Examining Emotional Responses to a Natural Disaster Using Pre-existing & Repeated Measures

by David V. Baldwin, PhD
Eugene, Oregon
(541) 686 2598

If a natural disaster occurs within the United States during the term of this grant's approval, affecting on-going research as described below, I will be able to collaborate with investigators in the affected area. Please contact me if you are a principal investigator and interested in possible collaboration. Thank you.

NOTE: The following proposal was originally approved for possible funding in fall 1996, by the Natural Hazards Center at the University of Colorado at Boulder, and has received annual extensions. In November 2004, approval for possible funding of this proposal was extended through 2005.

A Quick Response Research Proposal

Some people develop Post-traumatic Stress Disorder (PTSD) following a traumatic event, while others with similar experiences do not. We don't know why. Were these people already different in some ways, or do peri-traumatic factors (e.g., subjective perception, etc.) fully explain these differing emotional responses? The answer to this promises far-reaching benefits.

Summary Of Research Goals

This Quick Response application proposes to study subjects already involved in research where relevant data concerning pre-disaster characteristics or functioning are available, when they are affected by a natural disaster. These pre-disaster data will be used to predict individual differences in severity of emotional response (e.g., susceptibility to PTSD). Additionally, post-disaster repeated measures will be taken and analyzed to clarify whether previously observed group differences on similar variables predict or result from exposure to traumatic events.

Project Significance

Differential susceptibility to PTSD, given a similar objective traumatic experience, remains the most fundamental unresolved question concerning this disorder. Understanding individual differences in vulnerability and resilience to traumatic-stress requires that: (1) appropriate data are available describing the subjects' pre-trauma functioning (Reid, 1990); and that (2) exposure to the traumatic event is objectively similar for a broad range of affected individuals (as in natural disasters). Unfortunately, most studies of trauma responses begin collecting data only after the traumatic event has occurred, and focus on individual traumatic experiences (e.g., abuse, mugging, rape, even combat experience) that may differ widely across individuals. Studies lacking data prior to the traumatic event cannot distinguish causes from effects. Studies using variable traumas have difficulty comparing varied results across subjects.

This proposal seeks to overcome these two methodological flaws by selecting participants already involved in research studies where pertinent data were collected prior to a natural disaster. Selecting a discrete natural disaster standardizes the precipitating traumatic event in objectively measurable ways. Availability of baseline pre-disaster data permits a prospective design -- and assessment of the disaster's effect on individuals in relation to pre-existing variables. By examining the relationship between individual variations in disaster response with baseline and repeated post-disaster measures, we can learn potentially important clues about differential susceptibility to PTSD and we may uncover the causal nature of some previously observed group differences on similar variables. Understanding the predictive importance of pre-existing variables, and their changes at re-test after the trauma, increases our grasp of traumatic-stress responses. This knowledge may open possibilities for more effective preventive interventions -- and it is crucial for an adequate theory of PTSD. Thus, these results can have direct implications for applied clinical interventions as well as public health risk-reduction efforts.

Research Design (Sequential Tasks)

Criterion: A large scale, natural disaster, directly affecting a large metropolitan area within the United States. Uniform community impact is more important than disaster type. Promptly after the disaster, the affected city or area will be screened for suitability for this project. Particularly suitable sites would include university cities where medical, psychological or other longitudinal research studies are routinely conducted with large samples.

Initial Site Screening: Initial site screening will be conducted via the internet and by other means. A general announcement could be placed at my Trauma Information Pages web site; on-line searching for currently-funded research projects in the affected geographical areas is readily available from several links at this site (e.g., by city or zip code). Additional messages would be placed on pertinent electronic mailing lists (e.g., traumatic-stress, disaster research, FEMA, or PsyUSA), and on appropriate USENET news groups (e.g., sci.psychology.research). Telephone or email contacts will also be initiated with individual researchers in or near the affected area, as well as with major funding sources (NIMH, NIH, NSF, etc.).

Sample (Data) Selection: A disaster site will be judged suitable if it contains over 100 subjects involved in longitudinal or other psychological research on which appropriate baseline data were collected less than 1 year before the disaster struck. Principal investigator(s) cooperation with the aims of the proposed project would be essential, including repeated measures of selected data. However, suitable research studies need not be focused directly on disasters or on traumatic-stress responses - they need only have collected relevant pre-disaster data.

Suitable data are defined as either relevant measures concerned with the avoidance, hyper-arousal, and/or re-experiencing PTSD symptom clusters, or as characteristics believed to be affected by trauma where pre-trauma measures have never been available (e.g., volume of hippocampus). Specific examples might include brain imaging, psychophysiological, or neuroendocrine measures, clinically-relevant biographical interview or self-report data (e.g., known trauma or abuse histories; temperament; psychopathology), structured behavioral observations, certain types of data pertaining to memory or dissociation, MMPI or similar personality test batteries, WAIS full-scale intelligence scales, etc.

In the event more than one study (and cooperative principal investigator) meets these sample selection criteria, preference will be given to projects with large Ns using multi-modal measures collected shortly before the disaster, offering the most promising data, where repeated (post-disaster) measurement of selected variables would not interfere with existing research design.

Activation: I will promptly contact the Natural Hazards Center, once a suitable disaster-affected research sample is identified, requesting permission to activate this Quick Response project. Following early coordination with investigators at the site, my travel to the affected area would occur about six weeks following the disaster, for a period of up to two weeks; a single trip is planned. [Note: Since the DSM-IV-TR diagnosis of PTSD requires that symptoms persist at least one month following the precipitating traumatic event, earlier travel is not indicated.]

Subject Selection: Subjects currently involved in the selected research project will, with the permission and assistance of its principal investigator(s), be contacted to request participation in this disaster study. They will be informed by phone or letter of the purpose of this study and the reasons they were selected, and asked to sign an informed consent form signifying participation. It is anticipated that subjects involved in an on-going research project may be more willing to extend their participation to this study than non-subject individuals.

Assessment and Group Composition: Participants in the study will be tested by this investigator (a licensed psychologist) with a standardized structured interview schedule (e.g., CAPS-1, SCID-PTSD, etc.) or efficient battery of standard psychometric scales (e.g., Mississippi Scale, Penn Inventory, SCL-90-R, IES-Revised, TSI, etc.) -- as appropriate for the specific sample selected -- to assess both the presence and severity of their PTSD responses, and similarity of disaster exposure. These results will determine assignment of individual participants (matched for demographics, disaster exposure, etc.) to high-response and low-response PTSD groups. Participants will have similar experience of the disaster (i.e., present when it occurred; equivalent losses), and will be without serious physical injury.

Repeated Measures: Investigator(s) involved in the prior study will be asked to re-test the two selected participant groups (probably a subset of subjects from their study) with selected pre-disaster measures. Re-testing should be completed between two to four months post-disaster. These data are time-sensitive, requiring a Quick Response format. [All participants judged in need of therapy will be referred to local mental health resources offering effective treatment promptly after re-testing; with client permission, diagnostic data may be provided to clinicians.]

Statistical Analyses: This investigator will analyze obtained data with two specific questions in mind: (1) Do pre-disaster variables predict differential emotional responsiveness, or susceptibility to PTSD, after a disaster (significant between group pre-disaster differences)? (2) Do repeated measures data suggest that observed group differences result from trauma exposure or predispose participants to PTSD (significant group X repeated measures interactions)? The current version of SPSS (e.g., SPSS 12.0 for Windows, with additional modules, or later version) will be used for most data analyses.

Publication: As much as posible, writing will be collaborative with investigator(s) of the project collecting pre-disaster and repeated measures data. A completed preliminary report is expected promptly after re-test data collection, with the final report completed within five months of field work. The Quick Response report may form the basis of a later proposal for a larger follow-up study to examine the identified relationships more fully, possibly with the collaborating investigators. Submitted reports will be prepared for publication by the Natural Hazards Center at their web site and on paper, and I will link to the report from my web site.

Estimated Budget

Up to $ 2500.00 requested, primarily covering expenses surrounding travel to the disaster area, per diem, assessment materials, supplies, and minimal payments to selected subjects for re-testing.