Psychometric Properties of the Indonesian Version of Impact of Event Scale-Revised Questionnaire During COVID-19 Pandemic

Citation : Wijovi Felix, Kurniawan Andree, Hardjo Lugito Nata Pratama, Handy Agung Fransisca, Alfa Cipta Darien, Angelina Stella, Halim Devina Adella, Jodhinata Claudia, Orlin Sisilia, Hamdoyo Audrey, Evangelista Nadya Nathalia. Psychometric Properties of the Indonesian Version of Impact of Event Scale-Revised Questionnaire During COVID-19 Pandemic Medicinus. 2021 June; 9(2): 55-67


Introduction
Coronavirus Disease 2019 (COVID- 19), previously known as 2019 novel coronavirus is a respiratory infection that caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS CoV-2). 1,2 In late-December 2019, the first case of Coronavirus Disease 2019 (COVID- 19) was confirmed in Wuhan, China. 3 On 30 January 2020, the World Health Organization declares COVID-19 to be a public health emergency of international concern. 4 Since then, the number of cases continues to increase globally, including Indonesia and thus declared as a pandemic on 11 March 2020. 5 Based on the World Health Organization (WHO) data on 7 th August 2021, there were 3,607,863 cases, and 2.8% case fatality rate across the country. 6 The first case of COVID-19 in Indonesia was confirmed in 2 March 2020. 7 The physical distancing and local quarantine were commenced on 10 April 2020 in the capital city of Jakarta and followed by other provinces in Indonesia. 8 COVID-19 affects many aspects, including daily activities, healthcare sector, [9][10][11][12][13][14][15][16] and the most vital one, the country's economy. 17 In addition, due to the surge of COVID-19 case in Indonesia in early July, further measure called Community Activities Restriction Enforcement or Pemberlakuan Pembatasan Kegiatan Masyarakat di Indonessia is taken by the government by further limiting more activities in the society. This event later affected the mid-low economic status which relied on daily income for living.
Several policies were issued to break the transmission chain of COVID-19 which also lead to alteration of daily routine most people during COVID-19 pandemics. These include policy of physical distancing, selfisolation, local quarantine or the transition, tax relief, social assistance, business closure, safety protocol in public, postponed holiday, and worship regulation. 18 According to World health organization (WHO), continuous changes might affect mental health as social determinant is critical influence for an individual. This determinant however very directly corresponding with policy including income, education, and services. 19 A disaster (including health-related disaster) may result in a wide range of mental and physical consequences. 20 Based on DSM-IV, Post-Traumatic Stress Disorder (PTSD) characterized by 3 different cluster of symptoms, which includes re-experiencing, avoidance, and hyperarousal. 21 The study following MERS [22][23][24] and SARS 25-27 pandemic showed the occurrence of Post-Traumatic Stress Symptoms (PTSS) within health workers, quarantined patients in a hospital, and the society. One of the study during MERS outbreak in Korea, showed the occurrence of PTSS in 40% of the sample. 22 Thus, it is really important for early identification to prevent further mental health alterations, especially during quarantine time. 28,29 The growing worries and threats, especially in this COVID-19 situation, have heightened the awareness of disaster as a potentially important determinant of population mental health to highlight areas that need additional study in Indonesian population.
There were several tools available for PTSD screening using from an interview and selfreported instrument. 30-39 Impact of Event Scale-Revised (IES-R) is a self-reported instrument and it corresponds to DSM-IV symptoms of PTSD. 40 This instrument is widely used for PTSS screening in a lot of mental health study in similar setting during previous SARS and MERS outbreak. It has been translated as well as validated into different languages, including Malay, 41 Chinese, 42 Persian, 43 Japanese, 44 Swedish, 45 French, 46 and Korean. 47 The Malay versions of the IES-Rm has the internal consistency reliability ranged from 0.605 to 0.845 with internal validity ranged from 0.71-0.73. 41 There was scarce data about the validation of the Indonesian version of IES-R. Thus, the purpose of this study was to validate the IES-R in Indonesian adult population.

Sample and Procedure
This was a validity and reliability test of the Indonesian version of IES-R. Initially, the questionnaire was translated from English to Indonesian by two independent translators and afterwards, it was back translated to English by another two independent translator for health professional review. [48][49][50] The data of this review were collected within the restriction period in Indonesia from April 21 st to May 10 th , 2020. The survey was distributed via online questionnaire to Indonesian adults across the country. All health workers were excluded from this study to avoid biased results stressful conditions during the COVID-19 pandemic. 51,52 The survey was given in the Indonesian language with a total of 4 sections: informed consent, demographic data, and IES-R questionnaire in the Indonesian language. A total of 234 Indonesian-speaking adults completed the survey. Furthermore, we obtained estatement of informed consent from all participants.

Indonesian
Version of the IES-R compromises 22 items that measure the subscales such as intrusion symptoms (dreams about the event), avoidance symptoms (effort to avoid reminders of the event), and hyperarousal symptoms (feeling watchful and on guard) concerning a particular life-threatening event for PTSD screening. Participants rated on a 5-point linear scale to show their experiences during the preceding 7 days. The total score on the IES-R ranges between 0 and 88. Selected items were totalled to create the 3 subscales such as intrusion, avoidance, and hyperarousal which correspond with PTSD criteria in DSM-IV. To fulfil the purpose of this review, samples were requested to complete the Indonesian translation of IES-R concerning the COVID-19 pandemic in Indonesia. Appendix A provides the Indonesian version of IES-R and appendix B provides the English version of IESR.

Statistical analysis for Validity and Reliability
To determine its' construct validity, the existing data were checked for its distribution using the 1-sample K-S test. Pearson correlation was done using bivariate correlation to show correlation between each subscale in the questionnaire. r-value of 0-0.25; 0.26-0.5; 0.51-0.75; 0.76-1 were classified as not correlated; weakly correlated; moderately correlated; strongly correlated; perfectly correlated respectively. Data with abnormal distribution and had r-value 0.80 were excluded from the analysis. Measures of Sampling Adequacy (MSA) and the Bartlett test of sphericity were done to classify variables with a strong correlation between items using data reduction factor analysis which included KMO, Barlett's and Antiimage tests to calculate MSA of each item. Items with MSA < 0.05 were excluded from the analysis process. The next process was extraction by a principal component method and Cattell's scree test. In this step, items with eigenvalue  1 were extracted and proceed to the component matrix analysis to show where each item included in the three subscales, marked by loading factor  0.5. Rotation analysis was used to reassure the result. 53,54 Cronbach's alpha coefficients were calculated using the reliability analysis to measure the internal reliability for each subscale. The cut-off of high internal reliability was 0.7. 55,56

Description of the sample
The total of 234 Indonesian adults from all provinces across the country had completed the survey, 179 (60.9%) were men and 115 (39.1%) were women. Mean age of the samples was 37.19 ± 11.284 years old. In addition, the mean score of intrusion, avoidance, and hyperarousal subscales were 10.432 ± 6.75, 11.401 ± 6.12, and 7.874 ± 4.65 respectively. The mean score of depression, anxiety, and stress subscales were 11.03 ± 4.071, 10.55 ± 3.63, and 11.46 ± 4.314 respectively. The other demographics data of the samples are described in Table 1.

Data Distribution
Collected data were analysed before continuing to the next analysing process. All sub-scales data were distributed normally (p > 0.05).

Internal Validity and Reliability
During the process of translating back from Indonesian to English, no significant changes were found between the initial text and the translated text. Pearson correlations were done between the subscales and the total score was high and significant (p = 0.000) (

Principal Component Analysis of the IES-R
To assess the construct validity of the Indonesian version IES-R questionnaire, a principal component analysis (PCA) was done on the 22 items of the questionnaire. It was found that 3 components had eigenvalues > 1.0. Cattell's scree test was performed to determine the number of components to be extracted. Furthermore, a PCA using an orthogonal-varimax rotation was then performed (Table 3). Factor loading of 0.50 was considered as significant. The solution which explained 55.5 % of the variance, generated a hyperarousal subscale (items 4, 10,15,18,19,21), avoidance subscale (items 5,7,8,11,12,13,17), and intrusion subscale (items 1,2, 3,6,9,14,16). Factor items 20 and 22 did not load on any of the three factors and thus excluded from the questionnaire. Therefore, a PCA was performed without these items and increasing the variance explained to 59%. The final Indonesian version of IES-R is attached in appendix A.

Discussion
This study assessed the internal consistency, and construct validity of an Indonesian translation of the IES-R questionnaire in a sample of Indonesian adults in the COVID-19 pandemic situation. The result of the Indonesian translation of the IES-R was remarkable with good internal consistency and Cronbach's alpha ranging from 0.858 to 0.907. The test-retest data were not available for this study.
Three factors solutions were accepted for the IES-R in this study, which explains 55.5% of the total variance. Item 20 ("I had dreams about it") and item 22 ("I tried not to talk about it") did not load on any of the three factors of this study. In this study, we found six items in hyperarousal subscale, seven items in avoidance subscale, and seven items in intrusion subscale. In comparison with the theoretical eight items in intrusion subscale, eight items in avoidance subscale, and six items in arousal subscale. Other items loaded on the same factors in the theoretical model.
A similar study has been conducted in other countries and showed that IES-R is a reliable questionnaire and validly translated into those languages. [41][42][43][44][45][46][47] The Malay version of IES-R by Norhayati and Aniza. showed satisfactory results. 41 It has achieved content validity through the translation process. The confirmatory factor analysis showed a good fit and a good convergent validity, discriminant validity, internal reliability, and construct reliability. In comparison to Norhayati and Aniza study, this Indonesian version of IES-R shows similar result. The Indonesian IES-R has proven to be valid and reliable through internal consistency test and has good convergent validity. The item 20 and 22 were removed from the final questionnaire since they did not load any subscales.
The advantage of this study was it assesses principal component analysis for the Indonesian version of IES-R which was the first in Indonesia. This study also excluded medical personnel samples to avoid bias due to stressful conditions during the COVID-19 pandemic. 52 On the other hand, the limitation of this study was a small number of samples in this study. Since it was a cross-sectional study, we were unable to examine other important psychometric properties such as reassurance of reliability or sensitivity which could change over time. Thus, a study in the other population with a higher number of populations was recommended to confirm the structure and testing its invariance across samples which might be due to data retrieval methods using online questionnaire and were not easily accessible to all social level in the community. This might add further evidence to support the Indonesian version of IES-R items.
In conclusion, the Indonesian IES-R had proven to be a valid and reliable tool to measure post-traumatic stress disorder in the sample of Indonesian adults during COVID-19 pandemic. The translation and validation of the IES-R into the Indonesian language filled the important gap in healthcare's ability to screen for PTSD symptoms among Indonesian populations. Also, this study provides a principal component analysis of the IES-R Indonesian Version.