Impact of Event Scale (IES)

The IES was developed to measure the current degree of subjective impact experienced by a person, as anchored to a specific stressful event (Refs: 1 and 2). The IES is used to assess current distress for any life event. The wording is not specific to any particular occurrence but only to the qualities linked to the subjective experience. (Ref: 1) The IES taps dimensions that are similar to the defining symptoms of PTSD. The IES can be used to (1) observe individuals over time after a stressful event has occurred, (2) compare the degree of subjective distress experienced by subgroups after a stressful life event has occurred, and (3) compare the relative impact of different life events on different populations. (Ref: 1). Content related to the intrusion and avoidance subscales were based on evaluation and interviews with patients. Items were derived from the most frequent statements used to describe distress after recent life events. To make the items specific to each person, the life event experienced most recently was recorded at the top of the IES to be used as a referent for each item. Based on the clinical experience of the IES developers, the items were divided into the Intrusion and Avoidance subscales. Scores may be obtained for each subscale and for the total scale. (Ref: 1) The intrusion component includes items such as repetitive behavior, troubled dreams, and unbidden thoughts and images. The avoidance component includes items such as blunted sensation, ideational constriction, and emotional numbness (Ref: 1). Various forms of the IES were administered to psychotherapy patients with stress response syndromes and to non-patient volunteers who had experienced serious life events. The IES was revised during this pilot testing period to include 20 items (9 intrusion, 11 avoidance). The current IES includes 15 items that reference “the past seven days,” across the subscales of avoidance and intrusion. (Ref: 1) It was found that the scores for frequency and intensity of episodes were similar. Therefore, for a given item, only the higher score of either frequency or intensity is used to indicate the magnitude of the item.

Free;http://www.mardihorowitz.com

No information found. Based on number of items, less than 20 minutes.

1979

Flesch-Kincaid Grade Level 6.0; a person with a 6th grade level education should be able to read and comprehend the scale.

4-point scale: (0) not at all, (1) rarely, (3) sometimes, and (4) often. The specific life event and date are recorded at the top of the page. (Ref: 1)

Self-report

Add items 1, 4, 5, 6, 10, and 14 for the intrusive subscale score (score range is 0 to 35). Add items 2, 3, 7, 8, 9, 12, 13, and 15 for the avoidance subscale score (score range is 0 to 40). Sum all items for the total score, with higher scores reflecting “more stressful impact.” Total score range is 0 to 75. Scores below eight are subclinical, scores from 9 to 25 are mild to moderate, and scores above 26 indicate moderate to severe impact. (Ref: www.mardihorowitz.com) Recently, a cut-score of 35 has been suggested, which achieves a sensitivity of 0.89 and a specificity of 0.88. (Ref: 3)

French (Ref: Brunet A, et al., Can J Psychiatry. 2003 Feb;48(1):56-61. PMID: 12635566); Chinese (Ref: Wu KK, et al., Soc Psychiatry Psychiatr Epidemiol. 2003 Feb;38(2):94-8. PMID: 12563552). A revised version of the IES has also been developed, which is not reviewed here. (Ref: Weiss DS, Marmar CR. The Impact of Event Scale - Revised. In: Wilson JP, Keane TM, editors. Assessing psychological trauma and PTSD. New York: Guilford Press; 1997. 399-411)

Mardi Horowitz, MD; Nancy Wilner, BA; William Alvarez, MA

Department of Psychiatry, University of California School of Medicine, San Francisco.

No information found.

www.mardihorowitz.com

top

Principal components factor analysis with varimax rotation was performed with a two-factor forced solution. The Avoidance subscale had avoidance item loadings ranging from 0.391 to 0.860, while the Intrusion subscale had intrusion item loadings ranging from 0.578 to 0.745. (Ref: 2) The subscales correlated at 0.57 to 0.78 across two bereaved groups and three assessment periods, except at Time1 (pre-therapy) for psychotherapy patients (r=0.15, p<0.40). (Ref: 2) The subscales have also been found to correlate (r=0.49). (Ref: 3) In a sample of Vietnam veterans with a mean age of 46 years, principal axis factor analysis suggested one factor of general level of distress, on which 13/15 items had factor loadings > 0.50 and which accounted for 49% of the total variance. (Ref: 4)

Confirmatory factor analysis with varimax rotation was performed with a sample of 95 melanoma patients and revealed that the Intrusion factor was distinct from the Avoidance factor. The Intrusion subscale accounted for 42.4% of the total variance and the Avoidance subscale accounted for 13.2% of the total variance. (Ref: 5) Together, both items accounted for 55.4% of the total variance in IES scores.

Another confirmatory factor analysis produced a general factor (G-factor, all 15 items) and three specific residual factors of Intrusion (items 1, 4, 5, 8, 10, 11, 14), Avoidance (items 2, 3, 7, 8, 9, 12, 13, 15), and Sleep Disturbance (items 4, 6). Items 4 and 8 were each included in 2 different factors. The correlations between the factors were found to be highest when the original IES factor scales were used. (Ref: 7) A root mean square error of approximation (RMSEA) of 0.062 was obtained. All factor loadings were significant, with the exception of items 5 and 14 in the Intrusion factor and item 12 in the Avoidance factor. Determinacy coefficients were 0.94 for the General factor, 0.80 for the Intrusion factor, and 0.63 for a Sleep Disturbance factor. (Ref: 6)

A confirmatory factor analysis using three estimated models found that the third model afforded the best fit and explanation of the data (goodness-of-fit-index=0.97, adjusted goodness-of- fit-index=0.96, normed fit index=0.94, comparative fit index=0.96). Model 1 consisted of a single factor model. Model 2 consisted of 2 correlated factors in which the Intrusion items (1,4,5,6,10,11,12) loaded on the first factor and the Avoidance items (2,3,7,8,9,12,13,15) loaded on the second factor. Model 3 had the same specifications as Model 2 but included additional parameters, and items 2 and 12 were allowed to load on both factors (Ref: 7). This model used the same two-factor structure as the original IES, but allowed items 2 and 12 to load on both the Intrusion and Avoidance factors. All loadings on Model 3 were statistically significant (p<0.05). The factors correlated at 0.68. (Ref: 7) A confirmatory factor analysis using a 4-factor model including Intrusion, Sleep Disorder, Avoidance, and Numb showed adjusted goodness-of-fit of 0.92. (Ref: 8)

Confirmatory factor analysis was also performed using a sample of male combat veterans with chronic PTSD. In this group, a four-factor model was the best fit: Intrusion (items 1, 5, 10, 11, 14; alpha=0.79), Effortful Avoidance (items 2, 3, 7, 13; alpha=0.60), Sleep Disturbance (items 4, 6; alpha=0.77), and Emotional Numbing (items 8, 12, 15; alpha=0.49). Internal consistency for the entire scale was alpha=0.75. (Ref: 8)

No information found.

Test-retest reliability with a one-week interval was 0.87 for the total scale, 0.89 for the Intrusion subscale, and 0.79 for the Avoidance subscale. (Ref: 1)

Not applicable (self-report measure).

Internal consistency of 0.75, measured by Cronbach’s alpha, was found for the entire scale, in a sample of 195 male combat veterans (Ref: 7). Internal consistency coefficients, measured by Cronbach’s alpha, ranged from 0.79 to 0.90 for evaluations at 2 months, 7 months, and 13 months after the death of a loved one (Ref: 2). Cronbach’s alpha was 0.78 for the Intrusion subscale and 0.82 for the Avoidance subscale. The Intrusion and Avoidance subscales correlated at 0.42 (p>0.0002). (Ref: 1) Across two groups (patient or field subject) and three assessments, Cronbach’s alpha ranged from 0.79 to 0.92. Cronbach’s alpha was 0.86 to 0.90 across the pooled sample. (Ref: 2) Cronbach’s alpha in a group of Vietnam veterans was 0.93. (Ref: 4) In malignant melanoma patients, alpha was 0.88. (Ref: 5) Further, among a group of WWII and Korean War veterans, Cronbach’s alpha was 0.91 for the total IES, 0.87 for Intrusion, and 0.90 for Avoidance. (Ref: 7) A literature review of 23 studies found a mean Cronbach’s alpha of 0.86 for Intrusion and 0.82 for avoidance. (Ref: 9)

The split-half reliability of the total scale is 0.86 (Ref: 1)

The IES correlated with the PTSD MMPI scale at 0.79 and the SCL-90 at 0.78. (Ref: 3) The IES correlated significantly with the Hospital Anxiety and Depression Scale (HADS) anxiety subscale at r=0.33 (p<0.01), the GHQ anxiety subscale at r=0.37 (p<0.01), while the GHQ social dysfunction subscale at r=0.30 (p<0.01) in a group of malignant melanoma patients. (Ref: 5) The IES correlated with the total GHQ-28 scale at 0.55 (p=0.001). (Ref: 6 In a 4-factor model that included Intrusion, Sleep Disturbance, Avoidance, and Numb subscales, the Intrusion subscale had moderately large positive correlations with the Mississippi and MMPI PTSD scales, indicating convergent validity. (Ref: 7) In the 4-factor IES model, all of the correlations among the 4 IES subscales are small to moderate, indicating discriminant validity. (Ref: 8) In the 2-factor IES model, the correlation between Intrusion and Avoidance is moderate (r=0.63). (Ref: 9). The correlations between the IES Intrusion and Avoidance subscales and PTSD as diagnosed using six different instruments (Structured Clinical Interview for DSM-III-Revised, Mississippi Scale for Combat-Related PTSD, Structured Clinical Interview (CAPS-1), Intensity Score, PTSD-Minnesota Multiphasic Personality Inventory, and PTSD Inventory) range from 0.29-0.80. These results indicate that the relationships are moderate, and that the IES Intrusion and Avoidance subscales capture information that is not contributed by the other measures. (Ref: 9)

The IES correlated with the CAPS-1 endorsed symptoms score at 0.81 and the CAPS-1 intensity score at 0.78. (Ref: 3) To demonstrate sensitivity of distinguishing between groups, the IES was given to 110 medical students after their first cadaver dissection and compared to a patient group. There were significant differences between groups, with patients showing higher levels of distress as expected (Intrusion: F=212.1, p<0.0001; Avoidance: F=73.0, p<0.001; total: F=170.8, p<0.0001). (Ref: 1) A literature review, which assessed the correlation of the IES subscales with PTSD diagnosis in three studies, showed an average correlation of 0.68 for Intrusion and 0.59 for Avoidance. (Ref: 9) The 4-factor model of the IES had moderately positive correlations with the Dissociative Experiences Survey, indicating concurrent validity (Ref: 8). Concurrent validity is also shown by a small but significant correlation with a measure of combat exposure. (Ref: 7).

Content was based on evaluation and interviews with patients. Items were derived from the most frequent statements used to describe distress after life events. The IES was pilot-tested with psychotherapy patients with stress response syndromes and nonpatient volunteers who experienced severe life events. (Ref: 1) In a review of 12 research studies using the IES, Sundin and Horowitz found that the Intrusion and Avoidance subscales appear to be only moderately correlated with each other; the mean correlation of the subscales across the 12 studies was 0.63. (Ref: 9)

Patients completed the IES immediately before and after a brief stress response therapy, with a mean period of 11 weeks in between. IES scores agreed with clinical assessments indicating that 80% of patients achieved improvement at beneficial levels. (Ref: 1) Responsiveness was tested in 2 samples (1) patients undergoing therapy) and (2) field subjects, over three assessment periods. Patients showed significant improvement from Time 1 to Time 2 (from entry into the study and initiation of time-limited dynamic psychotherapy to 4 months after termination of the therapy), which plateaud from Time 2 to Time 3 (from 4 months after termination of the therapy to 12 months after termination of the therapy) (p<0.001). Field subjects showed significant improvement from Time 1 to Time 3 (p<0.025). (Ref: 2). The IES appears to be a sensitive measure of change in clinical status over time.

Yes. (Ref: 3-4,6-8)

Yes. (Ref: 1-3,5-6).