PROJECT SUMMARY
Posttraumatic stress disorder (PTSD) can be debilitating and is related to impaired occupational and social
functioning, chronic medical problems, disability, and increases in suicidality, violence, and criminal behavior.
Over 10 million Americans are treated annually for sudden, severe illnesses or injuries that require very urgent
treatment; and an estimated 20% of hospitalized traumatic injury patients have persisting mental health
problems. Identifying those at risk for later disorder is difficult, and no screens can accurately predict PTSD in
Americans exposed to traumatic stress. In research on adults exposed to traumatic injury, theory-driven pre-
trauma, time-of-trauma, and post-trauma risk factors were strongly related to outcomes and accurately
identified those who later had elevated PTSD symptoms. We propose to extend our work to a population level
to create a screen for use in U.S. hospitals. In diverse populations of sudden illness or injury patients at
hospitals sites in Baltimore, MD, Akron, OH, and Palo Alto, CA, data will be collected on theory-driven, highly
predictive pre-trauma (gender, education, socioeconomic status, ethnicity/race, childhood home life and
parental dysfunction, past trauma exposure); time-of-trauma (trauma intensity, pre-trauma home life; pre-
trauma life stress), and post-trauma (acute PTSD and dissociation symptoms, negative thinking; post-trauma
life stress; post-trauma social support; post-trauma social constraints) risk factors and later symptoms of
disorder in 1,500 patients, including 1,100 (76%) ethnic/racial minorities. Data will also be collected on five
variables found to relate to disparities in mental health or PTSD: discrimination/racism, competing needs,
community cohesion, stigma, and negative expectations about treatment. Data collected will include sufficiently
large samples of African American, Latino, Asians, and mixed ethnicity/race patients for separately analyses
for these groups. Relationships of all predictors and disparities variables to outcomes and covariance among
predictors and disparities variables will be examined to select risk factor and disparities variables that are
highly predictive, similarly related to outcomes across subgroups, and can be measured with the fewest items
and items that are briefer and less personal, stigmatizing, and difficult to understand. In Wave II of data
collection, predictive performance will be assessed in a new sample of 1,500 patients, including 1,100 (76%)
ethnic/racial minorities with a goal of identifying at least 80% (sensitivity) of those who later have high PTSD
symptoms and at least 70% (specificity) of those who recover well.
