Epidemiology and Treatment of Posttraumatic Stress Disorder in West-Nile Populations of Sudan and Uganda

Dissertation zur Erlangung des Doktorgrades

Eingereicht an der mathematisch-naturwissenschaftlichen Sektion der

Universität Konstanz

von

Dipl.-Psych. Frank Neuner

im Juli 2003 2

Acknowledgements/Danksagung

Schon seit einer Weile freue ich mich darauf, diese Danksagung zu schreiben. Nicht nur, weil das bedeutet, dass ich endlich die Arbeit drucken lassen kann. Vielmehr habe ich die Gelegenheit, mich nun einmal förmlich zu bedanken bei all den Personen, die zum Gelingen dieser Arbeit beigetragen haben.

Ich bedanke mich bei Thomas Elbert für die wissenschaftliche Weitsicht sowie den unermüdlichen Enthusiasmus, mit dem er mich bei diesen Projekten unterstützt hat. Maggie Schauer war von Anfang an eine tragende Kraft bei allen unseren Traumaprojekten, mit ihr zusammen wurde die Narrative Expositionstherapie entwickelt. Ich danke ihr für das gegenseitige Vertrauen und die tragfähige Zusammenarbeit auch in schwierigsten Situa- tionen.

Das unkomplizierte und höchst angenehme Klima der Arbeitsgruppe gab mir für die letzten Jahre ein anregendes Arbeitsumfeld. Bei Brigitte Rockstroh bedanke ich mich dafür, dass sie mich vom ersten Forschungsantrag bis jetzt zur Abgabe der Arbeit immer unterstützt hat und für all das, was sie mir im Studium und danach über klinische Psy- chologie und Forschungsmethoden beigebracht hat.

Diese Doktorarbeit baut auf dem “Demography of Forced Migration” Projekt auf, das von Unni Karunakara von der John Hopkins Universität unter Betreuung von Prof. Burnham geleitet wurde. Unni war letztlich derjenige, der uns nach Uganda gebracht hat, ohne ihn hätten wir keine der Studien durchführen können, vielen Dank dafür. Als weitere Institutionen waren die Makarere Universität, Kampala und Ärzte ohne Grenzen (MSF) Holland beteiligt, die uns auch in einem medizinischen Notfall sehr professionell geholfen haben. Die Flüchtlingssiedlung Imvepi wird vom Deutschen Entwicklungsdienst unter Lei- tung von Adi Gerstl verwaltet. Für seine zuverlässige Unterstützung bin ich sehr dankbar, zumal wir zur Befriedigung unserer lebensnotwendigsten Grundbedürfnisse wie Essen, Schlafen und e-mail völlig abhängig waren vom DED.

Bei der Therapiestudie waren außer Maggie, Thomas und mir noch Christine Klaschik und Elisabeth Kley als Untersucherinnen und Therapeutinnen und in der Nach- folgeuntersuchung auch Elisabeth Schauer beteiligt. Für all den unerschrockenen Einsatz 3 unter Kröten, Skorpionen und Rebellen bin ich sehr dankbar. Rebecca Horn und Barbara Meier können viel besser englisch als ich und haben diese Arbeit korrigiert. Ich danke für die endlosen formalen und inhaltlichen Anmerkungen (“strict but forgiving”).

Eine solche Studie ist nicht möglich ohne ein gutes Team lokaler Mitarbeiter. Hier sind vor allem Mary A. Martin (“Tall Mary”) und Nicolas W. Krispo vorzuheben. Ihrer Überzeugungskraft, Begeisterung, Zuverlässigkeit sowie ihrem hohen Ansehen in der loka- len Bevölkerung ist zu verdanken, dass wir die sudanesischen Flüchtlinge von unserer Ar- beit überzeugen und für die Nachfolgeuntersuchungen wieder aufspüren konnten.

Als ich diese Arbeit schrieb hatte immer wieder verschiedene ugandische Mitar- beiter des DED und sudanesische Flüchtlinge im Kopf, die ich in Imvepi kennengelernt hatte. Zu wenigen von ihnen habe ich noch sporadischen Kontakt, bei vielen frage ich mich, was wohl aus ihnen geworden ist. Ich bin ihnen sehr dankbar für all die bereichernden Begegnungen und die Erkenntnis, dass kulturelle Unterschiede im zwischenmenschlichen Kontakt völlig unerheblich sein können.

Schließlich bedanke ich mich bei meinen Eltern für die langjährige Unterstützung meines Studiums. Ganz besonders bedanke ich mich bei meiner Frau Nicole, die meine Be- geisterung für Afrika teilt aber dennoch oft (bei nicht immer zuverlässigen Kommunika- tionsmittlen) auf meine Anwesenheit verzichten musste. Ich freue mich über das, was wir gemeinsam erreicht haben, und vor allem auch auf das, was uns gerade erwartet.

Konstanz, im Juli 2003

Frank Neuner 4

Table of Contents

1 OVERVIEW 6

2 ORGANIZED VIOLENCE 9

2.1 Definition 9

2.2 Wars 10

2.3 Torture 16

2.4 Survivors of organized violence as refugees 19

3 PSYCHOLOGICAL CONSEQUENCES OF ORGANIZED VIOLENCE 21

3.1 The concept of posttraumatic stress disorder 21

3.2 PTSD in populations affected by organized violence 27

3.3 Criticism of PTSD concept 37

4 PTSD IN POPULATIONS AFFECTED BY THE SUDANESE WAR 45

4.1 The Sudanese Civil War 45

4.2 Sudanese refugees in Uganda’s West Nile region 53

4.3 Survey of West Nile populations 55

4.4 Dose-response effect for PTSD 68

4.5 Psychometric quality of the PTSD assessment 74

5 ETIOLOGICAL MODEL OF PSYCHOLOGICAL TRAUMA 84

5.1 Memory related features of PTSD 84

5.2 Traumatic event in memory 85

5.3 Sensory-perceptual representation 87

5.4 Autobiographic contextual memory 91

5.5 Neurobiological basis of memory and PTSD 95

5.6 Emotional Processing 97

5.7 The speechlessness of trauma: sociopolitical implications 101 5 6 PSYCHOTHERAPY OF PTSD 105

6.1 Overview 105

6.2 Acute Interventions/Debriefing 106

6.3 Psychodynamic Therapy 108

6.4 Anxiety management 109

6.5 Exposure oriented treatment 109

6.6 Cognitive Therapy 112

6.7 Combination of treatments 114

6.8 Treatment of survivors of organized violence 115

7 NARRATIVE EXPOSURE THERAPY (NET) 118

7.1 Basic principles of NET 118

7.2 Randomized controlled trial 120

7.3 Treatment 125

8 CONCLUSION 136

9 SUMMARY 141

10 ZUSAMMENFASSUNG 142

11 APPENDIX 144

12 REFERENCES 150 6

1 Overview

The term “trauma” has achieved unequaled popularity in recent years. The word originates from the Greek “trau=ma”, meaning an injury or a wound. At first, it was adopted by medicine as a technical term referring to tissue damage that was caused by external mechanisms. Later on, psychiatrists suggested that extremely stressful life events could be considered as traumas, as those events could contribute to the onset of mental disorders, even without any physical injury. In this context, trauma has become a metaphor to describe presumed wounds of the soul, caused by shocking events like combat experiences, sexual abuse and life-threatening accidents.

In recent years, more disciplines have begun to use the term in a metaphorical sense, using it to indicate emotionally upsetting personal events as well as radical social, cultural and historical changes. Manifestations of traumas have been identified in literature, music, architecture and film. For example, in his analysis “Das Trauma der verfehlten Melodie bei Robert Schumann” Zizek (1999) has used both personal traumatic events in Robert Schu- mann’s life and a proposed cultural trauma caused by the change from the classicistic to the romantic epoch to explain the melody in Schumann’s compositions.

Beyond different scientific disciplines, trauma has also found a place in every-day communications, referring to a wide variety of stressful personal and social events, like bereavement, unemployment, and poverty. The widespread use of the term “trauma” for all kinds of stressful events has led to an increasing fuzziness in the definition, and the lack of conceptual clarity risks misunderstandings and a loss of meaning of the term. The expan- sion of the trauma definition was even reflected in scientific literature. For example, psychologists and psychiatrists recently discussed the traumatic consequences of sexual harassment (Avina & O'Donohue, 2002) and even childbirth (Ayers & Pickering, 2001).

This thesis deals with the traumatic consequences of organized violence on mental health. Organized violence includes wars, torture and other severe human rights violations. Whereas it is straightforward that war and torture can wound the psyche and thus be considered as psychological traumas, organized violence is not first of all a psychological problem. It is important to be aware of the political context of wars and torture to comprehend the meaning of organized violence for the individual and the societies. As the characteris- 7 tics of current warfare and torture are very different from the traditional view of these phenomena I will present information on the context of organized violence in chapter 2.

As a response to the excessive generalization of the term “trauma”, a very specific and narrow “trauma” definition has evolved in recent years in psychological and psychiatric sciences. The psychiatric concept of posttraumatic stress disorder (PTSD) offers a framework for operational criteria of potential traumatic events and the characteristic symptoms of these events. I will explain these definitions in chapter 3.1. In this study I will refer to this understanding of “trauma”, as it applies to individuals. I will avoid the use of terms like “traumatized societies” and “collective trauma” as, contrary to the widespread use of these terms, there is no qualified definition of these concepts.

The introduction of the PTSD concept has stimulated much empirical research. One interesting finding was that not all survivors of wars and torture are traumatized. Many epidemiological studies have investigated the rate of traumatized individuals in populations affected by organized violence. Research in this field is confronted with many methodological challenges, I will discuss these problems and the current state of knowledge in chapter 3.2. Despite the vast research on PTSD, it has become popular to criticize the validity of the concept. This criticism is discussed in chapter 3.3.

The Sudanese civil war is one of the world’s oldest civil wars and offers a perfect example for the illustration of the characteristics of modern warfare (chapter 4.1). We carried out an epidemiological study of PTSD in this context. We compared Sudanese nationals who remained in the West-Nile region of Sudan with refugees who had fled to West-Nile Uganda to analyze the effect of war and forced migration on the mental health of survivors of the war. The Ugandan hosts provided a control group as they are culturally similar to the Sudanese but had a quite peaceful existence in the past years (chapter 4.3).

The epidemiological study offered an estimation of the size of the problem. The next question was what psychology can contribute to an alleviation of the suffering of traumatized individuals in this region. We aimed to develop a specific treatment approach. For the development of a treatment method, it is essential to understand the current knowledge of the psychological and neurobiological mechanisms behind the disorder (chapter 5) and previous treatment approaches (chapter 6). We applied this knowledge to the political background of organized violence and developed Narrative Exposure Therapy 8 (NET), the principles of which are explained in chapter 7. To examine the efficacy of NET, we carried out a randomized controlled trial with Sudanese refugees living in Uganda, and compared the efficacy of NET with other methods that have been used to treat traumatized refugees.

After a general conclusion, I present one example of the testimonies of refugees who participated in the treatment study (appendix). These testimonies can illustrate the meaning of traumatic war experiences within the life context of individual persons. Thus, they provide information that is usually hidden behind statistics. 9

2 Organized violence

2.1 Definition

Despite the destructive power of violence it remains a universal and enduring phenomenon. Human history can be structured along the series of wars between nations and bloody conflicts in states. Different religions and social movements have always tried to condemn violence and proposed a peaceful coexistence of people. Nevertheless, still today violence dominates many parts of the world and the danger of war and terror remains a continuous threat for most people.

Violence appears in different forms. Prominent classifications of violence distinguish between different forms with regard to the context (Derriennic, 1971). A major di- mension to qualify types of violence is their degree of organization. Examples of unorgan- ized types of violence are assaults, domestic violence, sexual abuse and other crimes that happen at an individual level. On the other hand, wars and political persecution provide the context for more organized forms of violence including torture, combat situations and bombardments.

The term organized violence is widely used in the political (Derriennic, 1971) as well as in the medical and psychological literature (Basoglu, 1993; Jensen, Schaumburg, Leroy, Larsen, & Thorup, 1989; Van Velsen, Gorst-Unsworth, & Turner, 1996). Notwith- standing the recent popularity of this concept, there is no consistent definition. It is mainly used in the study of the characteristics and consequences of torture, wars and forced migration. Other authors use terms like “state-sponsored violence” (S. M. Weine & Laub, 1995) and “severe human rights violations” (Silove, 1999) to describe related occurrences. Based on the usage of the term in literature and the classifications of violence offered by Derriennic (1971) and Galtung (1969) the following definition might describe the current understanding of organized violence:

Organized violence is violence that is directly and actually applied with a systematic strategy by members of a group with at least a minimum of centrally guided structure (police units, rebel organizations, terror organizations, paramilitary and military formations). It is applied with a certain con- tinuity against individuals and groups with a different political attitude, na- 10 tionality as well as racial, cultural or ethnical background. It is characterized by the violation of central human rights or other basic rights of people.

Organized violence encompasses three types of violence. The first type is the permanent state-sponsored persecution that is present in all dictatorships, and even in some countries that are considered democracies. This harassment includes different forms of violence like torture, extralegal executions, disappearances etc. The second type is the massive violence committed against people in an interstate war or a civil war. The third type of organized violence is characterized by violence committed by terror organizations. Whereas there is much literature on the consequence of torture and wars, terrorism did not attract much research until recently, so I will concentrate on the description of the characteristics and consequences of wars and torture.

2.2 Wars

2.2.1 Number and Locations

A war is a mass conflict that involves fighting between two or more armed parties. Probably most people could accept this definition, but it is not sufficient. In current wars it is increasingly difficult to distinguish wars from other conflicts. Consequently, different political scientists have developed various definitions of war. These definitions differ with respect to the questions of whether regular government forces have to be involved and the degree of organization in the forces that is required to classify for an army. In addition, most definitions have a cut off value for the number of battle-related casualties per year that are required to classify for a war (see Gleditsch, Wallensteen, Eriksson, Sollenberg, & Strand, 2002 for the discussion of the differences of the definitions).

The Hamburg working group for research into the causes of war (Arbeitsgemein- schaft Kriegsursachenforschung; AKUF) counted 31 wars in 2001 (Schreiber, 2002). Since the end of World War Two, the number of wars in the world was increasing more or less steadily until 1995, since when there was a small decrease (Gleditsch et al., 2002).

The public view of wars is dominated by knowledge about World Wars One and Two as well as subsequent wars with American or European participation, especially the Vietnam War. Despite the dominance of these inter-state wars in media reports, research 11 on current wars shows that wars that are characterized by two or more countries fighting each other are rather the exception than the rule. In 2001 more than nine out of ten wars (91%) were inner-state (or civil) wars (Schreiber, 2002). Whereas foreign armies may participate in the fights, these wars do not originate from conflicts between nations but within a country. There are two different reasons for civil wars. Currently, in about half of the inner-state wars a rebel army fights for the autonomy or secession of a region. In the other half of the wars, rebels aim at the overthrow of the ruling regime.

The wars in the world are not evenly distributed over the earth. There are certain major conflict zones that have generated most of the recent wars. The great majority of the 2001 wars (91%) were located in Africa, Asia and in the Near/Middle East. Figure 2.1 presents the location of wars in the period between 1989 and 1999. It shows that one major conflict zone included Central America and the northern part of South America with some single conflicts in Central South America. A global line of conflicts spanned from Eastern Europe through the Balkans and the Middle East and India to Indonesia. The third zone of conflicts was Africa, where almost the whole continent has a history of wars in the last decade.

Figure 2.1. Armed conflicts 1989-99 by location, the circles indicate interstate wars and the crosses inner-state wars. From Gleditsch, Wallensteen, Eriksson, Sollenberg, & Strand (2001). 12 2.2.2 “New Wars”

The traditional view of wars is well described by Clausewitz famous statement that war is the continuation of politics with different means. This view of wars dominates the accounts of the major European wars since Napoleon. It describes different countries and alliances that wage wars with professional armies, whereby the majority of casualties are soldiers who die in the battlefields. Wars start with a declaration and are terminated by the capitulation of one party or a peace agreement. The war parties aim for victory in these wars by concentrating their forces in major battles.

It was not until the Balkan War that an inner-state war raised questions about the validity of this traditional view. Kaldor (1999) introduced the term “New War” to describe this and other similar current conflicts. She suggested that the general characteristics of warfare underwent major changes in recent years in the Balkan as well as in most other war regions. Several characteristics of “New Wars” have been suggested, and other researchers (e. g. Münkler, 2002) adopted this term in their analyses. As there is little research in war areas, most of the following conclusions were drawn from the observations of war journal- ists. For an illustration of the typical characteristics of current warfare in the Sudanese civil war, see chapter 4.1.

Irregular forces

Contrary to the traditional view on wars, current conflicts involve more than regular armies. Instead, the fighting is dominated by irregular forces, including paramilitary units, rebel forces, mercenary troops and foreign armies that intervene in civil wars on one side. The majority of fighters on all sides of the conflicts have limited military training. As many characteristics of regular armies, like uniforms and regular salary, are not applicable to the majority of fighters, the clear separation between civilians and soldiers disappears. Most current wars lack clear declarations of war, and the fighting parties seem to hesitate to provoke decisions through major battles. This results in long-lasting conflicts, that have also been called “low-intensity conflicts. The clear difference between peace and war time dissipates in these conflicts.

A major indicator of development towards irregular forces is the fact that most current wars involve children as fighters. Whereas in traditional wars children were only recruited by weakened parties, forcibly recruited child soldiers belong to the usual repertoire 13 of many modern forces. The advantage of the usage of children as fighters is that they are considered to be easily controlled and manipulated to be unscrupulous fighters as they lack norms and values and cannot judge risks and dangers in the same way adults can (Schreiber, 2002).

Justification on identity

Competing ideologies played a major role in traditional wars. In contrast, the dominant justification of current wars is based on the different identities of the conflict partners, based on their membership of different ethnics groups, cultures or religions. For example, no different ideology separated Serbs, Bosnians and Croats in the Balkan wars. Instead, the ethnic identity discriminated between the war parties. Myths about ancient rivalries and wars between the ethnic groups were used to motivate the public for the war. Kaldor admitted that during the World Wars the national identities played a major role as well, but there was always a kind of ideological vision tied to the different parties.

Warfare targets civilians

Warfare in current wars cannot be described by the traditional view of two opposing armies fighting each other with heavy weapons at a clear frontline to gain territory. The main means of gaining power in new wars is controlling the civil population, by frightening the population and expelling civilians who do not belong to the powerful group. Heavy weapons are not necessary for this type of warfare. The current worldwide availability of small weapons was caused by the excessive equipment of many armies by the super powers during the cold war and the increasing trade in small weapons after the breakdown of many former socialist countries after the cold war.

Whereas former civil wars that were fought in the context of social revolutions involved guerilla fighters who aimed at winning the “hearts and minds” of the civil population in order to move amongst them like “fishes in the water”, this strategy is rarely to be found in current civil wars. Instead, all sides seem to use an anti-guerilla tactic, involving frightening the civilians and expelling opponent groups.

The strategy of new warfare thus includes systematic atrocities like massacres and mass rapes to frighten the civilians and to make regions uninhabitable for the group to be expelled. The widespread use of landmines and destruction of monuments are other means 14 of achieving this goal. Another reason for the prevalence of atrocities in current wars is the assumption that they help to unify the group committing the atrocities. Once a person has participated in committing war crimes, it is almost impossible for this person to leave the group, since he will always be rejected by others because of what he has done. In the beginning of the Rwandese genocide in 1994, which resulted in the killings of more than 800 000 Tutsis by the Hutus, there was a high pressure on every individual Hutu to participate in the killings. Many reports state that the children who were recruited as soldiers were forced to commit atrocities in their own village. This prevented the children from fleeing from the forces and returning to their home villages, since they would be rejected if they returned.

Economic factors

The observation of current wars suggests that rational motives cannot explain these excessively violent conflicts. Nevertheless, detailed analyses of these wars show that these wars are not wild and fanatic fighters killing each other randomly, but that tangible interests motivate the main actors to machinate these wars. In particular, economic factors play an increasing role in the onset and maintenance of wars. In a global economy, the war parties are usually not autarkic but get resources from supporting foreign countries and exile communities. Very often, the conflicts are fought to win and keep control over local resources like diamonds, minerals, oil and drugs. This consequently leads to powerful war- lords who do not depend on governments. As the war offers them the opportunity to keep power and to gain money without the control of any regulating institution, they do not have an immediate interest in a termination of the war. Consequently, many wars are extended by the deliberate delay of peace negotiations and the unwillingness of both war parties to fight deciding battles against each other.

“New wars” are not new

The term “New War” has become increasingly popular to describe modern conflicts. Some authors, especially experienced war researchers who have been observing conflicts for many years, criticized this development (e.g. Gantzel, 2002). Whereas there is no doubt that the warfare in most common wars is very different from the common public view on wars, it was pointed out that this kind of warfare is not at all a new phenomenon. Analysis of typical warfare since World War Two reveals that the “New War” characteris- 15 Table 2.1. Prevalence rates of exposure to events related to war and persecution in survivors of the Kosovar, Bosnian and Cambodian conflicts

Event Kosovar Albani- Bosnian refu- Cambodian refu- ans during war gees during war gees during Pol Pot (%) (%) regime (%)

Lack of food/water 67 28 96

Combat situation 67 35 43

Ill health/no medical care 48 23 87

Murder of family member 26 17 54

Shelling attacks 31 83 31

Lack of shelter 57 85

Being close to death 62 63

Murder of stranger 24 37

Rape/ sexual abuse 4 17

Hiding from snipers 75

Hiding outdoors 63

Forced labor 88

Brainwashing 87

Torture 36

tics just describe common warfare in inner-state conflicts. Most conflicts have presented these characteristics for decades but have typically gone unnoticed by the public in industrialized countries.

Table 2.1 illustrates the prevalence of exposure to different war events of representative samples of civilian war survivors. The studies represent Kosovar Albanians (Lopes Cardozo, Vergara, Agani, & Gotway, 2000), refugees from Cambodia who lived in Thai camps (Mollica et al., 1993) as well as refugees from Bosnia who lived in Croatian camps (Mollica et al., 1999; see chapter 3.2 for details about these studies). The table shows that whereas the distribution of specific events is clearly different between the groups, almost all survivors in all groups reported the experience of one or more severe events. The data reflects that the war and persecution in Cambodia as well as the Balkan war aimed at the 16 violation of the civil population. The experiences of those victims who did not survive the war have naturally not been assessed. As all three conflicts caused considerable numbers of casualties among the civil populations (20-40% of the total Cambodian population died during the Pol Pot regime), the prevalence rates still represent an underestimation of the actual violence that occurred during these conflicts.

2.3 Torture

Torture has often been called the “scourge of mankind”. The history of the Euro- pean juridical systems shows that torture has been a tool of persecution and punishment for centuries. It was widely applied in the Middle Ages and the most terrifying episodes of the history of torture included the Spanish Inquisition in 15th century and the witch-hunt in Germany that escalated in the 16th century. It was not before the 18th century that the Enlightenment philosophers turned against the use of torture. Unfortunately, Victor Hugo’s proclamation in 1874 that “torture has ceased to exist” (Jacobsen & Smidt-Nielsen, 1997) proved to be a rash conclusion, as torture is still evident in many countries of the world.

There are various definitions of torture in different international treaties. The most prominent definition was proposed by the UN convention against Torture and other Cruel, Inhuman or Degrading Treatment or Punishment (UN, 1984):

For the purposes of this Convention, the term "torture" means any act by which severe pain or suffering, whether physical or mental, is inten- tionally inflicted on a person for such purposes as obtaining from him or a third person information or a confession, punishing him for an act he or a third person has committed or is suspected of having committed, or intimi- dating or coercing him or a third person, or for any reason based on discrimination of any kind, when such pain or suffering is inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity. It does not include pain or suffering aris- ing only from, inherent in or incidental to lawful sanctions.

This definition relates to torture that is committed by government agents or by people who act with official sanctions. Other organizations have broadened this definition to make it applicable to other organized groups like rebel armies (Amnesty International, 2003). They point out that the traditional image of political prisoners being tortured in an interrogation chamber does not fit all the observations of human rights organizations that 17 increasingly document torture in refugee camps and public places, as well as in people’s homes.

Amnesty International noted that in more than 100 countries torture is systematically used by state organs (Amnesty International, 2002). The aim of torture is to generate a general high level of fear within a population to intimidate opponents of the ruling regime. In this context, torture is only one means of human right violations. Other common instruments are the imprisonment of nonviolent opponents, the conviction of opponents without the opportunity of a fair trial, secret imprisonments without contact with the world outside the prison, the application of death penalties, extralegal executions and disappearances.

Reports about many dictatorships and even democracies (like Turkey) indicate that torture has evolved into a systemized means of handling political prisoners. Sophisticated torture methods have been developed and torture specialists were trained. Hariotos- Fatouros (1988) has studied the practice of torture during the regime of the Greek Junta (1967-1974). The interviews with former professional torturers after the overthrow of the Junta showed that torture was not restricted to occasional excesses of interrogators and prison wards but that there were elaborated institutions for the selection and training of torturers. Trainees were carefully selected according to their beliefs and their socioeco- nomic status, making them susceptible to propaganda and monetary rewards. The trainees underwent a sophisticated training including their own violations and humiliations, and being fed propaganda, as well as receiving systematic rewards and punishments. In this way, they could be trained to be obedient torture specialists and to fulfill their roles within a systematic system of persecution.

The specialized torture methods have in common that they allow the infliction of maximum pain on the victim whilst and at the same time leaving a minimum of visible physical after-effects (Jacobsen & Smidt-Nielsen, 1997). The lack of enduring physical scars interferes with efforts to document the human rights abuses and makes it difficult for the victim to prove torture experiences when he applies for asylum in an exile country. It is well documented that in very many cases physicians and other health professionals participated in torture (British Medical Association, 1992). Their responsibility ranges from deliberately withholding necessary treatment in prisons and issuing false medical certificates, to the direct participation in torture sessions with the task of preventing the victim from dy- 18

Table 2.2. Forms of torture reported by Turkish torture survivors (from: Basoglu, Paker, Paker, Ozmen et al., 1994; selection from a total of 43 different torture forms assessed using a checklist)

Form of torture Prevalence (%)

Verbal abuse 100

Beating 100

Blindfolding 96

Alternating gentle/rough treatment 93

Forced standing 87

Threats of further torture 85

Electrical torture 78

Witnessing torture 75

Being stripped naked 71

Threats of rape 58

Threats against family 53

Hanging by wrists 51

Sham executions 38

Fondling of genitals 31

Submersion in water 15 ing. Medical treatment in torture prisons is often reduced to the single purpose of making the victim ready for further torture.

Resulting from their clinical experiences in the treatment of torture survivors, Vesti, Somnier and Kastrup (1992) have classified the torture techniques reported by survivors. They distinguish between deprivation techniques (depriving victims of fundamental bodily needs like sleep, nutrition, hygiene and health care), coercion techniques (forcing victims to take part in degrading activities, like witnessing torture, eating excrement, etc.), communication techniques (verbal abuse, alternating rough/gentle treatment), pharmacol- ogical techniques and sexual torture techniques. Usually a torture victim has to undergo a large variety of different techniques at unpredictable time-points during an imprisonment. 19 Often prisoners are forced to confess to deeds and to betray other opponents during torture. These coercions are a means of inflicting feelings of guilt and shame, and most torture victims report that giving in does not protect them from further torture. Instead, they are left in a completely helpless state without any means of stopping the humiliations.

Table 2.2 illustrates the occurrence of typical torture experiences as reported by Turkish torture victims (Basoglu, Paker, Paker, Ozmen et al., 1994). As it is not possible to select a representative sample of torture victims, these data may be biased, but they give an impression of what victims of systematic torture have to endure. The survivors reported a mean of 23 different forms of torture, with multiple exposures to each method. The survivors in this study rated a mean total of 291 exposures to torture during a mean of 47 months of imprisonment.

2.4 Survivors of organized violence as refugees

An obvious consequence of organized violence is that many people have to flee from their region of origin because of war or persecution. Very few war reporters dare to visit current war regions as it is usually not possible to guarantee security. Consequently, there are almost no photographs and films of combats, atrocities and torture. Media reports of overcrowded refugee camps are probably the most common confrontation with the consequences of organized violence for the western public. In 2001, the United Nations High Commissioner for Refugees (UNHCR) counted 21 million people who were fleeing worldwide (UNHCR, 2002b). The UNHCR differentiates between three groups of fleeing people. According to the UNHCR definition, refugees are persons who leave their home countries and find a more or less stable exile. During 2001, the size of the refugee population as estimated by the UNHCR remained unchanged at around 12 million. The other groups are asylum seekers who have submitted an application for asylum but have not yet received a decision and internally displaced persons (IDPs) who had to flee from their homes but did not leave their country of origin.

Most of the armed conflicts take place in non-industrialized countries. Conse- quently, the majority of refugees (86% in the past decade) originated from developing countries. At the same time, only a minority of refugees managed to flee to industrialized countries; 72% of all refugees stayed within the developing world, mostly in Asia or Africa. 20 About half of all refugees live in camps. As a result of the long duration of many conflicts, many camps have developed into refugee settlements that provide a home for many refugees who remain there for years or even decades.

No matter where the refugees flee to after war and persecution, most of them do not find a safe and accommodative exile. Many reports indicate that initial receptions by host government authorities and humanitarian agencies are impersonal and threatening, and that refugees assume roles of dependency and helplessness (Doná & Berry, 1999). Whilst developments of social networks, family reunions and permanent settlements do occur (Castles & Miller, 1993), harsh living conditions, continued anxiety about forced re- patriation and uncertainties regarding resettlement can cause considerable stress for the refugees. Host country refugee policies are often dictated by domestic concerns, usually of a foreign policy nature and not necessarily determined by security and protection concerns or by the wishes of host communities in receiving countries (Tandon, 1984).

There are many reports that refugee camps breed violence and refugees are often victims of violent acts perpetrated by the army, militias, humanitarian workers and by their hosts (Malkki, 1995; Turner, 1999; UNHCR, 2002a). For many women and children, the very act of going to communal latrines (Forbes Martin, 1991) or collecting firewood and water can be extremely dangerous.

Refugees who flee to industrial countries to apply for asylum also live under very stressful conditions (Baker, 1992; Kammerlander, 1997). Many organizations caring for refugees in these countries complain about living conditions below the level of people who receive social welfare, unfair and delayed acceptance proceedings and direct offences by rac- ist opponents of refugees. Baker (1992) introduced the term “triple trauma“ for refugees who apply for asylum in industrialized countries. According to this concept, the asylum seekers have to suffer not only from their traumatic experiences from their homeland, but also from stressful events and losses caused by the flight as well as stressful living conditions in their exile country. 21

3 Psychological consequences of organized violence

3.1 The concept of posttraumatic stress disorder

3.1.1 Historical overview

The idea that stress contributes to mental diseases is not new and was held by many clinicians and scientists long before the concept of posttraumatic stress disorder (PTSD) was introduced to the DSM III in the 1980s. Different authors postulated that stressful life events can contribute to mental diseases, but the common agreement in psychiatry was that the psychopathological impact of life events was restricted to vulnerable individuals. The main cause of the illness was seen as the genetic predisposition or developmental vulnerability and the role of stressful events was generally reduced to a nonspecific trigger that contributed to the onset or maintenance of a disease.

In the history of psychiatry, several authors opposed this view and suggested that traumatic events could cause mental diseases in normal individuals without the necessity of a predisposing vulnerability. Janet (1889) was the first to describe dissociation as a specific psychiatric consequence of traumatic events. In his early account of hysteria, Freud and Breuer (1895) suggested that the experience of sexual abuse was the main cause of this disorder. Later, Freud rejected this view in favor of developmental models.

Among the first victims of traumatic events to be systematically studied were victims of organized violence, the survivors of the Holocaust (see Levav, 1997, for an overview). In 1948, Friedman was the first to describe a specific disorder he called “Buchenwald- Syndrom“ that included sleeping difficulties, high arousal and affective numbing in survivors. In 1964, Eitinger pointed out symptoms like concentration difficulties, irritability and a chronic fatigue in survivors living in Israel and Norway. He still avoided, however, the notion that these symptoms indicated a psychiatric disorder caused by the experience of traumatic events. Instead, he assumed a physical condition to be responsible for the symptoms. This suggestion was supported by the fact that the survivors of concentration camps he studied had also been exposed to malnutrition, physical attacks and diseases dur- 22 ing their persecution. Some years later, the American psychiatrist Krystal defined the “survivor syndrome“ that included acute anxiety, nightmares, depression and hypochondria.

When the classification system diagnostic and statistic manual (DSM) was introduced by the American Psychiatric Association 1952 as a variant of the International Clas- sification of Diseases (ICD 6), the term “gross stress reaction“ was included to account for the immediate reaction to extremely stressful events. This diagnosis, like the consecutive diagnosis “transient situational disturbance” introduced in DSM II, was restricted to the acute upheaval that could be observed in individuals immediately after the experience of a traumatic event. Long lasting consequences of traumatic events were considered to be restricted to vulnerable individuals who developed an anxiety or depressive neurosis later on.

In the late 1970s there was a fundamental change in the view of the consequences of traumatic events. A lot of research was stimulated at this time mainly by the finding that a great proportion of the Vietnam veterans who returned home had major difficulties reinte- grating into their prewar roles. At the same time, researchers influenced by the women’s movement noted the prevalence of rape in the society and started to examine the resulting psychological symptoms in detail. This development was supported by the increasing acknowledgement of the suffering of traumatized individuals in society.

Different conceptualizations of the psychological symptoms related to specific events, like the “rape trauma syndrome“ (Burgess & Holmstrom, 1974) and the “delayed stress syndrome“ of former combatants (Horowitz, 1976) evolved. The major break- through in research into the consequences of traumatic events was when the psychiatric field realized that the different syndromes defined so far had much in common and should be seen as a single disorder. Consequently, the term posttraumatic stress disorder (PTSD) was introduced into the third version of the Diagnostics and Statistical Manual of Mental Disorders (DSM-III) to integrate the different classifications that had been offered before.

3.1.2 PTSD in DSM-IV

The current definition of PTSD in DSM-IV requires six criteria for the definition of PTSD. The first criterion (A) refers to the traumatic event. PTSD can only be diagnosed when the symptoms resulted from an experienced or witnessed event that involved the actual or perceived threat for the life or physical integrity of the person or another person. In 23 addition, the immediate reaction of the victim must involve fear, terror or helplessness. This event criterion was introduced to differentiate between traumatic events like rape and combat experiences from other stressful events like bereavement and the loss of a job. The idea between this discrimination is that normal stressful events may play a role in the onset and maintenance of psychiatric diagnoses like depression and schizophrenia but that the development of these disorders is restricted to vulnerable individuals whereas traumatic events could cause PTSD in anyone (Yehuda & McFarlane, 1995).

The second criterion (B) is related to intrusive symptoms. In the context of PTSD, intrusive symptoms describe the chronic re-experiencing of the traumatic event in the form of nightmares, flashbacks, stressful thoughts as well as the emotional and physiological re- activity to reminders of the event. Recent research showed that intrusive memories are not restricted to patients with PTSD, but that other stressful conditions could also cause the unwanted reliving of painful episodes (Brewin, Watson, McCarthy, Hyman, & Dayson, 1998). In contrast, reliving as flashbacks, in the form of multiple sensory qualities and including the feeling of being back in the traumatic situation, seems to be unique to PTSD (Brewin, 2001).

Criterion C relates to the avoidance behavior associated with PTSD. Contrary to the current classification, a factor analysis of PTSD symptoms suggested the subdivision of avoidance symptoms into two different factors (Foa, Riggs, & Gershuny, 1995). The first factor includes active avoidance of reminders of the traumatic event, like people and places that are associated with the event and the avoidance of talking and thinking about the event. In contrast to these phobic reactions are the symptoms of passive avoidance or numbing. These phenomena, which are also related to dissociation phenomena, include general emotional numbing as well as detachment from other people. According to the factor analysis, these symptoms of numbing are more directly associated with the other PTSD symptoms than is phobic avoidance.

The last group of symptoms consists of the arousal symptoms (criterion D). They include the consequences of a high general level of arousal, including sleeping and concentration difficulties, an exaggerated startle response and the enduring feeling of threat.

Criteria E sets a time frame for PTSD, as the symptoms must last for at least four weeks. The additional diagnosis of acute stress disorder (ASD) was introduced for the acute 24 reactions to traumatic events that could be diagnosed within the four weeks period after the event. The symptoms of ASD are similar to the PTSD symptoms but put more stress on dissociation symptoms as these seem to better predict long-term consequences. Finally, criterion F relates to the clinical significance of the disorder as it requires a remarkable reduction in functioning for the diagnosis of PTSD.

3.1.3 Assessment of PTSD

Qualified research and treatment of a psychiatric disorder depends on the availability of reliable and valid instruments for the diagnosis of the disease. Several standardized procedures have been developed for the assessment of PTSD. These instruments generally consist of a set of standardized questions about traumatic events and the symptoms of PTSD. There are two types of these instruments. Structured interviews are constructed to provide a standardized set of questions to an expert or a trained lay interviewer. They are generally considered to be the gold standard of psychiatric diagnosis. Some of these instruments, e.g. the SCID (Structured Clinical Interview for DSM-IV; First, Spitzer, Miriam, & Williams, 2001) and the CIDI (Composite International Diagnostic Interview; WHO, 1997) have been developed to cover a large range of different psychiatric diagnoses in one interview. The advantage of these instruments is that a more or less complete picture of comorbidity can be assessed in this form. Unfortunately the completion of the full interview lasts up to two hours, so in many studies on traumatic stress only the PTSD part of the whole interviews is selected.

The diagnoses obtained through these structured clinical interviews are considered to be the most reliable and valid forms of diagnosis. Unfortunately they do not offer an estimation of the severity of the disease. The severity of the disease is most important for treatment studies that aim at the reduction of the severity of certain symptoms. The CAPS (Clinician Administered PTSD Scale; Blake et al., 1995) is an instrument that fulfills both requirements, a reliable and valid diagnosis and an expert rating of the severity of symptoms and the disorder.

There are plenty of self-report forms for the assessment of PTSD. These instruments are filled in by the respondents themselves, usually in a paper-and-pencil form. They provide a severity rating of the symptoms and PTSD. Some of these instruments (e.g. the 25 Posttraumatic Stress Diagnostic Scale, PDS; Foa, 1995) can function as screening instruments for the diagnosis of PTSD with a satisfying accuracy, but the reliability and validity of the diagnoses are worse than those obtained through expert interviews. Other instruments, like the often used Impact of Event Scale (latest form: IES-R; Weiss & Marmar, 1996) don’t closely match the DSM-IV criteria and thus can not provide a diagnosis.

3.1.4 Epidemiology in industrialized countries

In industrialized countries, traumatic experiences like accidents, physical and sexual violence as well as war experiences of veterans can cause PTSD. In the US, epidemiological studies that used recent diagnostic criteria and instruments for the assessment of PTSD, found lifetime prevalence rates between 7.8% (Kessler, Sonnega, Bromet, Hughes, & Nel- son, 1995) and 9.2% (Breslau et al., 1998). There is no large-scale epidemiological study of PTSD in Europe. A study of adolescents carried out in the Munich region suggests that the prevalence of PTSD might be smaller in Germany than in the US, as the general level of violence is lower and fewer people have participated in recent wars (Perkonigg & Wittchen, 1999).

Epidemiological studies showed that not all persons who experienced a traumatic event develop chronic PTSD. The most adverse event seems to be rape as it leads to PTSD in about 56% of the victims, followed by war participation (39%) and childhood abuse (35%) (Kessler et al., 1995). Epidemiological studies have found that PTSD is accompanied by one or more comorbid psychiatric disorders in more than 80% of the cases (Kessler et al., 1995). Among the most prevalent lifetime comorbid disorders identified in this study were major depression (men: 48%, women 49%), alcohol abuse (52%, 28%), simple phobias (31%, 29%), social phobias (28%, 28%), and conduct disorder (43%, 15%).

The fact that not all victims of a traumatic event have posttraumatic stress disorder has stimulated a lot of research into predictors for the development of PTSD. In a recent meta-analysis of risk factors in trauma exposed individuals, Brewin, Andrews, and Valen- tine (2000) concluded that the risk factors identified so far only have a modest effect size for the prediction of PTSD. Pre-trauma vulnerability factors like education, previous trauma, childhood adversity, psychiatric history and family psychiatric history predicted PTSD consistently over several studies but only to a small extent. Factors operating during 26 the event (trauma severity) or immediately after the event (lack of social support) seem to have stronger effect sizes, but still their explanatory power is small. The fact that pre- trauma vulnerability has an effect on the development of PTSD challenges the original assumption that people could develop PTSD without a predisposing vulnerability (Yehuda & McFarlane, 1995). Current knowledge can neither confirm nor reject this assumption. We know that developmental factors can modulate the probability for PTSD but we do not know whether there is a type of trauma with such a high severity that all of the victims develop PTSD. Until now, no study has identified such an event and even among torture victims the prevalence of PTSD is well below 100%.

3.1.5 “Complex” PTSD

Looking at the diversity of traumatic events, Herman (1992) has suggested differen- tiating between two different types of traumas. Type I traumatic events are those events that lead to pathological consequences after a single exposure, like a car accident or rape. The more severe type II events include those events that either occur in childhood (like sexual abuse), or occur repeatedly over a long time-period, like torture experiences. Her- man suggested that whereas type I events cause “simple” PTSD, type II events can lead to a more complex pattern of symptoms. She suggested the introduction of a new diagnostic category of complex PTSD to refer to the symptoms that are thought to occur in addition to the usual PTSD symptoms in victims of type II events (van der Kolk, Roth, Pelcovitz, & Mandel, 1993). Complex PTSD was subsequently also called “disorder of extreme stress (DES)”. It includes difficulties in affect regulation, extended dissociation symptoms, somatization symptoms, changes in identity, interpersonal problems, the tendency to expose oneself to further threat and the general loss of meaning of life. A structured interview for the diagnosis of DES was developed (Pelcovitz et al., 1997), but evaluation showed that not all the new symptom categories could be reliably diagnosed. Until now the suggestion of DES has not stimulated much research and it cannot yet be considered to be a valid diagnostic category. Nevertheless, the similar category “personality changes under extreme stress” was included in ICD 10. 27 3.2 PTSD in populations affected by organized violence

One to two months after the September 11 terrorist attacks in New York City, the prevalence of PTSD related to the attacks among the inhabitants of the region near the World Trade Center (Manhattan south of the Canal Street) was 20%, whereas attacks- related PTSD in the other areas of New York was 7.5% (Galea et al., 2002). The investigators who arrived at this conclusion used telephone interviews for the assessment and selected the respondents using random-digit dialing, which may restrict the validity of the diagnosis and the random selection. Nevertheless, this study shows that organized violence in the form of terrorist attacks has severe consequences for the mental health of the population even in industrialized countries, at least in the immediate aftermath of the event.

Notwithstanding the attention this attack received in the industrialized world’s public, these kinds of terrorist attacks are still the exception to most organized violence. The dominant forms of organized violence are wars and persecution that can cause forced migration. Carrying out epidemiological research in these populations is a difficult en- deavor. Most victims of organized violence live in war regions that are dangerous to access. Drawing representative samples within war populations is complicated as there is usually no valid demographic background information on the households and inhabitants. For torture victims this problem is almost impossible to solve, as there are naturally no complete lists of torture survivors in their home or exile countries. The diversity of local languages in developing countries and the lack of fundamental mental health research in these countries poses a challenge for the application of standardized psychiatric instruments and the interpretation of the results.

Nevertheless, several studies have tried to overcome these problems. Table 3.1 summarizes studies that examined mental health in populations with experiences of war, persecution and forced migration. Prevalence rates of posttraumatic stress ranged between 2% (Hauff & Vaglum, 1995a) and 99% (K. de Jong, Mulhern, Ford, van der Kam, & Kle- ber, 2000) in these populations. These studies differ in the type of population studied, the selection procedure and the instruments used to assess PTSD. A closer look at the most serious studies is necessary to estimate the impact of organized violence. In the following review, a selection of the more influential publications that examined PTSD as a consequence 28 of organized violence are presented to illustrate the state of the current knowledge and the methodological difficulties in epidemiological refugee research.

3.2.1 Populations living in war regions

Most people who are affected by organized violence live in regions that are affected by wars or have a recent history of war. Only a few epidemiological studies have examined war-torn populations that remained in the war region. Among these studies, there are many articles with limited explanatory value. One example is the study of K. de Jong and coworkers (2000) which found a prevalence of 99% of severe posttraumatic stress in Sierra Leone. Although reported in a major medical journal, this finding cannot be taken seriously, as the instrument applied in this study (Impact of Event Scale, IES) does not allow a diagnosis of PTSD and there is almost no information about the sampling of the 245 respondents.

Somasundaram and Sivayokan (1994) examined PTSD and other psychological disorders in 101 randomly selected respondents in Sri Lanka. They found a prevalence rate of 27%. Unfortunately, there is no information about the selection procedure, the authors state that they were randomly selected but this statement requires further explanation as sampling is not straightforward in a war-torn population. Furthermore, they used the Stress Impact Questionnaire (SIQ) to diagnose PTSD, but this instrument has not been validated for use with the Sri Lankan population.

In a major study including data from different war-affected populations, J. T. de Jong and coworkers (2001) found PTSD prevalence rates of 37% in Algeria (n = 653), 28.4% in Cambodia (n = 610) and 17.8% in Gaza (n = 585). In this study, the assessments were carried out by local interviewers, who applied a translation of the PTSD part of the Composite International Diagnostics Interview (CIDI). The authors note that this instrument was carefully translated into the local languages, but there is no information about the training of the interviewers and no study confirmed the validity of the translation and the assessment procedure. Furthermore, the selection of populations within these countries was restricted by security factors as well as by the availability of a local team and psychosocial care. These factors limit the extent to which the prevalence rates can be generalized to the whole populations of the countries. The fact that the researchers could not access the 29 more insecure regions suggests that the prevalence rates might be underestimates. Never- theless, the study shows that PTSD is frequent and a severe mental health problem in conflict areas.

Lopes Cardozo and colleagues (2000) examined mental health in 1358 randomly selected Kosovar Albanians who lived in Kosovo in 1998. They found a prevalence rate of 17.1% PTSD. The method included a complex and innovative sampling procedure that used information from prewar population data as well as previous village surveys to identify representative rural and urban population clusters. Random sections were selected within these clusters and the households were selected by drawing a random direction (spinning bottle) from the center of the section. PTSD diagnoses were made in self-report form using the Harvard Trauma Questionnaire (HTQ), for illiterate respondents the questions were read aloud. The HTQ consists of a detailed checklist of traumatic events that was adapted to the requirements of victims of war. In addition, it contains a section on PTSD symptoms that allows a diagnosis of PTSD. The HTQ has been developed and validated for Cambodian refugees in the US. Although the instrument has only been validated with this population, it was widely used in refugee research in many different populations. In this study, there was no validation of the Albanian translation of the instrument. This is a severe limitation of the study as the validity of PTSD diagnoses obtained on the basis of self-report forms has not been demonstrated for victims of organized violence.

3.2.2 PTSD in refugees living close to the war regions

As carrying out epidemiological studies in war regions is very difficult, many researchers preferred to study the consequences of organized violence in refugee populations. As described above, most refugees do not flee to industrialized countries to apply for asylum there, but remain in regions close to their country of origin where a great percentage of them remain in camps for many years.

Mollica and colleagues (1993) examined Cambodian refugees who had fled to camps in Thailand. They used a “multistage probability area” sampling based on grid maps of a refugee camp. Within the sampling units, specific routes for the interviewers starting from a randomly assigned starting point determined the selection of single households. They achieved a response rate of 98%. The survey showed that most of the 993 respondents 30 (82%) had lived in the camp for more then 2 years. The HTQ was used as an instrument for the diagnosis of PTSD, which had been validated in the Cambodian translation before. They found a prevalence rate of PTSD of 15%.

In another study, Mollica et al. (1999) examined the consequences of war and forced migration for Bosnian refugees who had fled to a camp in Croatia. 26% of them suffered from PTSD according to HTQ. They randomly chose one adult from each of the 573 fami- lies living in the camp and thus achieved a sample size of 534 respondents (the remaining minority of refugees refused to participate). Although they state that culturally validated instruments were applied in this study, they report no validation of the Bosnian version of the HTQ.

3.2.3 PTSD in refugees in industrialized countries

A minority of refugees manages to flee to industrialized countries. Some studies addressed these populations. One study found a prevalence rate of 86% of PTSD among 50 Cambodians who resettled in the USA (Carlson & Rosser-Hogan, 1991). The respondents were randomly selected from a list of all refugees made by a nonprofit social services agency. They used expert interviews for the assessment of PTSD, but a non-validated checklist of PTSD criteria was used as an instrument. In addition, they excluded five of the 17 DSM-III-R criteria for PTSD because for some reason they did not consider these criteria to be appropriate for the group. Unfortunately, there is no explanation of which symptoms were excluded. As the criteria for a PTSD diagnosis were considerably reduced compared to the standard, the results probably greatly overestimate the true PTSD prevalence.

Another group (Silove, Sinnerbrink, Field, Manicavasagar, & Steel, 1997a) examined a sample of 40 refugees from different countries who applied for asylum in Australia. They used an opportunity sampling method, and selected the respondents from an English class. As expected, this resulted in a selective group with reasonable English skills, which cannot be assumed to be representative of asylum seekers. PTSD diagnoses were made using the PTSD part of the CIDI, which was in vivo translated into the respondent’s mother language if the English skills were not sufficient. Among the asylum seekers, 37% suffered from PTSD. 31 In a consecutive study in Australia, this group (Silove, Steel, McGorry, & Mohan, 1998) compared the levels of PTSD among 62 Tamil asylum seekers with 30 authorized Tamil refugees and 102 immigrants with Tamil background. The respondents were approached through aid services and refugee organizations, which may lead to uncontrollable selection biases. The HTQ was used for the assessment of PTSD symptoms, unfortunately the Tamil translation was not specially validated and no diagnosis of PTSD was made. As expected, both refugee groups presented with higher posttraumatic stress than the immigrants, but they did not differ from each other.

In a small sample of 20 Bosnian refugees in the immediate aftermath of resettlement S. M. Weine and coworkers (1995) found a PTSD prevalence of 65%. The refugees took part in a program that provided them with the opportunity to give testimony. The effect of this selection cannot be estimated, as there might be a bias either towards a specially morbid treatment-seeking group or towards a healthy group of those who present with less avoidance behavior. The assessments were carried out by experts using a translation of the Posttraumatic Symptoms Scale (PSS). An interesting result of this study is that one year later 44% of the original group still suffered from PTSD (S. M. Weine, Vojvoda et al., 1998), indicating a considerable stability of PTSD in refugees over time.

3.2.4 PTSD in patient groups

Confronted with a high psychiatric morbidity in refugees, several institutions in industrialized countries provide special medical and psychological assistance for refugees. Some of these institutions have studied their patients in detail. For example, Van Velsen et al. (1996) examined 60 patients with a history of organized violence who were referred to the London Medical Foundation for the Care of Victims of Torture. A self-constructed checklist was used by psychiatrists to diagnose PTSD. The subjects were from a wide variety of national backgrounds, the largest group were Kurds from Turkey. 52% of the patients suffered from PTSD. This result indicates that there is a high prevalence of PTSD among refugees who seek medical or psychosocial assistance. 32 3.2.5 PTSD in torture victims and former political prisoners

Torture is considered the most stressful form of organized violence. In several consecutive studies (Basoglu et al., 1997; Basoglu et al., 1996; Basoglu, Paker, Ozmen, Tas- demir, & Sahin, 1994; Basoglu, Paker, Paker, Özmen et al., 1994), Basoglu and coworkers tried to determine the mental health effects of torture for the survivors who remained in Turkey. With 55 respondents in each group, he compared a group of tortured political activists with a group of non-tortured activists. He found a rate of PTSD of 18% among the torture victims, compared to 4% among the non-tortured activists. Considering the severity of the torture experiences reported in this study (see chapter 2.3), the prevalence rate of PTSD is still remarkably low. In another group of Turkish torture victims, who were not political activists but were convicted because of criminal activities, the PTSD rate was 58%. A comparison of the torture experiences showed that the activists were tortured much more severely than the criminals, nevertheless their rate of PTSD was much lower. The interpretation was that the preparedness of the political activists who were aware of the risk of torture could be a protecting factor. In contrast, the torture in prison was not expected by the criminals and could less easily be justified on the basis of their beliefs. This interpretation was confirmed by a further analysis of the group’s cognitions and attitudes. Nevertheless, the convenient snowball sampling procedure for the group of activists limits the explanatory power of the results. It is probable that only the less severely affected individuals were ready to come to an investigation to talk about their suffering. So the 18% prevalence of PTSD in torture survivors is almost certainly an underestimate.

In two studies, Maercker and coworkers (Maercker, Beauducel, & Schutzwohl, 2000; Maercker & Schutzwohl, 1997) examined PTSD in former political prisoners from the German Democratic Republic after the German reunion. Again, a convenient sampling procedure was applied as respondents were recruited by newspaper articles and political prisoners’ organizations. In both studies about 30% of the former prisoners presented with PTSD. Assessments were reliable and valid as they were carried out by psychologists using a standard clinical interview validated in the German language (Diagnostisches Interview Psychischer Störungen; DIPS, extended German version of the Anxiety Disorders Sched- ule; ADIS). 33 Shrestha et al. (1998) studied the impact of torture on Bhutanese refugees living in Nepalese refugee camps. They compared 526 refugees with torture experiences with the same number of refugees matched for age and sex without torture experiences. The authors selected the patients from an existing list of tortured refugees prepared by a nongovern- mental organization. A self-constructed interview schedule was applied by local physicians, who received brief training in the diagnosis of PTSD. Unfortunately, there was no validation of the instrument and the assessment procedure. A prevalence rate of 14% was found among the tortured refugees. In comparison, only 3% of the non-tortured refugees suffered from PTSD. This result further supports the impact of torture on mental health even in a refugee population with presumably many war experiences. However, the relatively low prevalence rates have to be interpreted cautiously because of the methodological problems of the assessment. 34

3.2.6 Consequences of organized violence

Table 3.1 Studies examining the prevalence of PTSD in populations affected by organized violence

PTSD PTSD Instru- Preva- Author Year Population Selection ment Validation lence

Basoglu 94 55 tortured activists convenient SCID yes 18%

Cardozo 2000 1358 Kosovo Albanians Random HTQ not for population 17%

PTSD Carlson 91 50 Cambodian refugees in USA Random Checklist no 86% de Jong 2001 653 Algerians Random CIDI not for population 37.4% de Jong 2001 610 Cambodians Random CIDI not for population 28.4% de Jong 2001 1200 Ethiopians Random CIDI not for population 15.8% de Jong 2001 585 Palestinian refugees in Gaza camp Random CIDI not for population 17.8%

Gorst-Unworth 98 84 Iraqi refugees in treatment in UK Patients HTQ not for population 10.7%

Hinton 93 201 Vietnamese refugees in USA Random SCID Experts 3.5%

Maercker 97 146 former political prisoners in GDR convenient DIPS yes 30%

Maercker 2000 98 former political prisoners in GDR convenient DIPS yes 31%

Vietnamese refugees in USA: 51 ex- complete 90%/ Mollica 98 detainees, 22 non-detainees (81%) HTQ not for population 79%

Mollica 99 534 Bosnian refugees in Croatian camp random HTQ not for population 26%

Mollica 93 993 Cambodian refugees in Thailand camp random HTQ yes 15%

Peltzer 99 100 Sudanese refugees in Ugandan camps convenient HTQ not for population 32%

526 tortured Bhutan refugees in Nepalese PTSD Shrestha 98 camp/526 random Checklist no 14%

526 non-tortured Bhutan refugees in Nep- ! ! alese camp ! ! no 3%

Silove 97 40 asylum seekers in Australia convenient CIDI Experts 37%

PTSD Somasundaram 94 101 Sri Lankan residents living in war area random Checklist no 14%

Weine 95 20 Bosnian refugees in USA community convenient PSS not for population 65% 35

As expected, a review of the epidemiological studies with populations affected by organized violence revealed methodological difficulties in the research. First of all, the selection of respondents is a major problem in most of the studies. Organized violence is mainly found in war regions and in refugee populations. Usually there are no reliable lists of inhabitants in these areas, camps and settlements. The best studies use creative sampling methods, like a hut-to-hut procedure or select the whole population of a camp. The situation is even more complicated for the study of torture victims, as they are found as members of refugee populations or in their homeland. The only possible way to select suitable participants is to use a convenient selection procedure that depends on the voluntary identification of torture victims. The nature of the resultant sampling bias is unclear. It may be that those with more PTSD symptoms are less willing to volunteer for research because of their more severe avoidance symptoms. Alternatively, this group may be over-represented, because they want to obtain assistance.

Another difficulty is that organized violence is a worldwide problem, but is mainly found in developing countries. This poses problems for assessments, as assessment instruments are usually neither standardized nor validated in the local languages, and few qualified researchers are available to carry out the interviews. A lot of preparatory work is necessary to train a local team. The researchers should expect mostly uneducated respondents in the surveys, which is a special challenge for the translation of the instruments. The translation procedure involves a transfer of rather difficult explanations of symptoms into simple language with much less vocabulary, so the instrument will considerably change in this process. This makes a validation of each translation of an instrument into a different local language necessary. This validation should include a comparison of the results of the assessment procedure to the findings of a gold standard, which could be a detailed examination of a sub-sample of the respondents by experts. Of all the studies presented above, only one (Mollica et al., 1993) applied an instrument that was validated in the local language used in the assessment procedure.

The PTSD rates found in most studies are higher than the rates found in industrialized countries. When interpreting high prevalence of PTSD in war affected populations in Africa or Asia on the basis of figures obtained in Western Europe and North America, it must be borne in mind that a recent history of war and persecution is not the only difference between these societies. 36 As most modern conflicts and forced migrations occur in resource poor societies in developing countries, stressors caused by poverty, declining living conditions and a down- ward spiral of violence may contribute to the development and maintenance of PTSD in these countries. For refugee populations, the factor of forced migration in combination with the loss of cultural and societal bonds and acculturation problems in exile are further variables that could account for a higher psychiatric morbidity.

In addition, it has been argued that there may be cultural and societal biases in a population’s disposition to report posttraumatic symptoms that may contribute to higher prevalence of PTSD, even though there is no evidence to that effect.

One way to estimate the impact of single factors on the mental health outcome is to calculate the relative contribution of assumed predictors for the probability of PTSD in a regression analysis. Many studies showed that the main predictor of PTSD in populations affected by organized violence is the number and severity of the traumatic events (Carlson & Rosser-Hogan, 1991; Cunningham & Cunningham, 1997; J. T. de Jong et al., 2001; Gorst-Unsworth & Goldenberg, 1998; Hauff & Vaglum, 1993, 1995b; Hinton et al., 1993; Lopes Cardozo et al., 2000; Mollica et al., 1993; Mollica, McInnes, Poole, & Tor, 1998; Mollica et al., 1999). For refugees who live in industrialized countries, post migration stressors (e.g. caused by the application process) contribute to morbidity (Silove, McIntosh, & Becker, 1993; Silove et al., 1997a; Van Velsen et al., 1996).

The best way to isolate the impact of organized violence would be by comparing the prevalence of PTSD in a population affected by this type of violence with a culturally and economically equivalent but unaffected population. In this way, the isolated impact of torture experiences could be confirmed both for a refugee population (Shrestha et al., 1998) as well for victims of torture and former political prisoners living in their home countries (Basoglu, Paker, Paker, Ozmen et al., 1994; Maercker & Schutzwohl, 1997).

Until now, no-one has compared a war affected population with a culturally and economically equivalent population that was not affected by war. Therefore, the isolated influence of war experiences on the mental health of the affected population has not been empirically confirmed. At the same time, no-one has examined a population of refugees with the corresponding population in their country of origin, so again there is no proof of the influence of forced migration on mental health. These questions should stimulate fur- 37 ther research in this field, and our own study presents an attempt to clarify some of the outstanding questions.

3.3 Criticism of PTSD concept

The popularity of the PTSD concept among clinicians and researchers has rapidly increased within the last two decades. On the other hand, the concept has been subject to much criticism and debate. In recent influential articles and letters to major medical journals, Summerfield and others have criticized the usage of the PTSD concept in general mental health (Summerfield, 2001). In addition, the application of the PTSD diagnosis and the emphasis of mental health support in war-affected societies (Summerfield, 1997; Summerfield, 2002) has been questioned. The critiques have received much attention by the journals’ readers and initiated extended debates (e. g. Shalev, 2001; Van Ommeren, Sharma, & de Jong, 1997). Summerfield’s arguments have also had a major influence on donors, mental health consultants and representatives of different psychosocial aid organizations especially in the field of refugee mental health. Anyone who works in this field will sooner or later be confronted with this discussion.

3.3.1 The validity of the PTSD concept

The main criticism relates to the validity of the PTSD diagnostic category (Summerfield, 2001). According to Summerfield, PTSD is nothing but the medical description of normal suffering after a bad life event. In a historical analysis of the PTSD concept, he claimed that the main political and social background that supported the “invention” of PTSD was the antiwar movement in the USA in the years after the Vietnam war. At this time, the US society was confronted with many veterans who returned from the war thea- ter and had difficulties adjusting to their roles in a peacetime society. Confronted with a military psychiatry that used a classical understanding of psychiatric diseases, which im- plied a developmental and genetic vulnerability for the affected patients, the proponents of PTSD claimed that psychiatric disorders could be solely caused by war experiences and needed specialized care. Consequently, as Summerfield pointed out, the acceptance of the PTSD concept resulted in Vietnam veterans no longer being seen as perpetrators of war 38 atrocities but as victims of traumatic events who had a right to disability pensions and a moral “exculpation” for war crimes.

Summerfield further argued that classifying suffering from a stressful event as a mental disease is not justified, because no specific biological markers have been identified and no objective criteria are required for a diagnosis of PTSD. Instead, the invention of PTSD followed a recent social development towards an “expressive individualism”, which is also reflected in an increasing social disenchantment, including the emphasis of individual personal rights. According to Summerfield, a main motivation for the introduction of PTSD was the increasing tendency to claim for compensation and disability pensions after stressful events.

This argument is reminiscent of the long-going discussion on compensation neurosis (Kinzie & Goetz, 1996). German medicine especially was confronted with the necessity to discriminate between legitimately ill, those exaggerating complaints and those simulating a disorder after the introduction of damage claim laws in 1871 and compulsory insurance in 1884. At the height of this discussion in the 1920s, it became almost impossible to receive any compensation, as the main factor for the development of physical symptoms in claimants was seen in the secondary gain. Many claimants consequently received the diagnosis of rentenneurose (compensation neurosis) or even “psychopathy”. Contrary to Summerfield’s analysis, the problem of the potential relationships between secondary gain, exaggeration and poor treatment outcome is not at all a recent phenomenon. It may play a role in PTSD (Frueh et al., 2003), but only a minority of PTSD patients have the opportunity to claim for compensation and research into the influence of compensation on PTSD symptoms is still inconclusive.

Unfortunately, Summerfield’s analysis of the social and historical background of PTSD ignores previous notions of syndromes that strongly remind of the current PTSD concept (Kinzie & Goetz, 1996). Accounts of PTSD-like disorders were published as early as the American Civil War. Each major subsequent war, including World Wars I and II, triggered detailed examinations of chronic conditions that strongly resemble PTSD. When railroad traveling became increasingly popular in the 1870s, reports of psychological and physiological consequences of railway accidents appeared. There is no doubt that the relevance of the study of the consequences of traumatic events depends on the current social 39 context, as environments that cause many traumatic events will lead to an increased awareness of the consequences of these events.

It may be true that social developments in the past three decades fostered the development of PTSD. Summerfield’s analysis of the history of the PTSD concept mentions only the US antiwar movement as pressing for the acknowledgement of psychiatric consequences of traumatic events. He fails to mention the other influential social development of that time: the women’s liberation movement. This movement has shed light on the suffering of battered women as well as victims of rape and child sexual abuse. The psychological consequences of rape were examined by Burgess and Holmstrom (1974). The resulting rape trauma syndrome was one of the major precursors of PTSD.

Summerfield contrasts the proposed social developments towards an emphasis of individual suffering with traditional attitudes of stoicism. He points towards the British “stiff upper lip” attitude and the endurance and stoicism emphasized in the Soviet Union (Summerfield, 2001). The value of this change in attitudes depends on one’s philosophical viewpoint. However, the fact that something is not discussed and acknowledged in society does not mean that it does not exist. Before the effects of rape and sexual abuse were recognized, the stoicism in society caused great pressure on the victims to keep silent and deny their suffering. The fact that many rape victims have received adequate treatment since the introduction of PTSD does not mean that normal human suffering has been medicalized.

The fact that historical events and social changes formed the background for the development of PTSD does not undermine the validity of the PTSD concept. The matter of validity is an empirical question, and unfortunately Summerfield’s analyses fall short of a detailed discussion of the relevant studies in this field. Instead, he relies on his own anecdo- tal impressions, personal communications and neglects the presentation of the available good empirical data that could contradict his assumptions. In addition, he presents excep- tionally bad reports in support of his criticisms. The studies cited by Summerfield are not at all representative of the research, as they have serious methodological flaws and consequently had only a minor influence on the scientific discussion.

Several criteria can be used to decide whether a diagnostic classification is valid. However, the availability of specific biological markers for a disease as proposed by Sum- merfield (2001) can definitely not be such a criterion, since no specific biological markers 40 have been identified for other mental disorders, like schizophrenia, depression and anxiety disorders, and even for some medical disorders like migraine. It is unclear whether Sum- merfield would suggest discarding these diagnoses as well. The fact that he differentiates PTSD from “real psychiatric problems” (Summerfield, 1998) shows that he is rather incon- sequent with his own suggested criterion. Besides the problem with this criterion, a re- searcher who studies the literature carefully must acknowledge that the understanding of the neurobiological processes behind PTSD has made a considerable progress in recent years (see chapter 5.5).

As described above (chapter 3.1), the original conceptualization of PTSD has undergone several changes. It has become clear that the traumatic event itself is not the only factor that contributes to the development of PTSD as pre- and post-trauma mediators have been identified. This fact does not question the validity of the classification in general.

It is important to study the level of functional impairment that is associated with PTSD to decide whether the consequences of traumatic events classify as a mental disorder rather than present as normal suffering. Breslau (2001) compared the level of impairment associated with PTSD and other mental diseases (mainly other anxiety disorders and depression) on several indicators including current limitations in activities, missed work, self- assessed health as well as desire to die. The PTSD patients presented with the worst outcome in respect to all criteria, compared to both the individuals with other diagnoses and those without a diagnosis. A striking result from this study was that almost half (46%) of those who were diagnosed with PTSD reported that they have already thought about suicide and 17% reported a suicide attempt. The corresponding rates of suicidal thoughts and attempts were significantly smaller in those with other mental diseases (30%/7%) and no diagnosis (9%/1%). In an analysis of the National Vietnam Veterans Readjustment Study, a detailed study of a representative sample of male Vietnam veterans, Zatzick et al. (1997) compared different quality of life indicators for veterans with PTSD and without PTSD. They found that PTSD was associated with a poorer outcome on almost all variables including physical limitations and current work status, even when the effects of comorbid psychiatric and medical diseases were controlled.

Some research shows that PTSD also has an impact on the social functioning. Inves- tigating the quality of intimate relationships of Vietnam veterans, Riggs, Byrne, Weathers, and Litz (1998) found that 70% of veterans who had PTSD were in relationships character- 41 ized by clinically significant levels of relationship distress; this rate was much lower among veterans without PTSD (30%). In particular, veterans with PTSD presented more aggression towards their intimate partners and are at increased risk for perpetrating domestic violence (Byrne & Riggs, 1996).

Another argument for the functional impairment involved in PTSD arises from the fact that PTSD is associated with high levels of comorbid mental disorders (see 3.1) as well as physical symptoms (McFarlane, Atchison, Rafalowicz, & Papay, 1994; Friedman & Schnurr, 1995). The causal explanation of these associations is unclear as different factors can contribute to this association. These include the overlap of symptom criteria of different mental diseases, common vulnerability factors, increased probability of exposure to traumatic events for patients with certain diagnoses, comorbid symptoms resulting from the exposure to traumatic events rather than PTSD as well as the development of comorbid disorders resulting from the functional impairment caused by PTSD. Chilcoat and Breslau (1998) further analyzed the relationship between PTSD and substance abuse. They concluded that neither common vulnerability factors nor the assumed increased probability of substance abusers to experience traumatic events could explain the high levels of alcohol abuse and dependence among PTSD patients. Instead, they suggested that the most probable explanation for this effect was that of self-medication. This means that PTSD patients tend to consume more alcohol in an effort to reduce the suffering associated with the PTSD symptoms.

The findings presented above confirm that PTSD causes high levels of functional impairment. Another criterion to evaluate whether PTSD is a useful category is the question of whether successful treatment strategies have been developed. In recent years, knowledge about the treatment of PTSD has rapidly increased and several specialized treatment approaches have proven successful for the therapy of PTSD (see chapter 6). Many patients with PTSD symptoms who sought treatment before the introduction of the concept PTSD are likely to have received inadequate treatments resulting from a wrong diagnosis. The study of Breslau (2001) shows that 31% of men and 43% of women with PTSD have already taken medication to gain relief from their symptoms. Research into the treatment of PTSD showed that only special antidepressants are successful for PTSD, and the probability is high that those who received medication from doctors without knowledge about PTSD took ineffective and potentially harmful medication. 42 3.3.2 PTSD in war affected areas

Another major criticism of the PTSD concept relates to the application of the PTSD criteria in war affected countries and to refugees (Bracken, Giller, & Summerfield, 1995; Muecke, 1992; Summerfield, 1997, 1999, 2002; Pupavac, 2002). Aid organizations, including UN organizations and WHO, and related expert consultants were criticized for imposing western mental health concepts on diverse cultures and suggesting the need for extensive trauma programs in these populations. Whereas it is true that many organizations used non-validated checklists and other inappropriate research methods to confirm high prevalence of PTSD in these populations, it is incorrect to state that there is no hard data on PTSD in war affected communities (see 3.2). A high prevalence of PTSD has been found in different war affected populations all over the world, and studies confirmed the association of PTSD with disability (Mollica et al., 1999). There is no evidence for the hypothesis that the prevalence and validity of PTSD depends on cultural factors. Contrary to the statements of Summerfield and colleagues that there is no universal trauma response, current research into the epidemiology of PTSD and the neurobiological mechanisms behind PTSD offer strong arguments for the hypothesis that PTSD is a pathological consequence of traumatic events that occurs in all cultures. There are studies of PTSD from all continents, and local experts from all continents have contributed to the research in this field.

The main criticism does not relate to the empirical question of the occurrence and validity of PTSD in these societies. The argument is that the transfer of western concepts and techniques to war-affected societies in developing countries risks “perpetuating the colonial status of the non-western mind” as every “culture has its own frameworks for mental health, and norms for help-seeking at times of crisis” (Summerfield, 1997, p. 1568.) This argument is based on a clear distinction between western “Eurocentric” cultures and other cultures in non-industrialized countries.

This distinction looks straightforward, but it is doubtful that it is valid. The concept of homogeneous cultures is neither true for industrialized nor for developing countries. In all regions there is a wide diversity of attitudes, values and habits, and neither fron- tiers between countries nor between the industrialized and non-industrialized world can offer valid borderlines of cultural values. The same is true for health and mental concepts, many societies in developing countries have already chosen to adapt mental health concepts 43 developed by western psychiatry and prefer the corresponding treatment methods rather than traditional healing. At the same time, in many rural areas in European countries traditional healing techniques for physical and mental complaints are still popular.

Those clinicians and researchers who emphasize the differences between cultures and advocate noninterference in cultures that are considered to be “traditional” use an ethical argument. Terms such as “culturally sensitive” are now included in mental health proposals and articles as a matter of political correctness. Even on moral grounds it is not straightforward to favor the position of not interfering in cultural traditions, norms and beliefs in psychosocial work. All cultures are constantly changing and the idea that there are any cultures that fully rely on traditional norms and have not been affected by the modern world is nothing more than the romantic view of western minds. The consequences of not interfering in cultural norms would also include withholding knowledge about general scientific methods of objective assessment and evaluation from these cultures. It is these methods that have led to the development of treatment approaches that have proven to be effective and to the identification of less successful or even harmful methods. More generally, the same research methods have led to the industrialization and the leading positions of the industrialized countries. Withholding this knowledge from developing countries might help to leave cultural norms untouched, but at the same time the global discrepancies in development and power remain unchanged. Protecting societies that are considered to be traditional from modern influences risks building cultural reservations of societies that remain dependent on the goodwill of the powerful countries.

Proposals are often made to support traditional healing techniques and rituals in mental health programs in non-industrialized countries. However, whilst there is no thorough evaluation of each of these treatment approaches, there is no reason to favor this position. It is important to know who participates in these treatments, how many of the cli- ents improve and how long this improvement is maintained. The question of which mental health approaches should be supported by psychosocial organizations is a matter of efficacy rather than an ethical question.

There is knowledge about the epidemiology and validity of the PTSD concept in different cultures. Until now, however, there is no treatment study that evaluated the efficacy of different approaches in war-affected populations in developing countries. At this level of knowledge it is an ethical obligation of psychosocial organizations to concentrate 44 research on how the problem of mental health in these societies can be approached most effectively. Summerfield is right to reject the practice of many organizations to implement large-scale psychosocial projects with non-evaluated treatment approaches. The consequence is to demand for more research instead of rejecting the whole field of mental health approaches. 45

4 PTSD in populations affected by the Sudanese war

4.1 The Sudanese Civil War

4.1.1 Historical context

The Sudanese war is a perfect example of an inner-state war that fulfills all the criteria for “New Wars” as defined by Kaldor (1999; see chapter 2.2.2), but it is not at all new (for accounts on the history of the Sudanese war see for example Peterson, 2000; Petterson, 1999). The Sudanese borders were fixed to cover a huge country around the southern section of the Nile river at the Berlin Conference in 1885. The consequence was an artificial union of geographically and ethnically most diverse regions. The northern part of Sudan is covered by widespread desert regions and has been dominated by different Arab tribes and nations. In contrast, the southern part consists of fruitful land, forests, savannas and swamps. It is inhabited by animist and Christian Nilotic and Bantu tribes. Struggles between the north and the south have been going on for many years as the southern region has been the target of Arab slave hunters. During the British predominance in the colonial period, Egyptian troops under British control tried to gain control over the southern regions. As this region is very difficult to access and the southern tribes defended themselves in bloody revolts the British neither gained full control over the southern tribes nor could they stop the slave trade.

The stage for the Sudanese civil war was set at the time of independence from the British predominance in the Anglo-Egyptian condominium in 1956. Fear of exclusion from political and economic control prompted rebel armies from the south to wage war against the northern Arab dominated government of Sudan. In total, for all but 11 of the 46 years of independence, Sudan has been at war. The most recent period of fighting erupted in 1983 when the Sudanese President Ja’far Muhammad Numayri revoked the Southern autonomy that had been granted in 1979 and reintroduced the sharia in the south. Since then, the Sudanese People’s Liberation Army (SPLA) under the leadership of John Garang has been fighting in rivalry with other rebel armies for the independence of Southern Su- dan. 46 4.1.2 Regular armies and irregular forces

A wide variety of forces is active in the Sudanese war. The southern rebels are split into different factions, not all of them follow Garang’s command. Garang originates from the Dinka tribe and never managed to unite the other southern tribes in the SPLA. The fight against dissident rebel leaders has often resulted in severe atrocities among the civilian population ruled by an opposing faction. The southern tribes have their own ancient history of rivalry. In the context of war the traditional raids to take cattle from neighboring tribes have turned into bloody attacks carried out with assault rifles instead of traditional weapons. In 1993, a war between the southern Sudanese Nuer tribe under the command of the SPLA dissident Riek (SPDA faction) and the Dinka evolved after the massacre of several thousand Dinkas in Garang’s hometown Bor. The Sudanese government took advantage of this conflict among their enemies and even supported the Nuer tribe with weapons. A severe famine was the consequence of this conflict as the food stocks of opposing villages were often destroyed and stolen animals spread diseases among the own cattle.

Different more or less regular militias and armies fight on the government’s side as well. During the war it became increasingly difficult to recruit combatants for the National Army. Young Sudanese who originated from southern regions were forcibly recruited. Consequently, the national army included mainly unmotivated fighters. The additional Popular Defence Force (PDF) was formed in the context of the transformation of the Su- dan into a fundamentalist regime under the dicator Umar al-Bashir. The PDF consists of volunteer fighters who believe they are fighting a jihad against the southern rebels. The approximately 100,000 PDF fighters are known to be ready to be martyrs and to justify atrocities in the name of Allah.

In addition to these formations, the so-called murahaleen militias are supplied by the government. The murahaleen are tribesmen from the Arab Baggara tribe who have been in conflict with the southern tribes since ancient times. The murahaleen are known to attack villages on horses and camels. They loot and burn villages, murder the inhabitants and till today they are known to abduct women and children as slaves (International Eminent Per- sons Group, 2002, Martin, 2002). The murahaleen are often used to protect the railway transport of supplies to the southern garrison towns along the railway to Wau, terrorizing the villages near the railway and using the trains on their way back to transport slaves. The 47

Map 4.1. Political map of southern Sudan. From Michael S. Miller, Rightsmaps, © 2001. Reprinted with permission. following testimony of a child slave who was abducted by murahaleen was recorded by the International Eminent Persons Group (2002):

“Name: A. Age: Fifteen Village: Marial Bai. Ethnic Group: Dinka Taken in 1998 in Nyamlel. The murahaleen came with horses and started shooting. She did not see government soldiers. She was taken to the River Kiir on foot with many other girls and boys. The girls and boys were kept separate. Many died along the way. It was 17 days walk to the River Kiir. At the River Kiir, they were put in pens. They were given little food and beaten if asked for food. They were handed over to their masters at the River Kiir. Her master was M who took her to Omdurban village. She became a concubine. The master had five wives who treated her badly. She did not go to school and did housework. The master gave her hard work and when she refused, he would send her to the pens. After three years she was moved to another place. She was brought back to Marial Bai by AM who bought her from the last master. She does not know where he got the money.” (International Eminent Persons Group, 2002, p. 54) 48 4.1.3 International involvement

Foreign countries

The civil war has never been restricted to the Sudanese nation. When the Sudanese dictator Umar al-Bashir overthrew the democratically elected government of Sadiq al- Mahdi in 1989, he introduced a fundamentalist Islamic regime in Sudan. Many severe human rights violations followed this transformation. Al-Bashir’s regime found support from other Islamic countries, like Iran, Iraq and Libya. Beyond the moral support for the fighters, these countries provided money and arms. On the other side, the SPLA initially received support from the then socialistic Ethiopian government that supplied the rebels with weapons and held SPLA training camps in Ethiopia. When the socialistic Ethiopian president Mengistu was overthrown in 1991, the SPLA lost this sponsor. Later on the SPLA received supplies from the US and presumably the Israeli government as these governments had an interest in destabilizing the fundamentalist al-Bashir regime. Ongoing material support was also delivered by the Ugandan government. As a counterweight for the Ugandan assistance of the SPLA, the Sudanese government backed the Ugandan fundamental Christian “Lord’s Resistance Army” that turned the Ugandan East Nile region into a war area by committing severe atrocities among the Ugandan civilian population and Su- danese refugees.

Humanitarian assistance

Humanitarian organizations have played a controversial role in this war and may even have contributed to the stabilization of the conflict. In 1989, the UN initiated the Operation Lifeline Sudan (OLS) to bring help to the south. OLS was established as a consequence of the failure of the international community to respond to a major famine in 1988 (Human Rights Watch, 1999). This initiative has depended on the goodwill of the government to allow access to southern regions, but Khartoum has often prevented access to some of the most needy regions if the assistance of these areas did not meet their tactical objectives. As the government had great difficulties in supplying the isolated garrison towns in the South, the relief food was welcome to feed the civilians as well as the government troops. In the late 1990s, the SPLA demanded a “memorandum of consent” from the 49 aid organizations that were active in the regions ruled by the rebels. Consequently, aid organizations were forced to share their resources with all war parties.

During the war, relief food became one of the major goods to fight for. Rebel armies have been known to move their headquarters to famine areas to gain access to the relief food and turned these places into war theaters for rivaling factions. Reports indicated that whole communities were instructed to move to attract relief food to places under control of certain rebel leaders.

In addition, in many regions that have been controlled by rebels, aid organizations have taken on most of the tasks that should have been provided by a government bureauc- racy. In this way, aid organizations released the rebels from the responsibility of caring for their civil population and allowed them to concentrate on purchasing weapons and supplying the combatants. This dilemma is not easy to solve, as it is very unlikely that the rebels and the government would be ready to share their resources with the civil population if the aid organizations withdrew. But the humanitarian aid has been criticized for preventing a “Darwinistic” solution to the war that would mean accepting the starvation of many people in the short term but bringing the war closer to a termination by the “survival of the fittest” (Peterson, 2000).

4.1.4 Civilian targets

In the last 19 years of fighting, the Sudanese war has received little attention from the public in industrialized countries except for single periods of attention provoked by fundraising activities of aid organizations during severe famines. In the literature on developmental policy, the Sudanese war has had an unenviable reputation. Journals repeatedly reported severe human rights violations and atrocities committed by all war parties1. The reports indicated that the main targets of the fighting parties were civilians rather than armies. In recent years, there were ongoing credible reports by witnesses from aid and human rights organizations about mass rapes, child soldiers recruited by force, torture, slavery, concentration camps and the intended bombardment of schools and hospitals (Human Rights Watch, 1993; Amnesty International, 2002; Stieglitz, 1998, 1999). The aims of these 50 human rights violations are diverse. The violence is often used to frighten the civilian population and to punish them for supporting enemy forces. In recent years, the discovery of oil fields has led government forces to use any violent means to forcibly displace whole communities from the regions around the oil fields.

The number of casualties of this war is very difficult to count, as only a minority of the victims died in the battlefield. Instead, the war has caused many fatalities amongst civilians through bombardments and massacres as well as epidemic plagues and diseases. The major Sudanese famines that resulted in thousands starving during the past 15 years were not solely caused by unavoidable droughts. Without human rights violations committed by war parties there would probably not have been any severe famine (Human Rights Watch, 1999).

Estimates of the number of fatalities caused by the war are around two million, this would be more than the number of casualties caused by the wars in Angola, Bosnia, Chechnya, Kosovo, Liberia, the Persian Gulf, Sierra Leone, Somalia and Rwanda taken together (Martin, 2002). Current estimates place the number of Sudanese refugees at 460,000 and internally displaced at four million. As most of those people who were forced to flee came from the south, this means that almost the entire southern Sudanese population (five million) has been displaced (USCR, 2001).

1 For reports on Sudan see for example Internationales Afrikaforum: 2000, 36, 2, p. 218; 2000, 36, 1, p. 17, 1999, 35, 3, p. 219, 1998, 34, 4, p. 319, 1998, 34, 3, p. 214, 1998, 34, 2, p. 118, 1997, 33, 4, p. 319 51 4.1.5 Economic motivation

The Sudanese war has often been described as a religious war. Religion has indeed been used to motivate fighters on both sides and to prevent the troops from tiring of the war, but the real causes of the war are more complicated and include first of all economic factors. Southern Sudan is assumed to be rich with oil and mineral resources. In 1999, the Sudanese government managed to open a pipeline from southern oil fields to the Red Sea and consequently gained an internal source of revenue for the war expenses. Since then, international oil companies have been criticized for their cooperation with the Sudanese regime and their obvious contribution in financing the civil war (Martin, 2002). Major fights are now concentrated around the oil fields and the pipeline with the government at- tempting to displace the local population from the surrounding area and the SPLA sabotag- ing the oil production. Map 4.2 shows the concessions of oil and gas in Sudan that have al- 52 ready been licensed to local and foreign companies. Note that almost the whole area of presumed oil fields is located in regions currently controlled by southern rebels.

4.1.6 Termination of the conflict

At the time this thesis was written, the war was far from reaching a termination point. In 2002, the government still controlled the major garrison towns in the South (Martin, 2002), including the southern Sudanese capital Juba, whilst the SPLA or one of the factious rebel groups ruled the rest of the region. Over the years a continuous pattern of fighting has evolved with the government conquering large territories during the dry season which they lose again as their heavy weapons and supply transports get stuck during the rainy season. There were many peace negotiations and initiatives from surrounding countries, but none of them gained more than a short period of regionally restricted ceasefire (Lesch, 1999). In previous peace talks the Sudanese government had suggested a referendum for Southern independence, but this offer was refused by the SPLM (the political arm of the SPLA) as Garang insisted on the resignation of al-Bashir’s regime.

After the terrorist attacks of the September 11th 2001 in New York the Sudanese government has taken an opportunistic position towards the US and agreed to participate in the “war against terror”. The Sudanese regime clearly hoped to ease tensions with the US and with neighboring countries. The Ugandan army was even allowed to fight the LRA (which had previously been supported by the Sudanese government) deep in the Su- danese territory. At the same time, many factors oppose the hope for a peaceful solution to the civil war. Even if the war parties agreed on serious talks about independence, the fact that the major oil fields are south of the traditional divide would cause difficult negotiations about the borderline. As far as the SPLA leaders are concerned, it is unclear whether they are ready for peace as the war has prevented them from the necessity to legitimize their power in a democratic election and has provided most of them with a rather convenient life in exile offices. In addition, many observers fear a liberation of Southern Sudan as this might cause the outbreak of new wars among the southern tribes for predominance in southern Sudan. Nevertheless, in early 2003 a Kenyan peace initiative led to the agreement of a new ceasefire. The war parties have set up a framework for a peaceful solution including a referendum about the independence of southern Sudan. New hopes were elicited by 53 this initiative, but reports about violations of the ceasefire on both sides have already limited expectations.

4.2 Sudanese refugees in Uganda’s West Nile region

Steady streams of Sudanese refugees have crossed the border to Uganda since the lat- ter half of the 50s (van der Gaag, 1996). About 200,000 Sudanese have sought refuge in Uganda, about half of them in the Ugandan West Nile region (USCR, 2001). Among those, 12,000 have been settled in the UNHCR assisted Imvepi refugee settlement and 8,000 have opted to self-settle in the border town of Koboko.

The West Nile region of Uganda turned out to be anything but a safe exile for the Sudanese refugees. Uganda has its own history of cruel civil wars and quite often the Suda- nese refugees got roped into the struggles. The fights in Uganda sometimes forced Sudanese refugees to return to Sudan, and often they were accompanied by Ugandan refugees who had to flee as well. In addition, the Sudanese war had a major impact on the security situation in the West Nile region of Uganda as well, since the SPLA has been active in this region. At the same time, Ugandan rebel armies have been supported by the Sudanese government to pay the Ugandans back for their support of the SPLA.

In the peaceful period after 1972, approximately 78,000 Sudanese refugees returned to Sudan, fleeing a struggle for power between the Ugandan colonel Idi Amin and Milton Obote who was Ugandan president at the time. Amin, who originated from the West Nile region, overthrew Obote but was himself beaten in 1979. At this time the security situation in the Ugandan West Nile region deteriorated again. The Ugandans were severely affected, with over 163,000 seeking refuge in southern Sudan (Gersony, 1997). The insecurity that prevailed among the settled and self-settled Ugandan refugees in southern Sudan at this time has been well documented (Harrell-Bond, 1986). 54

After Yoweri Museveni’s National Resistance Army (NRA) took Kampala, Ugan- dan refugees started to return to their homes in Arua district in stages between 1986 and 1989. They were followed by over 100,000 Sudanese refugees who were fleeing the atrocities of the Sudanese government troops as well as the SPLA forces. By 1994, the Sudanese government increased support to the Ugandan rebel army West Nile Bank Front (WNBF) 55 led by Juma Oris, once Idi Amin’s Foreign Minister. The WNBF attacked Sudanese refugees settlements, looting, raping, mutilating and killing more than several hundred refugees (Gersony, 1997; Kearney, 1999; USCR, 2001; S. M. Weine et al., 1995). In 1997, the attack on Ikafe refugee settlement and the atrocities that followed precipitated the return of many refugees to southern Sudan, while others self-settled in Koboko or moved to the nearby Imvepi refugee settlement.

SPLA continues to be a source of insecurity for Sudanese refugees in Uganda. Armed soldiers have been known to put pressure on refugees to return, as well as to abduct and torture them. Young Sudanese males are also being forcibly conscripted into their ranks. Despite the continuous pressure of the SPLA on Sudanese refugees, the situation in the West Nile region was quite safe for the past decade. In contrast, the Ugandan East Nile region, which hosts many thousands of Sudanese refugees, is still very unsafe, as Ugandan rebels, most of all the Lord’s Resistance Army (LRA) continues to commit atrocities among the Ugandan nationals as well as the Sudanese refugee settlements2.

4.3 Survey of West Nile populations

4.3.1 The “Demography of Forced Migration“ Study

Between 1999 and 2001, a large survey was carried out in the West Nile regions of Sudan and Uganda. This “Demography of Forced Migration Study“ was a cooperation between John Hopkins University, Makarere University in Kampala and University of Kon- stanz. The study was supported by Medicins sans Frontieres (MSF) Holland, who were an active aid organization in this region. They provided logistical and medical support for the study. The first step of the study was a survey of Sudanese living in the war region This group was compared to Sudanese refugees in Uganda as well as Ugandan nationals living in this region. The survey studied demographic and health characteristics as well as the security situation in these populations. The study included the assessment of posttraumatic stress disorder. In this thesis, the focus is on the analysis of the data on PTSD, as the analy-

2 The ongoing attacks on refugee settlements still receive little attention even by the Ugandan public. For examples of remarkably short reports on those attacks during the period of this investigation see Ugan- dan Monitor, August 14, 2000, p. 3. 56 sis of demography and fertility is the subject of different theses. In a second step, the quality of the study, in particular the mental health variables, was examined in a validation study that was carried out with a sample of Sudanese refugees living in the Imvepi refugee settlement in Uganda. The third step was a treatment study that investigated the efficacy of different short-term treatment approaches for individuals suffering from PTSD in Imvepi.

4.3.2 Methods

Survey Population

A single-round demographic survey of residents of the sub-counties of Odupi, Midia and Yivu in the West Nile area of northern Uganda and of Otogo in the West Nile region of southern Sudan was conducted between September 1, 1999 and March 4, 2000. Cross- sectional and retrospective data on household and individual characteristics was collected.

A multi-stage sampling technique was used to select sites for the survey. The sub- counties were selected purposively based on the absence or presence of the refugees. Odupi hosts the Imvepi Refugee Settlement, while Midia, closer to the Sudanese border, is host to thousands of self-settled refugees who receive no assistance. Yivu, with no refugees, served as a comparison group in the study of the effect of post migration residential arrangements on the lives of refugees as well as their hosts. The relatively stable security situation that existed at the time of the survey, allowed access to Otogo, home to many of the refugees surveyed in Uganda. The different parishes, sub-parishes and villages — the primary sampling unit — to be surveyed in the sub-counties were selected randomly. Households — the ultimate sampling unit — were selected systematically on the day of the interviews. Typi- cally, one of the two or three possible directions from the village school or church was randomly selected. Starting with the first household, subsequent households were selected systematically (e. g. every third household) based on the estimated population of the village.

Sample size calculation was based on under-5 mortality figures made available by Uganda Demographic and Health Surveys (DHS) 1995 and UNHCR. Mortality rates were used for sample size calculations, as mortality required the largest sample size, since it is the rarest of the three events (fertility, mortality and violence) studied. As most of the survey’s topics referred to women’s issues, the sample size was calculated for women. One man per 57 three women was randomly chosen out of each household to receive information about men. A sample size of 2958 women and 888 men was calculated to provide sufficient power for statistical analyses.

Survey Questionnaire

A demographic questionnaire (demography of forced migration questionnaire, DFMQ) with a section on household composition and characteristics (Part A with 60 questions) and another section (Part B with 257 questions) on individual characteristics, repro- ductive and child health as well as migration and security histories was developed. Qualita- tive research provided an understanding of local cultural and social norms as well as perceptions of fertility, mortality and violence in refugee and national communities. An event calendar was created to support the respondent’s assessment of age.

Section B included several questions about physical symptoms and diseases. The respondents were asked if they had been suffered in the month before the interview from any of the following problems: cough, diarrhea/stomach pain, fever/shivering/malaria, tuberculosis, leprosy, asthma, diabetes, hypertension, epilepsy. On a four point scale respondents were asked to rate how many times they consumed alcohol in the past month.

The questionnaire was written in English and translated to Lugbara for the Ugan- dans and to Arabic (colloquial Juba version) for the Sudanese. This was achieved by using back-translation techniques (Flanagan, 1999).

PTSD in respondents was assessed using a Posttraumatic Stress Diagnostic Scale (PDS), modified for assessment by trained lay interviewers (Foa, 1995). PDS is widely used in industrialized countries as a screening instrument for the diagnosis and severity of PTSD based on DSM-IV Criteria.

Traumatic events were assessed using a checklist consisting of possible war and non- war related traumatic event types that turned out to be of relevance in qualitative research. Events included 19 experienced events and 12 witnessed events (see Table 4.2). Respondents were asked for each event type if they had experienced or witnessed such an event ever (i.e., lifetime experience) and if it happened in the past year.

As most respondents did not have regular incomes, socio-economic status was determined by household assets. The number of essential assets (blankets, pots, etc.) divided 58 by the number of adults living in the household was used as estimate for the economic status of a household. The number of meals on the previous day was chosen as indicator of the respondent’s nutritional status.

Interviews

Twenty-four interviewers, two supervisors and a project assistant were hired locally from the Ugandan national and Sudanese refugee communities. Project researchers and a consultant trained the team in quantitative research methodology and interviewing techniques. A workshop on sexual and gender-based-violence was conducted before the survey to increase awareness and sensitivity of the team towards respondents.

All interviews were conducted in Lugbara and Arabic within a six-month period. All female respondents were interviewed by female interviewers. Male respondents in the same household were interviewed simultaneously by male interviewers. All family members aged 15-55 years were eligible to be interviewed. One eligible respondent, usually the head of the household, responded to the household questionnaire. Individual questionnaires were administered to all eligible females and randomly selected eligible males (30% of eligible females) in a household. Signed informed consent was obtained. The respondents received no monetary reward for taking part in the survey. The survey was approved by the Uganda National Council of Science and Technology, Johns Hopkins University School of Public Health Committee on Human Research and the University of Konstanz Ethical Review Board.

4.3.3 Results

Demographic characteristics of 3323 individuals, categorized by citizenship status were analyzed. The average age of the respondent was 30. The majority of respondents were female (75.1%) and had never attended school (76.8%). Thirty-seven per cent of the respondents identified themselves as refugees in Uganda at the time of the survey and 76.8% of the respondents reported at least one experience of forced migration in their life. Characteristics of the three populations — Ugandan nationals (42.7%), Sudanese refugees (37.3%) and Sudanese nationals (20%) — are presented in Table 4.1. Chi square tests were calculated for categorical variables and ANOVAs for continuous variables to analyze differences between the three groups of people on demographic variables. 59

Table 4.1. Population Characteristics by Citizenship and Refugee Status

Ugandan Sudanese Sudanese Nationals Refugees Nationals (n = 1419) (n = 1240) (n = 664) Statistics p Age M (SD) 30.1 (10.0) 29.3 (9.8) 32.5 (11.3) F = 7.48 ns Sex n (%) c2 = 7.60 < 0.1 Male 379 (26.7) 274 (22.1) 164 (24.7) ! Female 1040 (73.3) 966 (77.9) 500 (75.3) ! ! Marital status n (%) c2 = 113.95 < 0.001 Single 149 (10.5) 284 (22.9) 94 (14.2) Married / Cohabiting 1077 (75.9) 845 (68.2) 455 (68.5) Separated / Divorced 64 (4.5) 25 (2.0) 25 (3.8) Widowed 34 (2.4) 52 (4.2) 51 (7.7) ! Missing 95 (6.7) 34 (2.7) 39 (5.9) ! ! Religion n (%) c2 = 490.22 < 0.001 Catholic 902 (63.6) 454 (36.6) 256 (38.6) Protestant / Other Christian 264 (18.5) 709 (57.3) 360 (54.2) Other 164 (11.6) 54 (4.4) 9 (1.4) ! Missing 90 (6.3) 22 (1.8) 39 (5.8) ! ! Schooling n (%) c2 = 22.35 < 0.001 None 1038 (73.2) 970 (78.2) 545 (82.1) Some 289 (20.4) 237 (19.1) 80 (12.1) ! Missing 92 (6.5) 33 (2.7) 39 (5.9) ! ! Occupation n (%) c2 = 804.34 < 0.001 None 122 (8.6) 334 (26.9) 48 (7.2) Agriculture 947 (66.7) 311 (25.1) 478 (72.0) Trade / Crafts 72 (5.1) 130 (10.5) 24 (3.6) Contract labor 12 (0.9) 210 (16.9) 9 (1.4) Alcohol brewing 84 (5.9) 120 (9.7) 44 (6.6) Other 64 (4.5) 79 (6.4) 19 (2.9) ! Missing 118 (8.3) 56 (4.5) 42 (6.3) ! ! Nr. of possessions M (SD) 2.08 (1.5) 1,24 (1.3) 0.75 (0.9) F = 260.90 < 0.001 Nr. of meals per day M (SD) 2.12 (0.7) 1.34 (0.7) 2.24 (5.4) F = 40.06 < 0.001 Experience of migration n (%) c2 = 499.01 < 0.001 Yes 938 (66.1) 1213 (97.8) 401 (60.4) ! No 481 (33.9) 27 (2.2) 263 (39.6) ! ! Years since last migration M (SD) 13.0 (11.9) 2.70 (3.2) 8.0 (11.9) F = 369.71 < 0.001 60 The three groups differed substantially in their socio-economic characteristics. The Ugandan (66.8%) and Sudanese nationals (72.0%) were mostly subsistence farmers while only 25.1% of Sudanese refugees were occupied with agriculture. Most refugees were unemployed (26.9%) or were temporary contract workers (16.9%). As expected, Ugandan nationals were better off, economically, than the Sudanese groups in terms of household possessions. Daily nutrition intake seemed to be worst among the Sudanese refugees, who stated to have fewer meals per day than Ugandan or Sudanese nationals.

Prevalence of traumatic events ever and in the past one year for the three population groups are presented in Table 4.2. Most of the traumatic events — ever or in the past year — were more prevalent in Sudanese refugees. The magnitude of exposure to traumatic events for the three population groups was estimated by calculating the mean number of the different types of traumatic events reported by respondents, separately for experiences ever or in the past year. The resulting values are presented in Table 4.3. The mean number of different types of traumatic events experienced differed significantly between the three populations. The Sudanese refugees had the highest exposure to different types of traumatic events. Post-hoc analyses revealed that each population group differed from the other groups (p < 0.001) for traumatic event exposure ever, with the Sudanese nationals reporting more events than Ugandan nationals. Post-hoc analyses also revealed that only the Su- danese refugees differed from the other populations with respect to traumatic event exposure in the past year, with the difference between the Ugandan and the Sudanese nationals not being significant. Traumatic event exposure, ever and in the past year, were further tabulated by experienced and witnessed events. The differences in the three population groups were significant with Sudanese refugees having reporting the highest numbers of experienced and witnessed traumatic events. 61 Table 4.2. Prevalence (%) of traumatic event types experienced ever and in the past year for the West-Nile populations

ever in the past year

Ugandan Suda- Suda- Ugandan Suda- Suda- Nationals nese nese Na- Nationals nese nese Na- Refu- tionals Refugees tionals gees Abduction 3.9 17.9 8.5 1.5 8.5 1.4 Accidents 34.6 25.8 12.2 15.6 12.0 4.0 Beatings from spouse 27.9 12.2 10.1 12.1 7.5 1.7 Beatings or torture 27.4 34.9 29.2 8.4 14.8 1.3 Child marriage 6.6 12.2 5.0 1.3 9.1 0.4 Combat situation 29.8 19.9 12.9 2.8 12.3 5.0 Confiscation of property by offi- 10.9 57.8 65.3 3.9 18.2 5.4 cials Dangerous evacuation 7.9 35.8 19.3 1.5 13.1 1.7 Experienced injury by weapon 11.8 15.6 4.1 3.3 9.9 0.2 Forced circumcision 1.9 7.3 1.1 1.6 7.5 0.6 Forced isolation 7.9 16.4 5.9 6.3 11.0 0.7 Forced marriage 6.8 12.3 3.6 1.9 9.3 0.4 Forced prostitution / sexual 1.0 6.3 1.1 0.7 7.0 0.4 slavery Harassment of armed personnel 32.9 27.0 16.3 4.8 11.8 1.5 Imprisonment 9.1 17.9 8.0 3.4 10.9 2.2 Poisoning/witchcraft 44.1 15.6 10.0 16.0 10.6 1.3 Rape 3.5 10.0 2.4 1.5 7.8 0.2 Robbery / extortion 23.8 29.4 25.2 5.7 17.7 4.8 Sex for food or security 1.1 6.5 0.5 0.5 6.8 0.4 Witnessed abduction 11.9 61.2 51.6 5.2 32.3 26.1 Witnessed accident 52.6 53.4 36.7 38.6 28.0 10.8 Witnessed act of suicide 45.5 44.5 24.1 24.3 22.2 12.6 Witnessed beatings or torture 55.7 66.8 56.4 25.7 31.5 11.8 Witnessed combat situation 37.9 40.8 23.2 5.6 26.0 9.1 Witnessed forced circumcision 6.8 31.5 15.1 4.1 21.4 10.4 Witnessed forced prostitution 10.4 45.0 35.3 5.5 29.6 26.8 Witnessed harassment of 50.9 63.9 54.6 11.8 27.4 14.2 armed personnel Witnessed injury by weapon 44.5 65.7 53.9 11.2 28.8 13.6 Witnessed murder 27.6 55.2 46.7 14.7 33.1 23.4 Witnessed rape of a woman 13.6 48.1 36.9 8.1 27.9 20.3 Witnessed robbery / extortion 35.4 56.1 45.5 12.4 31.0 15.5 62 Table 4.3. Mean number of different types of traumatic events in the three West- Nile populations

Ugandan Sudanese Sudanese Nationals Refugees Nationals Statistic P total ever (SD) 6.05 (4.18) 9.87 (7.14) 7.11 (5.13) F = 151.31 < .001 total in the past year (SD) 1.67 (2.10) 4.48 (7.04) 1.78 (2.32) F = 138.32 < .001 experienced ever (SD) 2.58 (2.07) 3.71 (4.60) 2.38 (3.01) F = 49.11 < .001 witnessed ever (SD) 3.47 (2.69) 6.16 (3.57) 4.73 (3.17) F = 242.79 < .001 experienced in the past year 0.60 (1.04) 1.70 (4.24) 0.27 (0.76) F = 80.05 < .001 (SD) witnessed in the past year (SD) 1.07 (1.42) 2.78 (3.43) 1.51 (2.02) F = 163.71 < .001

Prevalence rates of PTSD for the three population groups are shown in Figure 4.1. The prevalence of PTSD, assuming a 1/1 males/females sex ratio in all populations, was 19.7 % for the Ugandan nationals, 48.7% for the Sudanese nationals and 47.7% for the Su- danese refugees. The difference in PTSD prevalence between the groups is significant (p < 0.01) with a significant pair wise difference in prevalence between Ugandan nationals and the Sudanese refugees (p < 0.01), the Ugandan and the Sudanese nationals (p < 0.01). The prevalence of PTSD among males and females differ significantly in the Ugandan national (p < 0.01) Sudanese national (p < 0.01) and the Sudanese refugee (p < 0.01) population groups. However, the difference in the prevalence of PTSD among the sexes show dramatically different patterns in Uganda-resident populations and the Sudan-resident population. While more females suffer from PTSD than males in the Ugandan national and Sudanese refugee populations, there is a significant reversal in this ratio in Sudanese nationals. 63

males 30 females population

0 Ugandan nationals Sudanese nationals Sudanese refugees

Figure 4.1. Prevalence of PTSD for three population groups, by sex.

A binary regression analysis was carried out to determine which variables significantly contributed to the diagnosis of PTSD, controlling for the effect of other variables. The analysis revealed the independent significance of some socio-demographic, migration and trauma related variables in the prediction of PTSD, after controlling for the influence of the other variables. The variables were entered simultaneously. The categorical variables were entered using a deviation coding scheme. Consequently, the values for each category represented the deviation from the average effect of all categories. The results of the analysis are presented in Table 4.4. Adjusted R2 of the equation was 0.29.

The likelihood of PTSD increased with age, with women being more likely to suffer from PTSD than men. Likelihood of PTSD also increased with the number of traumatic events experienced and witnessed, most pronounced in those who witnessed traumatic events in the past year. Having a history of migration did not change the probability 64

Table 4.4. Odds ratios (including 95% confidence intervals), standard errors, and Wald statistics of possible predictors of current PTSD in three West-Nile populations

Predictor OR (95% C.I.) SE Wald p Sex (male) 0.35 (0.28-0.44) .118 75.964 <.001 Age 1.02 (1.01-1.03) .005 24.907 <.001 Some education 0.63 (0.49-0.82) .131 12.051 <.001 Migration factors History of migration 1.20 (0.95-1.52) .119 2.448 ns Years since last migration 0.99 (0.98-0.99) .005 4.760 <0.5 Economic factors Number of possessions 0.95 (0.89-1.03) .036 1.588 ns Number of meals 0.96 (0.89-1.03) .037 1.237 ns Religion 7.913 ns Catholic 0.95 (0.82-1.11) .750 0.383 ns Protestant 1.21 (1.04-1.41) .077 6.375 ns Islam 0.86 (0.68-1.09) .120 1.533 ns Marital status 6.652 ns Single 0.97 (0.75-1.26) .132 .053 ns Married 1.10 (0.91-1.32) .095 .983 ns Separated 0.66 (0.45-0.96) .191 4.669 <0.5 Widowed 1.42 (1.02-1.98) .171 4.198 <0.5 Occupation 18.319 <.01 No occupation 1.13 (0.90-1.42) .115 1.229 ns Agriculture 1.18 (0.99-1.39) .083 3.884 ns Trading 1.18 (.88-1.59) .150 1.274 ns Contract labor 1.43 (1.07-1.91) .148 5.885 <0.5 Brewing alcohol 0.75 (0.55-1.01) .152 3.563 ns Other 0.58 (0.40-0.84) .185 8.301 <.01 Number of traumatic events experienced in life 1.05 (1.01-1.10) .021 6.812 <0.1 witnessed in life 1.12 (1.08-1.16) .018 44.661 <.001 experienced last year 1.05 (0.97-1.13) .039 1.517 ns witnessed last year 1.32 (1.25-1.39) .025 121.508 <.001 of PTSD, whereas those with a recent migration did have a higher risk. Some education seemed to reduce the probability of PTSD in the sampled population. Widows seem to 65 have a higher risk for PTSD and people who live separately from their partners seem to have a lower probability, but the migration status did not turn out to be a significant variable. People who depended on contract labor presented with a higher risk for PTSD. Relig- ion and economic variables did not significantly contribute to the prediction of PTSD.

4.3.4 Discussion

In a demographic survey of three population groups in the West-Nile regions of Uganda and Sudan, we examined the prevalence of PTSD in Sudanese nationals, Sudanese refugees and their Ugandan hosts. The extent of PTSD in these population groups, affected by war and forced migration, and the impact of socio-economic factors and the experience of migration and traumatic events in the development and maintenance of PTSD were examined.

A higher prevalence of PTSD was observed in Sudanese nationals (48.7%) and refugees (47.7%), groups exposed to decades of insecurity and conflict compared to their Ugan- dan hosts (19.7%) who have had a relatively peaceful existence in the past decade. This is a clear indication of a substantial mental health burden especially in the Sudanese populations. Differences in prevalence between Sudanese and Ugandan populations are also large and significant enough to discount a potential cultural bias in the reporting of PTSD symptoms. In spite of ethnic and cultural similarities, there were significant socio-economic differences between the three groups. Important differences were found with regard to the respondents’ marital status, religion, occupation, education, economical situation and migration history. Even though a high number of Ugandans report having had at least one experience of forced migration (66.1%), political stability has helped improve their lives. They have more possessions and are better educated than the Sudanese.

War has wreaked havoc on the lives of the Sudanese. The breakdown of the economy has meant that the Sudanese nationals have very few possessions. Poverty combined with insecurity has led many Sudanese to seek refuge in Uganda where they have access to food, health care and education. However, there are signs that food aid has not had its intended effect. Sudanese refugees reported having just 1.34 meals a day. Both Ugandan (2.12) and Sudanese (2.24) nationals reported having more meals, highlighting the importance of land ownership, agriculture and nutrition. Consider also the fact that only 25.1% of the 66 refugees were able to grow food in their current situation, the figures being 66.7% for the Ugandans and 72% for Sudanese nationals.

There are significant differences in violent and traumatic events experienced by the three population groups. Not surprisingly, Sudanese groups, refugees followed by nationals, reported the highest exposure, both in terms of magnitude and forms of events. The respondents who continued to live in southern Sudan experienced a relatively safe year as the front line in the war between the government and the SPLA had moved elsewhere. Refugees, on the other hand, continue to experience a high level of insecurity in their country of exile. They have been threatened and attacked by Ugandan as well as Sudanese rebels during their stay in Uganda. Refugees also face considerable threats from within the settlements (Turner, 1999). Events most frequently reported were robberies, extortions, beatings, torture and harassment by officials and armed personnel. This data highlights the need for better protection guarantees for refugees in settlements as they continue to be targets of violence.

The most significant predictor for the development of PTSD, in these West Nile populations, is the witnessing of a traumatic event (particularly a recent one), whereas experiencing such an event was not a significant predictor. Regression analyses show that traumatic events experienced in the preceding year were more predictive than traumatic events ever experienced. Moreover, contrary to our expectations, witnessed events turned out to be more significant in predicting PTSD than experienced events. One explanation could be that the refugees continue to be in an insecure state and that witnessing increases anxiety levels due to the fear that the same could happen to them. Whereas, surviving a violent act may have the effect of limiting levels of anxiety in some of the survivors. On the other hand, it could also be that respondents were unwilling to admit to being a victim of the worst forms of traumatic experiences, like sexual violence. These survivors may in- accurately report having witnessed, rather than experienced, the traumatic event.

Sex, age, education and occupation proved to be significant predictors of PTSD. The finding that being female and being less educated increased the risk of developing PTSD is similar to observations made in North American populations (Kessler et al., 1995). In addition, being older increased the risk of suffering from PTSD. This could be attributed to the loss of family members and social support, subsequent to war and forced migration. The influence of occupation was minimal with the small residual group marked “other” 67 having a smaller probability of suffering from PTSD. This group consisted mainly of those in privileged positions such as politicians and village leaders who might be protected from stressors that affect the rest of the population. When experience of traumatic events is controlled for, a history of migration did not significantly increase the risk for PTSD, whereas those with a recent history of migration seemed to present with a higher probability. Hu- manitarian aid could be credited with softening the blow for refugees by providing food, health care and schools .

An interesting aspect of the picture of PTSD prevalence among Sudanese nationals and refugees is the sex distribution. Whilst Sudanese national and refugee populations showed similar prevalence, there is a significant difference in how the disorder affects males and females. Significantly higher numbers of refugee females suffer from PTSD than their male counterparts. This resembles the pattern observed in Ugandan nationals and other populations (Kessler et al., 1995). However, there is a marked reversal of this sex differen- tial in the Sudanese national population. This could be because women who are most at risk are more likely to seek refuge or because men who stay behind are more likely to be recruited by rebel forces and actively participate in the conflict, therefore experiencing, witnessing or committing more violence.

A major limitation of this study is that the only mental health variable considered by this study is PTSD. In populations that experience war, a high level of depression (Kessler et al., 1995; Mollica et al., 2001) anxiety and dysfunction can be expected in addition to PTSD, resulting in a considerable mental health burden. A previous study in a similar Sudanese refugee population (Peltzer, 1999) showed high levels of depression and its correlation with PTSD. The present study has also not explored the relationship between PTSD and the ability of those affected to cope and function as members of a household and a community.

This study highlights both the high prevalence of violence and the psychological consequences in populations that have been exposed to war and forced migration. Sudanese refugees are worse off than their compatriots who chose to stay in Sudan in terms of being targets of violent acts as well as mental health morbidity. Utmost priority should be given to the protection of refugee populations and much care should be spent on settlement location and planning. Refugees should be hosted in safe areas away from international borders to protect them from the conflict they have fled from. In addition to providing food and 68 health care, an effective policing and justice system within the camp or settlement should help reduce the number of stressors in the life of a refugee. This study also indicates the need for mental health programs directed towards helping those affected by conflicts. Given the number of those affected, mental health programs should make the leap from an individual therapy based approach to public mental health initiatives that are likely to reach more people who are traumatized. Possibilities include the training of paramedical personnel, chosen from the refugee community to assist traumatized individuals. These programs should seek to restore the day-to-day functioning and emotional well-being of the refugees, thereby helping the traumatized to lead productive lives.

4.4 Dose-response effect for PTSD

4.4.1 Introduction

The debate about the impact of traumatic life events on psychiatric disorders has a long tradition in psychiatry. The introduction of posttraumatic stress disorder (PTSD) into the Diagnostic and Statistical Manual of Mental Disorders manifested the general recogni- tion that a chronic condition consisting of characteristic symptoms including involuntary intrusions of the past, avoidance behavior and a condition of general hyperarrousal can be caused by traumatic exposure and must be viewed as mental disorder. Consequently, the original conceptualization of PTSD was based on the implicit assumption that the traumatic event is the main agent for the development of PTSD (Yehuda & McFarlane, 1995). The initial idea was that traumatic events could cause PTSD in anyone regardless of pre- trauma vulnerability.

Contrary to this assumption, the following research showed that the development of a chronic PTSD is rather the exception than the rule after the experience of a traumatic event. Community studies in the US showed that whereas more than 50% of the population reported the experience of a traumatic event, the prevalence of PTSD was not higher than 7.8% (Kessler et al., 1995). Among the different events studied, rape seemed to be the most adverse experience, as about 50% of victims developed a chronic PTSD. But even studies that researched PTSD in those who experienced events considered to be most adverse, like torture in prison, found PTSD prevalence rates of under 50% (Basoglu, Paker, 69 Paker, Ozmen et al., 1994). The realization that traumatic exposure is not a sufficient determinant of PTSD has stimulated vast research into risk and protection factors for the development of PTSD (Brewin et al., 2000; Ozer, Best, Lipsey, & Weiss, 2003). These studies could show that pre-trauma developmental vulnerability (adverse childhood, psychiatric history, etc.) (Brewin et al., 2000) as well as genetic factors (True et al., 1993) mediate the development of PTSD, although effect sizes were generally small.

A popular and intuitively plausible assumption in this context is the dose-response model of PTSD. This hypothesis predicts that the probability for the development of PTSD after the experience of a traumatic event mainly depends on the severity of trauma exposure. Some studies tried to test this hypothesis by relating the objective severity of the traumatic event to symptoms of PTSD. However, the empirical evidence for this model is scarce, with some findings supporting this hypothesis but many failing to confirm a relationship (Bowman, 1999; McNally, 2003).

The probability of detecting a relationship between trauma exposure and PTSD depends on the range and variance of traumatic exposure that is present in the population studied. Studies investigating the relationship between the objective severity of single events and PTSD are restricted to a narrow variance of traumatic exposure. Community studies that assess trauma exposure across different types of traumatic events should be more adequate to examine a dose-effect hypothesis. On a worldwide perspective, even community studies in industrialized countries are restricted to a relatively narrow range of trauma exposure. In contrast, community studies in civil populations affected by war enable the study of a much wider range of traumatic exposure. These populations present a continuum of subjects ranging from individuals without any history of traumatic events up to victims with a history of high numbers of severe events that are rarely to be found in communities without a history of war. Studying a community sample of Cambodian refugees who had fled the Pol Pot regime, Mollica et al. (1998) could actually confirm a clear linear relationship between the number of traumatic events and symptoms of PTSD and depression. Other studies with refugee populations are in line with this result (Fawzi et al., 1997; Lopes Cardozo et al., 2000; Shrestha et al., 1998; Silove, Sinnerbrink, Field, Mani- cavasagar, & Steel, 1997b). These studies suggests a specification of the dose-response model, i.e. that not the severity of a single traumatic event but the severity of previous cumulative trauma exposure is linearly related to symptoms of PTSD. 70 A consequent deduction from this model is the hypothesis that, from a certain threshold of traumatic exposure, each exposed individual will develop PTSD. As this threshold is probably very high, a large number of subjects exposed to a large variance of traumatic events is necessary to test this hypothesis. We examined the dose-response relationship in the context of a large survey in the West-Nile regions of Sudan and Uganda. The study included Ugandan nationals with a quite peaceful development in the last decade, as well as Sudanese nationals living in the Southern Sudan war region and Sudanese refugees who had fled to Uganda. Among these groups we expected a sufficient variance of traumatic exposure to test for the specified dose-response hypothesis, including an adequate number of subjects who had to experience a series of extremely severe traumatic events. Cumulative trauma exposure was estimated by assessing the number of different traumatic event types experienced or witnessed so far. We considered this measurement to be more reliable than assessing the frequency of traumatic events as many survivors of civil wars reported countless exposures to specific traumatic events. We also assessed the traumatic event types experienced or witnessed in the last year to examine the impact of recent traumatic exposure.

4.4.2 Method

As part of a study designed to better understand the impact of forced migration on fertility, mortality, violence and traumatic stress among Sudanese nationals living in southern Sudan and Ugandan nationals and Sudanese refugees living in northern Uganda, we interviewed 3371 individuals from 1842 households in the Ugandan and Sudanese populations in the West Nile. Interviews were structured and were administered in the native languages of Lugbara or Juba Arabic. The study’s design involved a multi-stage sampling design. Data were complete and analyzed for N=3179 respondents: 2,540 (75 %) of the respondents were women (15-50 years of age) and 831 (25%) were men (20-55 years of age).

Traumatic events were assessed using a checklist consisting of possible war and non- war related traumatic event types that turned out to be of relevance in qualitative research (i.e. witnessing or experiencing injury by a weapon or gun, beatings/torture, harassment by armed personnel, robbery/extortion, imprisonment, poisoning, rape or sexual abuse, beatings, abduction, child marriage, forced prostitution/sexual slavery, forced circumcision, etc.). Events included 19 experienced events and 12 witnessed events. Respondents 71

Figure 4.2. Scatterplot of relationship between number of traumatic event types reported for whole life and severity of PTSD symptoms. Note: A number randomly chosen in the interval between -.05 and +0.5 was added to both, the abscissa and the ordinate to visualize overlapping points. were asked for each event type if they had experienced or witnessed such an event ever (i.e., lifetime experience) and if it happened in the past year. PTSD in respondents was assessed using the Posttraumatic Stress Diagnostic Scale (PDS), modified for assessment by trained lay interviewers.

4.4.3 Results

We correlated the PDS score and its subscales, intrusion, avoidance and arousal with the number of event types to examine the relationship between continuous PTSD symptoms and the number of event types reported. The number of event types in life correlated with the frequency of intrusions (r = .49), hyperarousal (r = .41) and avoidance (r = .47), all p < 0.001. The PDS sumscore correlated significantly (p < 0.001) with the number of traumatic event types in the past year (r = .45) and for the whole life (r = .49, see scatterplot in Figure 4.2). 72

Figure 4.3 Prevalence of PTSD and number of individuals in groups of respondents pooled on the basis of number of traumatic event types reported for whole life and last year

In the whole group, 38% of the respondents (N=3179) fulfilled the DSM criteria for a PTSD-diagnosis. We divided the whole population studied in the survey into different groups based on the number of traumatic event types reported, separately for the events reported for last year and in life. The initial division was made as follows: the first group was constructed out of the respondents who had reported 0-3 event types, the second group were the individuals with 4-7 event types, and each following group four more event types. As the number of individuals in the groups of 12-15, 16-19, 20-23 and 24-27 event types were very small for the analyses of events reported last year (n=38, 14, 8, 13, respec- tively), these groups were merged to two groups of 12-19 and 20-27 event types. Figure 4.3 presents the number of individuals and the prevalence of PTSD in these groups, separately for the groups based on the events reported for the whole life and for last year. The presentation indicates a near linear rise for increasing psychological strain with the number of 73 traumatic event types ranging from 23% in respondents who reported three or fewer trau- matizing experiences to 100% prevalence of PTSD in those who report 28 or more traumatic event types in their past. Figures related to traumatic event types in the past year display an even more pronounced increase of PTSD symptoms with significantly higher prevalence rates for the first three categories of numbers of events (Figure 4.3).

4.4.4 Discussion

High prevalence rates of PTSD have been reported for three different population groups in the West Nile: Sudanese nationals (48.7%), Sudanese refugees (47.7%) and Ugan- dan residents (19.7%). Here we show that the exposure to traumatic events and the number of different types of traumatic experiences in particular can account for the different proportion of PTSD cases. The prediction of increased PTSD prevalence with increasing number of traumatic events is consistent with other studies investigating victims of organized violence (Fawzi et al., 1997; Lopes Cardozo et al., 2000; Shrestha et al., 1998; Silove et al., 1997b).

As demonstrated, the number of traumatic events correlated equally strong with avoidance and with re-experiencing symptoms but coefficients were weaker, although still significant, for the hyperarousal cluster. These results are in agreement with Allden et al. (1996), who also found a strong correlation between cumulative trauma and symptoms of re-experiencing and avoidance. Contrary to these findings, Mollica (1998) could not find a correlation with avoidance symptoms. Subtile differences in the formulation translation of the avoidance items in the PTSD instruments might be responsible for this difference, as the PTSD avoidance criteria are specially vulnerable to be understood as unspecific depressive items that are unrelated to a traumatic experience.

Typically, even severe single traumatic events produce PTSD in not more than half of those affected. Therefore, PTSD is not an inevitable consequence of potentially trauma- tizing events. Results from this study, however, suggest that there may be no ultimate resilience to ward off PTSD or that a psychobiological breaking point exists for even the most resistant individual. In the three population groups that were surveyed, each respondent experiencing 28 or more different traumatic event types developed the full set of symptoms of PTSD. This cumulative trauma threshold identified in this study is very high and af- 74 fected only a small minority of persons even in a war-torn population. Nevertheless, if the cumulative exposure to traumatic events is high enough, these results indicate that anybody will develop chronic PTSD. We conclude that there is no ultimate resilience to traumatic stress and that the repeated occurrence of traumatic stress has a cumulative damaging effect on the mental health of the victim. In these conditions, the effect of pre-trauma factors is reduced to the modulation of the probability of exposure to traumatic events itself. The factors that determine who is exposed to many traumatic events and who manages to flee to secure places may depend on pre-trauma psychological factors. Further studies with war- populations should examine whether the exposure to traumatic events only depends on uncontrollable external factors or whether individual factors contribute to a person’s ability to seek safe places.

The present study focused on PTSD as an indicator of severe psychological suffering in refugees. These findings highlight the need for reducing the frequent exposure to traumatic events by providing safe and stable living environments for refugees. At the same time, the presence of high numbers of chronic PTSD cases requires the implementation of individual and community based treatment programs. Given very limited resources in refugee communities, these centers must be created to provide short-term care and must be manageable by local personnel. The provision of appropriate mental health assistance is necessary to break the vicious cycle of violence and psychological morbidity.

4.5 Psychometric quality of the PTSD assessment

4.5.1 Introduction

Despite the common criticism related to the validity of the PTSD concept for war- affected populations, almost no studies have examined the validity of the variables they examined. Furthermore, most studies have used trained local interviewers for the assessments. The reliability of these interviews has rarely been studied. 75 4.5.2 Methods

Participants

A selection of 130 participants were re-interviewed with the same questionnaire about half a year after the first examination to assess the validity of the PTSD concept for the Sudanese refugees, as well as the psychometric quality of the PDS assessment in the epidemiological survey. The participants were randomly selected from the 512 Sudanese refugees from the Imvepi refugee settlement in Northern Uganda who had participated in the first round of the survey. A random selection of 77 of those 130 refugees were examined by European clinical psychologists in a validation study to investigate the validity of the PTSD diagnosis. In addition, two more mental-health instruments were administered in the form of assisted self reports by local interviewers.

Additional instruments in validation study

Assisted self-report. The Self Report Questionnaire 20 (SRQ-20) was used to measure comorbidity which was used as indicator of convergent validity. The SRQ-20 was developed to assess symptoms of anxiety and depression in primary health care in developing countries and proved to be valid and reliable for this purpose (Harding et al., 1980). Physi- cal and psychological functioning was measured using the 12 item version of the Medical Outcome Study Self-report Form (SF-12; Ware, Kosinski, & Keller, 1996). This questionnaire was developed to assess health-related quality of life in different cultures. Several items of the questionnaire had to be adapted to the living conditions in an African refugee settlement. As no comparison data was available for the Sudanese refugees, the usual standardization procedure was not used for scoring.

Expert rating. The PTSD-section of the Composite International Diagnostic Inter- view (CIDI, WHO, 1997) was applied by expert interviewers as a standard for PTSD diagnosis according to DSM-IV. In order to ensure that it was possible to examine the symptom structure, the CIDI-typical skip rules were not applied, instead each respondent was asked all questions. 76 4.5.3 Results

Validity of PTSD concept

Those refugees who received a PTSD diagnosis in the CIDI interview were compared with those refugees without the diagnosis on the other mental-health variables assessed with the SRQ-20 and the SF-12 to examine the concurrent validity of the PTSD diagnosis in the population of the Sudanese refugees. In addition, the functioning criteria included in the CIDI were compared between the groups. The results are presented in table 4.5.

Table 4.5. Comparison of refugees with and without PTSD diagnosis according to CIDI on different mental health variables and functioning criteria

No PTSD PTSD Statistics p

n 34 43

SRQ-20 Depression M (SD) 8.53 (3.07) 10.47 (2.57) t = -3.00 <0.01

SRQ-20 Anxiety M (SD) 6.97 (2.62) 8.28 (1.88) t = -2.58 < 0.05

SRQ-20 Sumscore M (SD) 13.32 (4.44) 16.23 (3.51) t = -3.13 < 0.01

SF-12 Physical Health M (SD) 0.36 (0.14) 0.34 (0.13) t = 0.52 ns

SF-12 Psychological Health M (SD) 0.30 (0.17) 0.28 (0.13) t = 0.67 ns

CIDI: Told a doctor about the symptoms n (%) 9 (27%) 22 (51%) c2 = 4.81 < 0.05

CIDI: Told other professional about symptoms 4 (12%) 15 (35%) c2 = 5.46 < 0.05 n (%)

CIDI: Took medication, drugs or alcohol more 7 (21%) 14 (33%) c2 = 1.37 ns than once because of symptoms n (%)

CIDI: Symptoms interfered with life a lot n (%) 14 (41%) 42 (98%) c2 = 30.6 < 0.01 77 Reliability of PDS assessment

Two variables were used as indicators of reliability: internal consistency and retest reliability over the 6-months period between the first and the second administration of the questionnaire. The internal consistency of the PDS calculated in the DFM survey was excellent (Cronbach’s a = .92). The stability of the variables over the six months retest period was considered to be another indicator of reliability. The stability was calculated as the correlation between the values of the first and the second administration of the DFMQ.

The stabilitiy of the PDS sumscore was rtt = .21 (p < 0.05). The stability of sociodemographic variables was rtt = .78 (p < 0.01) for age, rtt =.77 (p < 0.01) for the number of children in a household and rtt =.49 (p < 0.01) for the number of adults in a household.

Validity of PTSD assessment using PDS

Relationship with health indicators. The DFM epidemiological survey included the assessment of different physical symptoms and diseases of the respondents. The prevalence of physical symptoms and diseases in the month before the assessment reported by the respondents divided by the PTSD diagnosis is presented in Table 4.5.

Table 4.5. Relationship between PTSD diagnosis in PDS interview (N=3178) and health problems in the past month assessed with DFMQ

no PTSD PTSD c2 p

Cough 63% 79% 91.1 < 0.001

Diarrhoea/ stomach pain 37% 54% 93.5 < 0.001

Leprosy 0.5% 1.5% 8.2 < 0.01

Asthma 0.7% 1.3% 3.0 ns

Diabetes 0.3% 0.3% 0.2 ns

Fever/ shivering/ malaria 63% 82% 132.4 < 0.001

Tuberculosis 2% 3% 7.5 < 0.01

Hypertension 1% 1.7% 2.0 ns

Epilepsy 0.9% 2% 8.3 < 0.01 78 Several confounding variables may contribute to the relationship between PTSD diagnosis and physical health. Three binary logistic regression analyses were carried out for the three symptoms/diseases that were prevalent in more than 5% of the respondents (cough, diarrhoea and fever) to control for possible sociodemographic and stress related variables. The physical symptoms were the binary dependent variables in the analyses. The predictors were entered simultaneously either as continuous or as categorical variables. The variable group consisted of three categories (Ugandan nationals, Sudanese nationals, Suda- nese refugees). For the analyses, the Ugandan nationals were chosen as reference category as this group was expected to present with the best health outcome as they had the best access to medical health care. The results of the analyses are presented in Table 4.6.

Table 4.6. Binary logistic regression analyses of the influence of PTSD, stress and sociodemographic variables on physical symptoms and diseases

Cough Diarrhoea/stomach Fe- pain ver/shivering/malaria

OR (SE) Wald OR (SE) Wald OR (SE) Wald

Sex (female) 1.08 (0.12) 0.42 0.34*** (0.10) 112.13 1.09 (0.12) 0.50

Age 1.00 (0.01) 0.83 1.01** (0.00) 7.81 1.01** (0.01) 8.98

No education 0.89 (0.12) 0.94 1.48*** (0.11) 12.39 1.14 (0.12) 1.01

Nr. of possessions 0.93* (0.03) 4.56 0.89** (0.04) 10.21 0.93** (0.03) 4.23

Nr. of meals per day 0.98 (0.02) 0.83 1.00 (0.01) 0.05 1.01 (0.04) 0.10

Alcohol consumption 0.96 (0.07) 0.42 0.90 (0.07) 2.67 0.88 (0.06) 3.56

Sudanese refugees 2.86*** (0.11) 129.02 2.32*** (0.10) 99.98 3.57*** (0.12) 157.2 vs. Ugandans 0

Sudanese nationals 19.38*** (0.23) 174.31 1.23 (0.12) 2.88 10.86*** (0.19) 161.4 vs. Ugandans 2

Nr. of traumatic event 0.99 (.01) 0.79 1.02* (0.01) 5.66 1.01 (0.01) 0.95 types

PTSD diagnosis 1.59*** (0.11) 18.45 1.58*** (0.09) 25.75 1.73*** (0.11) 23.46

Nagelkerke R2 0.25 0.19 0.26

Note: * p<0.05; ** p<0.01; *** p<0.001 79 Relationship to CIDI diagnosis. The PTSD diagnoses obtained in the second administration of the DFMQ by the local interviewers were compared with the subsequent CIDI expert rating in the validation study to examine the convergent validity of the PTSD diagnosis of the PDS. The PDS achieved a sensitivity of 70.7%, specificity was 38.7%, indicating an overestimation of PTSD diagnoses using the PDS. According to the PDS, 68% of the participants in the validation study were diagnosed with PTSD, whereas the CIDI resulted in a prevalence of 58%. The agreement between PDS and CIDI diagnosis was not significant (k = 0.97, N=72).

Relationship to depression and anxiety. The concurrent validity of the PTSD diagnosis determined with the PDS was evaluated by comparing the levels of depression and anxiety (SRQ-20) as well as physical and psychological health (SF-12) for those who received a PTSD diagnosis and those who received no such diagnosis in the PDS. Independent samples t-tests found no significant differences between the groups on these variables. The PDS sum-score was correlated with levels of depression, anxiety (SRQ-20), physical health and psychological health (SF-12) to examine the concurrent validity of the PDS severity rating. No significant correlation was found in this analysis.

4.5.4 Discussion

Despite the vast amount of research into the psychological consequences of traumatic events, the validity of the PTSD concept has been repeatedly criticized as mental health diagnosis in general (Summerfield, 2001) and in particular as diagnosis for victims of war in non-industrialized countries (Summerfield, 1999). We carried out a validation study with 77 refugees who fled from the war in Sudan to a Ugandan refugee settlement to examine the validity of the PTSD diagnosis. Two different criteria were examined in this context. First, PTSD was related to mental health variables considered to be well established for African cultures. The findings showed that those refugees diagnosed with PTSD by European clinicians using the CIDI Interview also presented with significantly higher scores in depression and anxiety assessed with the SRQ-20, an instrument that was specifically developed for use in non-industrial countries and previously showed a high validity for this population (Harding et al., 1980). In contrast, no similar relationship was found for the physical and psychological health as measured with the SF-12. The proponents of this questionnaire claim that this instrument is able to assess the health related quality of life in 80 different cultures (Ware et al., 1996). Until now, this assumption has not been proven. We found many difficulties in the administration of the questionnaire, as many items did not fit the living conditions in a developing country and had to be modified. For example, the activities of climbing stairs, using a vacuum cleaner and playing golf are of no relevance in these countries. This might explain the fact that we found no relationship between scores on this questionnaire and other measures. The fact that we did not even find a correlation between SRQ-20 depression and SF-12 psychological health questions the general validity of this instrument.

The second indicator for the validity of the PTSD concept was the functioning criteria assessed in the CIDI interview. As the skip rules of the CIDI were not applied and all respondents were asked these questions, we could compare the PTSD related impairment between those who were diagnosed PTSD and those without PTSD. All patients who fulfilled the symptom and time criteria for PTSD also reported a decrease in functioning related to PTSD symptoms in at least one area of functioning.

More than half the refugees with PTSD reported that they had already told a doctor about the symptoms. Anecdotally we can report that we subsequently talked to a medical doctor in the Imvepi health center who confirmed that many refugees seek help because of symptoms related to PTSD. About one third of the refugees with PTSD reported that they had talked to other professionals, mostly missionaries, about the symptoms and the same proportion had already taken substances to gain relief from the PTSD symptoms. Al- though the proportions are significantly smaller among those without a PTSD diagnosis, still more than 40% of them reported a reduction in functioning related to PTSD symptoms. This finding indicates that, contrary to the claims of some authors, even refugees living in desperate conditions report that they have severe problems in mental health and in functioning because of PTSD symptoms. The fact that many refugees without a diagnosis of PTSD nevertheless indicated a reduction in functioning suggests that the high prevalence of PTSD among the refugees might still lead to an underestimation of the number who are suffering because of the mental consequences of traumatic events.

In the DFM survey we also asked several questions about physical symptoms and diseases. Respondents were asked whether they suffered from a list of symptoms to get an impression of health status without the possibility of conducting medical examinations. In general, we found high prevalence rates for symptoms of infectious diseases (cough, diar- 81 rhoea, fever). The diseases tuberculosis, leprosy, asthma, diabetes, hypertension and epilepsy were found only in a small minority of respondents. This does not necessarily mean that these illnesses are rare among the populations studied. Contrary to the symptoms of infectious diseases it is difficult or even impossible for a person to indicate that he has a specific disease without the diagnosis of a medical doctor. As it was difficult for most of the respondents to access medical assistance, it is likely that a significant proportion of these diseases went unrecognized.

We found a high relationship between the PDS-PTSD diagnosis and physical symptoms in the survey. Respondents with PTSD reported a higher prevalence of cough, diarrhoea, fever/malaria, leprosy, tuberculosis and epilepsy. We carried out regression analyses with the most prevalent physical symptoms cough, diarrhoea and fever as outcome variables to examine the influence of different confounding variables. The analyses showed that the group variable generally had the largest and most consistent influence on the symptoms, with the Sudanese nationals having the highest probability for physical symptoms (except for diarrhoea), followed by the refugees. Differences in the availability of medical care and access to medication might explain this effect, as medical care is almost absent for the Sudanese nationals and still difficult to gain for the refugees in Uganda, whereas the Ugandan public health care has made considerable advances in recent years. The relationship between PTSD and physical symptoms remained stable in the regression analyses and the effect was generally larger than the factors of poverty, age, sex, and the history of traumatic events.

The relationship between stress and physical illnesses, in particular infectious diseases, is well established and there are several assumptions about the immunological mechanisms that are responsible for this effect (Kiecolt-Glaser, McGuire, Robles, & Glaser, 2002). If PTSD is a chronic stressor for a person, it may have the same health effect as other chronic stressors. In a previous study McFarlane et al. (1994) found a similar relationship between PTSD and physical symptoms in a sample of Australian fire fighters. It is noteworthy that even in African populations exposed to constant stressors like poverty and threats, the stress caused by PTSD has an independent impact on physical symptoms.

It is unclear whether the relationship between PTSD and health would remain stable if medical examinations were applied instead of the self-report of symptoms. As the previous research in this field has been inconclusive (Friedman & Schnurr, 1995), it is too 82 early to conclude that PTSD causes a bad health outcome. Different hypotheses have been offered to explain the relationship between PTSD and self-reported physical health. These include both neurobiological (e.g. symptoms caused by increased autonomous arousal) and psychological factors (e.g. somatization and augmented lamenting). Our findings show that it could be worthwhile to carry out further research including detailed medical assessments to arrive at more conclusive evidence.

The psychometric evaluation of the PDS showed mixed results. Whilst the internal consistency of the PDS was excellent in the survey, a half-year retest showed only moderate stability of the symptoms. The stability over six months cannot be considered a indicator for retest reliability as the time interval between the two administrations is usually smaller and many factors can influence a change in symptoms in six months. Nevertheless, a correlation of r = .21 can be considered to be moderate at best and we cannot know whether the nature of the symptoms underwent much change in this population or if the retest-reliability of the instrument was low. The fact that the stability of the sociodemographic data, most of all age, was also much lower than expected showed that data that can be assessed very easily and reliably in industrialized populations is quite difficult to obtain for illiterate respondents in developing countries. Most respondents were not sure about their own age and the timeline tool that was constructed to overcome this problem has obviously had only a limited effect.

We compared the PTSD diagnoses with the diagnoses obtained with the CIDI expert interview to evaluate the convergent validity of the PDS. We found no significant convergences of these diagnoses and the levels of specificity and sensitivity were not satisfying. The evaluation of the concurrent validity was also disappointing. Although we found significant differences between the subjects with PTSD and without PTSD based on the CIDI interview on the measures of depression and anxiety, we did not find similar differences for the PDS diagnosis. As the original PDS presented with good psychometric properties in a study in the US (Foa, 1995), this problem is probably related to the translation and administration of the questionnaire in this study. The questionnaire had to be translated into the local language of the respondents. This was difficult to achieve as the only local language spoken by almost all southern Sudanese was an Arabic dialect called Juba- Arabic. The vocabulary of this language is much smaller than English so the items had to be reformulated and more extensive descriptions given for difficult items. The blind back- 83 translations of the final versions came very close to the meaning of the original questions. But the procedure of the study revealed that Juba-Arabic was the first language only for a small minority of the respondents, as a variety of tribal dialects were spoken by the Suda- nese. This complicated the correct understanding of the PDS items and might have caused problems for the quality of the instrument.

A subsequent validation study of the PDS with ex-combatants in Somalia Odenwald, Schauer, Neuner, and Elbert (2002) showed that a PTSD assessment can achieve good psychometric properties in an African culture with a language (Af-Somali) that consists of a similarly reduced vocabulary. Neuner, Onyut, Odenwald and Elbert (unpublished data) replicated these findings for a translation of the PDS into Kinyarwanda. The difference in these studies was that a much more extensive standardized training on the PTSD concept was provided for the local interviewers. In these studies, the interviewers were able to give further explanations about single items if the respondent had difficulties in understanding the content. In addition, the interviewers were instructed to ask for examples of the characteristics of the individual’s symptoms to ensure a correct understanding. Using this extended training the concurrence between the expert diagnoses and the PDS diagnoses achieved excellent sensitivity and specificity levels of around 90%, and significant group differences were found on different mental health indicators. These findings show that it is necessary for further epidemiological studies that rely on local interviewers in non- industrialized countries to evaluate the psychometric quality of the application of their instruments. It is not sufficient to rely on evaluations obtained in other populations. Exten- sive training of the local interviewers is also recommended. 84

5 Etiological model of psychological trauma

5.1 Memory related features of PTSD

For the development of effective treatment methods, it is essential to understand the psychological and biological processes that underlie PTSD. The classical behavioral theories of anxiety disorders, like Mowrer’s (1947) two-factor theory, cannot account for all PTSD symptoms. Whilst the emotional responses to cues which remind of traumatic experiences and the resultant avoidance behavior could be well explained by this theory, other key aspects of PTSD could not be explained (Steil & Ehlers, 1996). Instead, some key features of memories of traumatic events point to the relevance of memory processes for the explanation of PTSD.

The most distinct symptom of PTSD is the re-experiencing of the traumatic event in the form of flashbacks. These involuntary intrusions can be triggered by cues that remind a person of the traumatic situation. The reliving can include all kinds of sensory information, like pictures, sounds, smells and bodily sensations (Terr, 1993; van der Kolk, 1995). A special feature of these flashbacks is the feeling that this event is happening again right at that moment. This means that at the moment of a flashback they are not fully aware that they are remembering a past event. The memory of the traumatic event does not seem to be fixed in the temporal or spatial context in which it actually occurred (Ehlers & Clark, 2000).

The other feature of memories of traumatic events is that despite the fact that patients report frequent episodes of reliving, it is extremely difficult for them to narrate the event. In this context, recent theoretical formulations point to the importance of early observations made by Janet (1889, 1904). Janet stressed that the dissociation of the painful experience during a traumatic event can lead to a distortion of the memory and even the amnesia of the event. While a complete or partial amnesia has been reported for different kinds of traumatic events, most patients who suffer from PTSD do usually not report that they are unable to remember the event. Typically, the patients rather report that they can remember the event too well, as they suffer from painful involuntary recollections. 85 However, if the patient is asked to report the event, the narrations are typically dis- organized, fragmented and incoherent (van der Kolk, 1995). Harvey and Bryant (1999) showed that the disorganization of the narration is correlated with symptoms of acute stress disorder. Observations of clinicians and researchers show that there is particularly likely to be a breakdown in the ability to put into words the most emotional part of the traumatic event, which can last from several seconds to some hours.

For victims of more long-lasting events or survivors who were exposed to a series of traumatic events, like most victims of organized violence, this problem in narrating one’s history can cover a long period in time. This effect was described by Rosenthal (1997) who documented the life histories of survivors of the Holocaust:

“It is difficult to establish a relationship between the different stages of life – this means the time before the persecution, the time of persecution and the time after having survived. Within these stages, the relationship of the different events can be substantially broken into pieces. Whole stages can sink into the sphere of speechlessness and are accessible for the biographer only in single fragments, pictures and moods.” (Rosenthal, 1997, p. 40, translation by the author)

So the characteristic of the memory of a traumatic event is twofold: On the one side a person has very vivid recollections including many sensory details. Notwithstanding the detailed representation of the memory, it is very difficult for the victim to narrate the event.

5.2 Traumatic event in memory

On the one hand, traumatic events seem to be excessively represented in memory as patients with PTSD suffer from vivid and painful recollections. On the other hand, the representation seems to be weak as it is very difficult to talk about the traumatic event. These characteristics seem to be contradictory, but they can be explained on the basis of recent theories of memory research.

Memory theories of PTSD share the assumption that traumatic events are stored in memory in a different way to every day events. This pathological representation in memory is responsible for the symptoms of PTSD (Brewin, 2001; Brewin, Dalgleish, & Joseph, 1996; Ehlers & Clark, 2000; Metcalve & Jacobs, 1996; van der Kolk, 1994, 1996, 1997). 86 Before the pathological characteristics of traumatic memory can be explained, one must understand how normal past events are stored in memory. Based on neuropsy- chological research, Squire (1994) suggested the separation of different memory structures. He differentiated between declarative (explicit) memory, that can be deliberately retrieved, and non-declarative (implicit) memory that has an impact on a person’s behavior and experience but it is automatically activated by environmental or internal cues. According to Squire, declarative memory consists of memory of personal events as well as memory of facts and knowledge of the world. Declarative memory represents for example significant events like a person’s marriage or a graduation as well as knowledge about the history of the world that a person learned at school. This knowledge can be deliberately retrieved when a person tells others a story about his wedding or explains the French Revolution. On the other hand, non-declarative memory covers skills, habits, emotional associations and conditioned reactions. This form of memory is automatically activated when required. For most people, opening the door of a car is a highly practiced skill that is automatically activated when required without ever remembering how and when he learned that skill.

Single events have a simultaneous impact on different memory structures. For example, the experience of learning to ride a bike can affect declarative memory, as a person can later remember the day and the place when it happened. At the same time, non- declarative memory changes as the person learns the skill of riding a bike, which can be activated later on.

Tulving (2001) further differentiated between episodic and semantic memory. Ac- cording to this classification, episodic memory is about happenings in particular places at particular times and covers information about “what”, “where” and “when”. Another feature of episodic memory is the recollective experience. This means that episodic memory allows people to consciously re-experience one’s previous events. On the other hand, semantic memory covers all the knowledge about the world. In contrast to episodic memory, retrieval of semantic memory is not necessarily associated with a “mental travel in time” or the recollection about a personal past event. For example, when a person thinks about the sequence of emperors in the Roman history, he might recollect the very moment when he learned this knowledge in the 8th grade in school, but such a recollection is neither necessary nor typical for the retrieval of semantic knowledge. 87 Tulving’s concept of episodic memory cannot be easily integrated into Squire’s classification of declarative and non-declarative memory systems. It appears straightforward to classify episodic and semantic memory as substructures of declarative memory. However, research showed that semantic memory can be retrieved implicitly (Schachter, 1987), and retrieval of episodic memory can happen automatically and cue-driven. Furthermore, the concept of episodic memory also covers other non-declarative structures, like emotional associations, so the two different concepts cannot be translated into each other’s terms.

Unfortunately the previous research into episodic memory has led to a confusion about the usage of different terms like episodic, autobiographic, declarative, non- declarative, explicit and implicit memory (Conway, 2001; Tulving, 2001; Tulving & Mark- owitsch, 1998). For the understanding of pathological processes of PTSD, the representation of events in two different forms is relevant. One type of representation is a form of sensory-perceptual representation that is close to Tulving’s concept of episodic memory that allows recollective experiences. The other type of representation contains autobiographic context information in a form that is related to Squire’s concept of declarative memory about past personal events.

5.3 Sensory-perceptual representation

When a person thinks about past life events, it can be possible that he does not only retrieve abstract knowledge about what has happened, but can sometimes also imagine (Lang, 1979) the event in the form of a “recollective experience” (Tulving, 2001). This means that he can directly access visual and other sensory information about past events previously stored in his mind, including an awareness of the subjective time when the event happened.

Obviously, this form of vivid and detailed recollection is not possible for all events experienced in a life. For every-day events with a minor meaning for the person, these representations usually only last minutes or hours (Conway, 2001). They can only become more stable if they are integrated with other memory structures, then detailed images of this event can be retrieved even years later. This enduring storage of sensory-perceptual representations only happens for events stored in a highly emotional state, as they are significant for the achievement or failure of the individual goals of a person (Conway & Pley- 88 dell-Pearce, 2000). One form of these representations is called “flashbulb memories” (Brown & Kulik, 1977). These vivid and long-lasting memories contain information about the environmental conditions of a person when he learned of a significant public event like the death of President Kennedy.

Sensory Cognitive Emotional Physiological Elements Elements Elements Elements Sensory Sensory Sensory Sensory Elements Elements Elements Elements Freezing See road in home village I can’t do anything! Helplessness See Kosovo Smell dead landscape bodies I - Me Heart beats fast

Fear See soldiers This is with guns dangerous! Breath fast

Hear Rage screamings

See See media Think about Can’t do Rage Physical policemen in reports event anything because of activity exile town about children’s asylum misbehavior process

Figure 5.1. Schematic presentation of a hypothetical sensory-perceptual representation (fear network) including sensory, cognitive, emotional and physiological elements. The network represents a Kosovar refugee’s memory of the attack on his home village. The boxes below indicate environmental stimuli with the potential to activate the representation

Lang’s bio-informational theory of emotions (Lang, 1979) offers a good framework within which to understand the nature of sensory-perceptual representations and their relationship to emotions. In this view, emotions are mentally represented as propositional networks. These representations consist of sensory-perceptual information about the stim- 89 uli present in the past situation in different modalities (visual, auditory, olfactory, etc.). At the same time, this network contains information about the cognitive and affective evaluation of the stimuli and the corresponding physiological responses. They may even include motor responses but for healthy individuals these reactions are restricted to action disposi- tions rather than full behavioral responses. The items of these networks are connected in such a way that the activation of single items leads to the activation of the connected items (see Figure 5.1 for an example). This theory can explain the ability to retrieve representations of past events in the form of recollective experiences described by (Tulving, 2001), as single perceptual elements being interconnected and therefore likely to be activated together. In addition, it predicts that these recollections should occur alongside physiological and emotional responses similar to those which occurred when the event happened. These predictions have been confirmed in numerous studies using physiological measurements during mental imagery.

Conway (2001) stated that the retrieval of a sensory-perceptual representation is the endpoint of the reconstruction of a memory of a past event and takes some effort and several seconds of time. However, the impact of emotional structures on behavior is not limited to the deliberate retrieval of memories of past events during mental imagery. Elements of the structure can also be automatically retrieved by a cue and lead to emotional behavior. Foa and Kozak (1986) have related Lang’s theory to anxiety disorders and especially to PTSD. In this context, the sensory-perceptual representations of traumatic events have also been called fear structures. According to Foa, they differ from representations of normal events in several ways.

First of all the fear structures encoded during a traumatic event are unusually large and cover a wide variety of single elements. This means that they can be easily activated, as many cues in the environment resemble elements in the fear structure. Secondly, intercon- nections between the single elements are unusually strong. Consequently, the activation of only a few elements is sufficient to activate the whole structure. This explains why a person suffering from PTSD can have sudden flashbacks when reminded of the traumatic event. According to this theory, an environmental stimulus or internal cue (e.g. thinking about the event) that resembles the stimuli of the traumatic situation can cause the activation of the sensory information and the associated emotional, physiological and motor responses stored in the fear structure (See Figure 5.1). Southwick et al. (1993) showed that 90 nonspecific high arousal induced by the injection of yohimbine, a substance that increases autonomous arousal, can be sufficient to trigger a flashback in PTSD patients. The fear structure concept can explain this effect. In this context, high arousal is a typical element of the physiological response coded in the structure, and is thus closely tied to the sensory- perceptual elements.

According to this theory, a flashback is the activation of a fear structure. In addition to intrusive phenomena, this theory can also explain the typical PTSD symptom of avoidance behavior. The activation of a fear structure is experienced as a painful recollection, and consequently many PTSD patients learn to prevent this by avoiding cues that remind her or him of the traumatic event. They have to avoid both internal and external cues, so they try not to think about it, not to talk about it and keep away from persons and places that remind them of the event. In addition, avoidance and flight behaviors can be parts of the fear structure as motor responses themselves. The third major symptom is the permanent hyper-arousal of patients with PTSD. This can be explained by a partial activation of the fear structure that does not always include sensory-perceptual elements, but can be restricted to the activation of evaluative cognitions and physiological elements of the structure. Consequently, a person can perceive permanent threat even when he is safe, and remain in a state of elevated arousal in which he is always prepared to flee.

The theory of fear structure can explain most of the PTSD symptoms. On the other hand it is not sufficient to account for all features of traumatic memories described above. It cannot explain PTSD patients’ difficulties in narrating their event. One can assume that efforts to narrate the event always lead to the activation of the fear structure. An activated fear structure that is associated with high arousal may attract so many resources that the person is no longer able to talk about the event. This assumption is not consistent with the observations of clinicians, who state that PTSD patients can have a flashback whilst conducting a normal conversation with the therapist (Brewin, 2001). Furthermore, the experience of flashback seems to be more than an exaggeration of a normal recollective experience. Contrary to recollections of every-day events, flashbacks are experienced as if the event occurs again right at that moment (Ehlers & Clark, 2000). They do not fit the perception of subjective time but seem to be without context in time and space. It is necessary to look at theories about autobiographical context information to explain these phenomena. 91 5.4 Autobiographic contextual memory

The sensory-perceptual representation is only one type of memory about past events. When a person retrieves a memory about a past event, he usually falls back on an organized knowledge base that is called autobiographic memory (Conway & Pleydell- Pearce, 2000). Figure 5.2 presents an example of the organization of autobiographical memory.

At the top of the hierarchy of autobiographic memory organization is the memory of lifetime periods. It represents general knowledge of important other persons, locations, actions, activities, plans and goals that characterize a special period. They cover distinct time-periods with identifiable beginnings and endings. An example of a lifetime period is “when I lived with Marry”. Lifetime periods are typically organized along major themes of persons who are related to goals in life. Typical themes are relationships, occupations and places of living. Lifetime periods related to different themes can overlap with respect to the time period they refer to. For example, the time covered by “when I was working at the farm” may overlap with the time span of “when I lived with Marry”.

Another type of knowledge base that can be located at one level below the lifetime periods is the memory for general events. General events can be divided into repeated events, e.g. “having lunch at the canteen” and single events, e.g. “my first day in school”. These knowledge bases organize the sequence of events. Not all events a person experiences are represented with the same accuracy. Shum (1998) suggested that those events that de- marcate beginnings and ends of lifetime periods have an outstanding role in autobiographic memory. These landmark events (like the first date with a subsequent intimate partner or the struggle that indicated the end of a relationship) are thought to be represented in more detail and accessed more easily. Memories of activities that cover a series of events relating to a common theme can be connected to form so-called ‘mini-histories’ (Robinson, 1992). For example, the mini-history “my first love” can represent the association of the memories of general events like “my first date ” and “my first kiss”.

One level below the memory of general events, Conway locates the “Event-specific knowledge” (ESK). The ESK corresponds to the sensory-perceptual representations described above. ESK is usually linked to general event structures, and the activation of the Relationship Occupation War Theme Theme Theme 92

Lived with Went to Lived in my parents school peace

Worked at Upheavals in Lived with parent’s surrounding Mary farm villages alone Worked at War at own farm home village Had own familiy Was unemployed Peace in exile

First Bought own major army farm attack Met Mary at market

Figure 5.2. Structure of autobiographical memory base. Adapted from Conway and Pleydell-Pearce (2000). sensory-perceptual details of an event is accompanied by the activation of knowledge about the sequence of the event and the location of the event in lifetime periods.

It is assumed that patients who suffer from PTSD have a significant distortion in their autobiographic memory. The traumatic event does not seem to be clearly represented as a general event and it does not seem to be clearly positioned in a lifetime period. Even 93 though the sensory-perceptual representations of traumatic events are unusually strong and long lasting, there is no reliable autobiographical structure to tie them to. As a good autobiographical organization is necessary to narrate an event, the distortion in autobiographical memory leads to the failures of PTSD patients to narrate the traumatic experience. At the same time, this theory can explain the qualitatively distinct features of flashback experiences. The recollection of normal events is usually accompanied by the perception of subjective time, which means that even during the in-depth mental imagery of a past event a person always feels that this event took place at a different time and place. During a recollection of a normal event, the simultaneous retrieval of contextual information, relating to time and space, seems to be unavoidable. However, the sensory-perceptual representation of a traumatic event is not tied to corresponding autobiographical context information, resulting in the “here and now” perception of flashbacks. Metcalve and Jacobs (1996) described the characteristics of the isolated activation of a sensory-perceptual representation as is typical for patients with PTSD:

“Memories and reactions that are attributable to the isolated hot- system encoding may seem irrational both to the individual him- or herself, and to the therapist, since such fragments are ungrounded by the kind of narrative and spatio-temporal contextual anchors that tie our ordinary experience to reality. Such memories are disturbing, not only because of the direct fear they evoke but also because of their strangeness.” (Metcalve & Jacobs, 1996, p. 2.)

Different authors have proposed theories that resemble the principles described above (Brewin, 2001; Brewin et al., 1996; Ehlers & Clark, 2000; Metcalve & Jacobs, 1996; van der Kolk, 1994, 1996, 1997). The sensory-perceptual representations of the traumatic event have also been called “hot memory” (Metcalve & Jacobs, 1996) and “situationally accessible memory” (SAM) whereas the autobiographical context memory has been called “cold memory” and “verbally accessible memory” (VAM). Table 5.1 summarizes the features of sensory-perceptual and autobiographic representations in the context of the memory theory of PTSD. 94 Table 5.1. Characteristics of memory representations of normal and traumatic events

Sensory-perceptual representa- Autobiographic representation tion

Related terms hot memory, SAM, event specific cold memory, VAM, declarative memory, memory, fear structure, non- explicit memory declarative memory, implicit memory

Memory of normal events

Description Network of sensory, emotional, Memory that contains information about cognitive, physiological and be- lifetime periods and general events havioral elements

Organization Connected to related autobio- Organized in knowledge bases that are graphic memory related to personal goals, organization is active process that happens through thinking and talking about the event and leads to adjustments in memory or personal goals

Automatic retrieval Automatic activation of full structure Automatic activation by retrieval cues oc- is inhibited by access of autobio- curs graphic memory

Deliberate retrieval Deliberate retrieval needs effort but Deliberate retrieval happens when person is possible to get access to sensory thinks and talks about event and reaction information, detailed access possible in imagery

Relevance for nar- Provides sensory and reaction in- Provides context information for narration ration formation for narration to enable consistent and chronological report

Memory of traumatic events

Characteristics Large structure with strong inter- Incomplete representation with gaps and connections between sensory and inconsistencies reaction elements

Organization Organized without connection to Discrepancies to personal goals result in autobiographic memory disrupted organization

Automatic retrieval Automatic activation by a wide Activation of sensory-perceptual repre- range of retrieval cues happens sentation is not accompanied by activa- often and results in hyper-arousal tion of autobiographic memory which re- and flashbacks sults in “here and now” sensation of retrieval

Deliberate retrieval Possible but usually avoided, as Retrieval of fragments possible but al- activation does not remain under ways accompanied by activation of hot full control memory

Relevance for nar- Efforts to narrate the event are ac- Narration is inconsistent and lacks con- ration companied by strong sensory, cog- text in time and space nitive, emotional and physiological reactions and may be interrupted by flashbacks 95 5.5 Neurobiological basis of memory and PTSD

5.5.1 Traumatic stress and autobiographic memory

The neurobiological correlate of autobiographical memory cannot be identified as a single brain structure. Rather, long-term storage of autobiographic and other declarative knowledge depends on widespread neocortical neuronal activity. This complex neuronal network underlies a special organization including different rules and consistencies. This organization allows the effective storage of the vast amount of knowledge a human being can acquire. At the same time, this complex organization makes this network slow to integrate new incoming information, especially new events that are incompatible with previous knowledge.

The hippocampus is a brain structure that plays a major role in the construction of memories that contain autobiographic information, including the temporal and spatial context of an event (Tulving & Markowitsch, 1998). The hippocampus itself is probably not the place where this information is stored, as damage to this structure interferes with the acquisition of new autobiographical knowledge but does not necessarily have any effect on remote memories. Several neuroscientists assume that the function of the hippocampus is to bind single perceptual items into a conjunctive code. This code is constructed as a mean- ingful episodic representation containing information about the dynamics of the situation in a spatio-temporal context (Shastri, 2002). McClelland, McNaughton, and O'Reilly (1995) suggested that the hippocampus is especially important for the coding of information that contradicts the knowledge previously learned. This is typically the case for traumatic experiences, as they contradict basic adaptive assumptions about security, trust, etc. They assume that the neocortical system is slow to integrate contradictory information as it un- dermines strict rules and consistencies. In this context, the hippocampus may permit the formation of a rapid representation of the event and gradually expose the neocortical system to the new information. In this way the network can be slowly reorganized to integrate the new information.

Research with rodents showed that both the function and the structure of the hippocampus is strongly affected by stress hormones. Exposure to glucocorticoids increases the activity of the hippocampus, but if a certain threshold is exceeded the activity declines 96 again. Under very high levels of stress, the functioning of the hippocampus is severely impaired. It is assumed that very high doses of adrenal steroids can even cause permanent and irreversible atrophy of the hippocampus (McEwen, 1999; Kim & Yoon, 1998).

Consistent with these findings, several neuroimaging studies have found that the hippocampus is smaller in individuals with PTSD than it is in comparison subjects (Bremner et al., 1997). This decreased hippocampus volume could be a consequence of the excessive exposure to adrenal steroids during the traumatic event. Alternatively, this effect could also reflect a pre-trauma vulnerability for PTSD (Gilbertson et al., 2002) or result from chronic PTSD symptoms rather than the trauma itself (Pitman, Shin, & Rauch, 2001). The results of studies that measured the hippocampus volume in PTSD patients are inconsistent and the explanation of the findings is still a matter of debate. The fact that this structure, which relates to the encoding of spatio-temporal information, is vulnerable to stress strongly supports the assumption that a malfunction of the hippocampus during severe stress might be responsible for the distortion of the autobiographic memory.

5.5.2 Traumatic stress and sensory-perceptual representations

The fact that sensory-perceptual representations of traumatic events can remain stable over time, and that they seem to lack information about temporal and spatial context, leads to the assumption that they are constructed independently from the hippocampus. The sensory-perceptual representations of past events are probably not stored in a single brain structure. Neuroimaging studies suggest the complex involvement of areas considered to be responsible for visuospatial processing and emotion (limbic structures, especially pos- terior cingulate including the surrounding cortex as well as parietal cortex), and preparation for action (motor cortex) in the storage of sensory-perceptual representations (Bremner, 2002).

Studies of neural mechanisms of fear conditioning offer ideas about the coding of fearful memories (LeDoux, 1995). As animal studies can obviously not study the sensory and cognitive outputs of these memories, they have concentrated on the physiological and behavioral responses to conditioned fear cues. The suggestion is that the amygdala relates to the fast evaluation of cues that might indicate a potential threat. LeDoux proposed that a direct connection between the thalamus and the amygdala might offer a rapid pathway for 97 fear responses even before the stimulus underwent a detailed analysis in the visual cortex. Besides the input from the thalamus itself, the amygdala receives information from cortical areas and can trigger motor, peripheral and sympathetic stress responses via direct connections to the relevant brain structures. Contrary to the hippocampus, which shows decreased functioning under high levels of stress, the functioning of the amygdala seems to be enhanced as stress increases (Pitman, Shalev, & Orr, 2000). A recent study with rats showed that whereas chronic stress induced a dentritic atrophy in the hippocampus, the dendritic arborization of pyramidal and stellate neurons in the amygdala was enhanced in the same condition (Vyas, Mitra, Shankaranarayana Rao, & Chattarji, 2002). This finding supports the theory of a dissociation of stress effects on the hippocampus and amygdala. The enhanced activity of the amygdala during stress might cause an excess of fear conditioning to a large variety of stimuli during a traumatic event and thus explain the unusually large fear structures in PTSD patients. Neuroimaging studies with PTSD patients found inconsistent results, with some studies confirming the increased activity of the amygdala during symptom provocation in PTSD patients whereas other studies did not find this effect (Bremner, 2002).

5.6 Emotional Processing

5.6.1 Normal processing of stressful events

Whereas most people react with an intensive emotional upheaval, including PTSD symptoms, immediately after a traumatic event, only a minority develops chronic PTSD. The recovery from acute stress symptoms seems to be a normal process. Several authors have tried to identify the mechanisms behind this emotional processing, since they could offer valuable ideas for the treatment of chronic PTSD.

In an influential account of emotional processing, Foa and Kozak (1986) suggested that emotional processing involves the modification of the original fear structure. The activation of the fear structure in a safe context offers the possibility of including new elements into the structure that are incompatible with original connections. For example, the cognition “I am safe” contradicts main connections in the fear structure that usually lead to the sensation of a current threat. The prolonged availability of both the fear structure and cur- 98 rent incompatible information leads to modification of the structure in such a way that maladaptive associations between stimulus and response decrease over time. This process happens as natural processing after an emotional event but it may be inhibited when the activation of the fear structure is avoided or is prematurely terminated by the person. This inhibition of emotional processing consequently leads to the stabilization of the fear structure and the development of chronic PTSD.

Research into fear conditioning has challenged this view. Several investigators have shown that extinction of fear responses does usually not change the original stimulus-fear associations, as the original fear response can easily be reinstated in a context different to the context of extinction learning. Instead, extinction probably occurs through the inhibition of the fear response by cortical areas, especially the prefrontal cortex (LeDoux, 2000). Pathways from the prefrontal cortex, which receives input from the hippocampus, to the amygdala possibly allow the fear response to be modified, depending on the current cortical evaluation of the stimulus in the current environmental context. Brewin (2001) suggested that this cortical evaluation depends on the availability of declarative memory of the stimulus. This knowledge is necessary to enable the stimulus to be evaluated in terms of whether it indicates a current threat. As noted above, after a traumatic event declarative knowledge, especially autobiographic knowledge that could offer information about the context of the feared stimulus is fragmented or even absent. This leads to a lack of capability for the cortical inhibition of the fear response.

This means, that a person who experienced a traumatic event and who is confronted with a stimulus that reminds him of that event has to learn that this stimulus represents a fearful episode that took place a long time ago at a special place and does not indicate any threat right at the moment. In normal emotional processing, the intrusive memories provide an opportunity to build up declarative memory about the event as the intrusive memories present details of the event. When a person thinks and talks about this event and includes the stimuli presented by intrusive memories, he can construct autobiographic knowledge about the traumatic episode. This contextual knowledge offers material for the evaluation of reminders of the traumatic event and consequently a conditioned fear response can be inhibited as the person learns to distinguish between past and current threat. 99 5.6.2 Failures in normal processing in PTSD

One assumption is that PTSD develops only when the stress level during a traumatic event achieves a threshold that causes the malfunctioning of the hippocampus and other brain areas involved in PTSD. Brewin (2001) suggested that so-called peritraumatic dissociations reflect the failure of efficient supervisory control and memory construction during a traumatic event caused by such a deficiency of the hippocampus and the prefrontal cortex during extreme stress. This phenomenon, which is difficult to comprehend by non-affected people, describes strange experiences of detachment from the ongoing event, like the “out-of body phenomenon”, the sensation of external control, or a slowing down of the event. Consistent with this view, several studies showed that retrospective accounts of peritraumatic experiences are highly correlated with PTSD (Ozer et al., 2003).

As noted above, the construction of autobiographic knowledge about the traumatic event is no easy task as the organization of autobiographic knowledge is connected to personal goals and to basic beliefs about the self. This means that active emotional processing takes some time and effort. Secondary emotions like anger and guilt may indicate problems in the cognitive processing of the event and the adaptive placement of the traumatic experience in the context of preexisting beliefs about the self, which also interferes with the construction of a coherent autobiographic representation. As thinking about the traumatic event automatically causes painful emotions, many people avoid this process and try to terminate intrusive memories as soon as possible. The fact that victims with poor social support after a traumatic event have an increased risk of developing PTSD (Ozer et al., 2003) might be explained by the lack of support in talking about the event.

5.6.3 Consequences for treatment

Following the memory theory of PTSD, the main goal of therapy should be the construction of a consistent declarative (autobiographic) representation of the event that provides information which enables the cortical inhibition of fear responses. The consistency of the autobiographic knowledge about the event can be evaluated by the quality of the narration about the event. Foa and colleagues have developed exposure therapy for PTSD (Foa & Rothbaum, 1998) that has proven to be one of the most successful treatment approaches for this disorder (see chapter 6.5.2). In exposure therapy for PTSD, the patient 100 is instructed to repeatedly talk about the traumatic experience. Whereas initial formulations of the theory of exposure therapy were based on different assumptions about fear extinction (Foa, Hearst-Ikeda, & Perry, 1995), recent analyses of the treatment process have supported the importance of the construction of autobiographic knowledge. In particular, it was demonstrated that those patients who manage to construct a coherent narration of the event during exposure therapy profit most from treatment (Foa, Molnar, & Cashman, 1995).

The main focus of therapy should be on the part of the memory that is most fragmented in autobiography and most intensively represented in sensory-perceptual representations; this moment has often been referred to as “hot spot”. To form a consistent autobiographic narration of this moment, the sensory-perceptual representation is inevitably activated, as it provides detailed knowledge about the event that is not yet available in declarative structures. The autobiographic representation should cover the most salient stimuli of the sensory-perceptual representation with the highest probability of eliciting intrusive symptoms. As the activation of the sensory-perceptual representation is always accompanied by the emotional reactions coded in this structure, a high emotional involvement is necessary for therapy. Consistent with this view, Jaycox, Foa, and Morral (1998) showed that treatment success in exposure therapy is positively correlated with the level of fear initially experienced in treatment. The task of a therapist would be to encourage the activation of painful memories and to prevent the patient’s habitual strategies to avoid or terminate the activation. At the same time, the therapist should support the patient in the organization of declarative memory related to the traumatic event.

The question of how much emotional engagement is necessary and useful for therapy remains a matter of debate. Whereas Foa favored an initial maximum level of distress- ing emotions in exposure treatment indicating the full activation of the fear structure (Foa & Kozak, 1986), Brewin (2001) suggested a graduated approach to confrontation and the careful modulation of arousal during treatment. He pointed out that high stress levels might interfere with the activity of the prefrontal cortex and hippocampus and prevent the successful encoding of declarative knowledge. Dissociative experiences reported by the patient during therapy might indicate the malfunctioning of the encoding processes and should be avoided by regulating arousal. 101 Whilst the activation of the sensory-perceptual representation leads to a high emotional response, the increasing formation of autobiographical knowledge should lead to an increasing capability of cortical areas to inhibit the fear reactions. This effect can be observed as habituation of the emotional response. Habituation is closely monitored in exposure therapy by instructing the patient to report his current level of distress on a 0-100 SUD (“subjective units of distress”) scale. A decrease of the SUD value during exposure indicates habituation and is an indicator of treatment success (Foa & Rothbaum, 1998).

According to neurobiological knowledge of fear extinction, associations between stimuli and fear responses cannot be erased in therapy of PTSD patients. Instead, an alternative declarative knowledge should be constructed and tied to the stimuli that used to provoke intrusive symptoms. Several repetitions of the new associations may be necessary before the declarative knowledge is the preferred representation that is activated when reminded of the event. Only then can the fear responses and the associated intrusive symptoms be stably inhibited by the new knowledge base.

5.7 The speechlessness of trauma: sociopolitical implications

Beyond the fact that psychological trauma is a consequence of violence, there is increasing discussion of whether it can also be a cause of violent conflicts and wars. Organi- zations that provide psychosocial interventions in war affected societies have justified their interventions not only as a means of improving mental health care for individuals, but also by referring to sociopolitical factors. A common statement is that the treatment of “traumatized societies” is necessary to break a “cycle of trauma” (Tauber, 2003; UNICEF, 2001). This logic implies that traumatized individuals are more likely to become perpetrators later on. On the other hand, treatment of victims is considered to facilitate forgivingness and reconciliation within the society.

Contrary to the widespread use of this rationale, it is grounded on a weak empirical base. Whereas some investigations indicate that traumatized individuals are more likely to become perpetrators, there is not enough data to confirm a “cycle of violence”. For example, about 10% of male victims of child sexual abuse become abusers themselves as adults; this rate is considered to be much higher than the rate of abusers in the general population (Salter et al., 2003). Likewise, Vietnam veterans with PTSD are more likely to become per- 102 petrators of domestic violence than those without PTSD (Byrne & Riggs, 1996). Until now, there is no study that indicates that these findings can be transferred to political conflicts which result from the complex dynamics of different actors with interests at different levels, e.g. individual claims for power, economic interests, etc. It is unclear what role traumatized people play within these dynamics and whether they are more likely to contribute to the outbreak of wars. Feelings of revenge and hatred may result from violations and interfere with peaceful conflict resolutions, but these emotions are not pathological symptoms and they are not restricted to traumatized individuals. It is unclear whether a change of revenge feelings and an increase in forgiveness can be a desirable and realistic goal of mental health interventions.

The use of psychological reasoning to explain the outbreak of conflicts also risks re- introducing old myths about previous violations by the enemy as they are often used by war parties for propaganda purposes. For example, psychologists have explained the outbreak of the Balkan war as a consequence of the violence Croats perpetrated on the Serbs during World War Two, an explanation that strikingly resembles the Serbs’ propaganda.

Societies deal with matters of conflict resolution in different ways. One typical reaction of societies to political repression is to deny and minimize the violence and its consequences. In times of political suppression this silence can be understood as a reaction to the threat of the ruling regime and the general climate of intimidation. The situation is different in countries that have overcome war and dictatorship. Most often, there are incompatible interests in these societies. The victims’ demand for reparation and justice competes with ongoing dependencies from the perpetrators and the desire for peace. This conflict results in differing developments within these societies.

One way to offer justice after conflict is to set up international tribunals that deal with war crimes and human rights violations committed during conflicts and dictatorships. For example, the UN set up tribunals after the Balkan war and the Rwandese genocide. The experience of these tribunals shows that they have to deal with two major difficulties. One problem is that the capacity of the courts is limited and only a small minority of the perpetrators can be charged. The other problem is that the danger of juridical consequences might prevent the perpetrators from withdrawing their power and this might cause further conflicts. 103 Many countries (like Nigeria and most recently Kenya) have set up truth commissions following the example of South Africa after the Apartheid regime to overcome these problems. Truth commissions offer the opportunity for perpetrators to confess their deeds in public, but juridical consequences for perpetrators willing to speak out are neglected. The idea behind truth commissions was to find a compromise between the different interests, which means allowing a social processing of the past and the identification of the perpetrators in public. At the same time, the resistance of powerful parties is avoided as they don’t have to fear juridical consequences. Whilst truth commissions might offer a chance to balance justice and peace, the consequence is that neither the victims nor the offenders can be fully satisfied and their contribution to reconciliation is not yet proven.

The way countries and the international community deal with these matters depends on many different factors, like the power of the parties of perpetrators and victims, as well as the contribution of international communities in the termination of the conflict. Some conflicts, like the war in Mozambique, have resulted neither in tribunals nor in truth commissions. Until now, there is not enough experience to allow any recommendations to be made for the right way to deal with these issues. The variables of justice and peace after a conflict should be the major outcome variables by which the efficacy of social institutions in processing the violent past should be evaluated. Some researchers have pointed towards the importance of mental health variables in this context. The idea is that the political process after a conflict has an impact on victims’ well being, which is considered to be an important variable itself as well as a factor that might have an influence on the peace by initiating a “cycle of trauma”. Unfortunately, there is not much research into possible associations between political processes after conflict and mental health. The public acknowledgement of traumatic experiences might help survivors to cope with their experience, but there is little evidence for this effect. Examinations of mental health in victims who testi- fied at the South African Truth and Reconciliation Commission (TRC) (Kaminer, Stein, Mbanga, and Zungu-Dirwayi, 2001) concluded that there was no association between participation at the TRC and psychiatric status. They suggested that truth commissions are considered as only a part of therapeutic interventions for victims of human rights violations.

Whereas it is unclear whether political processes affect the victims’ mental health, it is very likely that the victims’ psychological status has an effect on social and political 104 processes beyond the assumed “cycle of trauma”. In every society there will be individuals who want to speak out about what has happened and to pass on their experiences to their children as well as to the public. Some of them will be able to do so on their own, and many local human rights groups have evolved to give these people a forum. Many survivors of the Holocaust have chosen to document their own experiences as a means of edu- cating the following generations (Frankl, 1946; Bettelheim, 1986).

A considerable number of people, especially those who suffer from PTSD, are unable to narrate their personal history because of the pathological effects of the traumatic events on their memory. This puts the victims at a disadvantage in comparison to the perpetrators and bystanders who usually have no difficulties explaining their position. Offer- ing these people a means of processing their traumatic events and documenting their history can help to give them a voice within their society.

During the Pinochet regime in Chile, Lira and Weinstein (published under the pseudonyms Cienfuegos and Monelli, 1983) developed Testimony Therapy (TT) as a specialized treatment approach for torture victims that directly addresses this issue. TT is an innovative approach that combines political as well as psychological goals. In TT, the biography of a survivor of human rights violations is documented in detail, with an emphasis on the persecution history and the traumatic events. The resulting documents have been used to accuse the regime of human rights violations and became a powerful tool in the resistance to the Pinochet dictatorship. Thus, this treatment approach offered both the emotional processing of the victim’s traumatic event, resulting in improved mental health, and a document which could be directly used for political purposes. Other therapists have followed this example and used TT for different groups of survivors of wars and torture (Agger & Jensen, 1990; Weine, Kulenovic, Pavkovic, & Gibbons, 1998), but, despite reports offering promising results, TT is not widely applied among psychosocial organizations that still favor non-political approaches, such as supportive counseling, in war affected populations. 105

6 Psychotherapy of PTSD

6.1 Overview

Different schools of psychotherapy have a tradition of treating trauma patients. For example, hypnosis has been widely used to support the resolution of conflicts resulting from “real“ traumatic events through encouraging the processes of catharsis and abreaction in a psychoanalytical framework (Spiegel, 1989). A variety of specialized approaches for the treatment of PTSD have been developed since PTSD was introduced as a formal diagnosis in DSM. The proponents of new treatment techniques usually claim that their approach is as effective, or even more effective, than other methods. In contrast to the statements of therapists, only a minority of these approaches have undergone a rigorous testing of their efficacy.

The standard test of the efficacy of a medical or psychological intervention is a treatment study that includes a reliable measurement of symptom severity before and after treatment. It is essential to compare the development of a group of patients who received treatment with other groups who either received no treatment or a placebo treatment to control for spontaneous remission and unspecific treatment effects. The participants should be randomly assigned to the different groups to prevent selection factors. This type of study, which has evolved as a standard in psychotherapy research, is called a randomized controlled trial.

The results of randomized controlled trials have become the main way of classifying therapy approaches as effective or not. However, contrary to the claims of some proponents of treatment approaches, a randomized controlled trial is not sufficient to prove the explanatory power of the method under examination. In addition to the random assignment of participants to different treatment conditions, other methodological demands are essential to achieve a satisfying evaluation of a treatment. Foa and Meadows (1997) have suggested the following “gold standards” for treatment outcome studies in addition to the randomized assignment to different treatment conditions:

First of all the target symptoms must be defined before treatment. This includes the proper definition of inclusion and exclusion criteria to determine the patients to whom this 106 study is relevant. The measures used to define these criteria as well as the severity ratings of the symptoms must have good psychometric qualities, and the assessors have to be trained in the application of the instruments. The evaluators who carry out the assessments should not be the people carrying out the therapy and they should be blind to the treatment the participant has received to avoid the influence of expectancy biases. To ensure consistent treatments across patients and therapists and to enable the transfer of the results to other therapists, the treatment should be standardized and detailed manuals should be available. Within the study, the treatment adherence should be rated to prove that the treatments of the different conditions were carried out according to the design of the study.

In a review of treatment trials for PTSD, Foa and Meadows (1997) revealed that only a minority of the studies fulfilled these methodological criteria. Less rigorous trials might inflate the outcome in favor of the proposed approach, so a close look at the methods is relevant to the evaluation of treatment research. In the following review of treatment approaches for PTSD, I will limit the studies presented to those that fulfilled at least part of the standard criteria.

6.2 Acute Interventions/Debriefing

Many people who experience very stressful events immediately present with an emotional upheaval. Some psychologists thought that it might be necessary to offer these people immediate assistance to prevent long lasting disturbances. Different structured models for the support of people in a crisis have been developed. The most prominent approach is the “Critical Incident Stress Debriefing” (CISD) that was created by Mitchell (Mitchell & Bray, 1990). CISD is a structured single session group intervention which takes about 1.5 to 3 hours and is typically offered within one week of the event. The goal of the intervention is to “provide some facilitation of the process of psychological ‘closure’ upon the traumatic, or critical, incident” (Everly & Mitchell, 2000, p. 212.) A CISD intervention consists of seven phases, including “recreating” the event by having all participants give their perspective on the worst part of the event for them and an educative intervention, where common reactions to traumatic events are discussed.

Since the introduction of CISD, several studies have evaluated the efficacy of debriefing interventions. Only a few of these studies included control groups with a random 107 assignment. These studies compared the development of a group of recent trauma victims who took part in a debriefing session with those who received no intervention. In a systematic review of six randomized controlled trials Rose and Bisson (1998) concluded that the studies presented diverse outcomes. In some trials there was an improvement, whereas in others the symptoms of the participants did not change in comparison to the control group and in others the intervention group developed worse symptoms than did the control group. One of the studies included a 3-year follow-up of victims of road traffic accidents and found that those who had been assigned to the intervention group were in a significantly worse condition with respect to PTSD symptoms compared to the control group three years later (Mayou, Ehlers, & Hobbs, 2000). A recent meta-analysis of randomized controlled trials of debriefing and other acute interventions (van Emmerik, Kamphuis, Hulsbosch, & Emmelkamp, 2002) found generally low effect sizes of debriefing. The effect sizes were close to zero when accumulated in the analysis.

The randomized controlled trials do not provide evidence to support the use of debriefing in the immediate aftermath of a trauma. These studies show that intuitively appeal- ing interventions do not necessarily have the desired effect and that rigorous research is necessary before one can apply interventions on a large scale. Nevertheless, the proponents of CISD have defended their approach (Everly & Mitchell, 2000). They criticized the trials mentioned above for not adhering to the CISD protocol, as all these studies applied debriefing in an individual rather than group format. In addition, they stressed that debriefing was not constructed as an isolated intervention and must not be evaluated as such. In- stead, they presented a wider context of crisis intervention now called “Critical Incident Stress Management” (CISM), that includes pre-crisis preparation, individual crisis intervention, pastoral crisis intervention, family CISM and other approaches into one framework. Still, the efficacy of these single interventions is not yet proven and it is doubtful that merging all these steps into one approach will simplify a rigorous test of their efficacy.

Several explanations have been put forward of the inefficacy of debriefing (van Emmerik et al., 2002). One major problem of CISD could be that it is offered to all victims of traumatic event, and does not discriminate between those who are at risk of psychological disturbances and those who are not affected. In this way, CISD might impede the natural emotional and social processing process for the majority of victims who would recover on their own. The education about psychological symptoms might be critical for those vic- 108 tims as it might lead to a sensitization towards intrusive and avoidance phenomena that could be interpreted as potentially pathological. For the minority of victims who are cannot manage the processing on their own and who are at risk of developing PTSD the procedure might be too short and the time offered to talk about the traumatic event within a single session might not be enough to allow a habituation of the emotional responses.

Other acute interventions have been suggested to prevent the development of PTSD. Brief cognitive behavioral procedures including exposure methods have shown promising results (Bryant, Harvey, Dang, Sackville, & Basten, 1998; Foa, Hearst-Ikeda et al., 1995). These approaches differ from the debriefing method as they are offered only to victims who present with acute stress disorder within four weeks after the traumatic event, and are therefore at high risk of developing PTSD. They consist of four to five weekly individual sessions. Imaginary exposure to the memory of the traumatic event seems to be the most effective intervention for the prevention of PTSD (Bryant, Sackville, Dang, Moulds, & Guthrie, 1999).

6.3 Psychodynamic Therapy

Horowitz and colleagues have developed a psychodynamic short-term therapy for the treatment of PTSD (Horowitz, 1986). They have developed a theoretical account to explain reactions to traumatic events. In this theory, which is close to cognitive theories, the trauma symptoms are seen as the result of discrepancies between internal and external information. Response to trauma is seen as a vacillation between the phases of denial (ex- pressed with avoidance symptoms) and intrusion. The intrusive phase is seen as an attempt to present information about the external event to the cognitive system to enable the integration. If this integration fails, the person remains at the stage of vacillation for a long time, which can be defined as the development of chronic PTSD. The treatment offered to these people is adjusted to the patient’s current stage of processing. During the intrusive phase, a person is instructed to avoid memories and to control anxiety through a supportive therapeutic environment. When a person is in the phase of avoidance, confrontation techniques are used to enable the integration and abreaction. The techniques used for this approach are similar to cognitive-behavioral interventions. 109 This treatment was evaluated in a controlled study that compared hypnosis, systematic desensitization and psychodynamic therapy with a waiting-list control group (Brom, Kleber, & Defares, 1989). All three treatments were effective in comparison to the waiting-list, but no significant differences were observed between the active treatment groups. The quality of the study was limited as the duration of treatments was not standardized across the conditions and evaluators were not blind to the treatment conditions.

6.4 Anxiety management

Anxiety management programs are combinations of standard cognitive behavioral procedures for the treatment of pathological anxiety. They are based on the assumption that maladaptive coping mechanisms are responsible for the maintenance of anxiety symptoms and that increasing the skills to deal with the symptoms leads to recovery (Meichenbaum, 1994). Veronen and Kilpatrick (1983) have adapted a standardized variant of anxiety management, the stress inoculation therapy (SIT), for victims of rape. It includes teaching breathing exercises, relaxation, thought stopping, self-guided dialogue, covert modeling and other coping skills. Two randomized controlled trials have included SIT groups for female assault victims (Foa et al., 1999; Foa, Rothbaum, Riggs, & Murdock, 1991). They found that, although it was less effective than exposure methods in the follow- up, SIT was effective in treating PTSD compared to a waiting-list control.

6.5 Exposure oriented treatment

Exposure techniques are a different cognitive behavioral approach for the treatment of anxiety disorders. The rationale behind these methods is to confront the patient with the feared stimuli in order to prevent avoidance behavior. Exposure can be classified with regard to the medium for confrontation. In in-vivo exposure, the patient is directly exposed to the feared object, whereas in imaginative or in-sensu exposure the patient is instructed to imagine the stimulus intensively. Exposure techniques differ with regard to the level of arousal that is induced during confrontation. In graduated exposure, only moderate arousal is necessary as confrontation starts with objects that provoke medium anxiety and is gradually increased towards the most feared stimuli. Systematic desensitization for example is an 110 imaginative procedure that requires minimum arousal to begin with, where as flooding is an in-vivo technique that start with the most feared stimulus and requires maximum arousal.

6.5.1 Systematic Desensitization

When cognitive behavioral treatment for PTSD began, several therapists applied systematic desensitization and reported good results. In one study described above, systematic desensitization proved as successful as hypnosis and psychodynamic therapy (Brom et al., 1989).

6.5.2 Prolonged Exposure

Instead of using an exposure approach that requires minimal arousal, other researchers combined prolonged imaginal exposure (requiring high levels of arousal) and in- vivo techniques. Prolonged exposure (PE) is a standardized exposure method for trauma victims that was introduced by Foa et al. (1991). It requires the patient to narrate the event in detail and to relive the emotions associated with the event as intensively as possible. The following instructions are used for imaginal exposure:

“I am going to ask you to recall the memories of the assault. It is best for you to close your eyes, so you won’t be distracted and so you can envi- sion these events in your mind’s eye. I will ask you to recall these painful memories as vividly as possible. […] What I would like you to do is describe the assault in the present tense, as if it were happening now, right here. I’d like you to close your eyes and tell me what happened during the assault in as much detail as you remember. We will work on this together. If you start to feel too uncomfortable and want to run away or avoid it by leaving the image, I will help you to stay with it […]” (Foa & Rothbaum, 1998; p. 162)

The patient reports the event for 60 min per session and he is instructed to start again from the beginning if he finishes before this time. During the exposure, the therapist regularly asks for the patient’s current anxiety level on a 0-to-100 subjective units of distress (SUD) scale. The narratives are recorded on audio-tapes and the patient’s homework is to listen to the tape at least once a day. Additional homework involves in-vivo exposure to feared and avoided objects.

Several methodologically sound randomized controlled trials have demonstrated the efficacy of PE. Exposure was superior to supportive counseling, relaxation training, SIT 111 and waiting list conditions (Foa et al., 1999; 1991; Marks, Lovell, Noshirvani, Livanou, & Thrasher, 1998; Tarrier, Sommerfield, Pilgrim, & Humphreys, 1999; Tarrier, Pilgrim et al., 1999).

6.5.3 Eye Movement Desensitization and Reprocessing (EMDR)

Eye Movement Desensitization and Reprocessing (EMDR) is a treatment approach that has become very popular among therapists since its introduction by Shapiro (1989). It was originally developed for the treatment of traumatic memories but recent formulations propose the applicability of EMDR for phobias, panic disorder and other psychiatric diagnoses.

EMDR is an imaginal exposure technique that has the unique feature of requiring the patient to execute eye movements whilst imaging the traumatic event. It differs from PE as exposure does not require a detailed narrative of a single event but follows associative memories that appear during treatment. The first formulations of EMDR proposed that a single session of EMDR would be sufficient to treat traumatic memories (Shapiro, 1989).

Many randomized controlled trials have been carried out to evaluate the efficacy of EMDR for the treatment of PTSD. Whilst only a few of these studies fulfilled the standard criteria for methodological sound studies, it is generally accepted that EMDR is more effective than a waiting-list control condition (Foa & Meadows, 1997).

Many studies have targeted other claims made by proponents of EMDR. The most distinct feature of the EMDR treatment that differentiates this approach from other exposure techniques is the eye movement component. Some studies found that EMDR is not superior to control conditions that use other alternating stimuli like finger tapping, therefore the conceptualization of EMDR was changed to the assertion that not the eye movements but the addition of alternating stimuli during exposure is the crucial element (Shapiro, 1995). However, other studies compared EMDR with a condition that used no alternate stimulation but required the patients to keep their eyes fixed during exposure. In a meta-analysis of EMDR treatment studies, Davidson & Parker (2001) concluded that neither eye movements nor other dual tasks increase the efficacy of exposure. The fact that studies have not found EMDR to be more effective than cognitive behavioral approaches, 112 including exposure, raises further doubts about the necessity of any additional stimulation during exposure.

In a review of the EMDR literature Herbert and coauthors (2000) heavily criticized EMDR proponents who exaggerate the expected outcome of the treatment. According to Herbert, the dissemination of EMDR follows principles of commercial promotion rather than a critical reflection of research results. Training in EMDR follows a rigid system with the requirement that each trainer has to be licensed with the EMDR institute. Full training by a licensed trainer has been claimed to be a prerequisite of effective EMDR therapy. Contrary to this assertion, Davidson and Parker (2001) found that studies that involved therapists who received formal training at an EMDR institute did not produce better results than those involving therapists who did not receive such training.

Research into EMDR has led to a considerable change in the procedure and theory of this approach. The Adaptive Information Processing (AIP) model now proposed by Shapiro (Shapiro & Maxfield, 2002) is strikingly similar to the theories of emotional processing formulated by Foa and Kozak (1986) and other authors. The suggested number of sessions required for therapy has changed from one session to a variable amount of sessions adapted to the treatment progress. As cognitive interventions as well as anxiety management techniques have been incorporated into the procedure, EMDR has evolved into an eclectic approach consisting of a combination of cognitive behavioral techniques. As a result of this development, the assertion that EMDR is a novel and unique treatment approach can no longer be maintained.

6.6 Cognitive Therapy

Cognitive therapy was originally developed by Beck (Beck, Rush, Shraw, & Emery, 1979) for the treatment of depression. Since then, several variants have been developed to treat different psychological disorders beyond depression, i.e. anxiety and personality disorders. Cognitive therapy is based on the theory that dysfunctional thinking patterns de- riving from maladaptive beliefs or schemas are responsible for pathological emotions and psychiatric symptoms. In cognitive therapy, the patient is instructed to detect, challenge and modify dysfunctional thinking patterns. 113 Resick and Schnicke (1993) have adapted cognitive therapy for the treatment of PTSD. Their approach was named cognitive processing therapy (CPT). The basic assumptions of CPT can be outlined as follows: Every person holds basic beliefs about oneself and the world. An example of such a belief would be “There are many dangers in this world but in general I am safe”. The experience of a traumatic event, for example a rape, contradicts many of these assumptions in a fundamental way (Janoff-Bulman, 1992). When a person is exposed to schema-discrepant information, two different things can happen: The information can be changed so it fits to the existing schemas, for example “It was not really a rape”. Intrusive memories are explained as attempts of integration when the cognitive processing cannot be resolved. In this example, the assimilation might not form a stable integration as the person still knows that she has been severely harmed.

The other alternative is that the existing schema can be altered to accommodate the discrepant information. An adaptive accommodation would be a modified schema that fits the experience but still allows functioning. For example, “This road is a dangerous place at night” can be an example of an adaptive accommodation. However, an adaptive accommodation needs some effort and social support as it happens through active thinking and discussion of the event and the consequences. Without good support, overaccomodation can happen as an extreme change of a schema that allows the integration of the event but severely hinders good functioning. An example of overaccomodation would be “I can’t be safe anywhere”. This schema would cause a person to feel frightened almost everywhere.

The goal of cognitive therapy is to assist traumatized people in the cognitive processing of the event, which means the adaptive modification of schemas. Some themes have been found to be of special relevance for rape victims, including safety, trust, power, es- teem, and intimacy (McCann, Sakheim, & Anbrahamson, 1988).

CPT is offered in a group format and consists of 12 sessions of 90 min. Techniques used in the program were adapted from other cognitive approaches. They include writing about the meaning of the event and affect protocols to understand the connection between beliefs, thoughts and emotions. Furthermore, the patients are instructed to write a detailed report of the event. As the reliving of emotions whilst writing the event is encouraged, CPT contains an exposure component. Most of the time is used for the discussion of the consequences of the event in terms of the person’s beliefs. Using the different cognitive 114 techniques, maladaptive beliefs are detected and changed into adaptively assimilated schemas.

CPT has proven to be effective in a randomized controlled trial against a waiting list comparison group (Resick & Schnicke, 1993). As CPT also includes an exposure component, it is difficult to extract the effective mechanism. The trial of Tarrier and coworkers (Tarrier, Pilgrim et al., 1999; Tarrier, Sommerfield et al., 1999; N=72) compared cognitive therapy with PE. They found no difference in effectiveness, but reported that more patients worsened in exposure therapy. This argument was criticized by Devilly and Foa (2001) who argued against Tarrier’s operational definition of worsening and pointed out the comparatively low effect sizes found in this study, suggesting problems with treatment integrity. Support for this was that, contrary to Foa’s suggestion of PE, Tarrier did not include in-vivo exposure homework. Another study that compared cognitive therapy with PE was carried out by Marks et al. (1998; N=87). This trial also included a relaxation control condition. Again, no significant difference between PE and cognitive therapy was found, while both treatments were effective in comparison to relaxation. In a recent study with 171 patients, Resick and coworkers (Resick, Nishith, Weaver, Astin, & Feuer, 2002) carried out another comparison between CPT and PE, and included an additional waiting list condition. They found both CPT and PE to be highly effective in comparison to the waiting list, but there was no major difference between CPT and PE.

6.7 Combination of treatments

Different procedures have been found to be effective in the treatment of PTSD. The fact that these procedures propose different mechanisms has led to the assumption that a combination of approaches might enhance the efficacy of psychotherapy. So far, two studies have tried to evaluate the efficacy of combinations of treatments. The study of Marks et al. (1998) compared the combination of exposure and cognitive therapy with each method alone and with relaxation training as control condition. The participants in each condition received the same time of 10 sessions of treatment. All three conditions showed good effects and were superior to relaxation training, but, contrary to the expectations, the combination of treatments brought no additional effect. Foa et al. (1999) tested the efficacy of a combination of SIT with PE against SIT and PE alone and a waiting list condition. Again, 115 all active treatments reduced severity of PTSD but did not differ significantly from each other and a combination had no additional effect.

Two studies show that although different treatments can be effective, a combination of treatments does not lead to a better outcome. This warrants caution for the tendency of many therapists to propose the superiority of eclectic treatment approaches. Several reasons have been suggested for the lack of superiority of the combined treatments (Foa, 2000; Foa et al., 1999). As the duration of the treatments was kept constant across the conditions, many different procedures were packed into the combination condition, and this may have led to an overload of information. The combined programs also required double the amount of homework from the participants, which might reduce the probability that the participants found enough time to complete the tasks.

6.8 Treatment of survivors of organized violence

In contrast to the considerable body of PTSD treatment research, knowledge of effective treatment of PTSD in populations of civilians affected by war is still scarce. Not- withstanding the considerable attention war-torn populations have received from psychosocial organizations in recent years, very little clinical research has been conducted into how to adequately support and treat these groups.

The majority of survivors of war and torture are unable to safely escape their countries, forced instead to flee to insecure places within their home country or in adjacent regions that are often equally affected by war and terror. In addition to living with violence, many of these refugees are also living in poverty, dependent on humanitarian aid and suffering from malnourishment. These living conditions question the applicability of psychotherapeutic treatment approaches that have been developed for western PTSD populations. First of all, it is unclear how many refugees suffering from PTSD seek and accept aid that might provide relief (Van der Veer, 1998). Looking to Maslow’s hierarchy of needs, for in- stance, it seems evident that treatment for psychological problems cannot be addressed as long as the basic needs of nutrition and safety are pressing. Furthermore, it is questionable whether any psychological assistance can be successful given that treatment methods for PTSD typically require the establishment of a safe and reassuring environment, conditions that are difficult to meet within the context of refugee camps and settlements. In addition, 116 any psychotherapeutic work that takes place must be brief given the large numbers of people and limited monetary resources. Any broad scale treatment program must be pragmatic and easy for local personnel to learn, even with little or no access to medical or psychological education or additional training. Consequently, the method must be adaptable to any cultural environment and easily implemented. The oral tradition is a common element among many cultures, thus narrative approaches seem ideally suited to cross cultural appli- cations.

Psychosocial organizations usually provide forms of counseling approaches to assist traumatized refugees (Van der Veer, 1998). These types of treatment, often referred to as trauma-counseling or cross-cultural counseling (Doherty, 1999), encompass a large variety of approaches, such as problem solving procedures (AMANI, 1997) as well as group discus- sions and individual assistance including variants of exposure treatment (WHO/UNHCR, 1996). In most cases, the counseling procedures and duration of treatment are not rigor- ously standardized. The lack of standardization creates serious difficulties for the evaluation of counseling. Little knowledge about the efficacy of counseling in general exists (Bower et al., 2002), and as of yet there is no clinical trial that has examined the efficacy of counseling approaches for traumatized refugees.

A common element in brief interventions for the assistance of large numbers of traumatized refugees is psychoeducation, thought to cause some relief for traumatized people by explaining the nature of the symptoms caused by traumatic experiences and by demonstrating that this response is normal. Psychoeducation has been used as an add-on element with various treatment approaches (Foa & Rothbaum, 1998; WHO/UNHCR, 1996). Some therapists suggest using psychoeducation as an economic and effective large- scale intervention to reduce the suffering of populations affected by war.

Only a minority of organizations providing psychological assistance have tried to adapt scientific knowledge from PTSD research to the conditions of traumatized refugees. As an exception, Paunovic and Ost (2001) show that exposure treatment is a promising treatment approach for refugees. They compared the efficacy of a 16-20 session exposure based treatment with a combination of cognitive and exposure interventions of equal duration for 16 refugees who had fled to Sweden. Both treatments were similarly effective. A limitation of this study was the rather rigid inclusion criteria. Therapy was only offered for refugees who had a lasting Swedish residence permit. In addition, the treatment was con- 117 ducted in Swedish and thus the patients had to be relatively well adapted to the Swedish culture. These findings can be applied only to the minority of refugees who manage to flee to Western countries and to adapt to the new culture. Furthermore, there was no control group in this study that controlled for spontaneous remission and unspecific effects. 118

7 Narrative Exposure Therapy (NET)

7.1 Basic principles of NET

We developed Narrative Exposure Therapy (NET; Neuner, Schauer, Elbert, & Roth, 2002) as a standardized short-term approach based on the principles of cognitive behavioral exposure therapy by adapting the classical form of exposure therapy to meet the needs of traumatized survivors of war and torture. In exposure therapy, the patient is re- quested to repeatedly talk about the worst traumatic event in detail while re-experiencing all emotions associated with this event. In the process, the majority of patients undergo habituation of the emotional response to the traumatic memory which consequently leads to a remission of PTSD symptoms.

As most victims of organized violence have experienced many traumatic events, it is often impossible to identify the worst event before treatment. To overcome this difficulty, we based our approach on Testimony Therapy (TT), a method of therapy created by Lira and Weinstein (Cienfuegos & Monelli, 1983) to treat traumatized survivors of the Pinochet regime in Chile and successfully applied in a uncontrolled trial to Bosnian refugees in the US (Weine, Kulenovic et al., 1998). Instead of defining a single event as a target in therapy, the patient constructs a narration of his whole life from birth up to the present date while focusing on the detailed report of the traumatic experiences.

The focus of the NET procedure is twofold. As with exposure therapy, one goal is to reduce the symptoms of PTSD by confronting the patient with memories of the traumatic event. However, recent theories of PTSD and emotional processing suggest that the habituation of the emotional responses is only one of the mechanisms that improves symptoms. Other theories suggest that the distortion of the explicit autobiographic memory of traumatic events leads to a fragmented narrative of the traumatic memories (Ehlers & Clark, 2000). Thus, the reconstruction of autobiographic memory and a consistent narrative should be used in conjunction with exposure therapy. NET focuses on both methods, the habituation of emotional responding to reminders of the traumatic event and the construction of a detailed narrative of the event and its consequences. 119 Narrative Exposure Therapy (NET) is a treatment approach that was developed for the treatment of PTSD resulting from organized violence. The basic procedure of the Nar- rative Exposure can be outlined as follows:

In a relatively small number of sessions, the client constructs a detailed and consistent narration of his biography in cooperation with the therapist. The focus of the therapy lies on the completion and integration of the initial fragments of the traumatic events into a whole, including the sensory, emotional and cognitive experiences of the incident. The testimony is written down and, depending on the willingness of the client, also used for documentary purposes. This procedure has been adapted to the special demands of the situation in a refugee camp and has evolved to the following standard:

After the assessment of PTSD a psycho-educational introduction is given to the survivor, focusing on the explanation of his/her disturbance and symptoms, as well as a statement about the universality of human rights, followed by a preparatory introduction to the therapeutic approach. Treatment starts immediately after a diagnostic assessment, which gathers demographic data, medical and psychiatric history, and current complaints.

Using a semi-structured interview, the following topics are explored in consecutive sessions:

Individual and family history prior to the persecution

Experiences from the beginning of persecution to the first terrifying event

Terrifying events

History of flight

Life in the refugee camp

Plans, hopes and fears concerning the future

The therapist structures the topics and helps to clarify ambiguous descriptions. S/he takes an empathic and accepting role. Inconsistencies in the client’s report are gently pointed out and often resolved by raising in-depth awareness about recurring body- sensations or thoughts. The client is encouraged to describe the traumatic events in as much detail as possible and to reveal the emotions and perceptions experienced at that moment, being assured that s/he is in control of the procedure at all times and will not be 120 pushed to do anything against his or her will. A translator, oriented beforehand to the psychological goals, may be necessary.

The therapist writes down the client’s narration (or the translation if necessary). In the subsequent session, this report is read to the client (simultaneously translated by the interpreter if necessary) and s/he is asked to correct it and to add further details. The procedure is repeated in subsequent sessions until a final version of the client’s biography is reached.

In the last session, this document is read again and the client, the translator and the therapist sign the written narration. One copy of the signed document is handed to the client; another is kept for scientific purposes. With the agreement of the client, another copy is passed on to human rights organizations for documentary and advocacy purposes or published in any other way.

7.2 Randomized controlled trial

7.2.1 Introduction

This treatment study was carried out with Sudanese refugees who fled from the civil war to Northern Uganda. Currently about 200,000 Sudanese refugees live in Uganda; many of them have lived in the country for several years. In the setting of a refugee settlement we compared the efficacy of three treatment conditions in a randomized controlled trial. One group received psychoeducation (PE) only which was completed in one session. This group served as a control group to monitor the development of refugees who only received a minimal intervention beyond the diagnosis procedure. A second group received four sessions of supportive counseling (SC) in addition to PE. The lack of standardized procedures of trauma counseling prevented the inclusion of a group that was representative for the current standard of trauma counseling. Thus, the main purpose of the SC group was to control for unspecific treatment effects. Finally, a third group was offered four sessions of NET, which also included a PE component. Standard clinical instruments were applied for the examination of treatment effects immediately after therapy as well as four months and one year after treatment. We predicted a significant impact of the treatment condition on the development of the psychological status of the subjects after treatment. Instruments of 121 posttraumatic stress were the main outcome measures. In addition, we applied measures of comorbid anxiety and depression as well as psychological health. Specifically, we expected the participants in the NET group to present with the best outcome at the 1-year follow- up.

7.2.2 Methods

Setting

This study followed a project that researched the demography of forced migration (Karunakara et al., submitted; Schauer et al., 2001). The first step of this project was to investigate the impact of war and forced migration on levels of fertility, child mortality, violence and stress in Sudanese and Ugandan populations in northwestern Uganda and southern Sudan. Following the wide scale survey, the Imvepi settlement in northern Uganda was selected for the treatment group as the refugees living there presented with the highest prevalence of PTSD (51%). In Imvepi, most refugees had experienced multiple traumatic events in the Sudanese civil war before they fled to Uganda. However, northern Uganda was not a safe exile for the refugees as the settlements were threatened and attacked by Su- danese and Ugandan rebel armies. At the time of this study (2000-2001), living conditions were still inadequate with most refugees suffering from poor nutrition and poverty.

Participants

We randomly chose 77 participants from a list of respondents who had previously been randomly selected in a hut-to-hut procedure for the survey. 43 (56%) of these respondents presented with PTSD according to DSM-IV criteria. All of them were offered the opportunity to participate in this treatment study, as no respondent met the exclusion criteria of mental retardation or psychosis as proven by a clinical examination. The objectives of randomized controlled trials and of this study in particular were fully explained to the patients. With the exception of only one respondent, all were willing to participate. Those who decided to participate gave written informed consent either with their own signature or the signature of a witness who was able to write. The informed consent and the study protocol were approved by the Ethical Review Board of the University of Konstanz and by the Ugandan National Council for Science and Technology, Kampala. 122

Table 7.1 Sociodemographic characteristics of patients in the different treatment groups

NET SC PE (n =17) (n =14) (n =12) Age (SD) 31.9 (6.7) 33.8 (7.9) 34.2 (6.9) Sex N (%) Male 7 (46.7) 6 (42.9) 3 (25.0) ! Female 9 (53.3) 8 (57.1) 9 (75.0) Marital Status N (%) Single 4 (25.0) 2 (14.3) 1 (8.3) Married / cohabiting 12 (75.0) 10 (71.4) 8 (66.7) Widowed 0 (0.0) 2 (14.3) 3 (25.0) Religion N (%) Catholic 4 (25.0) 5 (35.7) 5 (41.7) Protestant/ other Chris- tian 11 (68.8) 7 (50.0) 7 (48.3) Other 1 (6.3) 2 (14.3) 0 (0.0) Education N (%) None 13 (81.3) 14 (100.0) 11 (91.6) Primary school 3 (18.7) 0 (0.0) 1 (8.4) Occupation N (%) None 3 (18.8) 5 (35.7) 4 (33.3) Agriculture 8 (50.0) 7 (50.0) 5 (41.7) Other 5 (31.3) 2 (15.3) 3 (25.0) Nr. of possessions (SD) 1.38 (0.61) 1.51 (1.22) 1.86 (1.22) Nr. of meals per day (SD) 1.27 (0.59) 1.36 (0.17) 1.55 (0.82) Years since last migration (SD) 3.23 (0.93) 3.36 (0.67) 3.40 (0.52) Nr. of traumatic event types (SD) 9.27 (6.68) 11.42 (6.21) 9.58 (6.70)

The respondents found it difficult to identify one single worst event as they had usually experienced a series of traumatic events with similar high intensity. Instead of insisting on the identification of one event, we asked for the most severe type of events they had experienced or witnessed. The majority of participants (52.3%) reported the witnessing of people badly injured or killed as worst event type (which included the killing of relatives as well as witnessing massacres and mutilations), further worst event types were threats with weapons and kidnappings (16.7%), physical attacks (11.9%), torture (7.1%), combat 123 experiences (7.1%), sexual assaults (4.8%) and natural disasters (2.4%). The average time past since the worst period of traumatic events was 7.52 years (SD = 3.25). Table 7.1 shows the sociodemographic characteristics of the participants in the treatment study. The randomization procedure resulted in different group sizes. There were no systematic group differences in any of the sociodemographic variables, as confirmed by Fisher’s Exact and ANOVA testing.

The participants received a compensation of 2000 Ugandan Shilling (equivalent to 1.25 US$) for participation in the initial screening interview. They were told that they would receive no further monetary compensation or food for participation in the treatment study. However, to compensate for traveling to the location of the four months and 1-year follow-up examinations, we provided the equivalent of 3.00 US$. Participants were not informed that they would be receiving this money when they agreed to participate and therefore did not expect any payment when participating in the treatment study or for the follow-up interview.

Instruments

Expert Interviews. The DSM-IV diagnosis and severity of PTSD was assessed using the PTSD part of the Composite International Diagnostic Interview (CIDI; WHO, 1997). The items of the interview were translated before the interview and presented with the help of trained local translators. Further inquiries about details and examples of symptoms were made by the experts with the help of the translators to ensure a correct understanding of the symptoms and to validate the clinical significance of symptoms and the severity rating of PTSD.

Assisted Self-Report. The Demography of Forced Migration Questionnaire (DFMQ) was used to assess detailed information about demography, health and traumatic experiences. For the assessment of traumatic experiences, a checklist of 31 different types of traumatic events (witnessed events and experienced events) was administered. Each event was scored as present or not present in the respondent’s life. The number of different experienced and witnessed types of traumatic events was used for the estimation of the severity of trauma exposure. Economic status was estimated using a household asset list that included essential household items such as blankets, pots, etc. The relevant indicator was the number of possessions per adult in a household. Nutrition was evaluated by asking for the 124 number of meals eaten the previous day. Self rating of symptoms and functioning was carried out using standardized clinical instruments. Because of a high illiteracy rate, the translations of the self-rating instruments were administered in the form of standardized interviews by a trained local team. The frequency and severity of PTSD symptoms was assessed using the Posttraumatic Stress Diagnostic Scale (PDS; Foa, 1995), which was modified during the translation procedure to simplify the frequency rating of symptoms. The Self Re- port Questionnaire 20 (SRQ-20; Harding et al., 1980) was used to indicate comorbid symptoms of anxiety and depression. This instrument had been widely used in Africa previously. Psychological functioning was measured using psychological health scale from the 12-item version of the Medical Outcome Study Self Report Form (SF-12; Ware et al., 1996). This questionnaire was used to assess health related quality of life in different cultures. Sev- eral items of the questionnaire had to be adapted to the living conditions in an African refugee settlement. As no comparison data was available for the Sudanese refugees, the usual standardization procedure was not used for scoring.

All instruments were translated into the Arabic dialect spoken by the refugees in Imvepi (Juba-Arabic). Several steps of blind back translations and subsequent corrections were necessary to ensure a valid translation of the questionnaires.

Procedure

Each respondent received two initial interviews in a screening phase, one by a local interviewer (DFMQ, PDS, SRQ-20, SF-12) and one by experts from Konstanz University (CIDI). Those who presented with PTSD in the CIDI interview were offered participation in the treatment study. Each participant was randomly assigned (using a dice) to one of three treatment groups: NET, SC or PE only. The treatment conditions consisted of one (PE) or four (NET, SC) sessions of treatment. The first session started immediately after the initial expert interview. For the NET and SC groups, the subsequent three treatment sessions were scheduled within the following two weeks. The duration of a treatment session was approximately 90 min but could extend up to 120 min for exceptional sessions. Three to four weeks after the first interview, a post-test using the standard battery of PDS, SRQ-20 and SF-12 was carried out by local interviewers. This battery was also administered at follow-up examinations four and 12 months after treatment. The 1-year follow-up test included a further interview with the CIDI administered by trained researchers. In addi- 125 tion, the trauma event checklist was applied once again to record the individual’s history of traumatic events during the follow-up period.

The local and expert interviewers who carried out the post-tests, as well as the follow-up tests, were blind for the individual participant’s treatment condition. The respondents were instructed not to inform the interviewers or the trained researchers about the type of treatment or the number of sessions they had received. The initial interviews and the treatment sessions took place in huts or under trees around the settlement’s medical center, which consisted of three straw huts. The post and follow-up tests were carried out at the same place or in the respondents’ huts for those who had moved to places that were too far away to travel to the settlement.

7.3 Treatment

Treatment was carried out by three female and two male therapists from the Uni- versity of Konstanz and the aid organization Vivo with the help of interpreters trained in a one-week course. Therapists were PhD level psychologists or graduate students with education and experiences in other treatment approaches including counseling. Therapists were trained in NET and SC by M.S. and F.N, who had developed NET and were educated in counseling, through direct observation of treatments. All therapists carried out treatments in all conditions. Weekly meetings were appointed to supervise the quality of treatments. As audio recordings were not practicable in the setting, treatment adherence was monitored by requesting the therapists to give a detailed account either on the participant’s current life problems (SC) or the participant’s life history recorded so far (NET) at least once for each treatment in the supervision meeting. In addition, selected treatment sessions were directly observed by M.S. or F.N. No major deviations from treatment protocol were detected.

The first session always included psychoeducation about the nature and prevalence of PTSD symptoms independent of the treatment group. A standard written rationale was developed for this. The goal of the procedure was to explain that PTSD-related symptoms and dysfunction frequently occur after multiple traumatic experiences. For the NET and SC group, psychoeducation was followed by the explanation of the respective treatment rationale. For participants in the PE group, no further treatment was offered. 126 Narrative Exposure Therapy (NET). In NET, the participant constructs a detailed chronological account of his own biography in cooperation with the therapist. The autobiography is recorded by the therapist and corrected with each subsequent reading. Special focus of the therapy is on the transformation of the generally fragmented report of traumatic experiences into a coherent narrative. During the discussion of traumatic experiences, the therapist asks for current emotional, physiological, cognitive and behavioral reactions and probes for respective observations. The participant is encouraged to relive these emotions while reporting the events. The discussion of a traumatic event is not terminated until a habituation of the emotional reactions presented and reported by the patient takes place. In the last session, the participant receives a written report of his biography. A manual of NET is available from the authors.

Supportive Counseling (SC). Looking at the variety of mainly non-standardized approaches currently applied in the field, we could not aim at a SC procedure that was representative for the current standard. We applied a procedure that was open depending on the participants’ specific needs. The main goal of SC was to explore and strengthen the participants’ individual, social and cultural resources. The focus of the treatment was on current interpersonal problems, personal decisions, as well as plans and hopes for the future. To keep a minimum of standardization, a problem solving approach, specifically developed for trauma counseling in Africa (AMANI, 1997), was generally used. To contrast SC from NET, the basic restriction of SC was that talking about specific past traumatic events had to be avoided, a goal that was achieved easily given the severity of the symptoms and the high level of avoidance in these patients. This restriction was especially useful as some representatives of psychosocial organizations reject talking about traumatic events within short-term treatment approaches as they feared “retraumatization”.

7.3.1 Results

Participation

As a consequence of the randomization procedure 17 participants were offered NET, 14 SC and 12 PE. In the NET group one patient refused to participate, all other patients agreed. I In the SC group two patients did not complete the full treatment. The rea- 127 son for not participating or terminating treatment was generally due to a lack of time given the need to cultivate their fields at the beginning of the rainy season.

Symptom Scores

Table 7.2 presents the results of the symptom and health scales for the time points pre treatment, post treatment, 4-month follow-up and 1-year follow-up. All participants who were randomized to treatment, including the one refuser in the NET group and the two dropouts in the SC group, were included in the analyses.

As the randomization procedure resulted in different baseline levels for some measures, repeated measure ANOVAs were calculated with time as a four level within-subject factor and the treatment group as three level between-subject factor for each outcome measure. Significant interactions between time and treatment were considered as the relevant indicators of a different development of the treatment groups over time. Mauchly’s tests were calculated to test for the sphericity assumption of the univariate ANOVA. The sphericity assumption was met for SRQ, SF-12, and CIDI, but not for the PDS. As the multivariate approach’s assumption of a homogenous variance-covariance matrix was met for the PDS (verified with Box’s M test), a multivariate ANOVA was calculated for this measure instead of the univariate approach. To make use of a maximum of information in this study with a small sample size, missing data was estimated using a restricted maximum likelihood procedure. To control for a possible confounding effect of further traumatic life events after treatment, the number of traumatic events reported for the 1-year follow-up period was entered as covariate.

Differences between the treatment groups were examined with two contrast analyses for each outcome measure. In each of these analyses, we compared NET with each other treatment condition with regard to the changes in the mean of the individual outcome variable between the pre-test and the 1-year follow-up. As two calculations were carried out for each measure, significance level was Bonferroni corrected to a = 0.025 for the contrast analyses.

A significant Time x Treatment interaction was found for the PDS (Wilks’ L = 0.78), F(6, 54) = 4.3, p = 0.01, h2 = 0.31. The contrast analysis revealed that the NET group had a better improvement at 1-year follow up than the SC group, F(1, 106) = 5.65, p < 0.01, as well as the PE group, F(1, 106) = 14.0, p < 0.01. This pattern was confirmed by 128 the CIDI, with a significant Time x Treatment interaction, F(2, 34) = 4.87, p = 0.01, h2 = 0.21. In the contrast analysis, the NET group differed significantly from the SC group, F(1, 34) = 7.05, p = 0.01 as well as the PE group, F(1) = 7.03, p = 0.01. Contrary to our hypotheses, no significant Time x Treatment interactions were found for the SRQ, F(6, 106) = 0.2, h2 = 0.02, p > 0.90 and the SF-12, F(6, 106) = 1.8, h2 = 0.11, p = 0.10. Contrast analyses showed a significant difference between NET and SC for the SF-12, F(1, 106) = 7.2, p < 0.01, but not between NET and PE, F(1, 106) = 0.34, p = 0.54. For the SRQ, neither the comparison between NET and SC nor the comparison between NET and PE revealed a significant difference.

Clinical Significance

Effect sizes (Table 7.2) at post-test were calculated as the change of the mean between pre- and post test divided by the pooled standard deviation of the outcome variable at pre- and post-test. Effect sizes at 1-year follow-up resulted from the analogous comparison between the means at pre-test and 1-year follow-up.

The percentage of patients with a diagnosis of PTSD for those patients who participated in the 1-year follow-up was compared for the three treatment groups. Table 7.3 presents the results of this analysis, indicating that only in the NET group the majority of participants did not present with PTSD one year after treatment. The differences between the treatment groups were significant, c2(2, N = 38) = 9.48, p < 0.01. Differences between single treatment groups were analyzed with Fisher’s Exact tests (Bonferroni corrected a = 0.025). They showed that significantly less subjects in the NET group presented with PTSD than in the SC group, p = 0.01 and in the PE group, p = 0.02. 129

Table 7.2. Means, standard deviations and effect sizes of the outcome variables by treatment groups

Pre Post 4-month 1-year ES ES Pre - post Pre - 1-year follow-up n

NET 17 15 15 14

SC 14 13 13 13

PE 12 12 12 11

Posttraumatic Stress Diagnostic Scale

NET 25.2 (7.44) 19.1 (11.7) 24.5 (7.77) 16.0 (5.10) 0.6 1.6

SC 22.0 (8.0) 19.8 (10.9) 22.8 (10.1) 23.1 (7.7) 0.2 -0.1

PE 19.5 (8.0) 21.2 (9.4) 27.7 (6.6) 23.9 (7.0) -0.5 -0.9

Composite International Diagnostic Interview – PTSD part

NET 13.43 (2.10) 8.85 (2.74) 1.9

SC 13.93 (2.34) 12.64 (3.15) 0.4

PE 14.16 (2.85) 13.40 (3.34) 0.3

Self-Reporting Questionnaire-20 Score

NET 15.6 (2.9) 13.1 (5.1) 11.9 (4.9) 11.0 (5.1) 0.6 1.1

SC 16.5 (2.7) 14.3 (5.0) 12.8 (3.9) 12.4 (4.8) 0.5 1.0

PE 18.6 (2.0) 15.3 (3.2) 15.1 (2.6) 14.4 (4.1) 1.2 1.3

Medical Outcome Study Short-Form 12 - Psychological Health subscale

NET 0.27 (0.12) 0.36 (0.19) 0.38 (0.12) 0.44 (0.19) -0.6 -1.1

SC 0.34 (0.11) 0.33 (0.21) 0.33 (0.14) 0.36 (0.14) 0.1 0.1

PE 0.23 (0.15) 0.33 (0.19) 0.37 (0.14) 0.35 (0.17) -0.7 -0.8

Note. NET = Narrative Exposure Therapy; SC = Supportive Counselling; PE = Psychoeducation; ES = effect size; Composite International Diagnostic Interview was not assessed at post-test and 3-month follow-up 130 Table 7.3 Number and percentage of patients with a diagnosis of posttraumatic stress disorder according to Composite International Diagnostic Interview at 1- year follow-up

Treatment n PTSD at 1-year follow-up n (%)

NET 14 4 (28.6)

SC 14 11 (78.6)

PE 10 8 (80.0)

Note. NET = Narrative Exposure Therapy; SC = Supportive Counselling; PE = Psychoeducation

The SRQ-20 allows the classification of a patient as likely to have severe mental health problems. The cut off point for this decision is a score of 10 or above. Table 7.4 shows the percentage of persons in each group who were classified as likely to have severe mental health problems at the four time points. According to this index, before treatment all subjects were classified as suffering from severe mental problems. The difference in the distribution of cases and non-cases in the treatment group does not achieve significance, Fisher’s Exact Test p = 0.08 at 1-year follow-up.

Table 7.4. Percentage of patients classified as having severe mental health problems according to SRQ-20 cut-off

Treatment pre post 4-month 1-year

NET 100% 73% 73% 50%

SC 100% 85% 85% 77%

PE 100% 92% 100% 91%

Note. NET = Narrative Exposure Therapy; SC = Supportive Counselling; PE = Psychoeducation

At the 1-year follow-up assessment we observed that a large proportion of the refugees had left the settlement. We set up the post-hoc hypothesis that the migration pattern depends on treatment, and that more refugees from the NET group had left the settlement. In the NET group, 62% had left the settlement, whereas this was the case in only 7% (one) of the SC group and 17% (two patients) of the PE group. The difference was significant, c2(2, N = 40) = 12.34, p < 0.01. Fisher’s Exact tests (Bonferroni corrected a = 0.025) 131 showed that from the NET group significantly more refugees left than from SC group (p < 0.01) and the PE group (p = 0.01).

The incidence of one or more further traumatic events in the 1-year follow-up period was 92.5%. The participants reported a mean of 3.55 (SD = 2.36) event types for the time between treatment and the examination. The mean number of traumatic event types differed significantly between those refugees who stayed (M = 4.00, SD = 2.40) and those who left (M = 2.50, SD = 1.97) the settlement, t(39) = 2.20; p = 0.03. Figure 7.1 presents summarized the results of the indicators of clinical significance at the 1-year follow-up.

50 NET SC 40 PE

0 Recovered from PTSD Recovered from SRQ "Mental Left Settlement Case" Figure 7.1. Percentage of recoveries in treatment groups according to different criteria at the 1-year follow-up

7.3.2 Discussion

We carried out a randomized controlled trial of three different short-term treatment methods for survivors of war experiences who suffered from PTSD in a settlement for Su- danese refugees in North-Uganda. All but one of the participants who presented with PTSD in the screening interview agreed to participate in treatment and only in the SC condition two patients prematurely terminated treatment. This confirms that even in the desperate living conditions of a refugee settlement, without monetary or food assistance, 132 people who suffer from PTSD seek psychosocial assistance. Although hope was one motivator, curiosity concerning the possibility to talk to white researchers also may have played a role in the high rate of acceptance. The fact that there was no dropout in the NET group is noteworthy, as other treatment studies have reported high dropout rates for exposure therapy for PTSD of up to 28% (Foa et al., 1991). One main difference between those studies and this one is the small number of treatment sessions for NET (4), whereas other studies typically ranged from 9 to 12 sessions and included additional homework tasks. Furthermore, a unique motivator of NET is the anticipation of receiving a written biography upon completion. Anecdotally we can report that participants stated that they were looking forward to the document even if they could not understand English text. Many patients hoped to pass on their biography to their children once they were living in peace and their children were educated. They wanted the next generation to be informed about their suffering during the war and the liberation fights. In addition, they encouraged us to publish the testimonies to educate the world about their situation in a war that has been forgotten by the developed world.

We observed different developments of the participants’ symptoms after treatment. The PDS scores showed a worsening of PTSD symptoms for all groups between the post- test and the 4-month follow-up, possibly reflecting an increase in the occupation with the traumatic past as food rations were cut down dramatically in this period and the refugees were under pressure to leave the settlement to places where they might have experienced traumatic events.

At 1-year follow-up, the NET group presented with a significantly better outcome on measures of posttraumatic stress than both other groups. This shows that the improvement of the NET participants can not be attributed to spontaneous recovery, as the PE participants who received a minimal intervention did not improve in the same way. Like- wise, unspecific treatment effects, that are quite probable in the confrontation with white therapists, can not fully account for the efficacy of NET as the SC group did not benefit to the same extent. Despite the specific improvement of the NET participants, the symptom scores still remained high but the change was clinically significant as 71% of the NET patients no longer fulfilled the DSM-IV criteria for PTSD, a significantly larger proportion than in the two other groups. 133 The size of the treatment effect on posttraumatic symptoms at post-test (0.6) was lower than the effect sizes reported in other treatment studies using exposure techniques with traumatized populations in industrialized countries (e.g. Foa et al, 1999; Tarrier et al, 1999). However, when interpreting this value the fact that the PE group presented with a negative effect of -0.5 at post test must be accounted for. This result indicated a worsening of those who received minimal intervention during same time period. At 1-year follow-up, the NET effect sizes (1.6 for PDS and 1.9 for CIDI) indicated a clinically significant change for PTSD symptoms that did not occur in the other groups. These results indicate that the well established knowledge about the efficacy of exposure techniques for the treatment of PTSD (Foa, 2000) can be transferred to refugee populations living in unsafe conditions.

All groups improved somewhat during the one-year period after therapy, according to indicators of anxiety, depression, and overall mental health. However, the majority of participants was still classified as severe mental health case in the SC (77%) and PE (91%) groups. The NET group showed more improvement than the others, yet the difference was not significant as 50% of this group were still classified as severely psychologically dis- turbed. In addition, effect sizes did not indicate a superiority of the NET group on these measures. This outcome possibly reflects the limitations of short-term psychosocial interventions for traumatized refugees who continue to remain in desperate living conditions. Factors such as insecurity, poverty, malnutrition, lack of medical assistance and the hope- less political situation of the Sudanese refugees may restrict the impact of any psychosocial intervention.

Nevertheless, observation of the refugees’ migration that occurred after treatment indicates that the NET treatment may have had an astonishing effect on decisions to leave the settlement. Approximately two months after the last treatments were finished a dra- matic reduction of the food rations provided by UNHCR in Imvepi was implemented resulting in the departure of many refugees from the camp. Refugees who left the settlement often moved to more safe and fertile places in self-settlements closer to the Sudanese border, moved back to Sudan, or sought jobs in the surrounding region. The majority of the participants in the NET group managed to leave the settlement within one year following treatment, a tendency that did not occur for refugees in the other treatment groups. The comparison of the number of traumatic event types reported by the refugees for the one year period after treatment indicated that although almost all participants (93%) reported 134 the witnessing or experiencing of one or more further traumatic events, those who left the settlement did indeed move to safer places and reported a significantly lower number of event types. A reduction of avoidance symptoms caused by NET might have empowered these refugees to move to areas that were closer to the war regions and to abandon the in- sufficient assistance and putative security provided in Imvepi. This post-hoc finding may indicate a relationship between posttraumatic stress, treatment and migration patterns that should be examined with specific hypotheses in further studies.

To our knowledge, this is the first randomized controlled trial of treatment for traumatized survivors of war living in a developing country. The study shows that refugees who suffer from PTSD seek and accept psychosocial treatment even when living in difficult living conditions. NET was superior to counseling and psychoeducation with regard to significant long-term changes in PTSD symptoms and behavior. These changes even had direct effects on the life of the refugees. Unfortunately, comorbid symptoms were not suf- ficiently improved in the majority of patients as a result of treatment.

The lack of a long-term effect on both PTSD and comorbid anxiety and depression for the SC seriously questions the current practice of many psychosocial organizations to apply untested counseling approaches as a standard method for traumatized refugees. How- ever, the type of supportive counseling applied in this study might not be representative of counseling practices in other organizations, as talking about the traumatic event was avoided in our SC condition. More research about the efficacy of the different types of counseling is needed to specify effective mechanisms of this approach and to determine the conditions and duration of treatment that is necessary to achieve positive results.

Several limitations of the study indicate urgent need for further studies in this field. First of all, the treatment was carried out by well trained European psychologists. The procedure is costly, and thus cannot open the way to large-scale interventions for the size- able war-torn populations that exist. Future studies should examine possible interventions that can utilize local paramedical staff who can be trained to implement trauma focused treatments for PTSD. The pragmatic procedure of NET should be feasible for local staff without any psychosocial background.

Furthermore, the issue of the high prevalence of comorbid symptoms that remained for all groups following treatment must be investigated. It is unclear whether normal levels 135 of anxiety and depression can be achieved by any psychosocial treatment as long as the refugees remain to live in unsafe conditions. Further studies should examine if longer treatment approaches that extend the duration of exposure would be more effective. An- other possibility would be to follow NET treatment with a short-term symptom-oriented treatment to improve symptoms of depression and anxiety.

The possible social and political implications of a narrative approach such as NET requires further examination. Finding ways to effectively reduce, as well as prevent, the psychological suffering of victims of war is an important challenge for scientists and aid organizations and indeed remains an ethical obligation. Bringing forth the narratives of victims may not only benefit the survivors but may also increase public awareness of severe human rights violations in dictatorships and war areas, as well as the public’s ability and responsibility to make change on a global and local level. We maintain that the tradition of joint psychological and political approaches may offer opportunities to remedy mental suffering and to provide assistance to those in need in war torn populations. 136

8 Conclusion

Currently more than 30 nations, nearly all of them in developing countries, are plagued by civil war. The conflict parties mainly fight for tangible resources rather than opposing ideologies or the interests of hostile ethnicities or religions. As civilians are the main target of current warfare, violent expulsions and atrocities in rural villages have re- placed the battlefields as the major theatres of war. Civil wars have devastating effects on the affected countries as well as the surrounding nations, and are one of the main handicaps for development. The immediate as well as the long term consequences include the destruction of the economy, the social communities and the health care system. All this loss is in vain, as rebellions almost never cause any political improvements. Analyzing the costs of civil war, Collier and collegues (2003) conclude that civil war is “development in reverse” (p. 13). In a quantitative analysis of the health consequences of civil wars, Ghobarah, Huth, and Russett (2003) point out that “the direct and immediate casualties from civil wars are only the tip of the iceberg of their longer-term consequences for human misery” (p. 1).

A typical representative of current civil wars is the Sudanese war. At first glance this conflict looks like a religiously motivated secessionist war. A close look shows that the major reason for the insisting refusal of the conflict parties to turn down this conflict lies in the presumed oil fields in the south. In addition, both parties, the Arabic dominated northern government as well as the Animist and Christian southern rebels have been supported by foreign nations who fight a proxy war on Sudanese ground. The effects of this war are disastrous especially for the southern Sudanese civilian population. The majority of them are in flight, either within Sudan or in the neighboring countries. The southern Sudanese agricultural economy and the health care system has repeatedly broken down completely, leading to several severe famines and epidemics during the last years.

In our study of the survivors of the Sudanese war we could also find the devastating effect of this war. In comparison the West-Nile Ugandans who are culturally and ethnically closely related to the West-Nile Sudanese but went through a peaceful period during the last decade, the Sudanese are economically much worse off and have much fewer possessions. The situation is worst for the refugees who fled to Uganda, as they are clearly suffering from malnutrition. The data shows that, whereas most of the Sudanese who stayed in Sudan work as farmers, only a quarter of the refugees has the field and means to produce 137 their own food. Every fourth Sudanese refugee has no occupation at all and thus fully depends on food aid, which seems to be not enough to prevent hunger and malnutrition.

Surprisingly, the refugees are worst of not only in terms of nutrition, but also in terms of security. Compared to both, the Ugandans and the Sudanese who remained in the war-torn Sudan reported by far the greatest exposure to violent events not only for their whole life but also for their past year in exile. The inadequate security situation of the refugees in the Ugandan West-Nile was caused by the activities of rebel armies who deliberately targeted the Sudanese refugee communities. Many of those incidents have been reported in local newspapers, but the failure of UNHCR and the Ugandan government to protect the Sudanese refugees was almost completely unnoticed outside Uganda. This observation indicates that refugees might be an easy target for organized violence in many places of the world, beyond the sight of the international community. The Sudanese refugees are probably not the only refugee community that needs more attention from the outside world and more reliable protection.

Traumatic war events affect the psychological functioning of the victims. We could show that in both groups, the Sudanese refugees as well as the Sudanese nationals, almost every second civilian suffers from PTSD. This prevalence rate sounds incredibly high, but a close look at the severity and frequency of traumatic events reported by the Sudanese makes these figures understandable. We could confirm that the extent of exposure to different traumatic events is the main determinant of PTSD, whereas factors of poverty and education play only a minor role. Vulnerability and protection factors for PTSD, like personality, genetic factors, social support, etc. that have often been suggested and sometimes empirically found in studies in industrialized countries, seem to lose their importance as the exposure to traumatic events increases. We could find a linear relationship between the number of different traumatic event types reported and the probability of PTSD, with a 100% probability at a certain threshold. As this effect has not been found in studies in industrialized countries, it shows that the determinants of PTSD probably depend on the mean and variance of traumatic exposure that is present in the population. In this respect, the currently peaceful industrialized countries are not at all representative of the worldwide conditions.

Even if these figures might be a slight overestimation of the prevalence of PTSD, it indicates that war leads to a major mental health problem in these communities, as we 138 could show that PTSD is a valid concept for this African population that causes a significant reduction in individual functioning. The promising result of this study is that PTSD can be effectively treated with interventions that are based on scientific knowledge. Espe- cially Narrative Exposure Therapy is a psychotherapeutic tool that can fit the requirements of war-torn populations, as it is short, pragmatic and effective. We could show that sound research about treatment is possible and necessary, as not all psychosocial interventions are equally effective. The current lack of research in this field is in contrast with the considerable resources that are currently spent on psychosocial interventions in war areas. Shifting efforts from psychosocial ad-hoc intervention to research studying the long-term benefits of different programs should be the main focus of aid organizations and donors.

In this context, several findings of our study point at interesting questions that should be examined further. First of all, all psychosocial programs must be effective for single cases, but must also be feasible in the light of the high need for treatment found in communities that are affected by war. One solution could be to concentrate on group treatment procedures that can deal with many victims at once. This strategy is applied by many psychosocial organizations but awaits the prove of its efficacy. Another strategy to meet the high prevalence of PTSD is to train local staff from the communities themselves to treat their fellows in misery. NET could be a suitable procedure for this approach as the pragmatic procedure should be easy to learn even for local staff without any medical or psychological background. The finding that NET had a surprising impact on migration as those who received the treatment managed to leave the settlement indicates that effects on behavioral pattern and functioning might even be more pronounced than expected. It will be interesting to see whether this effect can be replicated in following studies.

Some limitations of our study can indicate valuable suggestions for further research. Psychological research crucially depends on the availability of sound instruments to measure symptoms and functioning. The verification of the translations of our instruments showed that the reliability and validity of the questionnaires was not fully satisfying. Re- search in communities with different languages should always start with the thorough development and translation of clinical questionnaires. We had to learn that a translation that meets international standards does not automatically produce highly reliable and valid instruments. An extensive training of local interviewers in the contents of the concepts could improve the quality of the assessments. Studying the validity and reliability of instruments 139 and the assessment procedure is an unavoidable prerequisite of all following clinical research. Unfortunately, even recent epidemiological studies in populations affected by war did not meet this requirement.

Two major conclusions can be drawn from this study. Firstly, refugees need more attention and protection to guard them from violence. Secondly, there is an urgent need to develop effective psychosocial interventions to remedy the impaired mental health of populations affected by war. But even if the refugees were to be perfectly safeguarded, assisted and treated, the outcome would still be unsatisfying, as this assistance remains restricted on handling the symptoms of wars. As long as the worldwide civil wars that cause the suffering can not be stopped and the outbreak of new wars can not be prevented, there is little hope for a sustainable development of large parts of the developing world.

Several authors have proposed ways for the international community to support the prevention and termination of civil wars. Following empirical research on the causes of civil wars, the World-Bank economist Collier (Collier, 2003; Collier et al., 2003) made several practical suggestions. Beyond the well known advices to support economic growth and the development of democracies, his focus is on ways to cut down the opportunities to finance rebel organizations. According to his analysis, civil wars are mainly financed through commodity trade (diamonds, minerals, timber, oil). Curbing rebel access to these markets through means of certification and observation could help to cut down their fi- nancial resources. The Kimberley process seems to be a positive example as international controls of the diamond trade have considerably increased the difficulties to sell conflict diamonds. This process has apparently contributed to the breakdown of two of the most prominent African rebel armies, the UNITA in Angola and the RUF in Sierra Leone. An- other way in which conflict parties finance their military activities include selling the future exploitation rights of natural resources that are not yet under their control to multinational corporations. Other means include the extortion of foreign companies and governments through kidnapping or sabotage, direct support from foreign governments, corrup- tion in the trade of natural resources and the collection of money from diaspora communities. Most of these and other ways to finance a civil war involve international trade and money transfer and can thus be influenced by the international community. Some single examples show that a joint effort of civil rights organizations, multinational corporations, governments and international organizations can achieve some transparency and control of 140 resource trade. But still the world economy is far of a general public control of the commodity trade with conflict regions.

Observers of civil wars agree that the international community misses a lot of opportunities to terminate and prevent these conflicts. In the last decades, international interventions were restricted to several efforts to end those wars that happened to gain international media attention through military, economic or diplomatic activities. The effect of these interventions was generally disappointing. Most civil wars never received any media attention in industrialized countries, but remain in the shade of those rare conflicts with international involvement. The public has habituated to reports of high numbers of casualties in conflict regions, and there are usually no TV pictures to show the atrocities. False beliefs about the nature and causes of these wars got in the way of the understanding of the responsibility of the international community.

Changing the way the international community deals with conflicts requires a better knowledge about the wars on the side of the policy-makers as well as well as the public. Research in conflict areas might be one possibility to increase the understanding of the causes and consequences of wars and to throw light on the suffering of the victims. The detailed documentation of the experiences of civil war victims can be important to supple- ment statistical figures about the dimensions of suffering with understandable reports about typical single fates. Joint psychological and human rights approaches, like NET, might be powerful make this contribution to a complex strategy in the fight against conflicts. 141

9 Summary

The consequences of civil wars are devastating for the affected countries. Wars im- pair the functioning of nations on social, economic, medical and political levels. We carried out an epidemiological study in the West-Nile area of Sudan and Uganda to study the psychological consequences of the Sudanese civil war. We compared three populations. Suda- nese, who remained in Sudan (n =664), Sudanese refugees who had fled to Uganda (n = 1240) and finally Ugandan nationals in the West Nile are (n =1419) who have been able to live in peace for more than 10 years now. The study focused on posttraumatic stress disorder (PTSD) as a characteristic consequence of extremely stressful experiences. A considerable proportion of the Sudanese groups was diagnosed with PTSD in this survey (in Sudan: 48.7%, refugees: 47.7%), whereas PTSD prevalence was 19.7% among the Ugandans. The analysis of predictors pointed at the extent of traumatic exposure as the most relevant predictor of PTSD, the correlation between the number of different traumatic event types and the severity of PTSD symptoms was r = .49.

In a second step we offered a part of those refugees, who were diagnosed with PTSD, a psychotherapy. To meet the requirements for psychosocial assistance in the context of the refugee settlement (short duration, cultural sensitivity, sociopolitical implications), we developed Narrative Exposure Therapy (NET). The assumed mechanisms of NET were explained on the basis of psychological and neurobiological theories of PTSD. In a randomized controlled trial we compared NET (n = 17, four sessions individual treatment) with Supportive Counselling (SC, n = 14, same duration) and, as minimal intervention condition, Psychoeducation (PE, n = 12, one session). With regard to posttraumatic symptoms NET turned out to be the most successful treatment condition. One year after treatment only 28% of the participants of the NET condition were still diagnosed with PTSD, whereas PTSD was still present in 79% (SC) and 80% (PE) of the participants in the control conditions. While NET could cause a significant reduction of PTSD symptoms, no similar results could be obtained with regard to comorbid depression and psychological health. The studies show a considerable demand for research and qualified assistance in the field of mental health of populations that suffer from the consequences of war. 142 10 Zusammenfassung

Bürgerkriege haben für die betroffenen Länder verheerende Auswirkungen auf soz- ialer, wirtschaftlicher, medizinischer und politischer Ebene. In einer epidemiologischen Studie, die im West-Nil Gebiet von Sudan und Uganda durchgeführt wurde, sollten die psychischen Folgen des sudanesischen Bürgerkrieges untersucht werden. Dabei wurden drei Populationen miteinander verglichen: Sudanesen, die im Sudan verblieben waren (n =664), Flüchtlinge, die aus dem Sudan nach Uganda geflohen waren (n = 1240) sowie ugandische Einwohner des West-Nil Gebietes (n =1419), die nun seit mehr als 10 Jahren in weitgehend friedlichen Verhältnissen leben können. Das Augenmerk der Untersuchung lag auf der posttraumatischen Belastungsstörung (PTBS), als charakteristischer Folge extrem belastender Erlebnisse. Es stellte sich heraus, dass bei den sudanesischen Gruppen bei einem erheblichen Anteil der untersuchten Personen eine PTBS zu diagnostizieren war (im Su- dan: 48.7%, Flüchtlinge: 47.7%), wohingegen die Prävalenz bei den Ugandern 19.7% be- trug. Eine eingehende Analyse der Prädiktoren stellte das Ausmaß der Exposition an traumatische Erlebnisse als wichtigsten Prädiktor für die Entstehung einer PTBS heraus, es er- gab sich eine Korrelation zwischen der Anzahl verschiedener Typen traumatischer Er- fahrungen und dem Schweregrad posttraumatischer Symptome von r = .49.

In einem zweiten Schritt wurde einem Teil der Flüchtlinge, bei denen eine PTBS di- agnostiziert wurde, eine Psychotherapie angeboten. Um den Anforderungen einer Psycho- therapie unter den gegebenen Umständen gerecht zu werden (kurze Dauer, kulturelle Sen- sitivität, sozialpolitischer Anspruch), wurde die Narrative Expositionstherapie (NET) entwickelt. Die vermutete Wirkungsweise von NET wurde auf der Grundlage psycholo- gischer und neurobiologischer Theorien der PTBS erklärt. In einer randomisiert kontrol- lierten Therapiestudie wurde NET (n = 17, vier Sitzungen individuelle Therapie) verglichen mit unterstützender Beratung (UB, n = 14, gleicher Umfang) und, als Kontrollbedin- gung mit minimaler Intervention, Psychoedukation (PE, n = 12, eine Sitzung). In Bezug auf posttraumatische Symptomatik stellte sich die NET als erfolreichste Therapieform heraus, nach einem Jahr war von denjenigen Teilnehmern, die mit NET behandelt wurden, nur noch bei 28% eine PTBS zu diagnostizieren, während bei den Kontrollbedingungen noch bei 79% (UB) bzw. 80% (PE) eine PTBS vorlag. Während NET posttraumatische Symptome deutlich reduzierte, konnten keine vergleichbaren Effekte auf komobide Symp- tome (Depression und psychische Gesundheit) nachgewiesen werden. Die Studien zeigen 143 einen erheblichen Bedarf von Forschung und qualifizierter Unterstützung im Bereich der seelischen Gesundheit von Populationen, die unter den Auswirkungen von Kriegen leiden. 144

11 Appendix

Example of one testimony documented with NET in the treatment study

This testimony documents the attack of the Ikafe refugee settlement in Northern Uganda. Many of the refugees in Imvepi had been affected by this attack. It is still unclear whether the Ugandan West Nile Bank Front or the Sudanese Peoples Liberation Front was responsible for this attack. This testimony illustrates the inadequate security situation of the refugees and the unreliable protection from UNHCR from the perspective of one refugee. Names and places were withheld to ensure anonymity.

Imvepi, August 2000

I was born in ##### in #####. I grew there. I have four brothers and five sisters. My father worked in a hospital as a medical assistant.

In 1980 he was transferred to ##### in the southern part of Sudan. We stayed well until the war started in 1983. By the time we had no chance to go back any more.

The war came to our town in a night. We heard gunshots and so we knew that there was a war. In the town were soldiers of the government. The town was assaulted by the rebels. There was a crossfire.

We were at home in this night. We were sleeping. I stayed with my brother in one room. He was sleeping in a bed next to my bed. I woke up by the cries of my brother. There were many bullets coming into our house. I was very frightened. I lay down on the ground. I felt the body of my brother and felt the blood in my hand. I knew he was injured. A short time later he stopped crying, so I knew that he was dead.

We left the body of my brother in the room and ran to #####. After us came people from ##### who told us that my brother was dead. They told us that there was a dead body in our house. 145 We stayed for about two years in #####. The rebels attacked the town all the time but they did not enter it.

In 1989 the rebels came to town and burned houses and killed many people. By that time the government soldiers recruited the young boys for training. By that time they wanted to pick me but I refused. So they thought that I might join the rebels. About 12 soldiers came to our house and tied my hands with ropes. The ropes made cracks, you can still see the scars. My father told them, that I was the only son that left at home. Two of my brothers were gone to the countryside. So they left me at home. They knew that my father was an important person in the hospital.

I was studying in secondary school. I wanted to become a technical engineer. To brothers stayed with another stepfather in the countryside. So I stayed with my father, my mother and my sister.

In 1991 the guerillas came to attack #####. They came to attack the government garrison. At this time my father was working in the hospital. In this night, the people from the hospital were running away from there. But my father was not among them.

We had to run away from #####. There was still fighting going on. We were running with the bare foot. We were hiding in the bush for about five days. A soldier from #####, who was a relative from us, told us that my father was shot in the hospital.

We stayed in the ##### area.

In 1993 the government bombed the area with a gunboat and with Antonows. The rebels started recruiting boys and young man to become soldiers. When they went to my home, I was at the market. My mother told me so when I came back. So I decided to run away from Sudan to Uganda. It was painful for me to leave my family behind. I ran to Morobo, Yei, Kaia and then to Uganda. At first I came to the Koboko transit camp. We stayed there for one year. I was very much afraid of the war and I worried about my family.

In 1995 we were taken to Ikafe for a settlement.

From 1995 till 1997 there was an Ugandan guerilla active. They began torturing refugees, looting their properties and sending the people away from settlements.

At this time the rebels captured 6-7 people. Among them was Ref. Paul with his wife, Ref. Jamba with his wife and the village leader Marlise. They captured them in 146 evening hours. They released them at midnight. Those people were tortured for about fife hours. They told them: „Tell your friends that we will be coming!“

In this time, the doughter of my neighbour #####, who was a close friend of mine, was also captured by the rebels. She was 12 years old. She was on the field together with her stepmother.

We told the UNHCR and the government officers that there was danger around. They did not believe us. When we went to the camp commander, he went out to see if it is true. He was shot on his way from Bidi-Bidi (base camp) to central Ikafe.

But the government did not want us to leave the place. Instead, the UNHCR staff ran away and left us alone there.

People moved from the settlement to the base camp (Bidi-Bidi). Inside the fences of the base camp were government troops to protect us. We stayed there for about two months. The camp was very crowded, as 47000 refugees were living in Ikafe.

Sometimes people went from there to get some things from their homes in the settlement. They came back and were beaten and tortured. So we knew that the guerilla was still around.

In one night, it was about three o’clock, we heard gunshots. The rebels were at- tacking the base camp. Among the 160 rebels there were some Ugandans who just came to loot the refugees. There was a fight between the government soldiers and the rebels. One group of rebels attacked the base camp. The others went around and terrorized the refugees.

So we ran into the bush to hide from the bullets and the rebels. We were hiding there for about three hours. In this time the rebels were taking bicycles and looting houses. All of us were afraid in this time.

About 300 of us were hiding not far away from the fence. Others went to the valley. I was told by a government officer to warn the refugees in the valley that the rebels would come into their direction. So I went there to warn these people.

After the fight we went back to the places where we settled. I was one people who came back first. In one tent I found the bodies of three refugees. One grandmother and two of their grandchildren were killed. The grandmother was shot in her head. Her brain was distributed in the tent. Her blood covered the ground. One child was shot in 147 the hip, the legs were torn off from the body. The other child was cut in his neck with a long knife. There was blood all over the place.

In another tent I found a pregnant woman who was shot into her belly. On the ground there was another dead body, it was a close friend of mine, a teacher. I was very sad. I’m left myself as a lost man in this world.

There was a shelter where many refugees put their properties and bicycles. The rebels burned this shelter and the things. It was still burning so we approached it with a stick. We found a child that was burned. It was mere bones. It looked like it had been roasted.

In the group that hided in the valley there were also four people killed. One village leader and three more people. There was one 12 years old boy who was killed.

We called the refugee leaders together and we wrote a memo to the UNHCR. There was a crowd of people, all of them were only walking up and down and were very afraid. We were fearing that the rebels might come again.

The government told us that we must not go anywhere but just stay where we are.

They called the UNHCR staff to find out about this incident. After one full day, the UNHCR came. It was the field assistant and the protection officer. We told them what happened and what we witnessed with our own eye. They said that they have no place to take us. They said that we had to stay there.

After two days a woman from the National came and she said that she loves us. She did not want any more bad things to happen to us.

We heard from the guerillas that they wanted to come back. We told the government: „Please let us go away!“

At this day five girls were captured by the guerilla. At this moment we decided to move from this area. We came to a town called Yumbe. We told the UNHCR staff that we were there in Yumbe. They said that they would not allow us to stay there. The wanted us to go back to Bidi-Bidi. But the commander of the soldiers said that we can not come back.

We stayed there a full month without food. UNHCR refused to give us anything. We just eat premature Mangos. By that time, UNHCR staff from Geneva came to visit us. 148 They asked us about our lives. We told them all that has happened. They said that we must be given food in the place where we were.

But the protection officer refused to give us food there. He told us we should go and get food in Bidi-Bidi, about 15 km away. We should walk there on bare feet.

In this time many people were beaten on their way to Bidi-Bidi to get food. Two boys were captured and one was killed. All people feared to go to get their food. So we just stayed without food.

Another group of researchers came. They told the UNHCR about our problems. The LC 3 of the town said to UNHCR and government: „If you don’t give food to these people, you must transfer them to another area.“

So they agreed to give us the food in that place. They thought about where to bring us. They offered us two places: Rhino-Camp and Imvepi. We went to Imvepi.

We came to the Imvepi transit camp at point J on May the 30th. My neighbor there was the chairman. He told me that a day before we arrived, the rebels came to Imvepi and cut off the ears of four boys. So we were frightened to be at this place from the very beginning.

On the next day we were in the transit camp. At about two o’clock two women were running naked towards us. They told the chairman that they had been raped by rebels. They were immediately taken to the hospital. The government soldiers went to the place to look for rebels.

Later we were told that the rebels drilled the mouths of some people and closed it with a lock. So Imvepi was no safe place at all.

After one month in the transit camp we settled in Imvepi. Since that we are here, just living. Three people were taken away by the guerilla in between. We still do not know anything about them.

Points H, I, E and F were displaced as they were not secure. So they were settled in B extension and D extension. We are here to stay but life is not OK.

Last year I heard that the daughter of my neighbor who was captured in Ikafe was still alive and stayed in #####. I told this to the settlement commander, but he said that it was too dangerous for him to go there. The girl was still living with the guerillas. So I went there myself with my brothers ##### and #####. A friend from ##### made 149 a plan for me of the place where these rebels lived. We entered this house. There were five boys and two girls in this house. When the friend’s daughter recognized us she was immediately crying. I asked the boys who took away the girl. Her husband said that he did not take her away. He went to call her father. I knew that her father was shot so I thought he would get a gun. So we just took the girl and ran as fast as we could. We ran about 12 kilometers and stayed than at the garnison in Oanga. From there we ran on to Rhino Camp at the next day. The girl is now living here in Imvepi together with her stepmother.

I was very frightened that the rebels might come back to take the daughter away and to take revenge until I finally heard that the husband was shot by a government soldier.

We have too much problems, there is no rain in this land. Many of us want to leave. There are many illnesses and not enough food. The food the UNHCR gives to us is finished in only one day. Many of us had good jobs in Sudan. But we not live here. 150

12 References

Agger, I., & Jensen, S. B. (1990). Testimony as ritual and evidence in psychotherpy for political refugees. Journal of Traumatic Stress, 3, 115-130.

Allden, K., Poole, C., Chantavanich, S., Ohmar, K., Aung, N. N., & Mollica, R. F. (1996). Burmese political dissidents in Thailand: trauma and survival among young adults in exile. American Journal of Public Health, 86, 1561-1569.

AMANI. (1997). A trauma counselling handbook. Harare, Zimbabwe: Amani Trust.

Amnesty International. (2002). Jahresbericht 2002. Frankfurt am Main: Fischer Taschen- buch Verlag.

Amnesty International. (2003). Definitions of torture. Retrieved 2003, January 7, 2003, from http://www.amnesty.org.uk/torture/definition.shtml

Avina, C., & O'Donohue, W. (2002). Sexual harassment and PTSD: is sexual harassment diagnosable trauma? Journal of Traumatic Stress, 15, 69-75.

Ayers, S., & Pickering, A. D. (2001). Do women get posttraumatic stress disorder as a result of childbirth? A prospective study of incidence. Birth, 28, 111-118.

Baker, R. (1992). Psychosocial consequences for tortured refugees seeking asylum and refugee status in Europe. In M. Basoglu (Ed.), Torture and its consequences. Cambridge: Cambridge University Press.

Basoglu, M. (1993). Prevention of torture and care of survivors. An integrated approach. Jama, 270, 606-611.

Basoglu, M., Mineka, S., Paker, M., Aker, T., Livanou, M., & Gök, S. (1997). Psychological preparedness for trauma as a protective factor in survivors of torture. Psychological Medicine, 27, 1421-1433.

Basoglu, M., Paker, M., Özmen, E., Tasdemir, Ö., Dogan, S., ceyhanli, A., et al. (1996). Appraisal of self, social environment, and state authority as a possible mediator of posttraumatic stress disorder in tortured political activists. Journal of Abnormal Psy- chology, 105, 232-236. 151 Basoglu, M., Paker, M., Ozmen, E., Tasdemir, O., & Sahin, D. (1994). Factors related to long-term traumatic stress responses in survivors of torture in Turkey. Jama, 272, 357-363.

Basoglu, M., Paker, M., Paker, O., Ozmen, E., Marks, I., Incesu, C., et al. (1994). Psycho- logical effects of torture: a comparison of tortured with nontortured political activists in Turkey. Am J Psychiatry, 151, 76-81.

Basoglu, M., Paker, M., Paker, Ö., Özmen, E., Marks, I., Incesu, C., et al. (1994). Psycho- logical effects of torture: A comparison of tortured with nontortured political activists in turkey. American Journal of Psychiatry, 151, 76-81.

Beck, A. T., Rush, A. J., Shraw, B. F., & Emery, G. (1979). Cognitive therapy of depression. New York: Guilford Press.

Bettelheim, B. (1986). Surviving the Holocaust. Fontana: Flamingo.

Blake, D. D., Weathers, F. W., Nagy, L. M., Kaloupek, D. G., Gusman, F. D., Charney, D. S., et al. (1995). The development of a Clinician-Administered PTSD Scale. J Trauma Stress, 8, 75-90.

Bower, P., Rowland, N., Mellor, C., Heywood, P., Godfrey, C., & Hardy, R. (2002). Ef- fectiveness and cost effectiveness of counselling in primary care. Cochrane Database Syst Rev, CD001025.

Bowman, M. L. (1999). Individual differences in posttraumatic distress: problems with the DSM-IV model. Can J Psychiatry, 44, 21-33.

Bracken, P. J., Giller, J. E., & Summerfield, D. (1995). Psychological responses to war and atrocitiy: the limitation of current concepts. Social Science and Medicine, 40, 1073- 1082.

Bremner, J. D. (2002). Neuroimaging studies in post-traumatic stress disorder. Curr Psychia- try Rep, 4, 254-263.

Bremner, J. D., Randall, P., Vermetten, E., Staib, L., Bronen, R. A., Mazure, C., et al. (1997). Magnetic resonance imaging-based measurement of hippocampal volume in posttraumatic stress disorder related to childhood physical and sexual abuse--a preliminary report. Biological Psychiatry, 41, 23-32. 152 Breslau, N. (2001). Outcomes of posttraumatic stress disorder. J Clin Psychiatry, 62, 55-59.

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 De- troit Area Survey of Trauma. Arch Gen Psychiatry, 55, 626-632.

Brewin, C. R. (2001). A cognitive neuroscience account of posttraumatic stress disorder and its treatment. Behav Res Ther, 39, 373-393.

Brewin, C. R., Andrews, B., & Valentine, J. D. (2000). Meta-analysis of risk factors for posttraumatic stress disorder in trauma-exposed adults. J Consult Clin Psychol, 68, 748-766.

Brewin, C. R., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of posttraumatic stress disorder. Psychological Review, 103, 670-686.

Brewin, C. R., Watson, M., McCarthy, S., Hyman, P., & Dayson, D. (1998). Intrusive memories and depression in cancer patients. Behav Res Ther, 36, 1131-1142.

British Medical Association. (1992). Medicine betrayed: the participation of doctors in human rights abuses. London: Zed Books.

Brom, D., Kleber, R. J., & Defares, P. B. (1989). Brief psychotherapy for posttraumatic stress disorders. J Consult Clin Psychol, 57, 607-612.

Brown, R., & Kulik, J. (1977). Flashbulb memories. Cognition, 5, 77-99

Bryant, R. A., Harvey, A. G., Dang, S. T., Sackville, T., & Basten, C. (1998). Treatment of acute stress disorder: a comparison of cognitive-behavioral therapy and supportive counseling. Journal of Consulting and Clinical Psychology, 66, 862-866.

Bryant, R. A., Sackville, T., Dang, S. T., Moulds, M., & Guthrie, R. (1999). Treating acute stress disorder: an evaluation of cognitive behavior therapy and supportive counseling techniques. Am J Psychiatry, 156, 1780-1786.

Burgess, A. W., & Holmstrom, L. L. (1974). Rape trauma syndrome. Am J Psychiatry, 131, 981-986.

Byrne, C. A., & Riggs, D. S. (1996). The cycle of trauma; relationship aggression in male Vietnam veterans with symptoms of posttraumatic stress disorder. Violence Vict, 11, 213-225. 153 Carlson, E. B., & Rosser-Hogan, R. (1991). Trauma experiences, posttraumatic stress, dissociation, and depression in Cambodian refugees. Am J Psychiatry, 148, 1548-1551.

Castles, S., & Miller, M. J. (1993). The Age of Migration: International Population Movements in the Modern World. New York: The Guilford Press.

Chilcoat, H. D., & Breslau, N. (1998). Posttraumatic stress disorder and drug disorders: testing causal pathways. Arch Gen Psychiatry, 55, 913-917.

Cienfuegos, J., & Monelli, C. (1983). The testimony of political repression as a therapeutic instrument. American Journal of Orthopsychiatry, 53, 43-51.

Collier, P. (2003). The market for civil war. Foreign Policy, May/June, 38-48

Collier, P., Elliot, L., Hegre, H., Hoeffler, A., Reynal-Querol, M., & Sambanis, N. (2003). Breaking the conflict trap - civil war and development policy. Washington: World Bank.

Conway, M. A. (2001). Sensory-perceptual episodic memory and its context: autobiographical memory. Philos Trans R Soc Lond B Biol Sci, 356, 1375-1384.

Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychol Rev, 107, 261-288.

Cunningham, M., & Cunningham, J. D. (1997). Patterns of symptomatology and patterns of torture and trauma experiences in resettled refugees. Aust N Z J Psychiatry, 31, 555-565.

Davidson, P. R., & Parker, K. C. (2001). Eye movement desensitization and reprocessing (EMDR): a meta-analysis. J Consult Clin Psychol, 69, 305-316. de Jong, J. T., Komproe, I. H., Van Ommeren, M., El Masri, M., Araya, M., Khaled, N., et al. (2001). Lifetime events and posttraumatic stress disorder in 4 postconflict settings. Jama, 286, 555-562. de Jong, K., Mulhern, M., Ford, N., van der Kam, S., & Kleber, R. (2000). The trauma of war in Sierra Leone. Lancet, 355, 2067-2068.

Derriennic, J. P. (1971). Theory and ideologies of violence. Journal of Peace Research, 8, 361- 374. 154 Devilly, G. J., & Foa, E. B. (2001). The investigation of exposure and cognitive therapy: comment on Tarrier et al (1999). J Consult Clin Psychol, 69, 114-116.

Doherty, G. W. (1999). Cross-cultural counseling in disaster settings. Australasian Journal of Disaster and Trauma Studies, 2.

Doná, G., & Berry, J. W. (1999). Refugee Acculturation and Re-Acculturation. In A. Ager (Ed.), Refugees: Perspectives on the Experience of Forced Migration (pp. 169-195). Lon- don & New York: Pinter.

Ehlers, A., & Clark, D. M. (2000). A cognitive model of posttraumatic stress disorder. Be- hav Res Ther, 38, 319-345.

Everly, G. S., Jr., & Mitchell, J. T. (2000). The debriefing "controversy" and crisis intervention: a review of lexical and substantive issues. Int J Emerg Ment Health, 2, 211- 225.

Fawzi, M. C., Pham, T., Lin, L., Nguyen, T. V., Ngo, D., Murphy, E., et al. (1997). The validity of posttraumatic stress disorder among Vietnamese refugees. J Trauma Stress, 10, 101-108.

First, M. B., Spitzer, R. L., Miriam, G., & Williams, J. B. W. (2001). Structured Clinical In- terview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition. (SCID- I/P). New York: Biometrics Research, New York State Psychiatric Institute.

Flanagan, A. Y. (1999). Three Ways of Translating Instruments in Cross-Cultural Research. E-Research Newsletter - Creating a Community Between Cross-Cultural Practitioners and Researchers, 1.

Foa, E. B. (1995). Post-traumatic Stress Diagnostic Scale (PDS). Minneapolis: National Com- puter Systems.

Foa, E. B. (2000). Psychosocial treatment of posttraumatic stress disorder. J Clin Psychiatry, 61, 43-48; discussion 49-51.

Foa, E. B., Dancu, C. V., Hembree, E. A., Jaycox, L. H., Meadows, E. A., & Street, G. P. (1999). A comparison of exposure therapy, stress inoculation training, and their combination for reducing posttraumatic stress disorder in female assault victims. J Consult Clin Psychol, 67, 194-200. 155 Foa, E. B., Hearst-Ikeda, D., & Perry, K. J. (1995). Evaluation of a brief cognitive- behavioral program for the prevention of chronic PTSD in recent assault victims. Journal of Consulting and Clinical Psychology, 63, 948-955.

Foa, E. B., & Kozak, M. J. (1986). Emotional processing of fear: Exposure to corrective information. Psychological Bulletin, 99, 20-35.

Foa, E. B., & Meadows, E. A. (1997). Psychosocial treatments for posttraumatic stress disorder. Annual Review of Psychology, 48, 449-480.

Foa, E. B., Molnar, C., & Cashman, L. (1995). Change in rape narratives during exposure therapy for posttraumatic stress disorder. Journal of Traumatic Stress, 4, 675-690.

Foa, E. B., Riggs, D. S., & Gershuny, B. S. (1995). Arousal, numbing, and intrusion: symptom structure of PTSD following assault. Am J Psychiatry, 152, 116-120.

Foa, E. B., & Rothbaum, B. O. (1998). Treating the trauma of rape: Cognitive-behavioral therapy for PTSD. New York: The Guilford Press.

Foa, E. B., Rothbaum, B. O., Riggs, D. S., & Murdock, T. B. (1991). Treatment of posttraumatic stress disorder in rape victims: a comparison between cognitive- behavioral procedures and counseling. J Consult Clin Psychol, 59, 715-723.

Forbes Martin, S. (1991). Refugee Women. London & New Jersey: Zed Books.

Frankl, V. (1946). Ein Psycholog erlebt das Konzentrationslager. Wien: Verlag für Jugend und Volk.

Freud, S., & Breuer, J. (1895). Studien über Hysterie. Leipzig: Franz Deuticke.

Friedman, M. J., & Schnurr, P. P. (1995). The relationship between trauma, post-traumatic stress disorder, and physical health. In M. J. Friedman, D. S. Charney & A. Y. Deutch (Eds.), Neurobiological and clinical consequences of stress: from normal adap- tion to PTSD. Philadelphia: Lippincott-Raven Publishers.

Frueh, B. C., Elhai, J. D., Gold, P. B., Monnier, J., Magruder, K. M., Keane, T. M., et al. (2003). Disability compensation seeking among veterans evaluated for posttraumatic stress disorder. Psychiatr Serv, 54, 84-91. 156 Galea, S., Ahern, J., Resnick, H., Kilpatrick, D., Bucuvalas, M., Gold, J., et al. (2002). Psy- chological sequelae of the September 11 terrorist attacks in New York City. N Engl J Med, 346, 982-987.

Galtung, J. (1969). Violence, peace, and peace research. Journal of Peace Research, 6, 167-191.

Gantzel, K. J. (2002). Neue Kriege? Neue Kämpfer? Hamburg: University of Hamburg.

Gersony, R. (1997). The Anguish of Northern Uganda: Results of a Field-Based Assessment of the Civil Conflicts in Northern Uganda (field-based assessment report). Kampala: United States Embassy & USAID.

Ghobarah, H., Huth, P., & Russett, B. (2003). Civil wars kill and maim people - long after the shooting stops. American Political Science Review, 97.

Gilbertson, M. W., Shenton, M. E., Ciszewski, A., Kasai, K., Lasko, N. B., Orr, S. P., et al. (2002). Smaller hippocampal volume predicts pathologic vulnerability to psychological trauma. Nat Neurosci, 5, 1242-1247.

Gleditsch, N. P., Wallensteen, P., Eriksson, M., Sollenberg, M., & Strand, H. (2001). Armed conflict 1946-2000: A new dataset. Oslo: International Peace Research Insti- tute.

Gleditsch, N. P., Wallensteen, P., Eriksson, M., Sollenberg, M., & Strand, H. (2002). Armed conflict 1946-2001: a new dataset. Journal of Peace Research, 39, 615-637.

Gorst-Unsworth, C., & Goldenberg, E. (1998). Psychological sequelae of torture and organised violence suffered by refugees from Iraq. British Journal of Psychiatry, 172, 90- 94.

Harding, T. W., de Arango, M. V., Baltazar, J., Climent, C. E., Ibrahim, H. H., Ladrido- Ignacio, L., et al. (1980). Mental disorders in primary health care: a study of their frequency and diagnosis in four developing countries. Psychol Med, 10, 231-241.

Hariotos-Fatouros, M. (1988). The official torturer: e learning model for obedience to the authority of violence. Journal of Applied Social Psychology, 18, 1107-1120.

Harrell-Bond, B. E. (1986). Imposing Aid: Emergency Assistance to Refugees. Oxford: Oxford University Press. 157 Harvey, A. G., & Bryant, R. A. (1999). A qualitative invesitgation of the organization of traumatic memories. British Journal of Clinical Psychology, 38.

Hauff, E., & Vaglum, P. (1993). Vietnamese boat refugees: the influence of war and flight traumatization on mental health on arrival in the country of resettlement. A community cohort study of Vietnamese refugees in Norway. Acta Psychiatr Scand, 88, 162-168.

Hauff, E., & Vaglum, P. (1995a). Organised violence and the stress of exile. British Journal of Psychiatry, 166.

Hauff, E., & Vaglum, P. (1995b). Organised violence and the stress of exile. Predictors of mental health in a community cohort of Vietnamese refugees three years after resettlement. Br J Psychiatry, 166, 360-367.

Herbert, J. D., Lilienfeld, S. O., Lohr, J. M., Montgomery, R. W., O'Donohue, W. T., Rosen, G. M., et al. (2000). Science and pseudoscience in the development of eye movement desensitization and reprocessing: implications for clinical psychology. Clin Psychol Rev, 20, 945-971.

Herman, J. L. (1992). Trauma and recovery. New York: Basic Books.

Hinton, W. L., Chen, Y. C., Du, N., Tran, C. G., Lu, F. G., Miranda, J., et al. (1993). DSM-III-R disorders in Vietnamese refugees. Prevalence and correlates. J Nerv Ment Dis, 181, 113-122.

Horowitz, M. J. (1976). Stress response syndromes. Northvale, NJ: Aronson.

Horowitz, M. J. (1986). Stress response syndromes (Vol. 2). New York: Jason Aronson.

HRW. (2000). Seeking Protection: Addressing Sexual and Domestic Violence in Tanzania's Refugee Camps. New York: Human Rights Watch.

Human Rights Watch. (1993). Civilian devastation - abuses by all parties in the war in South- ern Sudan. New York: Human Rights Watch.

Human Rights Watch. (1999). Famine in Sudan: the human rights causes. New York: Hu- man Rights Watch.

International Eminent Persons Group. (2002). Slavery, abduction and forced servitude in Su- dan. Karthoum: International Eminent Persons Group. 158 Jacobsen, L., & Smidt-Nielsen, K. (1997). Torture survivor - trauma and rehabilitation. Co- penhagen, Denmark: IRCT.

Janet, P. (1889). L'automatisme psychologique. Paris: Alcan.

Janet, P. (1904). L'amnesie et la dissociation des souvenirs par l'emotion. Journal de Psy- chologie, 1, 417-435.

Janoff-Bulman, R. (1992). Shattered assumptions. New York: Free Press.

Jaycox, L. H., Foa, E. B., & Morral, A. R. (1998). Influence of emotional engagement and habituation on exposure therapy for PTSD. Journal of Consulting and Clinical Psy- chology, 66, 185-192.

Jensen, S. B., Schaumburg, E., Leroy, B., Larsen, B. O., & Thorup, M. (1989). Psychiatric care of refugees exposed to organized violence. A comparative study of refugees and immigrants in Frederiksborg County, Denmark. Acta Psychiatr Scand, 80, 125-131.

Kaldor, M. (1999). New and old wars: organized violence in a global era. London: Blackwell.

Kaminer, D., Stein, D. J., Mbanga, I., & Zungu-Dirwayi, N. (2001). The Truth and Recon- ciliation Commission in South Africa: relation to psychiatric status and forgiveness among survivors of human rights abuses. Br J Psychiatry, 178, 373-377.

Kammerlander, A. (Ed.). (1997). Das Persönliche ist politisch. Frankfurt: Verlag für Interkul- turelle Kommunikation.

Karunakara, U., Neuner, F., Schauer, M., Singh, K., Hill, K., Elbert, T., et al. (submitted). War, Forced Migration and Post Traumatic Stress: Prevalence and predictors of PTSD in Sudanese and Ugandan Populations of the West Nile.

Kearney, C. (1999). World Mission Day: Surviving a Refugee School in Uganda. Retrieved 21 November, 2001, from http://www.catholicweekly.com.au/99/oct/76.html

Kessler, R. C., Sonnega, A., Bromet, E., Hughes, M., & Nelson, C. B. (1995). Posttrau- matic stress disorder in the National Comorbidity Survey. Arch Gen Psychiatry, 52, 1048-1060.

Kiecolt-Glaser, J. K., McGuire, L., Robles, T. F., & Glaser, R. (2002). Psychoneuroimmu- nology and psychosomatic medicine: back to the future. Psychosom Med, 64, 15-28. 159 Kim, J. J., & Yoon, K. S. (1998). Stress: metaplastic effects in the hippocampus. Trends Neurosci, 21, 505-509.

Kinzie, J. D., & Goetz, R. R. (1996). A century of controversy surrounding posttraumatic stress stress-spectrum syndromes: the impact on DSM-III and DSM-IV. J Trauma Stress, 9, 159-179.

Lang, P. (1979). A bio-informational theory of emotional imagery. Psychophysiology, 16, 195-512.

LeDoux. (1995). Emotion: clues from the brain. Annual Review of Psychology, 46, 209-235.

LeDoux, J. E. (2000). Emotion circuits in the brain. Annu Rev Neurosci, 23, 155-184.

Lesch, A. M. (1999). Sudan: the torn country. Current History, 97, 218-222.

Levav, I. (1997). Individuals under conditions of maximum adversity: the holocaust. In B. P. Dohrenwend (Ed.), Adversity, stress and psychopathology (pp. 13-33). New York: Oxford Universtiy Press.

Lopes Cardozo, B., Vergara, A., Agani, F., & Gotway, C. A. (2000). Mental health, social functioning, and attitudes of Kosovar Albanians following the war in Kosovo. Jama, 284, 569-577.

Maercker, A., Beauducel, A., & Schutzwohl, M. (2000). Trauma severity and initial reactions as precipitating factors for posttraumatic stress symptoms and chronic dissociation in former political prisoners. J Trauma Stress, 13, 651-660.

Maercker, A., & Schutzwohl, M. (1997). Long-term effects of political imprisonment: a group comparison study. Soc Psychiatry Psychiatr Epidemiol, 32, 435-442.

Malkki, L. H. (1995). Purity and Exile: Violence Memory, and National Cosmology Among Hutu Refugees in Tanzania. Chicago: The University of Chicago Press.

Marks, I., Lovell, K., Noshirvani, H., Livanou, M., & Thrasher, S. (1998). Treatment of posttraumatic stress disorder by exposure and/or cognitive restructuring: a controlled study. Arch Gen Psychiatry, 55, 317-325.

Martin, R. (2002). Sudan's perfect war. Foreign Affairs, 81. 160 Mayou, R. A., Ehlers, A., & Hobbs, M. (2000). Psychological debriefing for road traffic accident victims. Three-year follow-up of a randomised controlled trial. Br J Psychia- try, 176, 589-593.

McCann, I. L., Sakheim, D. K., & Anbrahamson, D. J. (1988). Trauma and victimization: A model of psychological adaptation. Counseling Psychologist, 16.

McClelland, J. L., McNaughton, B. L., & O'Reilly, R. C. (1995). Why there are comple- mentary learning systems in the hippocampus and neocortex: insights from the suc- cesses and failures of connectionist models of learning and memory. Psychol Rev, 102, 419-457.

McEwen, B. S. (1999). Stress and hippocampal plasticity. Annu Rev Neurosci, 22, 105-122.

McFarlane, A. C., Atchison, M., Rafalowicz, E., & Papay, P. (1994). Physical symptoms in post-traumatic stress disorder. J Psychosom Res, 38, 715-726.

McNally, R. J. (2003). Progress and controversy in the study of posttraumatic stress disorder. Annu Rev Psychol, 54, 229-252.

Meichenbaum, D. (1994). A clinical handbook/practical therapist manual for assessing and treating adults with post-traumatic stress disorder (PTSD). Waterloo, Ontario, Canada: Institute Press.

Metcalve, J., & Jacobs, W. (1996). A "hot-system/cool-system" view of memory under stress. PTSD Research Quarterly, 7, 1-3.

Mitchell, J. T., & Bray, G. (1990). Emergency service stress: guidelines for preserving the health and careers of emergency service personnel. Englewood Cliffs, NJ: Prentice Hall.

Mollica, R. F., Donelan, K., Tor, S., Lavelle, J., Elias, C., Frankel, M., et al. (1993). The effect of trauma and confinement on functional health and mental health status of Cambodians living in Thailand-Cambodia border camps. Jama, 270, 581-586.

Mollica, R. F., McInnes, K., Poole, C., & Tor, S. (1998). Dose-effect relationships of trauma to symptoms of depression and posttraumatic stress disorder among Cam- bodian survivors of mass violence. Br J Psychiatry, 173, 482-488. 161 Mollica, R. F., McInnes, K., Sarajlic, N., Lavelle, J., Sarajlic, I., & Massagli, M. P. (1999). Disability associated with psychiatric comorbidity and health status in Bosnian refugees living in Croatia [see comments]. Jama, 282, 433-439.

Mollica, R. F., Sarajlic, N., Chernoff, M., Lavelle, J., Vukovic, I. S., & Massagli, M. P. (2001). Longitudinal study of psychiatric symptoms, disability, mortality, and emi- gration among Bosnian refugees. Jama, 286, 546-554.

Mowrer, O. H. (1947). On the dual nature of learning - a re-interpretation of "conditioning" and "problem-solving". Harvard Educational Review, 61, 115-121.

Muecke, M. A. (1992). New paradigms for refugee health problems. Soc Sci Med, 35, 515- 523.

Münkler, H. (2002). Die neuen Kriege. Reinbeck bei Hamburg: Rowohlt.

Neuner, F., Schauer, M., Elbert, T., & Roth, W. T. (2002). A Narrative Exposure Treat- ment as intervention in a Macedonia's refugee camp: a case report. Journal of Behav- ioural and Cognitive Psychotherapy, 30, 205-209.

Odenwald, M., Schauer, M., Neuner, F., & Elbert, T. (2002). Validity and reliability of a PTSD assessment in Somali ex-combattants (Unpublished manuscript). Konstanz: Vivo.

Ozer, E. J., Best, S. R., Lipsey, T. L., & Weiss, D. S. (2003). Predictors of posttraumatic stress disorder and symptoms in adults: a meta-analysis. Psychol Bull, 129, 52-73.

Paunovic, N., & Ost, L. G. (2001). Cognitive-behavior therapy vs exposure therapy in the treatment of PTSD in refugees. Behav Res Ther, 39, 1183-1197.

Pelcovitz, D., van der Kolk, B., Roth, S., Mandel, F., Kaplan, S., & Resick, P. (1997). De- velopment of a criteria set and a structured interview for disorders of extreme stress (SIDES). Journal of Traumatic Stress, 10, 3-16.

Peltzer, K. (1999). Trauma and mental health problems of Sudanese refugees in Uganda. Cent Afr J Med, 45, 110-114.

Perkonigg, A., & Wittchen, H. U. (1999). Prevalence and comorbidity of traumatic events and posttraumatic stress disorder in adolescents and young adults. In A. Maercker, 162 M. Schützwohl & Z. Solomon (Eds.), Post-traumatic stress disorder: a lifespan developmental perspective. Seattle: Hogrefe & Huber.

Peterson, S. (2000). Me against my brother - at war in Somalia, Sudan and Rwanda. New York: Routledge.

Petterson, D. (1999). Inside Sudan. Colorado: Westview Press.

Pitman, R. K., Shalev, A. Y., & Orr, S. P. (2000). Posttraumatic stress disorder: emotion, conditioning and memory. In M. S. Gazzaniga (Ed.), The new cognitive neurosciences (Vol. 2). Cambridge: MIT Press.

Pitman, R. K., Shin, L. M., & Rauch, S. L. (2001). Investigating the pathogenesis of posttraumatic stress disorder with neuroimaging. J Clin Psychiatry, 62, 47-54.

Pupavac, V. (2002). Therapeutising refugees, pathologising populatins: international psychosocial programmes in Kosovo (Working paper no. 59). Geneva: UNHCR.

Resick, P. A., Nishith, P., Weaver, T. L., Astin, M. C., & Feuer, C. A. (2002). A comparison of cognitive-processing therapy with prolonged exposure and a waiting condition for the treatment of chronic posttraumatic stress disorder in female rape victims. J Consult Clin Psychol, 70, 867-879.

Resick, P. A., & Schnicke, M. K. (1993). Cognitive processing therapy for rape victims. New- bury Park, CA: Sage.

Riggs, D. S., Byrne, C. A., Weathers, F. W., & Litz, B. T. (1998). The quality of the intimate relationships of male Vietnam veterans: problems associated with posttraumatic stress disorder. J Trauma Stress, 11, 87-101.

Rose, S., & Bisson, J. (1998). Brief early psychological interventions following trauma: a systematic review of the literature. Journal of Traumatic Stress, 11, 697-710.

Rosenthal, G. (1997). Traumatische Familienvergangenheiten. In G. Rosenthal (Ed.), Der Holocaust im Leben von drei Generationen. Gießen: Psychosozial-Verlag.

Salter, D., McMillan, D., Richards, M., Talbot, T., Hodges, J., Bentovim, A., et al. (2003). Development of sexually abusive behaviour in sexually victimised males: a longitudinal study. Lancet, 361, 471-476. 163 Schachter, D. L. (1987). Implicit memory: history and current status. Journal of Experimen- tal Psycholoy: Learning, Memory and Cognition, 13, 501-518.

Schauer, M., Karunakara, U., Neuner, F., Klaschik, C., Kley, E., Rockstroh, B., et al. (2001). High prevalence of PTSD in African victims of organized violence: a demographic survey in the West-Nile region of Uganda and Sudan. Paper presented at the 7th European Conference on Traumatic Stress Studies, Edinburgh, United King- dom.

Schreiber, W. (2002). Das Kriegsgeschehen 2001. Opladen: Leske und Budrich.

Shalev, A. Y. (2001). Disorder takes away human dignity and character. British Medical Journal, 322, 1301.

Shapiro, F. (1989). Eye movement desensititazion: A new treatment for post-traumatic stress disorder. Journal of Behaviour Therapy and Experimental Psychiatry, 20, 211- 217.

Shapiro, F. (1995). Eye movement desensitization and reprocessing: Basic principles, protocols and procedures. New York: Guilford Press.

Shapiro, F., & Maxfield, L. (2002). Eye Movement Desensitization and Reprocessing (EMDR): information processing in the treatment of trauma. J Clin Psychol, 58, 933- 946.

Shastri, L. (2002). Episodic memory and cortico-hippocampal interactions. Trends Cogn Sci, 6, 162-168.

Shrestha, N. M., Sharma, B., Van Ommeren, M., Regmi, S., Makaju, R., Komproe, I., et al. (1998). Impact of torture on refugees displaced within the developing world: symptomatology among Bhutanese refugees in Nepal. Jama, 280, 443-448.

Shum, M. S. (1998). The role of temporal landmarks in autobiographical memory processes. Psychol Bull, 124, 423-442.

Silove, D. (1999). The psychosocial effects of torture, mass human rights violations, and refugee trauma: toward an integrated conceptual framework. J Nerv Ment Dis, 187, 200-207. 164 Silove, D., McIntosh, P., & Becker, R. (1993). Risk of retraumatisation of asylum-seekers in Australia. Aust N Z J Psychiatry, 27, 606-612.

Silove, D., Sinnerbrink, I., Field, A., Manicavasagar, V., & Steel, Z. (1997a). Anxiety, depression and PTSD in asylum-seekers: assocations with pre- migration trauma and post-migration stressors. Br J Psychiatry, 170, 351-357.

Silove, D., Sinnerbrink, I., Field, A., Manicavasagar, V., & Steel, Z. (1997b). Anxiety, depression and PTSD in asylum-seekers: associations with pre-migration trauma and post-migration stressors. Br J Psychiatry, 170, 351-357.

Silove, D., Steel, Z., McGorry, P., & Mohan, P. (1998). Trauma exposure, postmigration stressors, and symptoms of anxiety, depression and post-traumatic stress in Tamil asylum-seekers: comparison with refugees and immigrants. Acta Psychiatr Scand, 97, 175-181.

Somasundaram, D. J., & Sivayokan, S. (1994). War trauma in a civilian population. British Journal of Psychiatry, 165, 524-527.

Southwick, S. M., Krystal, J. H., Morgan, C. A., Johnson, D., Nagy, L. M., Nicolaou, A., et al. (1993). Abnormal noradrenergic function in posttraumatic stress disorder. Arch Gen Psychiatry, 50, 266-274.

Spiegel, D. (1989). Hypnosis in the treatment of victims of sexual abuse. Psychiatric Clinics of North America, 12, 295-305.

Squire, L. R. (1994). Declarative and nondeclarative memory: multiple brain systems supporting learning and memory. In D. L. Schacter & E. Tulving (Eds.), Memory systems 1994. Cambridge: MIT Press.

Steil, R., & Ehlers, A. (1996). Die posttraumatische Belastungsstörung: Eine Übersicht. Verhaltensmodifikation und Verhaltensmedizin, 17, 169-212.

Stieglitz, K. (1998). Zur Lage der Menschenrechte im Sudan vor dem Hintergrund des an- haltenden Bürgerkrieges. Internationales Afrikaforum, 34, 71-79.

Stieglitz, K. (1999). Zur Problematik der Konzentrationslager im Sudan. Internationales Af- rikaforum, 35, 177-182. 165 Summerfield, D. (1997). Legacy of war: beyond "trauma" to the social fabric. The Lancet, 349, 1568.

Summerfield, D. (1998). "Trauma" and the experience of war: a reply. Lancet, 351, 1580- 1581.

Summerfield, D. (1999). A critique of seven assumptions behind psychological trauma programmes in war-affected areas. Soc Sci Med, 48, 1449-1462.

Summerfield, D. (2001). The invention of post-traumatic stress disorder and the social use- fulness of a psychiatric category. Bmj, 322, 95-98.

Summerfield, D. (2002). Effects of war: moral knowledge, revenge, reconciliation, and medicalised concepts of "recovery". Bmj, 325, 1105-1107.

Tandon, Y. (1984). Ugandan refugees in Kenya: A community of enforced self-reliance. Disasters, 8, 267-271.

Tarrier, N., Pilgrim, H., Sommerfield, C., Faragher, B., Reynolds, M., Graham, E., et al. (1999). A randomized trial of cognitive therapy and imaginal exposure in the treatment of chronic posttraumatic stress disorder. J Consult Clin Psychol, 67, 13-18.

Tarrier, N., Sommerfield, C., Pilgrim, H., & Humphreys, L. (1999). Cognitive therapy or imaginal exposure in the treatment of posttraumatic stress disorder. Twelve-month follow-up. Br J Psychiatry, 175, 571-575.

Tauber, C. D. (2003). Psychological trauma, physical health and conflicht resolution in Croa- tia, Serbia and Bosnia: lessions for the future. Retrieved 17.02, 2003, from www.conflictres.org/vol184/tauber.htm

Terr, L. (1993). Unchained memories. New York: Basic Books.

True, W. R., Rice, J., Eisen, S. A., Heath, A. C., Goldberg, J., Lyons, M. J., et al. (1993). A twin study of genetic and environmental contributions to liability for posttraumatic stress symptoms. Arch Gen Psychiatry, 50, 257-264.

Tulving, E. (2001). Episodic memory and common sense: how far apart? Philos Trans R Soc Lond B Biol Sci, 356, 1505-1515.

Tulving, E., & Markowitsch, H. J. (1998). Episodic and declarative memory: role of the hippocampus. Hippocampus, 8, 198-204. 166 Turner, S. (1999). Angry Young Men in Camps: Gender, Age and Class Relations Among Bu- rundian Refugees in Tanzania (Working Paper No. 9). Roskilde: Roskilde Univer- sity, Institute of Development Studies.

UN. (1984). Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment. Retrieved 2003, January 7, from http://www.unhchr.ch/html/menu3/b/h_cat39.htm

UNHCR. (2002a). Extensive abuse of West African refugee children reported. Retrieved 9 June 2002, from http://www.unhcr.ch/

UNHCR. (2002b). Statistical yearbook 2001. Geneva: UNHCR.

UNICEF. (2001). The state of the world's children 2001. New York: UNICEF.

USCR. (2001). Uganda: 2000 Country Report. Retrieved July 10, 2001, from http://www.refugees.org van der Gaag, N. (1996). Field of Dreams: Life in a Refugee Settlement. New International- ist. van der Kolk, B. A. (1994). The body keeps the score: memory and the evolving psychobiology of posttraumatic stress. Harvard Review of Psychiatry, 1, 253-265. van der Kolk, B. A. (1995). Dissociation and the fragmentary nature of traumatic memories: overview and exploratory study. Journal of Traumatic Stress, 8, 4. van der Kolk, B. A. (1996). Trauma and memory. In B. A. van der Kolk, A. C. McFarlane & L. Weisaeth (Eds.), Traumatic stress. New Yoek: Guilford Press. van der Kolk, B. A. (1997). The psychobiology of posttraumatic stress disorder. Journal of Clinical Psychiatry, 58 Suppl 9, 16-24. van der Kolk, B. A., Roth, S., Pelcovitz, D., & Mandel, F. (1993). Complex PTSD: Results of the PTSD field trials for DSM-IV. Washington, DC: American Psychiatric Associa- tion.

Van der Veer, G. (1998). Counselling and therapy with refugees and victims of trauma (Vol. 2). West Sussex: Wiley. 167 van Emmerik, A. A., Kamphuis, J. H., Hulsbosch, A. M., & Emmelkamp, P. M. (2002). Single session debriefing after psychological trauma: a meta-analysis. Lancet, 360, 766-771.

Van Ommeren, M., Sharma, B., & de Jong, J. (1997). Culture, trauma, and psychotrauma programmes. Lancet, 350, 595.

Van Velsen, C., Gorst-Unsworth, C., & Turner, S. (1996). Survivors of torture and organized violence: demography and diagnosis. J Trauma Stress, 9, 181-193.

Veronen, L. J., & Kilpatrick, D. G. (1983). Stress management for rape victims. In D. Meichenbaum & M. E. Jaremko (Eds.), Stress reduction and prevention (pp. 341-479). New York: Plenum Press.

Vesti, P., Somnier, F., & Kastrup, M. (1992). Psychotherapy with torture survivors. Copen- hagen, Denmark: IRCT.

Vyas, A., Mitra, R., Shankaranarayana Rao, B. S., & Chattarji, S. (2002). Chronic stress in- duces contrasting patterns of dendritic remodeling in hippocampal and amygdaloid neurons. J Neurosci, 22, 6810-6818.

Ware, J., Jr., Kosinski, M., & Keller, S. D. (1996). A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med Care, 34, 220-233.

Weine, S. M., Becker, D. F., McGlashan, T. H., Laub, D., Lazrove, S., Vojvoda, D., et al. (1995). Psychiatric consequences of "ethnic cleansing": clinical assessments and trauma testimonies of newly resettled Bosnian refugees. Am J Psychiatry, 152, 536- 542.

Weine, S. M., Kulenovic, A. D., Pavkovic, I., & Gibbons, R. (1998). Testimony psychotherapy in Bosnian refugees: A pilot study. American Journal of Psychiatry, 155, 1720-1726.

Weine, S. M., & Laub, D. (1995). Narrative constructions of historical realities in testimony with Bosnian survivors of "ethnic cleansing". Psychiatry, 58, 246-260.

Weine, S. M., Vojvoda, D., Becker, D. F., McGlashan, T. H., Hodzic, E., Laub, D., et al. (1998). PTSD symptoms in Bosnian refugees 1 year after resettlement in the United States. Am J Psychiatry, 155, 562-564. 168 Weiss, D. S., & Marmar, C. R. (1996). The Impact of Event Scale - Revised. In J. Wilson & T. M. Keane (Eds.), Assessing psychological trauma and PTSD (pp. 399-411). New York: Guilford Press.

WHO. (1997). Composite International Diagnostic Interview (CIDI). Geneva: World Health Organization.

WHO/UNHCR. (1996). Mental health of refugees. Geneva: World Health Organization.

Yehuda, R., & McFarlane, A. C. (1995). Conflict between current knowledge about posttraumatic stress disorder and its original conceptual basis. Am J Psychiatry, 152, 1705-1713.

Zatzick, D. F., Marmar, C. R., Weiss, D. S., Browner, W. S., Metzler, T. J., Golding, J. M., et al. (1997). Posttraumatic stress disorder and functioning and quality of life outcomes in a nationally representative sample of male Vietnam veterans. Am J Psychia- try, 154, 1690-1695.

Zizek, S. (1999). Das Trauma der verfehlten Melodie bei Robert Schumann. In E. Bronfen, B. R. Erdle & S. Weigel (Eds.), Trauma - zwischen Psychoanalyse und kulturellem Deutungsmuster. Köln: Böhlau.