Psychosocial adjustment of directly exposed survivors 7 years after the Oklahoma City bombing

Abstract

Objective

The aim of this study was to prospectively examine the long-term course of psychiatric disorders, symptoms, and functioning among 113 directly exposed survivors of the Oklahoma City bombing systematically assessed at 6 months and again nearly 7 years postbombing.

Methods

The Diagnostic Interview Schedule/Disaster Supplement was used to assess predisaster and postdisaster psychiatric disorders and symptoms and other variables of relevance to disaster exposure and outcomes.

Results

Total prevalence of posttraumatic stress disorder (PTSD) was 41%. Seven years postbombing, 26% of the sample still had active PTSD. Delayed-onset PTSD and new postdisaster alcohol use disorders were not observed. PTSD nonremission was predicted by the occurrence of negative life events after the bombing. Posttraumatic symptoms among survivors without PTSD decayed more rapidly than for those with PTSD, and symptoms remained at 7 years even for many who did not develop PTSD. Those with PTSD reported more functioning problems at index than those without PTSD, but functioning improved dramatically over 7 years, regardless of PTSD or remission from PTSD. No survivors had long-term employment disability based on psychiatric problems alone.

Conclusions

These findings have potentially important implications for anticipation of long-term emotional and functional recovery from disaster trauma.

Introduction

Recent decades of disaster mental health research have yielded a sizeable literature on the mental health consequences of exposure to disaster. Much less is known, however, about the effects of direct exposure to large-scale terrorist events. Posttraumatic stress disorder (PTSD) prevalence among the few available studies of directly exposed survivors of terrorist incidents ranges from 18% to 50% [1].

General population research on PTSD has shown that PTSD follows an often chronic course [2], [3], [4]. A comprehensive review of disaster research [5] noted that few studies have chronicled the long-term course of PTSD among disaster survivors, and even fewer have used true panels (ie, prospective assessment of the same individuals repeatedly over time). The longest major prospective study of disaster survivors, conducted many years ago, examined a sample of litigants at 2 and 14 years after the Buffalo Creek dam disaster [6]. In that study, the current prevalence of PTSD was 44% at 2 years, dropping to 28% at 14 years.

Most longitudinal disaster studies have found that PTSD prevalence diminishes with time [5], [7]. Few consistent predictors of recovery from PTSD have been described, however. One negative predictor may be the amount of early postdisaster symptoms, which has been shown to predict PTSD chronicity in several studies [5]. Remission of symptoms of the avoidance/numbing (group C) symptom cluster of PTSD seems to be pivotal to recovery from PTSD [8], [9].

The long-term course of psychopathology in populations directly exposed to terrorist attacks has not been studied. One research group retrospectively examined adult survivors of a terrorist attack on their school in Israel 17 years earlier when they were adolescents. The majority still had traumatic stress symptoms, most commonly hyperalertness [10]. Because this study examined symptoms rather than PTSD and no early assessment was conducted, the longitudinal course of PTSD in these survivors of terrorism cannot be traced.

At the time, the Oklahoma City bombing was the deadliest act of terrorism ever perpetrated on American soil. Previous research by our team found that one third (34%) of directly exposed survivors of the Oklahoma City bombing developed PTSD in relation to it [11]. Approximately 6 months after the Oklahoma City bombing, our group conducted a study of a random sample of directly exposed survivors from a comprehensive state registry, systematically collecting data using a structured diagnostic assessment. The strength of the sampling, severe exposure level of the participants, and the full diagnostic assessment methods of the Oklahoma City bombing study provided a benchmark for research on psychiatric effects of extreme trauma. Its findings informed the efforts of mental health authorities to estimate psychiatric consequences and develop interventions after the September 11, 2001, attacks [12].

The most highly disaster-exposed groups are likely to experience the most profound mental health effects [1], [13], but paradoxically, the most highly exposed groups are typically the most difficult to access and engage in research. Despite the difficulties inherent in studying highly exposed groups, research on those with the greatest disaster impact is needed to determine the maximally anticipated mental health needs as they evolve from the early phases to long-term outcomes. Observations on the long-term course of the highly exposed Oklahoma City bombing sample may help anticipate outcomes and needs for long-term interventions for high-exposure survivors of the September 11, 2001, terrorist attacks, who were not systematically studied in the early postdisaster period.

The systematic data collected in the Oklahoma City bombing study in the early postdisaster period provide an important foundation for longitudinal study of postdisaster psychopathology in these survivors who were carefully assessed in the early postdisaster phases [9], [11]. Approximately 7 years after the bombing, a follow-up study of the directly exposed bombing survivors used measurement instruments consistent with those in the index study. The purpose of this study was to describe and identify predictors of the long-term course of psychiatric disorders, emotional distress, and social and occupational functioning after the bombing. The main research questions to be investigated by this study were as follows: (1) how much recovery from the most prevalent psychiatric disorders (PTSD and major depression) would occur during the course of 7 years; (2) how the recovery curve for PTSD would compare with the recovery curve for PTSD symptoms in the absence of full PTSD; (3) how the recovery curve for PTSD would compare with the recovery curve for new postdisaster major depressive episodes; (4) determination of the course of social and occupational functional impairment over time and their relationships to PTSD and recovery from PTSD; and (5) identification of predictors of remission from PTSD.