The relationship between demographic and medical characteristics and the development of post-traumatic stress disorder in children following emergency department attendance: A prospective cohort study

Objectives: This study adopted a prospective longitudinal design to assess the utility of demographic and medical characteristics routinely available to emergency medicine clinicians to predict the development of posttraumatic stress disorder (PTSD) in children exposed to death or serious injury 2 months following emergency department (ED) attendance.

Methods: A sample of children (8-17 years; N = 231) were recruited from 4 EDs in the East of England between 2010 and 2013. Within 2 weeks of attendance, research nurses screened records for appropriate cases and recorded information on relevant variables from ED attendance notes. At 2 months, a research assistant carried out a structured clinical interview to assess their Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) PTSD symptoms. Univariate analyses were conducted to compare ED characteristics between children who developed PTSD and those who did not. Logistic regression models were used to identify variables associated with increased risk of PTSD.

Results: Ten percent of children met the criteria for PTSD at 2 months. Systolic blood pressure, pulse, number of injuries, being subjected to interpersonal violence, and having a head injury were variables that distinguished PTSD and non-PTSD groups. Logistic regression models showed that being assaulted was predictive of PTSD (Odds ratio = 5.07, 95% CI [1.51, 17.00]); although these models had excellent specificity (0.96), the sensitivity was poor (0.30)—that is, there were a number of cases who developed PTSD but were not assaulted.

Conclusion: PTSD is a complication of exposure to death or injury that occurs in a significant minority of children. Children who are victims of interpersonal violence are more likely to develop the disorder.