**2. The current study**

The purpose of the present study was to determine the predictive power of ASD diagnosis, clusters, and symptoms on PTSD diagnosis based on semi-structured interviews according to gender in victims of violent crimes. To do so, four operational objectives were defined. The first objective was to assess the percentage of individuals with or without ASD (i.e., full, partial, or no diagnosis) who will have developed PTSD (i.e., full and partial) 2 months after a violent crime. The second objective was to assess the predictive power of overall ASD diagnosis and each ASD cluster (i.e., B, C, D, and E) on PTSD diagnosis (full or partial compared to no diagnosis). Moreover, we predicted that dissociation would better classify PTSD compared to other clusters. The third objective was to evaluate the predictive power of each ASD symptom on PTSD diagnosis. The fourth objective was to examine the predictive power of each ASD cluster and symptom according to gender.

#### **2.1. Method**

#### *2.1.1. Participants and procedure*

The study was part of a larger longitudinal study that aimed to examine the evolution of distress and well-being after a violent crime, as well as predictive factors of PTSD. Individuals were recruited between 2009 and 2014 through the Trauma Studies Center (TSC) with the collaboration of the Montreal Crime Assistance Center (CAC). The inclusion criteria were the following: (a) being exposed to a violent crime (e. g., physical or sexual aggression and armed robbery) during the previous 30 days (i.e., the timeframe to screen for ASD), (b) being aged between 18 and 65 years old, and (c) being able to communicate in French or English. Exclusion criteria included a past or present psychotic episode, bipolar disorder, traumatic brain injury, an organic mental disorder or active suicidal ideations requiring an intervention.

Individuals were informed of the project by the CAC up to 30 days post crime. Interested individuals contacted the coordinator to clarify the circumstances of the crime and to be screened according to the research criteria. Then, symptoms of individuals admitted to the study were assessed with a semi-structured clinical interview (T0) conducted by a trained assistant to evaluate the ASD diagnosis and to fill out questionnaires. Individuals were reassessed for PTSD diagnosis after 2 months (T1) and questionnaires were sent by mail to be completed at home. For each completed assessment, individuals received a 20\$ compensation. For the present subset of the larger study, only quantitative information of semi-structured interviews was used (i.e., ASD diagnosis, clusters, and symptoms at T0 and PTSD diagnosis at T1). In total, 69% of individuals completed both assessments.

#### **2.2. Sociodemographic characteristics at baseline**

Globally, studies on victims of violent crimes have supported the relevance of ASD diagnosis to predict PTSD. Predictive ability of partial ASD has only been documented in one study. Overall, all ASD clusters have shown high negative predictive power and low positive predictive power. However, it remains unclear which cluster better predicts PTSD diagnosis. The predictive power of each ASD symptom has not been examined in victims of violent crimes. Moreover, the differential predictive power of ASD clusters and symptoms according to gender has not been studied in these victims. Finally, previous studies have used self-report measures to evaluate ASD and PTSD, which suggests that responses were influenced by the

134 A Multidimensional Approach to Post-Traumatic Stress Disorder - from Theory to Practice

Moreover, it is well known that women are twice as likely as men to develop PTSD, but the reasons underlying this discrepancy remain unclear [5]. If ASD is considered a predictor of PTSD, it appears pertinent to determine to what extent ASD clusters and symptoms predict

The purpose of the present study was to determine the predictive power of ASD diagnosis, clusters, and symptoms on PTSD diagnosis based on semi-structured interviews according to gender in victims of violent crimes. To do so, four operational objectives were defined. The first objective was to assess the percentage of individuals with or without ASD (i.e., full, partial, or no diagnosis) who will have developed PTSD (i.e., full and partial) 2 months after a violent crime. The second objective was to assess the predictive power of overall ASD diagnosis and each ASD cluster (i.e., B, C, D, and E) on PTSD diagnosis (full or partial compared to no diagnosis). Moreover, we predicted that dissociation would better classify PTSD compared to other clusters. The third objective was to evaluate the predictive power of each ASD symptom on PTSD diagnosis. The fourth objective was to examine the predictive power of each ASD

The study was part of a larger longitudinal study that aimed to examine the evolution of distress and well-being after a violent crime, as well as predictive factors of PTSD. Individuals were recruited between 2009 and 2014 through the Trauma Studies Center (TSC) with the collaboration of the Montreal Crime Assistance Center (CAC). The inclusion criteria were the following: (a) being exposed to a violent crime (e. g., physical or sexual aggression and armed robbery) during the previous 30 days (i.e., the timeframe to screen for ASD), (b) being aged between 18 and 65 years old, and (c) being able to communicate in French or English. Exclusion criteria included a past or present psychotic episode, bipolar disorder, traumatic brain injury,

an organic mental disorder or active suicidal ideations requiring an intervention.

victims' perceptions.

PTSD according to gender.

**2. The current study**

**2.1. Method**

cluster and symptom according to gender.

*2.1.1. Participants and procedure*

**Table 1** presents the sociodemographic characteristics of individuals. In total, 39 female and 36 male victims of violent crimes participated in the study. The mean age was 35 years old for women and 42 years old for men. Most participants were Caucasians. Twenty-three percent of women and 38% of men were currently in a relationship. Thirty-nine percent of women and 50% of men were employed. Most often, women and men reported having been victims of a physical assault (i.e., 77 and 84%, respectively).


**Table 1.** Sociodemographic characteristics at baseline.

#### **2.3. Measurement**

The acute stress disorder interview (ASDI) [26] is a semi-structured interview administered at T0 that assesses ASD symptoms according to the DSM-IV. Individuals meeting all criteria were classified as having full ASD. Partial ASD was diagnosed when individuals met all criteria except for criteria B, for which only one out of three dissociative symptoms was required [18]. The ASDI has shown good internal consistency (*α* = 0.90), temporal stability (*r* = 0.90), sensibility (91%), and specificity (93%) based on clinicians' judgment [26].

The structured clinical interview for DSM-IV axis I disorders (SCID-I) [27] was used at T1 to assess the presence or absence of PTSD (i.e., full, partial, or no diagnosis) and other axis I diagnoses. Partial PTSD was diagnosed when individuals met all criteria for each cluster except one (i.e., B, C, or D) [18]. This semi-structured interview showed good convergent validity according to clinicians' judgment (*k* = 0.69), as well as an inter-rater reliability ranging from 0.77 to 0.92 [28].
