The International Trauma Questionnaire: An assessment of the psychometric properties of its Spanish version

ABSTRACT


INTRODUCTION
Several mental disorders have been associated with traumatic experiences during the lifetime, which is conceptualized as psychotrauma [1].The 5 th edition of the Diagnostic and Statistical Manual of Mental Disorders of the American Psychiatric Association (DSM-5) and the 11 th edition of the International Classification of Diseases of the World Health Organization (ICD-11) have proposed new perspectives in the classification of disorders related to psychological stress or psychotrauma, particularly on posttraumatic stress disorder (PTSD) and complex posttraumatic stress disorder (CPTSD) [2].
The international trauma questionnaire (ITQ) is a measure designed to detect stress-related disorders or psychotrauma and to assess the response to related treatments [3,4].This instrument is a brief and simple measure, focusing on the key features of PTSD and CPTSD.The ITQ was developed in accordance with ICD-11 principles proposed by the World Health Organization, such as maximizing clinical utility and ensuring international applicability in detecting the core symptoms of the disorder.The ITQ is freely available and focuses especially on 1. functional impairment related to both PTSD and CPTSD and 2. predicting differential treatment outcomes [3].
The ITQ has also been employed among Syrian refugees in Lebanon [5], in China [6,7], military and police populations in the United Kingdom [8,9], Norway [10], in academic populations in South Africa [11], Austria, Lithuania, Scotland and Wales [12], the United States [13], Israel [14], Portugal and Angola [15], and Nigeria, Kenya, and Ghana [16].Although this questionnaire has been translated into a Latin American Spanish version, as provided by The International Trauma Consortium [17], there are no evidence on its psychometric properties.Consequently, the aim of this research has been the validation of the Spanish version of the ITQ, as well as to test its psychometric properties and score subjects with PTSD or CPTSD criteria from the Paraguayan population.

Participants
An online survey was launched to recruit participants.This survey was shared via social networks (Twitter, Facebook) and messaging applications (Telegram, WhatsApp) from November 15 to December 15, 2021.Each participant was informed about the privacy and data processing of the study, as well as about the research objectives.Individuals aged ≥18 years, who selfreported at least one traumatic life event (assessed through the life events checklist for DSM-5) [18] were included.
The sample size was calculated using Epidat software, taking into account an expected frequency of 3.9% of anxiety disorders in the adult population of Paraguay [19], a confidence level of 95% and a precision of 3%.The minimum sample was thus set at 138 participants [20].A total of 189 subjects were surveyed, of which 141 were selected considering the previously mentioned inclusion criteria.
There is evidence to suggest that responses to online surveys are capable of providing similar results to those reported through "in-person" samples [21].This justified the use of an online survey approach, which has also proven useful in times of social distancing, such as those experienced during the COVID-19 pandemic.

International trauma questionnaire
The ITQ is an 18-item self-report measure to assess ICD-11 PTSD and CPTSD in adults.Six items represent three clusters of PTSD: Re-experiencing in the here and now (Re_dx: Re1 and Re2), Avoidance (Av_dx: Av1 and Av2) and sense of current threat (Th_dx: Th1 and Th2), and six items represent three clusters of DSO: Affective dysregulation (AD_dx: AD1 and AD2), Negative self-concept (NSC_dx: NSC1 and NSC2) and Disturbances in relationships (DR_dx: DR1 and DR2).Additionally, three items measure functional impairment (social, occupational, and other key areas of life) for PTSD and DSO clusters.Respondents must indicate how much they have been bothered by each symptom over the past month on a 5point Likert scale ranging from 0 ('not at all') to 4 ('extremely').Scores ≥2 ('moderately') indicate the presence of a symptom.PTSD diagnosis requires endorsement of one symptom in each PTSD cluster and associated functional impairment.CPTSD diagnosis requires a PTSD diagnosis, one symptom in each DSO cluster and associated functional impairment [3].
The original English version of the scale good psychometric properties [3].In this study we used the two-factor version reporting better psychometric properties [6,22].

Trauma questionnaire
The TQ is a screening tool for PTSD.It includes 44 items divided into three groups: The time of reference for the evaluation is any time after the event.In the symptoms listing section, a score is obtained by adding up each item (1 as yes and 0 as no), with higher severity for higher scores [23].

International trauma exposure measure
The international trauma exposure measure (ITEM) is a checklist developed to capture traumatic life events and their associated characteristics according to the ICD-11 criteria.The ITEM measures exposure to different traumatic life events across various stages of life (childhood, adolescence, adulthood, and across the lifespan), frequency of exposure to the traumatic event, and the main emotion associated with the traumatic event.The ITEM is freely available and can be used without specific permission [24].

Translation Process and Validation
We followed the procedures for the cross-cultural adaptation of self-report measures, using the back-translation method [25] for the translation and validation of the ITQ from English to Spanish.First, the original English version was translated into Spanish; this version was then backtranslated into English by a bilingual expert.Then, a native English speaker compared the back-translated version with the original English version to verify meaning equivalences.Minor changes were introduced after the comparison and the Spanish version was pilot tested with 15 people to verify its comprehensibility.After pilot test, final Spanish version was approved (available upon request to corresponding author).

Statistical Analysis
We assessed the pertinence of performing a factorial analysis (SPSS software version 23) through the Kaiser-Meyer-Olkin (KMO) test for sample adequacy and the Bartlett's test of sphericity.Confirmatory factor analysis (CFA) was performed using Jeffrey's amazing statistics program [26].For CFA (taking into consideration the sample size), we used the diagonally weighted least squares (DWLS) estimation procedure.Chisquare (χ 2 ), the comparative fit index (CFI), the normed fit index (NFI), the Tucker-Lewis index (TLI), the root mean square error of approximation (RMSEA), and the standardized root mean square residual (SRMSR) were used to test the model fit.These indices are used to assess if the fit model is acceptable (RMSEA and SRMSR between 0.05 and 0.08, and CFI and TLI between 0.90 and 0.95) or good (RMSEA and SRMSR <0.05 and CFI and TLI >0.95) [27].
Cronbach's alpha was used to measured reliability [28].Convergent validity was measured computing correlations between the ITQ and TQ using Pearson's method in SPSS [29].

Ethical Considerations
The Department of Medical Psychology of the National University of Asuncion, School of Medical Sciences (Paraguay), ethically approved the study.We followed the Helsinki principles regarding data processing.In case any participant requested information on the survey results, he/she was invited to write his/her e-mail address to receive information.

Participants
A total of 141 subjects were surveyed, of whom 76.6% were men.Age ranged from 19 to 69 years old with a mean of 36.32±9.76 years and a median of 34 years (IQR=12.5).Of participants, 84.4% reported a university education, 81.6% were employed and 39.7% were single.These characteristics are shown in detail in Table 1.

Preliminary Analysis
According to the ITQ, 22.7% of participants reported a PTSD, while 22.7% CPTSD.The ITQ demonstrated an excellent internal consistency: Cronbach's alpha of the total scale was α=0.95, for PTSD symptoms and DSO clusters were α=0.91 and α=0.93, respectively [28].Acceptable corrected item-total correlations (range=0.577to 0.801) [30] was reported by each of the 18 items.A Cronbach's alpha of α=0.975 was found for the TQ, indicating excellent internal consistency [28].According to this scale, 59.6% of participants reported a diagnosis of PTSD.

Factorial Analysis
The results of the KMO test and the sphericity test were adequate (KMO=0.878)and significant (p<0.001),respectively.Based on the responses to the 12 core symptom items, the twodimensional model was evaluated with a confirmatory factor analysis.According to all fit indices, χ 2 =207 (df=53, p<0,001).RMSEA=0.144(IC90 % 0.123-0.164),CFI=0.868,TLI=0.835,BIC=4,935, SRMR=0.0631, and AIC=4,826), the model adjustment was acceptable.These results confirm that the model of the Spanish version of ITQ replicates the two-factor model of the original English version, since all items had standardized factor loadings > than 0.40 (p<0.001).
Table 2 summarizes items-means and standard deviations, factor loadings, and communalities (h2) for the ITQ.

Convergent Validity
Convergent validity of the ITQ was assessed by correlating the ITQ with the TQ.A good construct validity was found, since the correlation between the ITQ and the TQ was direct and significant (r=0.807;p<0.001) [29].

Trauma Exposure
Table 3 reports the results of the ITEM, summarizing the contextual characteristics contributing to the psychotrauma among responders.65.2% of them reported as a traumatic event: "You were humiliated, belittled, or insulted by another person".Of participants, 29.8% reported fear as the prevalent emotion associated to the traumatic, while anger has been reported in 16.3% (Figure 1).

DISCUSSION
The aim of the study was to assess the psychometric properties of Spanish version of the ITQ in a sample from Paraguayan general population.
A confirmatory factor analysis was conducted, considering that the two-dimensional structure has been associated to good psychometric properties.Our research allowed us to determine that the two-dimensional structure correctly explained the construct analyzed (as in the other versions of the scale).This was demonstrated through the results reported by the fit indices [6,10,22].
Factor loadings were high on all items, which means that were equally valid as in the English version.In terms of internal validity, the Spanish version of the ITQ reported an excellent Cronbach's alpha value (α=0.95), while for PTSD and DSO subscales clusters were α=0.91 and α=0.93, respectively.In a community sample, reliabilities of the English version for all PTSD and DSO subscales were satisfactory (all α's ≥0.79) [3].Our study determined that the construct was adequately measured, taking into consideration the direct and significant convergence found with the TQ.
Our sample mainly included male patients.Epidemiology of PTSD in general shows a sex ratio 2:1 in favor of females [31].Although this study did not aim to test the association of sex with the incidence of PTSD or CPTSD, it is striking that in a sample mostly including males, a PTSD percentage of 59.6% is observed according to the TQ and 22.7% according to ITQ.It has been proposed that differences at the level of neuronal circuits or neurobiological processes between male and female individuals might play a role in the explanation of sex differences [32], as well as the involvement of hormones such as testosterone, estradiol, and progesterone.Traumatogenic factors and epigenetic changes may be also involved [33,34].
More than 80% of participants reported that they were from urban areas even if no evidence has been collected in the previous studies PTSD and urbanicity [35].Among American war veterans, it has been observed that those living in rural areas reported lower access to the mental health services and lower scores of PTSD, depression, substance use and global mental health [36].However, similar evidence in veterans have shown no significant differences in traumatic characteristics between rural versus urban veterans [37].
Employment is an impacting factor on the outcome of patients suffering from PTSD [38].Our findings have shown that most of participants were employed with a partly preserved functional outcome.In a study conducted by the US Veterans Administration reporting on the follow up of a program called individual placement and support (IPS), the authors found that those with PTSD who participated in the program have shown greater improvements in total, interpersonal and lifestyle functioning [38].This evidence has been replicated in other studies on the implementation of employment programs in veterans with PTSD symptoms [39,40].
In our study, the frequency of PTSD and CPTSD was 22.7% according to the ITQ.These frequencies are high since the sampling has been drawn from the general population.In fact, in Israel, frequencies of PTSD and CPTSD among subjects exposed to different types of psychotrauma (war conflicts, terrorism) were 9% and 2.6%, respectively [14].Also, in United Kingdom frequencies of PTSD and CPTSD were 10.9% and of 53.6% in a population exposed to various forms of psychotrauma [4].In addition, prevalence of CPTSD and PTSD were 36.1% and 25.2% among Syrian refugees living in Lebanon [5], with similar percentages to those from our study.Differences in the frequency of PTSD between the TQ (59.6%) and the ITQ (22.7%) are supposed to be based on the evidence that TQ directly uses the DSM-IV criteria whereas the ITQ employs the ICD-11 criteria.Similar studies have confirmed that the use of ICD-11 criteria leads to lower diagnosis rates [4].
Considering the emotions associated with the traumatic event in the sample, fear scored 29.79% followed by no emotions in 18.44% of cases: fear is common in PTSD and is part of diagnostic criteria [41] as well as is related to brain circuits of fear conditioning and extinction [42].Intrusive memories may elicit intense fear responses in PTSD patients leading to intense stress with cognitive overload and somatic dissociation [43].
Regarding the temporality of the traumatic event, was mostly reported between 10 and 20 years ago (33.3%).The timing of the reported traumatic event may impact on various aspects of PTSD and CPTSD: if it occurs in childhood, it is more likely to lead to CPTSD [24,44], as well as may be associated to other co-occurring conditions such as substance abuse [45].
Finally, the frequency of sexual aggression among traumatized subjects in our study is of concern: 20.6% reported having been victims of sexual aggression by someone other than a parent or guardian, as well as 53.9% of the sample claimed to have suffered some form of sexual harassment.This must be thoroughly investigated to provide victims with comprehensive care and support programs.
Limitations of this research may include: 1. the lack of data on sociodemographic or clinical factors, 2. the use of self-report measures to include/exclude participants, 3. the overrepresentation of college-educated men from urban areas, and 4. failure to calculate test-retest reliability (since contact information was not collected from subjects recruited for a second evaluation).
All these limitations could bias the results.
We may conclude that the Spanish version of the ITQ shows good psychometric properties and satisfactorily replicates the two-dimensional model of the original English version of the scale.This allows us to consider our research as relevant, since it offers a scale that could be useful for Spanish-speaking patients affected by PTSD or CPTSD.

Table 2 .
ITQ: Items-means and standard deviations, factor loadings, and communalities of the 12 core symptom items of the ITQ

Table 4
reports about the temporality of traumatic events according to the ITQ.