The Relationship between Lifetime Exposure to Potentially Traumatic Events, Peritraumatic Dissociation, and PTSD in a Sample of Sexually Assaulted Women in Sao Paulo, Brazil.

Sexually assaulted women represent a particularly high-risk group for developing post-traumatic stress disorder (PTSD). Potentially traumatic events (PTEs) and peritraumatic dissociation (PD) are known risk factors for PTSD. However, little is known about how previous trauma affects PD and how this relationship affects PTSD. We aimed to investigate whether PD acts as a mediator between PTEs and PTSD severity in a sample of recently sexually assaulted women in Sao Paulo, Brazil. Seventy-four sexually assaulted women aged 18-44 completed questionnaires and structured interviews on PTSD, PD, and PTEs. We examined direct and indirect effects of variables using causal mediation analysis. Lifetime exposure to PTEs was a risk factor for PD, but PD was not a risk factor for PTSD symptom severity. Also, PD was not a mediator between PTEs and PTSD severity. We provided recommendations on how to further explore the relationship between lifetime traumatic exposure, PTSD, and peritraumatic dissociation.

Identifying posttraumatic stress disorder staging from clinical and sociodemographic features: a proof-of-concept study using a machine learning approach.

This proof-of-concept study aimed to investigate the viability of a predictive model to support posttraumatic stress disorder (PTSD) staging. We performed a naturalistic, cross-sectional study at two Brazilian centers: the Psychological Trauma Research and Treatment (NET-Trauma) Program at Universidade Federal of Rio Grande do Sul, and the Program for Research and Care on Violence and PTSD (PROVE), at Universidade Federal of São Paulo. Five supervised machine-learning algorithms were tested: Elastic Net, Gradient Boosting Machine, Random Forest, Support Vector Machine, and C5.0, using clinical (Clinician-Administered PTSD Scale version 5) and sociodemographic features. A hundred and twelve patients were enrolled (61 from NET-Trauma and 51 from PROVE). We found a model with four classes suitable for the PTSD staging, with best performance metrics using the C5.0 algorithm to CAPS-5 15-items plus sociodemographic features, with an accuracy of 65.6% for the train dataset and 52.9% for the test dataset (both significant). The number of symptoms, CAPS-5 total score, global severity score, and presence of current/previous trauma events appear as main features to predict PTSD staging. This is the first study to evaluate staging in PTSD with machine learning algorithms using accessible clinical and sociodemographic features, which may be used in future research.

Validation of the Brazilian-Portuguese Version of the Clinician Administered Post Traumatic Stress Disorder Scale-5.

Objectives: The aim of this study was to validate CAPS-5 for the Brazilian-Portuguese language on a sample of 128 individuals from two centers (from the cities of São Paulo and Porto Alegre) who have been recently exposed to a traumatic event. Methods: We performed a reliability analysis between interviewers (with a subset of 32 individuals), an internal consistency analysis, and a confirmatory factorial analysis for the validation study. Results: The inter-rater reliability of the total PTSD symptom severity score was high [intraclass correlation coefficient =0.994, 95% CI (0.987-0.997), p < 0.001]. Cohen's Kappa for individual items ranged between 0.759 and 1. Cronbach's alpha coefficients indicated high internal consistency for the CAPS-5 full scale (α = 0.826) and an acceptable level of internal consistency for the four symptom clusters. The confirmatory factorial analysis for the 20-item original CAPS-5 did not fit the data well. A 15-item model with better results was then established by excluding the following CAPS-5 items: dissociative amnesia, recklessness, distorted cognitions, irritability, and hypervigilance. Conclusion: Despite the limitation of the predominance of female victims, and the high number of sexually assaulted women in our sample, the model with only 15 items provided a good fit to the data with high internal consistency (α = 0.835).