Dimensional structure of posttraumatic stress disorder symptoms after cardiac arrest

Background: Considerable evidence suggests that posttraumatic stress disorder (PTSD) is a heterogeneous construct despite often being treated as a homogeneous diagnostic entity. PTSD in response to cardiac arrest is common and may differ from PTSD following other medical traumas. Most patients are amnesic from the cardiac event, and it is unclear if and how certain PTSD symptoms may manifest. Methods: We examined the latent structure of PTSD symptoms in 104 consecutive cardiac arrest survivors who were admitted to Columbia University Medical Center. PTSD symptoms were assessed via the PTSD Checklist-Specific at hospital discharge. We performed a confirmatory factor analysis (CFA) to compare 4-factor dysphoria, 4-factor numbing, and 5-factor dysphoric arousal models of PTSD with our data. Results: The CFA showed that each of the models had good fit. We chose the 4-factor numbing model (chi(2) (113)=151.59, p <.01, CFI=0.94, RMSEA=0.057, 90% CI: [0.032, 0.081]) as most representative of the data, after considering a between-factor correlation of 0.99 in the 5-factor dysphoric arousal model, and greater fit statistics than the 4-factor dysphoria model. Limitations: Certain factors were defined by only two items. Additionally, PTSD was assessed at discharge (median=21 days); those assessed before 30 days could be displaying symptoms of acute stress disorder. Conclusions: Our findings suggest that PTSD symptoms after cardiac arrest are best represented by a 4-factor numbing model of PTSD. PTSD assessment and intervention efforts for cardiac arrest survivors should consider the underlying dimensions of PTSD.