The Psychological Impact of September 11 Terrorism on Australian Inpatients

METHOD

Subjects were included if they were hospital inpatients, and had free access to television, newspapers or other media between 11 and 14 September 2001. The study protocol received ethical approval.

All available psychiatric inpatients on September 14 were approached and asked to provide written consent for the study. Thirty individuals agreed, with three declining or being too unwell to participate.

A comparison group of medical and surgical inpatients from the same hospital were interviewed during the same 24 h period, once they had provided written consent. The comparison group was matched to the psychiatric group by age and gender. Individuals with a comorbid psychiatric condition were excluded. One patient declined to participate.

Every subject had a semistructured interview conducted by an experienced psychiatrist during the 24 h period of the study. The semistructured interview was based on the Impact of Event Scale.[3] The interviewing clinician had the opportunity to assess the patients’ mental state and make other observations. Demographic data were obtained, along with details of diagnosis and treatment. Additionally, self-report (including any sleep disturbance and nightmares) and observer report of the distress arising from the televised tragedies were noted, as well as the subjects’ accuracy of recall and causal attribution for the events. The psychiatrists interviewing were not blinded to diagnosis, although there was no a priori hypothesis. Multiple interviewers saw the psychiatric group, leading to possible interrater discrepancies, although only one psychiatrist evaluated the comparison group. However, all interviewers were experienced psychiatrists trained in the same facility.

The raw data were independently rated by both authors, with disagreement being low (14%). Disagreement was resolved through mutual re-analysis of the raw data. Descriptive statistics were applied to the data. Chi-squared tests of significance were used for between-group comparisons.

RESULTS

Table 1 shows characteristics of the psychiatry and medical/surgical patients. In the whole sample, 14 subjects (25%) reported serious distress or incorporated the events into their delusional beliefs, with 20 individuals (36%) reporting minor distress. The psychiatrist noted three cases (5%) of serious observable distress and 18 cases (32%) of minor distress when assessing the subjects’ reaction to the televised events. All patients with observed serious distress fulfilled Diagnostic and Statistical Manual of Mental Disorders (4th edn; DSM-IV) criteria[4] for an acute stress disorder, and had pre-existing psychiatric problems.

Women reported significantly more distress overall than men (x[2] = 13.5, p < 0.01), but no objective difference was detected. There was no relationship between age and any of the variables examined.

The psychiatric group reported, and was observed to have, more distress than the medical and surgical patients, but these differences did not achieve statistical significance. Seven patients (29%) with a preexisting psychotic illness exhibited delusional beliefs surrounding the terrorist attacks (e.g. ‘I did it’). The psychiatric group also had a significantly more varied explanation as to the cause of the events (X[2] = 15.1, p < 0.01).

There was no difference in sleep or dreaming, by self-report, between the two groups.

DISCUSSION

The present study confirmed our clinical impressions that a large number of hospital inpatients (from Alfred Hospital, Melbourne) at the time were adversely affected by the graphic media footage of this terrible series of events. Factors reported as being linked to later post-traumatic states include being young and female.[5] In the acute situation, we found that women reported more distress than men, although this may be due to the women involved being more verbal or more likely to report symptoms rather than being more sensitive to the coverage of 9/11. It is also perhaps unsurprising that psychotic patients reported more varied and illogical possible causes of 9/11, given that active psychosis is well known to be associated with difficulties in processing information.

Major world events have been shown to impact negatively on the public's mental health. Hawton et al. examined the rates of suicide and deliberate self-harm in the weeks after the death of Diana, Princess of Wales.[6] They found an increased rate of suicide in the month following the funeral, and observed from case notes that the influence of Diana's death was largely through amplification of personal losses or exacerbation of existing distress. In other words, the most severely affected were already in a weakened psychological state. Similarly, we found that the three individuals who developed an acute stress disorder (ASD)[4] had a pre-existing mental disorder. The symptoms of ASD are largely subjective, and the present study indicated a disparity between subjective and objective distress. This calls into question the validity of observable or objective distress, and the traditional mental state examination has been shown to have only moderate interrater reliability.[7]

Psychological adaptation to uncontrollable stress is thought to be linked to physical well-being,[8] but we found that patients with severe mental illness, as opposed to those with major medical or surgical problems, were more vulnerable and adversely affected, albeit to a non-significant extent. Some 29% of individuals with a psychotic illness became explicitly delusional regarding these events, while others in the psychiatric group appeared unaware of or blunted to the relentless reportage. There was no evidence of a ‘dose-response’ relationship between severity of reaction and extent of exposure to the media coverage of 9/11, although the extent of exposure was not strictly quantified beyond establishing prolonged daily viewing.

Limitations of the present study include the population being hospital based, and hence confined and unwell. Thus, our findings could not necessarily be generalized to the community. The study groups were comparatively small in number, but this was a time-sensitive project.

The impact of television violence has principally been researched in children[9] but little consensus emerges, and our review revealed a paucity of data on the effect of frightening or violent television on the mentally ill. Similar studies in the future could attempt to measure the type and amount of media exposure. Such studies should also consider using a community-based comparison group, even though our medical and surgical inpatients were a random group who had been screened for pre-existing mental disorder.