Hi everyone, my name is Priya Srivastava, and today I will be talking about predictors of PTSD severity in a primary care setting, mental health, and well-being factors. Post-traumatic stress disorder, or PTSD, is common in primary care settings in as large as 12.5% of presenting patients. The gap between where patients first present in primary care and where they receive treatment creates a significant barrier to patients suffering from an already burdensome disease. To gain an understanding of what factors constitute the development of PTSD in patients, we have introduced an interprofessional behavioral health model in our academic primary care setting at Rowan Family Medicine. Research members study how symptoms of PTSD vary in CAPS-5, the Hopkins checklist total, dissociative symptoms, trauma experiences, and total trauma each patient has experienced. Our overall objective is to determine what factors of a patient's history are associated with PTSD severity in patients as evaluated in a primary care setting, in hopes of informing treatment guidelines for PCPs and ultimately improving patient outcomes. Our workflow that we created consists of an intake phase, a 30-day app usage phase, and a follow-up phase. Initially, physicians detect or sense traumatic symptoms or signs in patients that come to them for their primary care visits. The physicians will then initiate a warm handoff to a researcher that is on-site, and the researcher will then screen the patient with the DSM-CAPS-5, which is the gold standard for PTSD diagnosis. The researcher will then download the PTSD Coach application on the patient's mobile phones, and the PTSD Coach app is an app that is used to help patients with their symptoms and provide the necessary resources. Patients are then instructed to use the application at least three times a day for 30 days, and then they follow up 30 days after with the researchers, and they are re-evaluated using the CAPS-5. To examine salient predictors of PTSD severity, a random forest was performed using the PARTI package in R. RF aggregates fitted predictions across decision trees to determine the best predictor performance. Our results are shown here. The best predictors of PTSD severity were looked at with variable importance. The HSC scores, or the well-being scores, had a variable importance of 1.157. The life satisfaction scores had a variable importance of 0.990 in predicting PTSD. Patients that had experienced sexual assault had a 0.343 variable importance in predicting PTSD, and patients that experienced serious injury had a variable importance of 0.212 in predicting PTSD. The mean number of total PTSD symptoms was 7.75 and the mean total severity was 22.4. 4% of the variance in PTSD symptoms was explained by the RF model. Participants who had directly experienced serious injury endorsed increased levels of PTSD symptoms, whereas participants who had just witnessed or heard about an event exhibited less symptom severity. The higher levels of the HSC-well-being scores are demonstrated in this diagram to show how predictive they were of increased PTSD severity. Similarly, the life satisfaction scores for patients that had lower levels of it were predictive of increased PTSD severity, as seen in this diagram. In this diagram, we are demonstrating how patients who experienced sexual assault had increased levels of PTSD severity, and that is also a predictive factor. In conclusion, increased symptoms of the following may be associated with PTSD symptom severity – depression, anxiety, decreased life satisfaction, sexual assault, and serious injury. These findings align with the presentation of PTSD, as mood and hyperarousal symptoms are often present in these patients, along with effects on global functioning. Findings supported that personally-directed events tend to be associated with PTSD severity broadly. On the other hand, a total number of traumatic events was not associated with PTSD severity. Patients in community mental health and veteran populations supports a dose-response relationship. In the future, research may continue to investigate correlates of PTSD severity that are specific to primary care populations. We strive to continue recruiting patients to understand how factors in a patient's history impact their symptoms and overall disease severity. I would like to thank my co-authors and everyone who helped on this project, including Danielle Schweitzer, Faith Schenk, Alexandra Groom, Salma Saifuddin, Deep Patel, and Dr. Jones. Thank you so much.

PTSD severity

primary care

life satisfaction

trauma experiences

behavioral health model