grief, and the experience of violent loss: Toward a mediational model.  Death Studies,  30(5), 403-428. doi:10.1080/07481180600614351 Currier, J. M., Holland, J. M., & Neimeyer, R. A. (2010). Do CBT-based interventions alleviate distress following bereavement? A review of the current evidence. International Journal of Cognitive Therapy, 3(1), 77-93. doi:10.1521/ijct.2010.3.1.77 Davis, C. G., Wortman, C. B., Lehman, D. R., & Silver, R. C. (2000). Searching for meaning in loss: Are clinical assumptions correct?  Death Studies,  24(6), 497-540. doi:10.1080/07481180050121471

19

Chapter 1 Ehlers, A., & Clark, D. M. (2000). A cognitive model of posttraumatic stress disorder. Behavior Research and Therapy, 38(4), 319-345. doi:10.1016/S0005-7967(99)00123-0 Ehlers, A., & Steil, R. (1995). Maintenance of intrusive memories in posttraumatic stress disorder: A cognitive approach. Behavioral and Cognitive Psychotherapy, 23(3), 217-249. doi:10.1017/S135246580001585X Foa, E. B., & Kozak, M. J. (1986). Emotional processing of fear: Exposure to corrective information. Psychological Bulletin, 99(1), 20-35. doi:10.1037/0033-2909.99.1.20 Foa, E. B., & Rothbaum, B. O. (1998). Treating the trauma of rape: Cognitive-behavioral therapy for PTSD. New York, NY, US: Guilford Press. Freedy, J. R., Resnick, H. S., Kilpatrick, D. G., Dansky, B. S., & Tidwell, R. P. (1994). The psychological adjustment of recent crime victims in the criminal justice system. Journal of Interpersonal Violence, 9(4), 450-468. doi:10.1177/088626094009004002 Freeman, L. N., Shaffer, D., & Smith, H. (1996). Neglected victims of homicide: The needs of young siblings of murder victims.  American Journal of Orthopsychiatry,  66(3), 337-345. doi:10.1037/h0080184 Ginzburg, K., Geron, Y., & Solomon, Z. (2002). Patterns of complicated grief among bereaved parents. Omega: Journal of Death and Dying, 45(2), 119-132. doi:10.2190/XUW5-QGQ9KCB8-K6WW Grobbink, L. H., Derksen, J. J. L., & van Marle, H. J. C. (2015). Revenge: An analysis of its psychological underpinnings. International Journal of Offender Therapy and Comparative Criminology, 59(8), 892-907. doi:10.1177/0306624X13519963 Horowitz, M. J. (1997). Stress response syndromes: PTSD, grief, and adjustment disorders (3rd ed.). Lanham, MD, US: Jason Aronson. Jacobs, S. (1993).  Pathologic grief: Maladaptation to loss. Arlington, VA, US: American Psychiatric Association. Jacobs, S., Mazure, C., & Prigerson, H. (2000). Diagnostic criteria for traumatic grief.  Death Studies, 24(3), 185-199. doi:10.1080/074811800200531 Janoff-Bulman, R. (1989). Assumptive worlds and the stress of traumatic events: Applications of the schema construct. Social Cognition, 7(2), 113-136. doi:10.1521/soco.1989.7.2.113 Janoff-Bulman, R. (1992). Shattered assumptions: Towards a new psychology of trauma. New York, NY, US: Free Press. Kaltman, S., & Bonanno, G. A. (2003). Trauma and bereavement: Examining the impact of sudden and violent deaths.  Journal of Anxiety Disorders,  17(2), 131-147. doi:10.1016/ S0887-6185(02)00184-6 Kersting, A., Brähler, E., Glaesmer, H., & Wagner, B. (2011). Prevalence of complicated grief in a representative population-based sample. Journal of Affective Disorders, 131(1-3), 339-343. doi:10.1016/j.jad.2010.11.032 Kessler, R. C., Sonnega, A., Bromet, E., Hughes, M., & Nelson, C. B. (1995). Posttraumatic stress disorder in the national comorbidity survey. Archives of General Psychiatry, 52(12), 10481060. doi:10.1001/archpsyc.1995.03950240066012 Kristensen, P. å., Weisæth, L., & Heir, T. (2012). Bereavement and mental health after sudden and violent losses: A review. Psychiatry: Interpersonal and Biological Processes, 75(1), 7697. doi:10.1521/psyc.2012.75.1.76 Kristensen, P. å., Weisæth, L., Hussain, A., & Heir, T. (2015). Prevalence of psychiatric disorders and functional impairment after loss of a family member: A longitudinal study after the 2004 tsunami. Depression and Anxiety, 32(1), 49-56. doi:10.1002/da.22269 Kunst, M. J. J. (2011). PTSD symptom clusters, feelings of revenge, and perceptions of perpetrator punishment severity in victims of interpersonal violence. International Journal of Law and Psychiatry, 34(5), 362-367. doi:10.1016/j.ijlp.2011.08.003

20

General Introduction

Leistra, G. (2015) Aantal moorden blijft dalen, maar nog veel liquidaties. Elsevier, 31. Lobb, E. A., Kristjanson, L. J., Aoun, S. M., Monterosso, L., Halkett, G. K. B., & Davies, A. (2010). Predictors of complicated grief: A systematic review of empirical studies.  Death Studies, 34(8), 673-698. doi:10.1080/07481187.2010.496686 Maccallum, F., & Bryant, R. A. (2011). Autobiographical memory following cognitive behavior therapy for complicated grief.  Journal of Behavior Therapy and Experimental Psychiatry, 42(1), 26-31. doi:10.1016/j.jbtep.2010.08.006 Maercker, A., & Znoj, H. (2010). The younger sibling of PTSD: Similarities and differences between complicated grief and posttraumatic stress disorder.  European Journal of Psychotraumatology, 1 Mancini, A. D., Prati, G., & Bonanno, G. A. (2011). Do shattered worldviews lead to complicated grief? prospective and longitudinal analyses. Journal of Social and Clinical Psychology, 30(2), 184-215. doi:10.1521/jscp.2011.30.2.184 McDevitt-Murphy, M., Neimeyer, R. A., Burke, L. A., Williams, J. L., & Lawson, K. (2012). The toll of traumatic loss in african americans bereaved by homicide. Psychological Trauma: Theory, Research, Practice, and Policy, 4(3), 303-311. doi:10.1037/a0024911 Morina, N., Rudari, V., Bleichhardt, G., & Prigerson, H. G. (2010). Prolonged grief disorder, depression, and posttraumatic stress disorder among bereaved kosovar civilian war survivors: A preliminary investigation. International Journal of Social Psychiatry, 56(3), 288297. doi:10.1177/0020764008101638 Murphy, S. A., Johnson, L. 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(1), 39-61. Murphy, S. A., Lohan, J., Braun, T., Johnson, L. C., Cain, K. C., Beaton, R. D., & Baugher, R. (1999). Parents’ health, health care utilization, and health behaviors following the violent deaths of their 12- to 28-year-old children: A prospective longitudinal analysis. Death Studies, 23(7), 589-616. doi:10.1080/074811899200795 Nieuwbeerta, P., & Leistra, G. (2007). Dodelijk geweld. Moor den doodslag in Nederland. Amsterdam, Nederland: Uitgeverij Balans. Orth, U. (2004). Does perpetrator punishment satisfy victims’ feelings of revenge? Aggressive Behavior, 30(1), 62-70. doi:10.1002/ab.20003 Orth, U., Montada, L., & Maercker, A. (2006). Feelings of revenge, retaliation motive, and posttraumatic stress reactions in crime victims.  Journal of Interpersonal Violence,  21(2), 229-243. doi:10.1177/0886260505282286 Parkes, C. M. (1993). Psychiatric problems following bereavement by murder or manslaughter. The British Journal of Psychiatry, 162, 49-54. doi:10.1192/bjp.162.1.49 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, United States: Guilford Press. Prigerson, H. (2004). Complicated grief. Bereavement Care, 23:3, 38-40, DOI: 10.1080/02682620408657612 Prigerson, H. G., Maciejewski, P. K., Reynolds, C. F., III, Bierhals, A. J., Newsom, J. T., Fasiczka, A., . . . Miller, M. (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 Riches, G., & Dawson, P. (1998). Spoiled memories: Problems of grief resolution in families bereaved through murder. Mortality, 3(2), 143-159. doi:10.1080/713685897 Rosner, R., Pfoh, G., Kotoučová, M., & Hagl, M. (2014). Efficacy of an outpatient treatment for prolonged grief disorder: A randomized controlled clinical trial.  Journal of Affective Disorders, 167, 56-63. doi:10.1016/j.jad.2014.05.035

21

Chapter 1 Rynearson, E. K., & McCreery, J. M. (1993). Bereavement after homicide: A synergism of trauma and loss. The American Journal of Psychiatry, 150(2), 258-261. Rynearson, E. K., Schut, H., & Stroebe, M. (2013). Complicated grief after violent death: Identification and intervention. In M. Stroebe, H. Schut, & J. van den Bout (Eds), Complicated grief. Scientific foundations for health care professionals (pp. 278-292). New York, United States: Routledge/Taylor & Francis Group. Rynearson, E. K. (1984). Bereavement after homicide: A descriptive study.  The American Journal of Psychiatry, 141(11), 1452-1454. Rynearson, E. K. (1988). The homicide of a child. In F. M. Ochberg (Ed.), Post-traumatic therapy and victims of violence (pp. 213-224). Philadelphia, United States: Brunner/Mazel. Rynearson, T. (1994). Psychotherapy of bereavement after homicide. Journal of Psychotherapy Practice & Research, 3(4), 341-347. Salloum, A., & Rynearson, E. K. (2006). Family resilience after violent death. In E. K. Rynearson (Ed.), Violent death. Resilience and intervention beyond the crisis (pp. 47-63). New York, United States: Routledge/Taylor & Francis Group. Schultz, J. M., Tallman, B. A., & Altmaier, E. M. (2010). Pathways to posttraumatic growth: The contributions of forgiveness and importance of religion and spirituality. Psychology of Religion and Spirituality, 2(2), 104-114. doi:10.1037/a0018454 Shahani, P. J., & Trish, H. M. (2006). Healing after September 11: Short-term group intervention with 9/11 families. In E. K. Rynearson (Ed.), Violent death. Resilience and intervention beyond the crisis (pp. 335-355). New York, United States: Routledge/Taylor & Francis Group. Shear, K., Frank, E., Houck, P. R., & Reynolds, C. F. III. (2005). Treatment of complicated grief: A randomized controlled trial. JAMA: Journal of the American Medical Association, 293(21), 2601-2608. doi:10.1001/jama.293.21.2601 Shear, M. K. (2015). Complicated grief. The New England Journal of Medicine, 372(2), 153-160. doi:10.1056/NEJMcp1315618 Shear, M. K., Simon, N., Wall, M., Zisook, S., Neimeyer, R., Duan, N., . . . Keshaviah, A. (2011). Complicated grief and related bereavement issues for DSM-5. Depression and Anxiety, 28(2), 103-117. doi:10.1002/da.20780 Simon, N. M. (2013). Treating complicated grief.  JAMA: Journal of the American Medical Association, 310(4), 416-423. doi:10.1001/jama.2013.8614 Smid, G. E., Kleber, R. J., de la Rie, S.M., Bos, J. B. A., Gersons, B. P. R., & Boelen, P. A. (2015). Brief eclectic psychotherapy for traumatic grief (BEP-TG): Toward integrated treatment of symptoms related to traumatic loss. European Journal of Psychotraumatology, 6 Solomon, R. M., & Rando, T. A. (2007). Utilization of EMDR in the treatment of grief and mourning.  Journal of EMDR Practice and Research,  1(2), 109-117. doi:10.1891/19333196.1.2.109 Solomon, R.M., & Shapiro, F (1997). Eye movement desensitization and reprocessing: A therapeutic tool for trauma and grief. In C.R. Figley, B. E. Bride, & N. Mazza (Eds.), Death and trauma: The traumatology of grieving (pp. 231-247). Washington, DC: Taylor and Francis. Sprang, G. (2001). The use of eye movement desensitization and reprocessing (EMDR) in the treatment of traumatic stress and complicated mourning: Psychological and behavioral outcomes.  Research on Social Work Practice,  11(3), 300-320. doi:10.1177/104973150101100302 Stuckless, N. (1996). The influence of anger, perceived injustice, revenge, and time on the quality of life of survivor‐victims. ProQuest Information & Learning. Dissertation Abstracts International: Section B: The Sciences and Engineering, 58, 1996‐9500 2‐008. Utzon-Frank, N., Breinegaard, N., Bertelsen, M., Borritz, M., Eller, N. H., Nordentoft, M., . . . Bonde, J. P. (2014). Occurrence of delayed-onset post-traumatic stress disorder: A

22

General Introduction

systematic review and meta-analysis of prospective studies. Scandinavian Journal of Work, Environment & Health, 40(3), 215-229. doi:10.5271/sjweh.3420 Valdimarsdóttir, U., Helgason, Á. R., Fürst, C., Adolfsson, J., & Steineck, G. (2004). Awareness of husband’s impending death from cancer and long-term anxiety in widowhood: A nationwide follow-up. Palliative Medicine, 18(5), 432-443. doi:10.1191/0269216304pm891oa Van Ameringen, M., Mancini, C., Patterson, B., & Boyle, M. H. (2008). Post-traumatic stress disorder in canada. CNS Neuroscience & Therapeutics, 14(3), 171-181. doi:10.1111/j.17555949.2008.00049.x Wagner, B., Knaevelsrud, C., & Maercker, A. (2006). Internet-based cognitive-behavioral therapy for complicated grief: A randomized controlled trial. Death Studies, 30(5), 429-453. doi:10.1080/07481180600614385 Weinberg, N. (1994). Self-blame, other blame, and desire for revenge: Factors in recovery from bereavement. Death Studies, 18(6), 583-593. doi:10.1080/07481189408252702 Zinzow, H. M., Rheingold, A. A., Byczkiewicz, M., Saunders, B. E., & Kilpatrick, D. G. (2011). Examining posttraumatic stress symptoms in a national sample of homicide survivors: Prevalence and comparison to other violence victims. Journal of Traumatic Stress, 24(6), 743-746. doi:10.1002/jts.20692 Zisook, S., & Shear, K. (2009). Grief and bereavement: What psychiatrists need to know. World Psychiatry, 8(2), 67–74 Zisook, S., Simon, N. M., Reynolds, C. F., Pies, R., Lebowitz, B., Young, I. T., … Shear, M. K. (2010). Bereavement, Complicated grief, and DSM, Part 2: Complicated grief. American society of clinical psychopharmacology, 71(8), 1097-1098

23

Chapter 2 Psychopathology among homicidally bereaved individuals: A systematic review

Van Denderen, M., de Keijser, J., Kleen, M., & Boelen, P. A. (2015). Psychopathology among homicidally bereaved individuals: A systematic review. Trauma, Violence, & Abuse, 16(1), 70-80. doi:10.1177/1524838013515757

Chapter 2

Abstract In the literature on bereavement, claims are made that homicidal loss is associated with post-traumatic stress reactions, depression, and other severe mental health problems. It is surprising that only a few studies have investigated the nature and prevalence of emotional symptoms following homicidal bereavement and a reference to systematic, empirical research is seldom provided. This paper reviews the available literature to investigate whether these claims have empirical evidence. Three databases were searched to identify relevant studies. This approach was supplemented with a bibliography search. Eligible studies included English-language peer-reviewed articles that assessed psychopathology in the homicidally bereaved, as defined in the Diagnostic and Statistical Manual of Mental Disorders (DSM) (American Psychiatric Association, 2000). Of the 360 potentially relevant articles, eight studies (thirteen references) met predefined inclusion criteria. Homicide-related psychopathology among the bereaved assessed in these studies includes Post-Traumatic Stress Disorder (PTSD), depression, Complicated Grief, and substance abuse. Prevalence of lifetime homicide-related PTSD varied from 19.1% to 71% across studies. Current PTSD varied between 5.2% and 6%. The reviewed literature was inconclusive regarding the course of symptoms over time and the severity of psychopathology among the homicidally bereaved, compared to individuals bereaved by other causes of death. A comparison of the nature and prevalence of psychopathology between studies was complicated by unequal sample sizes and type, recruitment strategy, study design and time since loss. Limitations of the included studies are discussed, as well as implications for clinical practice, policy, and future research. Keywords homicide, murder, bereaved, grief, review, prevalence

Key points of the research review • This review suggests that homicidal bereavement is associated with an increased risk of adverse mental health outcomes. Results are not easily generalizable, because of differences between studies in ethnicity, sex and relationship with the deceased. • Homicide-related psychopathology assessed includes PTSD, depression, Complicated Grief and substance abuse. • Prevalence of lifetime homicide-related PTSD varied from 19.1% to 71% across studies. Current PTSD varied between 5.2% and 6%. • The reviewed literature is inconclusive regarding the course of psychopathology over time and the severity of psychopathology among the homicidally bereaved, compared to non-victims, victims of interpersonal violence or individuals bereaved by suicide or accident.

26

Psychopathology among homicidally bereaved individuals

• Included studies differed greatly in sample size and type, recruitment strategy, study design, time since loss and the relationship of the bereaved and the victim. Instead of a generalization across studies, the studies are compared and contrasted to each other on an individual level. • When reviewing the literature, at least eleven synonyms were found in the literature to describe a population of individuals bereaved by homicide, including loved one, surviving family members, co-victims of homicide and secondary victim.

Introduction Murder and manslaughter are among the most severe criminal acts. Surviving family members, partners, and friends of the victim, have to deal with the emotional and psychological consequences of losing someone. In addition to the loss itself, these individuals must cope with the specific cause of the death, characterised by the violent and intentional nature of the act, involvement with the criminal justice system, media attention, and investigation by the police (Amick-McMullan, Kilpatrick, Veronen, & Smith, 1989; Kaltman & Bonanno, 2003; Parkes, 1993; Riches & Dawson, 1998; Rynearson, 1994; Sprang, 2001). The loss of a loved one and additionally the violent nature of the death may be associated with subsequent mental health problems. Some studies firmly state that homicidal bereavement is related to posttraumatic stress reactions, depression or other severe mental health problems: “there is ample evidence that loss due to homicide is associated with post-traumatic stress phenomena” (Freeman, Shaffer, & Smith, 1996; p. 337), “the few studies on the consequences of homicide for surviving family members are consistent in showing that homicide survivors, as the literature terms them, are traumatized greatly by the murder of a loved one” (Thompson, Norris, & Ruback, 1998, p. 224), and “Rather, losing a loved one to homicide is a chronic, never-ending trauma (…) Not only is the congruence in reported symptoms across studies striking, but so too is the gravity of symptomatology” (Thompson, 1995, pp. 3-4). Considering the assumptions made in the literature, it is surprising that few studies have investigated the nature and prevalence of emotional symptoms following homicidal bereavement and reference to systematic, empirical research is seldom provided. For the bereaved community, it may be of considerable importance whether or not such claims are based upon empirical evidence. If these claims are not based upon such evidence, they may serve as a reinforcement for pathological behavior and catastrophizing cognitions. In theory, this may increase the subjective psychological pain or pathological behavior and may thereby even prolong the time in which the bereaved suffer from psychological complaints. If these claims do indeed reflect objective reality, this information can serve as a base for the development of clinical interventions for this population, such as developing individualized or more specialized treatments. Recently, a review was published regarding Post-Traumatic Stress Disorder (PTSD), depression and Complicated Grief (CG) after sudden and violent losses, including wars, natural disasters, airline disasters, accidents, suicide, homicide and natural death (Kristensen, Weisæth, & Heir, 2012). However, a

27

Chapter 2 systematic review specifically on the psychopathology following homicidal loss has not yet been conducted. Consequently, it remains largely unclear if homicidal bereavement is associated with psychopathology. The major aim of this article is therefore to critically review the available evidence regarding psychopathology experienced by the homicidally bereaved. The main questions which will be addressed are: which symptoms are reported, what is the prevalence rate and course of symptoms over time?1 By defining the search terms, we did not specify any distinct disorder beforehand but chose to search as broadly as possible for any psychopathology classified in the DSM-IV2. Factors not classified as psychopathology in the DSM-IV, such as perceived social support, self-esteem and mental distress were not included. This choice was made to avoid pathologizing normal processes, such as the experience of mental distress after the loss of a loved one. An exception was made for dysfunctional grief, also termed complicated or prolonged grief (see for a review Boelen & Prigerson, 2012). While sharing some symptoms with disorders such as depression and PTSD, there is increasing recognition that symptoms of dysfunctional grief (including yearning, difficulty accepting the loss, avoidance of reminders) are distinct from these syndromes and are predictive of health impairments (Newson, Boelen, Hek, Hofman, & Tiemeier, 2011). While some symptoms of CG overlap with symptoms of PTSD, the separation distress component is unique for persons with CG. Moreover, individuals with PTSD or CG can both experience anxiety, but the form is different; PTSD typically includes threat-related anxiety, whereas CG includes mainly separation anxiety (Lichtenhal, Cruess, & Prigerson, 2004). Although depressive and CG symptoms are frequently comorbid, some symptoms are different. Preoccupation with thoughts of the deceased can be a symptom of CG but not of depression (Prigerson et al., 1995). Furthermore, CG and depression can occur independently. One study found that 46% of individuals diagnosed with CG did not meet criteria for major depressive disorder (Lichtenhal et al., 2004). Accordingly, a number of researchers propose to include a new CG diagnosis in DSM. Awaiting this discussion and possible inclusion in DSM, CG was included in our search.

Method Sources Three topics structured the search terms entered in the databases: (1) cause of death, (2) the bereaved, and (3) prevalence of homicide-related psychopathology. This resulted in the following search terms: homicide or murder or manslaughter AND *surviv* or *bereav* or grie* AND psychopathology or epidemiology or prevalence or health. The electronic databases Psychinfo, Medline and Cochrane were searched in October 2012. English articles published until this point in time were taken into account. If the databases or RefWorks did not provide an online version of an article, it was searched utilizing Google.com. The following inclusion- and exclusion criteria were applied. The main focus of this review is psychopathology among the homicidally bereaved individuals. When studies compared psychopathology among the homicidally bereaved with the bereaved of natural or other violent loss, these results are mentioned in a subparagraph. 2 Some studies date from 1991 and refer to criteria in the DSM-III. References to disorders in the DSM-III are included in this search as well. 1

28

Psychopathology among homicidally bereaved individuals

Inclusion Criteria: 1. The cause of death was homicide; 2. The bereaved were identified as partners, parents, children, brothers, sisters, grandparents, uncles, aunts, nieces, nephews and friends. Step- and adoptive family were also included; 3. Peer-reviewed studies, containing empirical, quantitative information about homicide-related psychopathology among the bereaved, using structured assessment methods; validated questionnaires and/or structured interviews following DSM-III or DSM-IV criteria. Exclusion Criteria: 1. Causes of death other than homicide; 2. Studies regarding homicide committed in the context of collective violence against Groups (e.g., war victims). Because collective violence may differ from individual violence on the psychological impact on the bereaved, studies examining violence against groups was excluded. Search Strategy The search strategy consisted of two steps. First, using the aforementioned search terms, articles that met the inclusion criteria were selected from the databases. Second, relevance was based upon close reading of the abstract and article. The selection of studies was extended by screening reference lists of the included studies. The search was done by two researchers (MD, MK). The researchers independently classified the studies based upon the above mentioned steps. Articles found by only one of the two researchers were discussed with a third independent researcher (JK).

Results Search Results The search resulted in 360 articles. Eight studies (thirteen references) met inclusion criteria. Two additional references were included by screening reference lists of the original eight studies. These were not found by entering the search terms in the databases, probably because the title and abstract included several of the search terms, but not all3. Table 1 lists the eight studies with information about their sample sizes, study population, sample type, and recruitment strategy, study design, comparison groups, average time since homicide, method, and quality of the study, outcome measure, social or demographic risk factors4, and main results. Three topics structured the search terms entered in the databases; cause of death, the bereaved, and prevalence of homicide-related psychopathology. Every topic included several synonyms. Eligible studies had to include at least one synonym from every topic. 4 Risk factors that influence the chance of psychopathology are mentioned. Risk factors that influence the likelihood to experience a homicide or seek help are not mentioned. Furthermore, (significant) risk factors were only mentioned when they applied to homicidally bereaved. Risk factors which applied to a group of bereaved, for example homicide, suicide and accident together, are not included. 3

29

30 National Cross sectional 16.6 years Interview, using PTSD representative Alcohol related DSM III criteria sample, vehicle A1 recruited by a accidents random digit (n = 91) dialing telephone survey to identify surviving family members and close friends.

Outcome measure and Risk factors

No significant difference between both groups or between time and PTSD were found.

Homicide bereaved: 19.1% met all three criteria for lifetime PTSD and 5.2% met current PTSD (alcohol related vehicle accidents respectively 27.5% and 4.4%)

Results

a

A1 are quantitative studies, which report percentages regarding psychopathology. A2 are quantitative studies with report mean scores, but do not give a percentage of participants which met criteria for a diagnose. A3 are quantitative studies which compare psychopathology among homicidally bereaved with other bereaved, but without mentioning actual numbers. b Family members is not further being defined by the authors. c Per ethnicity: Black: 29.6%, white: 66.1%, native American: 2.6%, other: 0.9%, Hispanic: 0%







Amick- 12.500/ 115 Adult McMullan (0.9%) family- Kilpatrick membersb Resnick and close (1991) friends Male: 2.2% Female: 67.8% Ethnically diversc

Authors N total / Population Sample Type and Study Design and Average time Method and N homicide Recruitment Comparison Group since homicide Qualitya (%) Strategy

Table 1. Description of selected studies.

Chapter 2

Outcome measure and Risk factors

18.5% screened positive for PTSD. 53.7% scored positive for at least mild depression. 54.5% screened positive for CG, 6 month post loss or more. > 6 month post loss; nearly all PTSD positive cases screened also positive for mild level depression and CG. < 2 years post loss; participants scored higher on PTSD and anxiety, than > 2 years. Time since loss was not associated with depression or complicated grief. Older people have lower PTSD rates. Lower income and frequent contact with the deceased relates to higher CG scores. Significant decrease in depression and CG scores over the 6-month study period, but not in PTSD symptoms.

Large support system is associated to lower levels of CG. More actual negative relationships is related to higher levels of PTSD and CG. Grief specific support is associated with reduced PTSD and depression.

PTSD: M = 36.59, SD = 15.33 Depression: M = 15.43, SD = 11.27 CG: M = 79.61, SD = 24.46

Results

d

47 participants of the 54 grievers completed T2. Only data from study completers (T1 and T2) were included in this study. The distribution male/female and relation with the deceased between the 54 participants and 47 participants differs only slightly.

Burke 54 / 54 Adults; Community Cross sectional 1.75 years PTSD CheckList- PTSD Neimeyer (100%) parents: based sample, Non Civilian Version Depression McDevitt- 63% African Complicated Murphy siblings: American adults Beck Depression Grief (2010) 13% (19-71), Inventory II Extended recruited by a Size of family: religious Inventory of support 13% organization Complicated Grief- system spouse: that offers Revised Number of 9.3% assistance to A2 negative Remaining survivors of relations participant: homicide in a Grief McDevitt- 1.9% large city in the A1 specific Murphy Mid-South of support Neimeyer Male: 11.1% US. Burke Female: Age Williams 88.9% Income Lawson Frequency (2012) Ethnically of contact homogenous Williams Burke 47/47d Longitudinal 6 month A2 McDevitt- (100%) follow up Murphy Neimeyer (2012)

Authors N total / Population Sample Type and Study Design and Average time Method and N homicide Recruitment Comparison Group since homicide Quality (%) Strategy

Table 1. (continued)

Psychopathology among homicidally bereaved individuals

31

32 Community based sample, recruited by a review of court cases prior to 1988 in South Carolina and family members of victims from cases resulting in conviction and incarceration.

Cross sectional 3 yearsg Physical assault Sexual assault

Structured PTSD telephone interview A1

Outcome measure and Risk factors 71% report lifetime PTSDh No significant differences between the groups were found.

Results

f

e

The total number of participants in this study is 251. The sample consists of direct victims and co-victims of crime. Co-victims of homicide are a subsample of the total number of co-victims (120) Therefore, 120 is chosen as the total n. Age of the sample is not specified. g This year relates to the average time since different types of crime, it is not specified to homicide. h This percentage includes pre-homicide PTSD. Homicide-related PTSD could therefore be considered lower than 71%. i This percentage concerns the whole sample, including crime victims. The percentage male/female among bereaved of homicide could be different. j Per ethnicity: Caucasian: 62.5%, African American: 35.5%. This percentage also concerns the whole sample, including crime victims.

Adolescents Freedy 120e / 62 Resnick (52%) and adults; Kilpatrick familyf Dansky Tidwell Male: 36.7% (1994) Female: 63.3%i Ethnically diversj

Authors N total / Population Sample Type and Study Design and Average time Method and N homicide Recruitment Comparison Group since homicide Quality (%) Strategy

Table 1. (continued)

Chapter 2

k

Community Cross sectional 5.4 months based sample, Healthy siblings of classmates homicide victims aged 19 or under, identified by homicide reports of the New York City police department

Per ethnicity: African American or Latino (black: 67%, other: 33%).

Ethnically diversk

Freeman 15 / 15 African Shaffer (100%) American or Smith Latino (1996) siblings, age 7-18 years (M = 14.5), who lived at home with the non- caregiving victim Male: 53% Female: 47%

Outcome measure and Risk factors

Diagnostic Depression interview PTSD schedule for children (DISC) A1

Authors N total / Population Sample Type and Study Design and Average time Method and N homicide Recruitment Comparison Group since homicide Quality (%) Strategy

Table 1. (continued)

80% of the bereaved developed a disorder, compared to 10% of the control group. Most common were comorbid depressive, PTSD and anxiety disorders.

Results

Psychopathology among homicidally bereaved individuals

33

34 Mothers are more likely to develop PTSD than fathers.



l The three studies of Murphy et al. are based on one dataset and therefore grouped together. Authors categorized by article are: Murphy, Braun, Tillery, Cain, Johnson & Beaton, 1999; Murphy, Johnson, Chung & Beaton, 2003a; Murphy, Johnson, Wu, Fan & Lohan, 2003b. m This percentage concerns the whole sample, including parents whose child died through suicide or accident. The percentage male/female among bereaved of homicide could be different. n Per ethnicity: Caucasian: 86%, other: 14%. This percentage concerns the whole sample.

Twice as many parents whose children were murdered met PTSD criteria two years post-loss compared to the control group.

60% of the mothers and 40% of the fathers met PTSD criteria 4 months post death. PTSD prevalence among homicide bereaved is significantly higher than bereaved of suicide or accident.

Results

The hypothesis that homicide bereaved report a slower reduction of PTSD complaints in a 5 year time-span than bereaved of suicide or accidents is not confirmed.

Outcome measure and Risk factors

Murphy et 171/17 Adult; Community Longitudinal 4 months Traumatic PTSD al., (1999; (10%) parents based sample, Suicide Follow-up: 1, Experience 2003a; age 32 – 61 Accident 2 and 5 years Scale (TES) Gender 2003b)l Male: 34.5% (M = 45 years) A1 Female: Recruited by a review of 65.5%m official death Ethnically records in diversn Washington state and Oregon, U.S.

Authors N total / Population Sample Type and Study Design and Average time Method and N homicide Recruitment Comparison Group since homicide Quality (%) Strategy

Table 1. (continued)

Chapter 2

Outcome measure and Risk factors

Treatment Cross sectional 6-7 months RIES PTSD seeking sample Adults who A3 The Support refused Project for treatment Unnatural Dying was initiated in Seattle to offer support and early intervention for bereaved of homicide.

p

o

This percentage concerns the whole sample, including bereaved of vehicular homicide. Per ethnicity: Caucasian: 47%, African American: 33%, Hispanic: 16%, Native American: 1%, Asian American: 1%. q Sex and ethnicity are not mentioned.

Rynearson 237 / 32 Adult family (1995) membersq

Rheingold 3614 / 333 Young National Cross sectional Not mentioned Modified version PTSD Zinzow (9.2%) adults; representative Vehicular of the National Depression Hawkins family sample of 1753 homicide (7% Women’s Study Drug use Saunders members young adults, of 3614) PTSD module. Alcohol use Kilpatrick and friends age 12-17, (2012) participating in NWS Depression Male: 28% the 2005 Module Female: National Survey A1 72%o of Adolescents Ethnically diversp

Authors N total / Population Sample Type and Study Design and Average time Method and N homicide Recruitment Comparison Group since homicide Quality (%) Strategy

Table 1. (continued)

Treatment seeking adults score higher on the RIES than non-treatment seeking adults; intrusion (29.4 ± 6.3 treatment / 15.3 ± 8.4 non-treatment), avoidance (24.2 ± 7.6 treatment / 9.8 ± 8 non-treatment).

Prevalence of PTSD or depression did not differ per ethnicity.

Prevalence rates of homicide bereaved on all four outcome measures were lower than among bereaved of vehicular homicide.

Current PTSD: 6% Past 6 month Depression: 8% Drug Use: 14% Alcohol Use: 10%

Results

Psychopathology among homicidally bereaved individuals

35

36 National Cross Not mentioned Structural interview Past six representative sectionals items that follow month PTSD sample, Victims of other the DSM-IV-TR follow-up young personal criteria adults from the violence A1 original (n = 653) National Survey of Adolescents.

Outcome measure and Risk factors

Homicide bereaved were significantly more likely than non-victims to report past year PTSD symptoms. (OR = 1.88), were at greater risk for past year depression (OR = 1.64) and drug abuse/dependence (OR = 1.77).

Homicide bereaved were more likely than victims of other violence to meet criteria for 2 or 3 symptom clusters.

15% of homicide bereaved met criteria for all 3 PTSD symptom clusters, compared to 8% among violence victims.

Results

r

The authors assume that the number of 268 participants represents homicide survivors, as well as vehicular homicide. In their article, Zinzow et al. ask participants whether their family member or close friend is murdered or killed by a drunk driver. The results of this question, in terms of number of homicide survivors and number of vehicular homicide is not mentioned. In their article of 2009, they do report the number of homicide survivors (n = 169) and vehicular homicide (n = 99). This adds up to n = 268. s While the sample type is a follow-up of young adults from the original National Survey of Adolescents, the study examines PTSD symptoms at one point in time and is therefore crosssectional. t This percentage concerns the whole sample. u Per ethnicity: Caucasian: 67%, African American: 17%, Hispanic: 9%. This percentage concerns the whole sample. v Per ethnicity: Caucasian: 41%, African American: 41%, Hispanic: 9%, Native American: 2%, Asian American: 4%.

Zinzow 1753 / 169 Male: 41% Vehicular A3 PTSD Rheingold (9.6%) Female: homicide Depression Hawkins 59% (n = 99) Drug use Saunders and non-victims Alcohol use Kilpatrick Ethnically (n = 1485) (2009) diversv

Ethnically diversu

Zinzow 1753 / 268r Close friend: Rheingold (15.3%) 64% Byczkie- Immediate wicz Family Sauders member: Kilpatrick 11% (2011) Male: 55% Female: 45%t

Authors N total / Population Sample Type and Study Design and Average time Method and N homicide Recruitment Comparison Group since homicide Quality (%) Strategy

Table 1. (continued)

Chapter 2

Psychopathology among homicidally bereaved individuals

Homicide-related psychopathology assessed across studies includes PTSD, depression, CG, and substance abuse. As can be seen in Table 1, the included studies differed greatly in terms of sample size and type, such as national representative sample or treatment seeking sample, recruitment strategy, study design, time since loss and method of assessing psychopathology. Direct comparisons between studies are therefore not possible and would lead to an oversimplification of the data presented. Instead of a generalization across studies, the studies are compared with each other on an individual level. Before the results will be presented, the methodological differences between studies are described in more detail. Methodological Differences Between Studies Mean time since loss ranged from 4 months to 16.6 years across studies. The number of participants in individual studies ranged from 15 to 333. Seven studies utilised a cross-sectional design and one was longitudinal. Regarding the samples, three national representative samples were used as was one treatment-seeking sample and four community based samples. Six studies used a mixed sample of family members, friends and spouses and two described a subpopulation of parents or siblings. Seven studies used control groups; non-victims, victims of interpersonal violence or individuals bereaved of other violent loss such as suicide, vehicular homicide or accidents. Psychopathology among adults was assessed, as well as among children. Six samples were ethnically diverse; the majority of the participants were Caucasian or African American, one sample was homogenous (African American) and one sample did not report ethnicity. To assess psychopathology, five studies used interviews and three used questionnaires. PTSD Prevalence The prevalence of current PTSD ranged from 5.2% to 6%. Lifetime prevalence of PTSD ranged from 19.1% to 71% across studies. The study with the highest prevalence of PTSD was conducted in a community based sample of bereaved adolescents and adult family members, recruited by a review of court cases prior to 1988 in South Carolina, U.S. In this sample, 71% reported lifetime PTSD (Freedy, Resnick, Kilpatrick, Dansky, & Tidwell, 1994). Lifetime PTSD could include symptoms of PTSD prior to the homicide, caused by another trauma. In the study with the longest time span since loss, 16.6 years post loss, a prevalence of 19.1% was found among an adult national representative sample, recruited by a random telephone survey as part of an epidemiological study (Amick-McMullan, Kilpatrick, & Resnick, 1991). The shortest time span measured was four months post loss (Murphy et al., 1999). In a community based sample of 17 predominantly Caucasian parents, PTSD prevalence was 60% among mothers and 40% among fathers. When measuring PTSD four months post loss, only individuals with acute and chronic, but not those with delayed PTSD were included. The percentage found by Murphy et al. (1999) could therefore be an underestimation of PTSD prevalence, because individuals with delayed onset of PTSD were not included. Burke, Neimeyer, and McDevitt-Murphy (2010) assessed PTSD 1.75 years post loss

37

Chapter 2 using the PTSD CheckList-Civilian Version in a community based sample of 54 African American parents, siblings, adult children, and spouses, recruited by a religious organisation that offers assistance to homicidally bereaved individuals. The average scores of the participants (36.59) is below the cut off score of 50 for caseness of PTSD as proposed by Weathers, Litz, Herman, Huska, and Keane (1993). In the same sample, a prevalence of 18.5% for PTSD was found (McDevitt-Murphy, Neimeyer, Burke, & Williams, 2012). Among a national representative sample of 333 ethnically diverse family members, close friends, and other relatives, aged 12-17 years, 6% met criteria for current PTSD (Rheingold, Zinzow, Hawkins, Saunders, & Kilpatrick, 2012). Time since loss was not reported in this study. In a national representative sample of 268 family members and friends, 15% met all three PTSD clusters (Zinzow, Rheingold, Byczkiewicz, Sauders, & Kilpatrick, 2011). The only study with a treatment seeking sample found treatment seeking individuals to score higher on the Impact of Event Scale-Revised (IES-R) than non-treatment seeking individuals (Rynearson, 1995). Prevalence of Complicated Grief, Depression, and Substance Abuse Using a community based sample of 54 bereaved African American parents, siblings, adult children, and spouses, Burke et al. (2010) assessed CG 1.75 years post loss using the Inventory of Complicated Grief-Revised (Prigerson & Jacobs, 2001). The average scores of the participants (79.61) is below the cut off score of 90 for CG as proposed in 2003 (Boelen, van den Bout, de Keijser, & Hoijtink, 2003). Burke et al., (2010) found a mean score of 15.43 on the Beck Depression Inventory II in the same sample (Beck, Steer, & Brown, 1996). Compared to cut-off scores established for psychiatric outpatients (Beck, Steer, & Garbin, 1988), this score is higher than a minimal depression (cut off 10.9), but lower than a mild depression (cut off 18.7). In the same sample, 53.7% scored positive for at least mild depression and 54.5% for CG (McDevitt-Murphy et al., 2012). Furthermore, a pattern of comorbidity was found; six months or more post loss, nearly all PTSD positive cases also screened positive for mild level depression and CG (McDevitt-Murphy et al., 2012). In a national representative sample of young homicidally bereaved (12-17 years), a prevalence of 8% was found for past year depression (Rheingold et al., 2012). Mean time post-loss was not reported in this study. In a community based sample of 15 African American or Latino siblings, aged 7-18 years, 80% had developed a psychological disorder 5.4 months post-loss, compared to 10% in the healthy control group. Most common were comorbid depression, PTSD, and anxiety disorder (Freeman et al., 1996). Very few studies have reported data on substance abuse. In a national representative sample of young adults (12-17 years), a prevalence of 10% past year alcohol- and 14% past year drug use was found (Rheingold et al., 2012). Course of Symptoms over Time In a cross-sectional study with family members and close friends of the victim, no relation was found between PTSD symptom severity and time since loss, on

38

Psychopathology among homicidally bereaved individuals

average 16.6 years post loss (Amick-McMullan et al., 1991). In a longitudinal study in a community based sample of African American parents, spouses, siblings, and other family members on average 1.67 years post loss, a significant decrease was found in symptom-levels of depression and complicated grief over the 6-month study period, but not in PTSD symptoms (Williams et al., 2012). In the same sample, participants scored higher on PTSD and anxiety within the first two years post loss, than later in time. Time since loss was not associated with symptom-levels of CG or depression (McDevitt-Murphy et al., 2012). Homicidal Bereavement Compared to Other Violent Loss Six studies compared psychopathology levels between homicidally bereaved individuals, individuals confronted with other violent losses, victims of violence, and non-victims. Almost 17 years post loss, no significant difference was found in PTSD scores between the homicidally bereaved and individuals bereaved of alcohol related vehicle accidents in a national representative sample (Amick-McMullan et al., 1991). In a community based sample three years post loss, no significant difference was found in PTSD scores between the homicidally bereaved and victims of physical or sexual assault (Freedy et al., 1994). In a longitudinal study, twice as many parents whose children were murdered met PTSD criteria two years post loss compared with parents who lost their child due to accident or suicide (Murphy et al., 1999). Five years post-loss, no difference in the number of individuals meeting PTSD criteria was found between individuals bereaved due to homicide, accident or suicide (Murphy, Johnson, Chung, & Beaton, 2003a). In a national representative sample, young homicidally bereaved individuals (1217 years) were compared with victims of vehicular homicide, victims of personal violence, and non-victims. Homicidally bereaved victims were significantly more likely than non-homicidally bereaved victims to report past year PTSD symptoms (OR=1.88) and were at greater risk for past year depression (OR=1.64) and drug abuse/dependence (OR=1.77) (Zinzow, Rheingold, Hawkins, Saunders, & Kilpatrick, 2009). Prevalence rates of PTSD, depression, and drug and alcohol abuse were all lower for the homicidally bereaved than for individuals bereaved due to vehicular homicide (Rheingold et al., 2012). The homicidally bereaved were significantly more likely than victims of other violence to meet criteria for two or three PTSD clusters (Zinzow et al., 2011).

Discussion This article reviewed evidence regarding the nature, prevalence, and course of psychopathology among homicidally bereaved individuals. Eight studies (thirteen references) were found, describing prevalence-rates of PTSD, depression, CG, and substance abuse in homicidally bereaved individuals. Prevalence of lifetime homiciderelated PTSD ranged from 19.1% to 71% across studies. Current PTSD ranged from 5.2% to 6%. Whereas a cross-sectional study found no relation between PTSD

39

Chapter 2 symptoms and time since loss, a longitudinal study found a significant decrease in levels of depression and CG over the 6-month study period, but did not find changes in PTSD symptoms. The reviewed literature is inconclusive regarding the severity of psychopathology among the homicidally bereaved compared to non-homicidally bereaved victims, victims of interpersonal violence or individuals bereaved by suicide, accident or vehicular homicide. Based on the findings in this review, it cannot be stated what form of psychopathology is most experienced by homicidally bereaved individuals. Included studies suggest that PTSD is most frequent. Yet, this conclusion may represent an overestimation since PTSD was measured in all eight studies, whilst depression, CG, and substance abuse were measured in only four, two and one study, respectively. As pointed out in the introduction, many studies suggest that homicidally bereaved individuals represent a highly traumatized population. There seems to exist a large discrepancy between estimated rates of psychopathology related to homicidal loss, as suggested in scientific literature, and the low number of studies devoted to this problem. This review suggests that homicidal loss is generally associated with increased risk of adverse mental health outcomes, including PTSD and CG. The extent to which these results can be generalized to the population of homicidally bereaved individuals is limited, due to diversity in sex, age, and ethnicity across studies. The samples in six studies were predominantly composed of women, which is of specific interest given the literature showing that women are more likely to experience CG (Kersting, Brähler, Glaesmer, & Wagner, 2011) and PTSD (Komarovskaya, Loper, Warren, & Jackson, 2011). The prevalence of psychopathology among men could therefore be lower and should be further examined. In one ethnically homogenous sample, a relatively high prevalence of CG and mild depression among African Americans adults was found. Women with an African American background, living in a large city in the Mid-South of the United States were especially at risk for negative mental health outcomes if they had a low income, frequent contact with the victim before the homicide, and reported more negative relationships with others. Older African American women, with large support systems and grief specific support had a lower risk of PTSD and CG. In one ethnically diverse sample of Americans children (12-17 years), no difference in prevalence of PTSD or depression was found between Caucasian and African Americans. Replication of this study would imply a better generalizability of the prevalence rates of these disorders found among young Caucasian or African American populations living in the United States. Other studies did not compare different ethnicities or describe the ethnic distribution for the whole sample, including non-homicidally bereaved individuals. Based on the differences in study populations, results regarding mental health outcomes after loss by homicide cannot as such be generalized to other homicidally bereaved populations. Limitations The current review has a number of limitations, including those related to the included studies. The study with the highest prevalence rate made no distinction between pre-homicide and post-homicide related psychopathology (Freedy, et al.,

40

Psychopathology among homicidally bereaved individuals

1994). This limitation is common in bereavement research, but has to be taken into account when interpreting the results. Part of the psychopathology presented here, such as lifetime PTSD, could have developed following other, pre-homicide, events. Therefore, it is unknown whether the psychopathology mentioned by Freedy et al. is homicide-related. Several studies did not distinguish between psychopathology in individuals bereaved by homicidal death and individuals bereaved by other causes of deaths. For example, individuals bereaved by homicide, suicide, and accidents are aggregated in a number of studies (Murphy et al., 1999, 2003a, 2003b). Yet, it cannot be ruled out that the various types of violent deaths result in meaningful differences in the types of symptoms that are elicited in the bereaved. In addition, few studies have compared the homicidally bereaved to individuals bereaved from natural losses. In addition to the limitations of the particular studies, several limitations of this review have to be taken into account when considering the conclusions. First, this review was restricted to research from English speaking countries. It is possible that the bereaved in non-Western cultures deal differently with homicide and bereavement and therefore experience different psychopathology. Second, only studies regarding individual violence were taken into account, excluding homicide committed in a context of collective violence against groups. Future research should examine the consequences of collective violence for bereaved individuals, such as war victims. Recommendations Our review suggests important considerations for further research. To enhance knowledge regarding psychopathology experienced by homicidally bereaved individuals in comparison to bereavement from natural causes, future research should not only differentiate between various causes of violent death but also compare bereavement following homicidal loss versus natural deaths. To improve insight in the course of psychopathology over time, more longitudinal studies are needed. Ideally, an international cooperation of researchers working in this field should use the same time interval and methodology, making comparisons between studies and subpopulations possible. As in any field of enquiry, researchers should use validated instruments to assess psychopathology. This is especially important in a research field like this, where every one can imagine the emotional burden of homicidal loss and claims about the degree to which individuals must be suffering are easily made. During the search, five potentially eligible studies had to be excluded because they used questionnaires which do not provide a diagnosis included in the DSM-IV but instead used assessments, such as the Impact of Event Scale or Symptom Checklist-90 (Amick-McMullan et al., 1989; Mezey, Evans, & Hobdell, 2002; Range & Niss, 1990; Rynearson & McCreery, 1993; Thompson et al., 1998). For future studies, researchers are advised to use structured clinical interviews for DSM-IV, or, from 2013 onwards, DSM-5, in order to assess psychopathology. During this search, five dissertations regarding psychopathology of the homicidally bereaved were found (Gerber, 1995; Sharpe, 2007; Stiehler, 1995, Stuckless, 1998;

41

Chapter 2 Thompson, 1995). These studies could have provided relevant information, but were excluded because they were not published in peer-reviewed journals. Authors are therefore advised to publish their findings in peer-reviewed journals, making results available to a larger public. When reviewing the literature, many synonyms used to describe a population of homicidally bereaved were found. To cover as many synonyms, we used the words *surviv* or *bereav* or grie*, thereby including terms such as family-survivor(s), survivor-victim(s) and grieving family(ies). During the search, at least eleven synonyms for individuals bereaved of homicide were found; bereaved; victim; secondary victim (Peay, 1997); loved one (Zinzow et al., 2009); homicide victim(s); homicide survivor(s) (Zinzow et al., 2011); co-victims of homicide (Armour, 2002); surviving family members (Amick-McMullan et al., 1991); homicidally bereaved individuals (McDevitt-Murphy et al., 2012); families of homicide victims (Horne, 2003) and family survivors of homicide victims (Amick McMullan et al., 1989). Use of many synonyms makes it difficult to find relevant studies. Moreover, some terms lack specificity and clarity and could easily be misinterpreted. The first four (bereaved, victim, secondary victim, loved one) do not cover the death cause. The next four (homicide victim(s), homicide survivor(s), co-victims of homicide, surviving family members) can refer to the victim of the homicide itself, and not to the individuals left behind, or to the victims who survived a homicide attempt. The last two (families of homicide victims, family survivors of homicide victims) cover only family members and not spouses or friends of the victim. To avoid different interpretations and to facilitate the search of relevant articles, we suggest an agreement on a single term to describe this population and recommend the term homicidally bereaved. In our opinion, this term is the least ambiguous and accounts better for the cause of death, as well as the relationship to the victim.

Implications for Practice, Policy, and Research • As in every research area, researchers need to use valid and reliable outcome measures. This is especially important in a research field like this, where most can imagine the emotional burden of homicidal loss and statements about how much individuals must be suffering are easily made. Researchers are therefore recommended to use structured clinical interviews for DSM-IV or equivalent instruments to assess disorders. • Studies cannot be easily compared because of different sample types, outcome measures, measuring instruments, sample sizes, average time since loss, and type of relation between the bereaved and victim. Policymakers and researchers should therefore give a nuanced image of homicide-related psychopathology, instead of an oversimplified one. • At least eleven synonyms were found in the literature to describe a population of homicidally bereaved. To avoid different interpretations of a synonym and to make it easier to find relevant articles, we suggest to use one term to describe

42

Psychopathology among homicidally bereaved individuals

this population and recommend the term homicidally bereaved. Homicidal bereavement can be used to refer to the event. • Research should not only differentiate clearly between the bereaved of different violent causes of death, but also compare psychopathology after a violent loss with natural loss. • Research and knowledge about the nature of psychopathology experienced by the homicidally bereaved can inform decision making regarding psychotherapy. • PTSD does not seem to diminish over time. More follow-up studies are needed to investigate the influence of time on psychopathology. Declaration of Conflicting Interests The author(s) declared no potential conflicts of interests with respect to the authorship and/or publication of this article. Funding The research project ‘Emotional consequences of homicide’ is funded by Victim Support Fund. Acknowledgement Special thanks to Prof. Dr. Peter J. de Jong, Dr. Brian Ostafin and Dr. Martin Brock for their comments.

43

Chapter 2

References American Psychiatric Association (2000). Diagnostic and Statistical Manual of mental disorders (4th Edition, Text Revision). Washington DC: Author Amick-McMullan, A., Kilpatrick, D. G., & Resnick, H. S. (1991). Homicide as a risk factor for PTSD among surviving family members. Behavior Modification, 15(4), 545-559. doi:10.1177/01454455910154005 * Amick-McMullan, A., Kilpatrick, D. G., Veronen, L. J., & Smith, S. (1989). Family survivors of homicide victims: Theoretical perspectives and an exploratory study. Journal of Traumatic Stress, 2(1), 21-35. doi:10.1002/jts.2490020104 Armour, M. P. (2002). Experiences of covictims of homicide: Implications for research and practice. Trauma, Violence, & Abuse, 3(2), 109-124. doi:10.1177/15248380020032002 Beck, A.T., Steer, R.A., & Brown, G. K. (1996). Manual for Beck Depression Inventory-Second Edition. San Antonio, TX: The Psychological Corp Beck, A. T., Steer, R. A., & Garbin, M. G. (1988). Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology Review, 8(1), 77-100. doi:10.1016/0272-7358(88)90050-5 Boelen, P. A., Van den Bout, J., Keijser, J. D., & Hoijtink, H. (2003). Reliability and validity of the Dutch version of the Inventory of Traumatic Grief (ITG). Death Studies, 27(3), 227-247. doi:10.1080/07481180302889 Boelen, P.A. & Prigerson, H.G (2012) Commentary on the Inclusion of Persistent Complex Bereavement-Related Disorder in DSM-V. Death Studies, 36, 771-794. DOI:10.1080/0748 1187.2012.706982 Burke, L. A., Neimeyer, R. A., & McDevitt-Murphy, M. (2010). African American homicide bereavement: aspects of social support that predict complicated grief, PTSD, and depression. Omega, 61(1), 1-24.* Freedy, J. R., Resnick, H. S., Kilpatrick, D. G., Dansky, B. S., & Tidwell, R. P. (1994). The psychological adjustment of recent crime victims in the criminal justice system. Journal of Interpersonal Violence, 9(4), 450-468. doi:10.1177/088626094009004002* Freeman, L. N., Shaffer, D., & Smith, H. (1996). Neglected victims of homicide: The needs of young siblings of murder victims. American Journal of Orthopsychiatry, 66(3), 337-345. doi:10.1037/h0080184* Gerber, F. (1995). The relationship of posttraumatic stress disorder and grief in family survivors of homicide victims. ProQuest Information & Learning. Dissertation Abstracts International: Section B: The Sciences and Engineering, 56 (5-). (1995-95021-100). Horne, C. (2003). Families of Homicide Victims: Service Utilization Patterns of Extra- and Intrafamilial Homicide Survivors. Journal of Family Violence, 18(2), 75-82. Kaltman, S., & Bonanno, G. A. (2003). Trauma and bereavement: Examining the impact of sudden and violent deaths. Journal of Anxiety Disorders, 17(2), 131-147. doi:10.1016/ S0887-6185(02)00184-6 Kersting, A., Brähler, E., Glaesmer, H., & Wagner, B. (2011). Prevalence of complicated grief in a representative population-based sample. Journal of Affective Disorders, 131(1-3), 339-343. doi:10.1016/j.jad.2010.11.032 Komarovskaya, I. A., Loper, A. B., Warren, J., & Jackson, S. (2011). Exploring gender differences in trauma exposure and the emergence of symptoms of PTSD among incarcerated men and women. Journal of Forensic Psychiatry & Psychology, 22(3), 395-410. doi:10.1080/147899 49.2011.572989 Kristensen, P., Weisæth, L., & Heir, T. (2012). Bereavement and mental health after sudden and Note. The references with an * are included in Table 1.

44

Psychopathology among homicidally bereaved individuals

violent losses: A review. Psychiatry: Interpersonal and Biological Processes, 75(1), 76-97. doi:10.1521/psyc.2012.75.1.76 Lichtenhal, W. G., Cruess, D. G., & Prigerson, H. G. (2004). A case for establishing complicated grief as a distinct mental disorder in DSM-V. Clinical Psychology Review, 24(6), 637-662. doi:10.1016/j.cpr.2004.07.002 McDevitt-Murphy, M., Neimeyer, R. A., Burke, L. A., Williams, J. L., & Lawson, K. (2012). The toll of traumatic loss in African Americans bereaved by homicide. Psychological Trauma: Theory, Research, Practice, and Policy, doi:10.1037/a0024911* Mezey, G., Evans, C., & Hobdell, K. (2002). Families of homicide victims: Psychiatric responses and help-seeking. Psychology & Psychotherapy: Theory, Research & Practice, 75(1), 65. Murphy, S. A., Braun, T., Tillery, L., Cain, K. C., Johnson, L. C., & Beaton, R. D. (1999). PTSD among bereaved parents following the violent deaths of their 12- to 28-year-old children: A longitudinal prospective analysis. Journal of Traumatic Stress, 12(2), 273-291. doi:10.1023/A:1024724425597* Murphy, S. A., Johnson, L. C., Chung, I., & Beaton, R. D. (2003a). The prevalence of PTSD following the violent death of a child and predictors of change 5 years later. Journal of Traumatic Stress, 16(1), 17-25. doi:10.1023/A:1022003126168* Murphy, S. A., Johnson, L. C., & Lohan, J. (2002). The aftermath of the violent death of a child: An integration of the assessments of parents’ mental distress and PTSD during the first 5 years of bereavement. Journal of Loss and Trauma, 7(3), 203-222. doi:10.1080/10811440290057620 Murphy, S. A., Johnson, L. C., Wu, L., Fan, J. J., & Lohan, J. (2003b). 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(1), 39-61. doi:10.1080/07481180302871* Newson, R. S., Boelen, P. A., Hek, K., Hofman, A., & Tiemeier, H. (2011). The prevalence and characteristics of complicated grief in older adults. Journal of Affective Disorders, 132(1-2), 231-238. doi:10.1016/j.jad.2011.02.021 Parkes, C. M. (1993). Psychiatric problems following bereavement by murder or manslaughter. British Journal of Psychiatry, 162, 49-54. doi:10.1192/bjp.162.1.49 Peay, J. (1997). Clinicians and inquiries: Demons, drones or demigods? International Review of Psychiatry, 9(2-3), 171-177. doi:10.1080/09540269775367 Prigerson H. G., Frank E., Kasl S. V., Reynolds C. F., Anderson B., Zubenko G. S., Houck P. R., George C. J., & Kupfer D. J. (1995) Complicated grief and bereavement-related depression as distinct disorders: preliminary empirical validation in elderly bereaved spouses. The American Journal of Psychiatry 152(1), 22-30. Prigerson, H. G., & Jacobs, S. C. (2001). Traumatic grief as a distinct disorder: A rationale, consensus criteria, and a preliminary empirical test. In M. S. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of bereavement research: Consequences, coping, and care (pp. 613–645).Washington, DC: American Psychological Association. Range, L. M., & Niss, N. M. (1990). Long-term bereavement from suicide, homicide, accidents, and natural deaths. Death Studies, 14(5), 423-433. doi:10.1080/07481189008252382 Rheingold, A. A., Zinzow, H., Hawkins, A., Saunders, B. E., & Kilpatrick, D. G. (2012). Prevalence and mental health outcomes of homicide survivors in a representative US sample of adolescents: Data from the 2005 National Survey of Adolescents. Journal of Child Psychology and Psychiatry, 53(6), 687-694. doi:10.1111/j.1469-7610.2011.02491.* Riches, G., & Dawson, P. (1998). Spoiled memories: Problems of grief resolution in families bereaved through murder. Mortality, 3(2), 143-159. Rynearson, E. K., & McCreery, J. M. (1993). Bereavement after homicide: A synergism of

45

Chapter 2 trauma and loss. The American Journal of Psychiatry, 150(2), 258-261. Rynearson, T. (1994). Psychotherapy of bereavement after homicide. Journal of Psychotherapy Practice & Research, 3(4), 341-347. Rynearson, E. K. (1995). Bereavement after homicide. A comparison of treatment seekers and refusers. The British Journal of Psychiatry: The Journal of Mental Science, 166(4), 507-510.* Sharpe, T. L. (2007). Coping with family member homicide: The African American experience. ProQuest Information & Learning). Dissertation Abstracts International Section A: Humanities and Social Sciences, 69 (2-). (2008-99151-236). Sprang, G. (2001). The use of eye movement desensitization and reprocessing (EMDR) in the treatment of traumatic stress and complicated mourning: psychological and behavioral outcomes. Research on Social Work Practice, 11(3), 300-320. Stiehler, J. Q. (1995). The aftermath of death: A study of parents of murdered children and other bereaved parents. ProQuest Information & Learning). Dissertation Abstracts International: Section B: The Sciences and Engineering, 56 (6-). (1995-95023-045). Stuckless, N. (1998). The influence of anger, perceived injustice, revenge, and time on the quality of life of survivor-victims. ProQuest Information & Learning). Dissertation Abstracts International: Section B: The Sciences and Engineering, 58 (7-). (1998-95002-008). Thompson, M. P. (1995). System influences on post-homicide beliefs and distress. ProQuest Information & Learning). Dissertation Abstracts International: Section B: The Sciences and Engineering, 56 (9-). (1996-95005-336). Thompson, M. P., Norris, F. H., & Ruback, R. B. (1998). Comparative distress levels of innercity family members of homicide victims. Journal of Traumatic Stress, 11(2), 223-242. doi:10.1023/A:1024494918952 Weathers, F. W., Litz, B. T., Herman, D. S., Huska, J. A., & Keane, T. M. (1993). The PTSD Checklist (PCL): Reliability, validity, and diagnostic utility. Paper presented at the 9th Annual Conference of the ISTSS, San Antonio, TX. Williams, J. L., Burke, L. A., McDevitt-Murphy, M., & Neimeyer, R. A. (2012). Responses to loss and health functioning among homicidally bereaved African Americans. Journal of Loss and Trauma, 17(3), 218-235. doi:10.1080/15325024.2011.616826* Zinzow, H. M., Rheingold, A. A., Byczkiewicz, M., Saunders, B. E., & Kilpatrick, D. G. (2011). Examining posttraumatic stress symptoms in a national sample of homicide survivors: Prevalence and comparison to other violence victims. Journal of Traumatic Stress, 24(6), 743-746. doi:10.1002/jts.20692* Zinzow, H. M., Rheingold, A. A., Hawkins, A. O., Saunders, B. E., & Kilpatrick, D. G. (2009). Losing a loved one to homicide: prevalence and mental health correlates in a national sample of young adults. Journal of Traumatic Stress, 22(1), 20-27.*

46

Chapter 3 Prevalence and Correlates of Self-Rated Post-Traumatic Stress Disorder and Complicated Grief in a Community-Based Sample of Homicidally Bereaved Individuals

Van Denderen, M., de Keijser, J., Huisman, M., & Boelen, P. A. (2016). Prevalence and correlates of self-rated posttraumatic stress disorder and complicated grief in a community-based sample of homicidally bereaved individuals. Journal of Interpersonal Violence, 31(2), 207-227. doi:10.1177/0886260514555368

Chapter 3

Abstract People confronted with homicidal loss have to cope with separation distress, related to their loss, and traumatic distress, associated with the circumstances surrounding the death. These reactions are related to Complicated Grief (CG) and Posttraumatic Stress Disorder (PTSD). The psychological effects for people who have lost someone through homicide, in terms of PTSD and CG, are largely unclear. This cross-sectional study (1) examined the prevalence of self-rated PTSD and self-rated CG in a community based sample of 312 spouses, family members and friends of homicide victims, and (2) aimed to identify socio-demographic, loss-related and perpetrator related correlates of PTSD and CG. Participants were recruited via support organizations for homicidally-bereaved individuals in the Netherlands (i.e. support group), and by case managers of a governmental organization, which offers practical, non-psychological, support to bereaved families (i.e. case manager group). Prevalence of self-rated PTSD was 30.9% (support group) and 37.5% (case manager group), prevalence of CG was 82.7% (support group) and 80.6% (case manager group). PTSD and CG severity scores varied as a function of the relationship with the victim; parents were at greater risk to develop emotional problems, compared to other relatives of the victim. Time since loss was negatively associated with PTSD and CG scores. Keywords Posttraumatic Stress Disorder, Complicated Grief, Homicide, Murder, Bereaved

48

Prevalence and Correlates of Self-Rated Post-Traumatic Stress Disorder

Introduction The death of a partner or close family member can lead to intense grief and distress. After homicide, people left behind face additional difficulties, associated with the violent and intentional nature of the act, involvement with the criminal justice system, media attention and investigation by the police (Amick-McMullan, Kilpatrick, Veronen, & Smith, 1989; Kaltman & Bonanno, 2003; Parkes, 1993; Riches & Dawson, 1998; Rynearson, 1994; Sprang, 2001; Van Denderen, De Keijser, Kleen, & Boelen, 2015). People confronted with homicidal loss are assumed to deal with both separation distress (e.g., yearning) related to the loss and traumatic distress (e.g., intrusive images) associated with the circumstances surrounding their loss. The preceding reactions are related to Complicated Grief (CG) and Posttraumatic Stress Disorder (PTSD), respectively (Rynearson & Sinnema, 1999). Persons with CG are assumed to be different from other bereaved individuals because they exhibit additional symptoms, such as intense yearning, searching, and permanent disbelief about the death of a loved one (Prigerson, Frank et al., 1995). While some symptoms of CG overlap with depression and symptoms of PTSD, cq. re-experiencing, avoidance and hyperarousal, the separation distress component is unique for persons with CG (Van Denderen et al., 2015). Furthermore, individuals with PTSD or CG can both experience anxiety, but the form is different: where PTSD complaints typically include threat-related anxiety, CG include mainly separation anxiety (Lichtenhal, Cruess, & Prigerson, 2004). Although depressive and CG symptoms are frequently comorbid, some symptoms are different. Preoccupation with thoughts of the deceased, for example, are a symptom of CG but not of depression (Boelen, Van de Schoot, Van den Hout, De Keijser, & Van den Bout, 2010; Prigerson, Maciejewski et al., 1995; Van Denderen et al., 2015). In preceding years, standardised diagnostic criteria have been proposed for CG, although the condition is not included in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5, American Psychiatric Association [APA], 2013). In the DSM-IV, bereavement was excluded as a mental illness, but was considered as a V-code, a condition which requires clinical attention when present with another illness. Prevalence of PTSD and CG Prevalence rates of PTSD after homicidal loss vary greatly between studies (Van Denderen et al., 2015). Among 333 young adults (aged 12-17) who lost a friend or family member due to homicide, the prevalence of current PTSD was 6%; having lost someone through homicide was further associated with an higher risk at depression (prevalence of past 6-month depression was 8%), drug use and alcohol use (Rheingold, Zinzow, Hawkins, Saunders, & Kilpatrick, 2012). Among 268 close friends and family members of homicide victims from a national representative sample, 15% of homicidally-bereaved met criteria for past six month PTSD (Zinzow, Rheingold, Byczkiewicz, Sauders, & Kilpatrick, 2011). Homicidally-bereaved individuals were significantly more likely than other bereaved participants to report past year PTSD

49

Chapter 3 symptoms (OR = 1.88) and were at greater risk for past year depression (OR = 1.64) and drug abuse/dependence (OR = 1.77) (Zinzow, Rheingold, Hawkins, Saunders, & Kilpatrick, 2009). Among a national representative sample of 115 homicidallybereaved individuals, 16.6 years post loss, lifetime prevalence of PTSD was 19.1%, whereas current PTSD was 5.2% (Amick-McMullan, Kilpatrick, & Resnick, 1991). In a small community based sample of 17 parents of murdered children, Murphy et al. (1999) found 60% of the mothers and 40% of the fathers to meet PTSD criteria 4 month post-loss. Because most studies have a cross-sectional nature, the effect of time on complaints could not be examined. In a longitudinal study among a community based sample of 47 homicidally-bereaved individuals, significant decrease in symptom-levels of depression and CG were observed over the 6-month study period, whereas PTSD symptom-severity remained stable (Williams, Burke, McDevitt-Murphy, & Neimeyer, 2012). Empirical, well-designed studies examining CG after violent death are currently lacking (Rynearson, Schut, & Stroebe, 2013). In a recent review on homicide-related psychopathology (Van Denderen et al., 2015), only one study (two references) was found concerning CG in homicidally-bereaved individuals, among a small number of participants (Burke, Neimeyer, & McDevitt-Murphy, 2010; McDevitt-Murphy, Neimeyer, Burke, Williams, & Lawson, 2012). The psychological consequences s of homicide for people left behind, in terms of PTSD and CG, are largely unclear. The present cross-sectional study was designed to (1) examine the prevalence of selfrated PTSD and CG in a community based sample of 312 spouses, family members and friends of homicide victims, and (2) identify socio-demographic, loss-related, and perpetrator related correlates of PTSD and CG. Hypotheses Although little research has been performed on the emotional impact of homicidal loss, we had a number of expectations related to variables associated with the magnitude of this impact. Based on prior research, we expected females to experience higher levels of PTSD and CG than males (Kersting, Brähler, Glaesmer, & Wagner, 2011; Kessler, Sonnega, Bromet, Hughes, & Nelson, 1995) (Hypothesis 1). In addition, we expected symptom-levels of PTSD and CG to be negatively related with time since loss (Applebaum & Burns, 1991; Prigerson & Jacobs, 2001) (Hypothesis 2). It has long been assumed that, more than other losses, the death of a child is perceived as untimely and unjust and therefore requires a major reconstruction of identity and worldview. In particular, the loss of a child appears to be a significant predictor for CG (Hibberd, Elwood, & Galovski, 2010; Kersting et al., 2011; Newson, Boelen, Hek, Hofman, & Tiemeier, 2011). In a study among 150 homicidally-bereaved individuals, varying in kinship relationship with the deceased, mothers scored significantly higher on PTSD than other relatives (Thompson, 1996). Accordingly, we expected higher PTSD and CG levels among homicidally-bereaved individuals who lost a child compared to other bereaved individuals (Hypothesis 3). The relationship between the bereaved individual and the perpetrator has been studied infrequently. There is some literature suggesting that bereaved individuals experience a more difficult

50

Prevalence and Correlates of Self-Rated Post-Traumatic Stress Disorder

bereavement process when the perpetrator is an acquaintance of the bereaved and victim, or when the homicide is intra familiar, for example a child whose mother was killed by the father (Harris-Hendriks, Black, & Kaplan, 1993; Hatton, 2003). We expected that participants who knew the perpetrator would report higher CG scores than participants for whom the perpetrator was someone unknown (Hypothesis 4). To our knowledge, no prior studies investigated the relation between the bereaved individual and the perpetrator with regard to PTSD. Therefore, we had no prior expectations about this association. Little research has been performed regarding the influence of the juridical status of the perpetrator on bereavement and PTSD. In a study among 15 bereaved adults, bereaved individuals reported less psychological complaints in cases where the perpetrator was arrested (Rynearson, 1984). Because, to our knowledge, no other studies examined the association between the juridical status of the perpetrator and PTSD and CG, we had no specific expectations about this issue.

Method Participants and Procedure A cross-sectional questionnaire study was conducted among homicidally-bereaved individuals in the Netherlands. Participants were 312 spouses, family members (parents, siblings, children, aunts, uncles, and grandparents) and friends of homicide victims. To be included, participants had to be 18 years or older and understand the Dutch language. The 312 participants were related to 255 different homicide victims. Demographic characteristics of the sample are shown in Table 1. Participants were recruited via three support organizations for homicidallybereaved individuals in the Netherlands, (N = 188) (hereafter called support group), and via case managers from the governmental organization Victim Support The Netherlands, which offers practical, non-psychological, support to homicidallybereaved families, (N = 124) (hereafter called case manager group). In the support organizations, people have contact with other individuals who have lost a loved one through homicide, and which they did not know before the homicide. The organizations have a supportive, informal and non-caregiving, non-professional and non-commercial character: they organize casual meetings in which individuals can share their experiences. Now and then, professionals (e.g., politicians, lawyers) are invited to inform members about juridical procedures. Victim Support is a governmental organization which offers practical and legal, non-psychological, support to homicidally-bereaved families since 2007. Their core aim is to inform these families about their rights in court, to help arrange the funeral, and to give advice in dealing with the media. Data collection took place between June 2011 and March 2013. Cohabiting participants received paper questionnaires individually addressed. The material contained an information letter, the questionnaire packet, and a stamped return envelope. Questionnaires were sent minimally six months post loss, to allow time

51

Chapter 3

Table 1. Demographic and Loss-Related Characteristics of the Homicide Sample (N = 312). Support group Case manager (N = 188) group (N = 124) Category

Subcategory

Sex Female Age participant (years) Time since loss (years) Witnessed the murder Experienced one loss Convicted perpetrator Participant is … of the victim Spouse Parent Child Sibling Non-immediate family member Other (friend, acquaintance) Perpetrator is… of the (ex)-spouse participant Direct family member (i.e. parent, child or sibling) Non-immediate family member Other (friend, acquaintance) Unknown

% or M SD

% or M SD

% or M SD

64.7% 53.4 15.5 6.9 6.5 4.6% 93.7% 68.5% 6.9% 49.5% 12.5% 15.8% 9.2%

61.7% 58.2 13.3 9.4 6.4 4.9% 92.9% 78.4% 5.5% 65.9% 5.5% 10.4% 7.7%

69.4% 45.7 3.1 4.1% 95% 53.2% 9.1% 24.8% 23.1% 24% 11.6%

5.9% .7%

4.9% .6%

7.4% .8%

6.2%

3.5%

10.2%

9.3%

6.4%

13.6%

28.3%

29.1%

27.1%

55.5%

60.5%

48.3%

15.2 4.9

for normal grief. Questionnaires were numbered and kept separately from the addresses, to which the first author had only access to. Case managers from Victim Support handed out the questionnaires to their clients. They knew which of their clients participated in the study, but had no access to questionnaire data from their clients. We handed out 504 questionnaires to Victim Support, resulting in a response rate of 24.6%. However, we have the indication that not all 504 questionnaires were handed out by the case managers. Therefore, the response rate is minimally 24.6%, but is probably higher. We sent 333 questionnaires to members of support groups, resulting in a response rate of 56.5%. The study was approved by the Ethical Commission Psychology board of the University of Groningen and consent was provided by the boards of the three support organizations and by Victim Support. Measures

Socio-demographic and Perpetrator Related Variables

We collected information about the following demographic, loss-related and perpetrator related variables; gender, time since loss (in years), the relationship between participant and victim, the relationship between participant and perpetrator, and juridical status of the perpetrator. The relationship between participant and

52

Prevalence and Correlates of Self-Rated Post-Traumatic Stress Disorder

victim was divided into six categories: i.e. the participant was a (i) spouse; (ii) parent; (iii) child; (iv) sibling; (v) non-immediate family member; or (vi) friend/acquaintance of the victim. The relationship between participant and perpetrator was divided into five categories: the perpetrator was an (i) (ex)-spouse; (ii) direct family member (i.e. parent, child or sibling); (iii) non-immediate family member; (iv) other person known to the participant (i.e. colleague, friend, business partner, or acquaintance); or (v) someone unknown to the participant. For scoring the juridical punishment of the perpetrator, four categories were distinguished: (i) convicted; (ii) the perpetrator was not found; (iii) the legal process is not yet completed; (iv) the perpetrator was discharged from punishment.

Post-Traumatic Stress Disorder

Symptoms of current PTSD were measured with the PTSD Symptom Scale, SelfReport (PSS-SR) (Foa, Riggs, Dancu, & Rothbaum, 1993; Dutch version Engelhard, Arntz, & Van den Hout, 2007). The PSS-SR is a 17 item self-report questionnaire to assess the symptoms of PTSD as defined in DSM-IV. The frequency of each symptom during the previous week was rated on 4-point scale ranging from 0 = not at all to 3 = five or more times per week/almost always. PTSD prevalence was determined using the scoring rule that symptom scores were at least 2 (‘two to four times a week/half of the time’) for at least one re-experiencing symptom, three avoidance symptoms, and two hyper arousal symptoms (cf. Brewin, Andrews, & Rose, 2000). In this sample, Cronbach’s alpha for all 17 items was .93. The alpha for the subscales were .87 (reexperiencing, e.g., “In the past week, have you had bad dreams or nightmares about the traumatic event?”), .85 (avoidance, e.g. “In the past week, how often did you feel distant or cut-off from people around you?”), and .83 (hyperarousal, e.g., “In the past week, how often were you overly alert?”). The PSS-SR is not limited to victims of war or migration, and has frequently been used in samples of other kinds of traumas, also in studies in which the trauma happened long ago (see for example Mol et al, 2005). Therefore, we found it suitable to use in our sample of homicidally-bereaved individuals. Before answering the PSS-SR, participants were instructed to report which of any of past events have been bothered them the most in the last month. Half of the participants, 50%, reported the homicide as event, 4 % reported a non-homicide related event (i.e. “sexual abuse in my childhood”), 8% a homicide related and a non-homicide related event (i.e. “the murder of my mother and the divorce from my husband”), and 38% did not report any event. For the analysis with PTSD as outcome measure, we excluded the 4% participants which reported a non-homicide related event. The participants which did not reported any event were not excluded because the context of the research, psychopathology following homicide, was made clear throughout the questionnaire package and cover letter. We assume it to be reasonable that participants completed the PSS-SR while bearing the homicide in mind. Since PTSD was not formally assessed and diagnosed by a structured interview, and the PSS-SR cannot diagnose PTSD, when PTSD is used throughout the paper, probable PTSD can be read.

53

Chapter 3

Complicated Grief

Current CG severity was measured using the 19-item Inventory of Complicated Grief (ICG) developed by Prigerson, Maciejewski et al., 1995 (Dutch version, Boelen, Van den Bout, De Keijser, & Hoijtink, 2003). Respondents rate the frequency of symptoms in the preceding month on scales ranging from 0 = never to 4 = all the time. Following Prigerson, Maciejewski et al. (1995), a cut-off score of 25 at the ICG was used to differentiate between complicated (greater than 25) and non-complicated grievers (lower than or equal to 25). Among 97 bereaved elders, they found respondents with ICG scores >25 to be significantly more impaired in social, general, mental, and physical health functioning than those with ICG scores ≤25. In the current sample, Cronbach’s alpha was .92. Examples of items are “I have the feeling that part of me has died with him or her” and “I feel tense, irritable or shocked since his or her death”. Statistical Analyses First, we examined whether participants from the so-termed support group and case manager group differed in terms of time since loss and age, based on recruitment strategy). To this end, t-tests were used. Depending on the outcome, participants were treated as one group or as two groups when reporting prevalence rates and analyzing the correlates of PTSD and CG severity. Prevalence of PTSD and CG was assessed using the above-mentioned scoring rule and cut-off score. Proportion tests were used to test the differences in prevalence rate between both groups (Z-scores and p values reported). When looking at the data, the assumption of independence of observations was not met: the 312 participants were related to 255 victims. To control for this dependency, we used multilevel analysis. The bereaved participants were nested in the victims. To test the differences in CG and PTSD scores between groups based on gender, relationship between participant and victim, relationship between participant and the perpetrator, juridical punishment, F-tests were reported. Where appropriate, significant results were followed by multiple comparisons t-tests to test differences between pairs of groups, using Bonferroni correction (hypothesis 1 - 4). Missing Data Missing items were encountered in the ICG for 46 participants (14.7%); 27 participants had one item missing (8.7%), 5 participants were missing more than half of the items (1.6%). For the PSS-SR, 53 participants (17.0%) had missing items; 24 participants had one item missing (7.7%) and 17 participants (5.4%) were missing more than half of the items. Participants with missing scores were compared with participants who completed all items. Both groups were compared for CG and PTSD on demographic and loss-related characteristics presented in Table 1. Only small and statistically non-significant differences were found. To retain as much of the item scores as possible, scale scores were calculated by averaging over the observed items when less than 50% of the scale items were missing. Because the two scales

54

Prevalence and Correlates of Self-Rated Post-Traumatic Stress Disorder

consist of a large number of items (19 items for the ICG and 17 items for the PSSSR), calculating scale scores on half of the items was deemed to give reliable results. Participants with more than 50% of the scale items missing were excluded from the analyses.

Results Preliminary Analysis Using independent sample t-tests, it was found that participants from the support group and those from the case manager group differed significantly in terms of time since loss; 3 (SD = 4.9) years for the case manager group, compared to 9 (SD = 6.4) years for the support group; t(299) = 9.88, p < .001. Groups also differed significantly in age; 46 (SD = 15.2) years (case manager group), compared to 58 (SD = 13.3) years (support group); t(238) = 7.4, p < .001. Holding age and time since loss constant across analyses (by including these variables as covariates) was deemed inappropriate because the overall estimated means on these variables are not representative for both subgroups. Therefore, we decided to treat both groups independently. We reported the prevalence rates of PTSD and CG for both groups together and separately, and performed all analyses on both groups separately. We found the two variables age and time since loss to be strongly related to each other. Individuals who experienced the homicide 15 years ago for example, were more likely to be older. In the analysis examining the correlates of PTSD and CG severity (see Table 3 and Table 4), the presented mean scores, SE and tests for differences were corrected for age and time since loss. Prevalence of Current PTSD and Current CG In Table 2, the prevalence rates for current PTSD and current CG were presented for the support group and case manager group, and proportion tests were reported (Z-scores and p values). As can be seen in Table 2, prevalence rates of PTSD and CG did not differ significantly between both subgroups. Table 2. Prevalence of Current Complicated Grief (CG) and Current Posttraumatic Stress Disorder (PTSD) among the Support Group and Case manager Group and Tests of Difference in Prevalence Between the Support and Case manager Group. CG PTSD

Total Sample 81.9% 33.7%

Support group 82.7% 30.9%

Case manager group 80.6% 37.5%

Test Z = 0.59, p = .55 Z = 1.12, p = .26

55

Chapter 3

Correlates of CG-severity In Table 3, socio-demographic correlates of CG were presented for both groups. Among the support group, CG scores were generally lower among participants for which the murder took place longer ago, b = -.40, p < .05. CG severity differed as a function of relationship to the victim, F(5,172) = 4.96, p < .001. Multiple comparison t-tests showed that parents and spouses reported significantly higher CG scores than non-immediate family members. Gender, the juridical status of the perpetrator and the relationship between participant and perpetrator did not correlate significantly with CG severity in the support group. For the case manager group, female participants reported significantly higher CG scores than male participants, F(1,103) = 9.53, p < .01. In terms of Cohen’s d effect size, difference in complicated grief scores between females (M = 40.25, SD = 13.51) and males (M = 32.38, SD = 14.52) was moderate (d = 0.56).1 Among the case manager group, scores were also lower when the homicide happened longer ago, b = -.66, p < .01. CG severity also differed as a function of relationship with the victim, F(5,115) = 5.82, p < .001; parents reported significantly higher CG scores than children, siblings, non-immediate family members and friends/acquaintances. CG severity was associated with the juridical status of the perpetrator, F(3, 100) = 3.20, p < .05; participants for which the legal process was still on-going, reported significantly higher CG scores compared to cases in which the perpetrator was convicted. CG severity did not vary as a function of the relationship between the participant and perpetrator in the case manager group.

When calculating Cohen´s d, the SD was used instead of the SE. To obtain the mean scores and SD, scores were not corrected for the covariates age and time since loss. Mean scores reported here differ therefore slightly from the mean scores reported in Table 3 and Table 4.

1

56

Prevalence and Correlates of Self-Rated Post-Traumatic Stress Disorder

Table 3. Socio Demographic and Loss-related Correlates of Complicated Grief (CG). Subcategory

Complicated Grief; Support group (N = 188)

N M SE

Complicated Grief; Case manager group (N = 121)

Test for N M SE difference#

Sex (N total) Male 69 38.23 1.65 35 31.98 2.21 Female 111 41.29 1.31 F(1,116) = 86 39.86 1.44 2.27, p = .14 Time since loss b = -.40, p < .05* Perpetrator

1. Convicted 128 39.85 1.21 57 35.32 1.78 2. Is not found 14 34.80 3.65 7 33.79 4.78 by the police 3. The legal 12 45.69 4.12 38 43.74 2.23 process is still on-going 4. Is discharged 6 44.31 5.53 F(3,140) = 4 36.78 6.30 from punishment 1.48, p = .22

Participant is … 1. Spouse 9 41.44 4.33 5 (+) 11 37.87 3.70 of the victim 2. Parent 115 42.24 1.37 5 (+) 29 47.37 2.42 3. Child 10 41.11 4.84 26 33.42 2.57 4. Sibling 18 37.11 3.66 29 37.70 2.29 5. Non-immediate 14 35.89 4.41 14 31.16 3.35 family member 6. ‘Others’ 9 F(5,172) = 9 25.02 4.44 (friend/ 4.96, colleague/ p < .001*** acquaintance) Perpetrator is… 1. (ex) Spouse 1 48.85 13.69 1 40.43 of the participant 2. Direct family 6 35.49 5.53 11 34.51 member (parent. child or sibling) 3. Non-immediate 11 48.53 4.17 16 36.32 family member 4. Other (Friend. 49 40.94 2.00 32 34.71 colleague. acquaintance) 5. Unknown 99 38.82 1.42 F(4,160) = 55 40.21 1.53, p = .20

Test for difference# F(1,103) = 9.53, p < .01** b = -.66, p < .01**

1 (+) F(3,100) = 3.20, p < .05*

3, 4, 5, 6 (+)

F(5,115) = 5.82, p < .001***

13.57 4.20

3.48 2.42 1.95 F(4,111) = .94, p = .44

Note. *p