HADStress: a somatic symptom screen for posttraumatic stress among Somali refugees.

OBJECTIVE

This study assessed whether a simple, readily implemented four-symptom somatic screen would be able to effectively identify current posttraumatic stress symptoms in victimized populations.

METHODS

The sample consisted of 622 Somali community-dwelling refugees who fled widespread violence and trauma occurring in East Africa during 1990-1992. Data were collected during 2000-2003 and included demographic characteristics, number of types of torture and nontorture trauma experienced earlier in Africa, and current self-rated posttraumatic stress symptoms, as measured by the Posttraumatic Stress Disorder Checklist (PCL). The sample was also assessed with the HADStress screen, which was developed for this study, to determine whether the screen was effective in detecting current posttraumatic stress symptoms. The HADStress screen assessed for the presence of four somatic symptoms: Headaches, Appetite change, Dizziness, and Sleep problems. All items were given equal weight. Possible scores on the screen range from 0 to 4, with higher scores indicating more somatic symptoms.

RESULTS

Univariate analysis showed that persons who experienced more types of trauma (both torture and nontorture trauma) and persons who had higher PCL scores (indicating more current posttraumatic stress symptoms) had significantly higher HADStress scores. Negative binomial regression analysis showed that PCL scores were the most effective variable in predicting HADStress scores. On the Tukey-B post hoc analysis, a HADStress score of 0 or 1 was associated with a mean PCL score of less than 30, a score of 2 was associated with a mean PCL score of 40.28, and a score of 4 was associated with a mean PCL score of 51.07 (suggesting that over 50% of this group would have active posttraumatic stress disorder).

CONCLUSIONS

A score of 2 or higher on the HADStress scale among refugees warrants additional evaluation for posttraumatic stress symptoms in clinical settings. For communitywide efforts at early recognition and treatment, a cutoff score of 4 may be more practical and cost-effective.