This study validates a new instrument for the assessment of peritraumatic symptoms. To our knowledge, PBQ-SR is the first instrument specifically designed to measure several components of combat-related peritraumatic stress. The questionnaire was assessed using retrospective data on 688 Marine infantry service members with respect to the most stressful event during their last deployment. PBQ-SR demonstrated satisfactory psychometric properties with good internal consistency and discriminant validity as to PANAS-Positive Score. Descriptive analysis of each item, inter-item correlation and Cronbach's α stability after item deletion indicate that all 15 items of the questionnaire could be retained. The statistically significant positive correlation between both the PBQ-SR total score and all of its 15 items to the PDEQ confirms the ability of PBQ-SR to reliably assess peritraumatic reactions as a general construct.
Consistent with prior results research , there was a significant positive correlation between the PBQ-SR total score and measures of general anxiety, depression, negative affect and lower general health after deployment. PBQ-SR also showed high concurrent validity with respect to posttraumatic symptoms after deployment and their changes over time. PTSD severity strongly correlated not only with PBQ-SR total scores and the two subscales, but also with most of the fifteen individual PBQ-SR items. In order to compare the concurrent psychometric properties of PBQ-SR to PDEQ, we recalculated all correlations with respect to PDEQ (cf. Table 3). Our results suggest similar psychometric properties, while the PBQ-SR shows slightly better concurrent validity to almost every other measure assessed (cf. Table 3).
Linear and logistic regressions have shown that PBQ-SR total score remained a significant predictive factor of PTSD symptom severity, even after controlling for depression, general anxiety, negative affect and general health. Our results suggest a PBQ-SR cut-off score of 12 points for screening purposes. This score has been shown to correctly classify respondents at risk for PTSD even after controlling for other psychopathologies. In addition, participants exceeding this cut-off score also showed significantly higher anxiety, depression and negative affect scores, lower general health scores and a greater increase of PTSD symptoms after deployment when compared to the control group.
As literature suggests that peritraumatic symptoms partly rely on pre-existing factors, such as trait dissociation and may be an important risk factor for PTSD development and resilience [24, 72, 73], we also investigated the correlation between peritraumatic symptoms (PBQ-SR), trait dissociation assessed by the DES and posttraumatic symptoms (CAPS). Our results showed a statistically significant correlation between all three scores and confirm prior findings. However, even though the DES total score also correlated significantly with the CAPS total score (r = .188, n = 394, p < .001), the close correlation of PBQ-SR to the CAPS total score remained significant even after controlling for trait dissociation (r = .370, n = 393, p < .001).
A principal component analysis suggested the existence of two underlying factors: physical/dissociative (PAS) and perceptive/emotional (EDS) aspects of peritraumatic stress reaction. Both factors have demonstrated satisfactory convergent validity to PTSD specific measures and significant concurrent validity correlations to other measures such as general anxiety, depression, negative affect and general health (cf. Tables 2 and 3). At present, the two pre-existing scales assessing peritraumatic reactions (PDEQ and PDI) seem to each focus separately on different dimensions of peritraumatic symptoms. PDEQ’s underlying structure is shown to reflect more dissociative symptoms such as altered awareness and derealization (reflecting the PAS), while PDI focuses more on emotional peritraumatic distress symptoms (reflecting the EDS). Hence, our results suggest that PBQ-SR is a new scale that may have incremental validity by unifying and assessing the two major underlying peritraumatic symptom dimensions, which would otherwise require the administration of both PDEQ and the PDI. Assessing both dimensions also affirms the clinical relevancy of PBQ-SR and we therefore support the use of both subscales in clinical practice. The separate use of EDS and PAS could, however, be considered in psychiatric research.
In summary, our study replicates and extends the findings of other studies reporting peritraumatic dissociation to be a robust correlate of PTSD symptoms in adults. Our findings suggest that peritraumatic symptoms as measured by the PBQ may be useful indices of risk and need for early intervention for PTSD in service members deployed to combat zones, should, however, be prospectively validated through immediate peritraumatic assessment through in-theater administration of the PBQ-SR in the future.
Limitations and strengths
Because the content development focused on the behavioral indicators of the phenomenology of service members under great strain in a war-zone, the PBQ stands out as a peritraumatic index uniquely suitable for military members. The PBQ is a questionnaire that is easy to administer and score due to its simple rating and scoring instructions and clearly specified areas of evaluation. The 5-point-Likert scale structure promotes the comparability of our results to prior studies using already existing scales. Concerning the psychometric properties of the PBQ-SR, excluding Marines with prior deployments from the predictive analyses has contributed to the reduction of bias risk due to prior stressors, while the use of the trauma-specific CAPS score likely contributed to a higher specific predictive power of our results. The size, as well as the homogeneity of our assessed sample (Marine, infantry) in terms of gender, age and kind of exposure type is a specific feature and further strength of our study, compared to validation studies designed for other scales. This may be an advantage for military-use of the questionnaire, but may limit its generalizability to other populations. In addition, as the optimal cutoff score was empirically established in our study, cross-validation in an independent sample appears necessary. Further limitations also include the lack of assessment of prior traumatic life events or childhood trauma, combat exposure level and test-retest variability of the measure. In addition, we evaluated the PBQ-SR on male Marine Corps infantry units who deployed to Afghanistan in a period of time that was relatively quiescent. It is an empirical question whether these results would be generalizable to more highly exposed troops, in other branches of service, and with women. However, the most significant limitation of the study is the retrospective, self-report assessment of peritraumatic symptoms, which precludes a conclusion on predictive validity of the PBQ-SR. Retrospective assessment of symptoms may also lead to distortion of recollections or bias due to current symptoms  and subjective assessment of combat-related symptoms could introduce bias and distortions related to cognitive barriers (i.e. fear of stigma, warrior ethos, criticism, fear of removal from unit, etc.) and adaptive denial coping mechanisms of Service Members [2, 75]. Despite those concerns, our results suggest good reliability, validity and applicability of the PBQ-SR.