The Psychological Impact of the Ebola epidemic among Survivors in Liberia: a retrospective cohort study

Background The 2013- 2016 West Africa Ebola epidemic resulted in an unprecedented number of survivors. The outbreak has contributed to many survivors, who have shown to have suffered from physical and mental sequelae – sigma, physical violence, diminished quality of life, ostracism, and isolation. The goal of this study was to evaluate the relationship between Ebola virus disease and long-term psychological problems among Ebola survivors in Liberia. Method In January 2019, quantitative data from 232 respondents (EVD survivors, n = 116; non-survivors, n = 116) were gathered by structured face-to-face interviews. We used Davidson Trauma Scale for PTSD (DTS), the Generalized Anxiety Disorder Scale (GAD-7) to assess anxiety, the Patient Health Questionnaire (PHQ-9) to assess depression, and the Suicide Behavior Questionnaire-Revised (SBQ-R) to assess suicidality of both the survivors and their controls. Data were analyzed by linear regression analyses, chi-square, or t-tests to determine the prevalence, Ebola virus infection, and factors associated with long term psychological problems among EVD survivors and non-survivors. Results We have identied 116 survivors of EVD and 116 non-Ebola survivors. Of the 116 people who survived, 76 (66%) recorded posttraumatic stress disorder, 61 (53%) depression, 43 (37%) anxiety, and 39 (34%) attempted suicide. Relative to EVD survivors, the non-survivors had very signicantly lower levels of posttraumatic stress disorder, 7(6%), depression, 5 (4%), anxiety, 5 (4%), and 2 (2%) attempted suicide reported three years after the outbreak. In a logistic regression analysis, EVD survivors had higher odds of observing posttraumatic stress disorder ( β =3.39, OR, 29.59 [95% Cl 2.53, 4.24]; p<.001), depression ( β = 3.20 OR, 24.62 [95% Cl 2.24, 4.17]; p <.001); anxiety ( β = 2.57, OR, 13.08 [95% Cl 1.60, 3.54]; p <.001), and suicidality (β = 3.36, OR, 28.87

exclusion, and isolation [3]. It is known that the prejudice and discrimination encounter by the survivors receive less attention [3].
As in most epidemics, Ebola virus disease control strategies are directed at managing the transmission, but issues related to the disease's mental/psychological impact on patients have gained less focus [3]. The EVD has resulted in many victims of Ebola who need psychological support. The EVD-related stigma impacts the survivors ' social and psychological aspects. A stigma has been described in several ways, but with regard to EVD, it refers to negative actions, attitudes, and activities aimed at people living with and accused of being infected [3]. A research study conducted among survivors of the 1995 Ebola epidemic in Kikwit, Democratic Republic of Congo, described the feelings and experiences of survivors of the Ebola epidemic [4]. They described survivors ' psychological consequences to include fear of serious illness, denial, fear of being blamed by neighbors, and guilt. Others included public condemnation, thinking the disease was a divine punishment, income loss, and deep sorrow for colleagues who did not survive the outbreak [4,5].
The Ebola outbreak's psychosocial consequences mirror those of other disasters, particularly epidemics, where populations and health workers were subjected to infection and psychological stressors, exacerbating exiting challenges [6][7][8][9][10][11]. The previous study that evaluated the psychological impact and depressive symptoms of the 2003 epidemic of the severe acute respiratory syndrome on hospital employees showed that about 10% of the participants had a high level of depressive symptoms and three years later had a high level of depressive symptoms [8,11]. SARS patients still had high levels of anxiety and alarming levels of psychological distress one year after the epidemic [7].
Survivors have also reported the mental health effects of other disasters, such as bioterrorism [12].
Psychological distress, fear, depression, and post-traumatic stress disorder (PTSD) have been recorded in communities subject to mass violence and relocation, including those affected by Sierra Leone and Liberia civil con ict between 1992 and 2002 [13].
Given the magnitude of the epidemic, adverse mental health outcomes could be expected in the general population. Among Ebola survivors in Guinea, Sierra Leone, and Liberia [14,15], and health workers in the three countries affected, high levels of psychological distress have been documented [16]. The risk of long-term psychological problems associated with Ebola virus disease has not been assessed in Liberia; moreover, most studies have excluded the general population despite people have been faced with civil wars, constant human rights violations and have already dealt with numerous other deadly diseases such as malaria, diarrhea, and HIV / AIDS [17].
Our primary goal was to investigate the association between the Ebola virus infection and long-term psychological problems among survivors of the Ebola Virus Disease in Liberia. Our secondary goal was to describe the types and prevalence of psychological problems experienced by survivors three years after the outbreak. We hypothesized that the Ebola virus is a risk factor for long-term psychological problems.
If true, this would suggest the need to develop and reinforce national diagnostic and care systems for mental health in the post-epidemic context.

Study design
We did a retrospective cohort study of EVD survivors of the 2013-2016 Ebola virus disease outbreak in Liberia. The study ran from January to March 2019. Our main goal was to evaluate the association between the Ebola virus infection and long-term psychological problems among survivors of the Ebola Virus Disease in Liberia. The secondary goal was to describe the types and prevalence of psychological problems experienced by survivors three years after the outbreak. The study was conducted in Monrovia, the capital city of Liberia, with a population of 1,010,970 inhabitants as of 2008 National Census.
The sample size of the cohort of 232 participants included in this study have been included in previous research studies: A sample of 150 participants in the Ebola Treatment Unit survivors study [18]; 112 survivors in a study of persistent viral excretion [19]; 151 survivors in a study of long term mortality [20], and 123 survivors in a series of household transmission studies [21][22][23].

Participants
The study sample was EVD survivors recorded in Monrovia from the beginning of the epidemic, May 2014 -September 3, 2015. All quali ed survivors ' names and contacts were identi ed from the database of Ebola virus disease handled by the Liberian Ministry of Health and collaborators. This database contains complete epidemiological information available to the Ministry of Health and information on the disease of all suspected, probable, and con rmed cases of Ebola virus disease in Liberia. Inclusion for cases selection were as follows: (1) Persons con rmed as a case of EVD in the present outbreak response and had been managed, cured, and discharged by the case management team (2) EVD survivors ≥ 18 years of age (3) EVD survivors should be a resident of Monrovia. Exclusion criteria included the following: (1) EVD survivors without an original discharge certi cate or copy. (2) EVD survivors with no membership card of the Survivors Network, Liberia. (3) EVD survivors refuse consent from being part of the study. Individuals who had no prior history of Ebola infection and unrelated to cases were included as controls. The cases and controls were matched 1:1 by age (± 5yrs) and sex.

Procedure
We randomly selected the cases using the contact tracing list of all survivors living in Monrovia from the National Ebola Survivor Network and randomized controls from the 2008 National Census of Liberia.
Participants were tracked using their telephone number. A team of 10 local health volunteers carried out a structured face-to-face interview with the participants. The interviewers randomly selected intersections; from there, they selected the households according to the protocol de ned by Hoffmeyer-Zlotnik [24].
Each interview took about an hour. Only a person per household was interviewed. The interviewers attended a three-day intensive training program prior to the data collection, during which they learned about the methodology of the study, its objectives, and the conceptual background of the study questionnaire. Also, they learned how to ll the questionnaire and practiced the translation of questions into local languages. The study procedure and instruments were approved by the Xiangya School of Public Health Institutional Review Board and obtained approval from the Mental Health Bureau, Ministry of Health -Liberia. The meeting was conducted with the Ebola Survivor Network to explain the objectives of the study before the participation of the survivors. After signing the consent form by each participant to ensure the willingness, privacy, and con dentiality of information, after which, each participant was asked to complete a face-to-face interview.

Questionnaire and measures
Data was collected through interviews with participants using a socio-demographic questionnaire that was designed to collect information on the respondents' age, gender, marital status, employment status, religion, and educational level. A semi-structured self-assessment questionnaire was comprising of Davidson's Trauma Scale, GAD-7, PHQ-9, and SBQ-R to collect information on the respondent's psychological health.
Davidson Trauma Scale -is designed to evaluate symptoms of PTSD in individuals with a history of trauma [25]. The17 item scale measuring each DSM-IV symptom of PTSD on a 5-point frequency and severity scale [25]. The DTS consists of 17 items that relate to each of the 17 symptoms of DSM-IV. For each item, the subject rates both frequency and severity during the previous week on a 5-point (0 to 4) scale for a total possible score of 136 points [25]. A participant with a DTS score of 40 or more was considered to have PTSD [25].
Generalized Anxiety Disorder (GAD-7) -used for screening and severity measuring of a generalized anxiety disorder (GAD) [26]. GAD-7 has seven items, which check measure the severity of various signs of GAD according to reported response categories with assigned points. The evaluation is shown by the total score, which consists of adding the scores for the scale of all seven items [27]. The threshold selected as an indicator of anxiety was 11 or more [26,27].
The Patient Health Questionnaire (PHQ-9) is a self-administered version of the PRIME-MD diagnostic instrument for common mental disorders. The PHQ-9 is the depression test, scoring each of the nine DSM-IV criteria from "0" (not at all) to "3" (nearly every day) [28]; PHQ-9 is a reliable and valid measure of depression severity, that is, a score of 11 or more was used; [28] higher scores correspond to more severe depression.
The Suicide Behavior Questionnaire-Revised (SBQ-R) -a validated questionnaire to identify patients at risk of attempted suicide. [29] In this research, we use a cutoff point of 8 or more to be considered a high symptom of suicide attempt [29].
It is established that study participants are different from the general population. The self-selection bias will exist. A randomized sampling was conducted to eliminate selection bias. Those who were not involved in the study may have quite different results; therefore, the generalizability of the study is limited.

Statistical analysis
Data were entered, cleaned, coded, and analyzed with SPSS version 25. Descriptive statistics explained the characteristics of the study participants and presented continuous variables as mean and range.
Categorical variables were presented in tables showing frequencies and percent distributions. Characteristics among EVD survivors and non-survivors were assessed for differences using Fischer exact or t-tests. We assessed the association between EVD survivorship (EVD survivors vs. non-survivors) and health outcomes and used linear regression models in bivariate and multivariate analyses adjusting for potential confounders of age, sex, educational level, marital status, employment status, region and being a health care worker. We calculated ORs with 95% con dence intervals for association with Ebola virus disease and PTSD, depression, anxiety, and suicidality. Our threshold for the statistically signi cant nding was p < 0.05. The Chi-square test was also used to compare the cases and controls ' categorical demographic characteristics. There were no missing data.

Results
One hundred twenty-ve EVD survivors were identi ed. One twenty (96%) agreed to participate, were deemed eligible, and enrolled. Five survivors (4%) refused or were unreachable, ineligible after screening, or incapacitated. Among these 120 survivors, four (4) were disquali ed because the interviews were not completed. Compared to the 116 non-survivors (controls) enrolled. No signi cant difference was seen between the cases and controls enrolled (p > 0.05).

Characteristics of the sample
One hundred EVD survivors and 116 non-survivors were enrolled. Among these 116 EVD survivors, 66 (57%) were women, and 50 (43) were men. Age range from 18-60 years, whereas 24-29 years reported the most frequently observed. The majority, 74 (64%), were single, and 13 (11%) were health care workers during the outbreak. Christianity was the largest religion, 89 (77%). Comparing these characteristics among EVD survivors and non-survivors, there was no signi cant difference for age, gender, marital status, religion, and being a health care worker (Table 1). However, there was a trend of signi cant differences in educational level and employment status ( Table 1). Most of the survivors had an educational level of secondary education, 37 (32%) or above, and were unemployed, 39 (34%). About 66% (n = 76) of the EVD survivors reported having high levels of PTSD symptoms (that is, a DTS score of 40 or more) at some time during the three years following their exposure to the Ebola outbreak.
The DTS scores in this sample ranged from 0-17, with a mean of 38.83, and the standard deviation was 14.68. It was found that among the EVD survivors, 61 (53%) had high symptoms of depression (that is a score of 11 or more). This scale had a range from 0-9; the mean was 9.76, with 5.89 standard deviations. It was further observed among the EVD survivors, nearly half had anxiety 43 (37%) (that is, a GAD-7 score of 11 or more), and score ranged from 0-7 (mean was 7.67 and Standard Deviation was 5.41). Also, the suicide attempt was 39 (34%) reported among the survivors with SBQ-R scores ranged from 0-4 with a mean of 6.30, and the standard deviation was 4.02. In this research, we use a cutoff point of 8 or more to be considered a high symptom level of a suicide attempt. In comparison, the nonsurvivors, 7 (6%) had low posttraumatic stress disorder symptoms, 5 (4%) reported to have depressive symptoms, and 5 (4%) had anxiety with just 2 (2%) had suicide attempts. Tables 2 and 3 summarize the results.  to 3.42, p < 0.001) more likely to have high PTSD symptom level. The overall model was signi cantly based on alpha 0.05. For depression, anxiety, suicidality, the overall model (multinomial regression analysis) was not signi cant; therefore, we did not observe a statistically signi cant difference in depressive, anxiety, and suicidal symptom levels with sociodemographic factors.   [4,6,9,14,30]. However, little is known, however, about the long term mental health implications of the EVD exposure. Our study is the rst to investigate the psychological problem in Ebola survivors three years after the outbreak to address the question of whether exposure to EVD a risk factor for long term psychological problems in survivors.
Earlier studies on catastrophe survivors have shown that about one year after more than three-fourths of those with psychological problems shortly after the tragedy still have it [31,32]. This study provides data on the psychological problems in EVD survivors, showing that they have higher odds of developing PTSD, depression, anxiety, and suicidality three years after recovery in comparison to non-survivors. Our results show a prevalence of PTSD (66%), depression (53%), anxiety (37%), and suicidality (34%) among EVD survivors three years after the outbreak. Such ndings contribute to our understanding of Ebola survivors' experience of persistent psychological symptoms. It is established that psychological symptoms that persist after an incident for more than six months are likely to persist over the long term [33,34].
Our research ndings indicate that most participants scored above the critical threshold with DTS, GAD-7, PHQ-7, and SBQ-R scores. In addition, the mean difference of depression, PTSD, anxiety, and suicidality scores were signi cantly higher in EVD survivors as compared to controls. The psychological consequences of EVD are often ignored in the acute setting but would be likely to persist into convalescing and may compound physical disabilities.
Earlier studies identi ed speci c instances of major depression with suicide attempts after the EVD outbreak [35,36]. Due to the shortage of therapists and social workers in the region, most survivors are still unable to receive adequate psychiatric care [17]. In order to offer medical and psychological services and the provision of appropriate medications, it is vital to have adequate skilled health professionals available. Currently, this is a signi cant limitation in Liberia. The availability of psychiatrists is rare in Liberia with just one psychiatrist [17]. Counseling of the Ebola survivors was performed by aid organizations and was done slowly [17].
Females account for most of our study participants. Other research studies have also reported more female survivors [37,38]. It is due to the proportion of females admitted was more than males.
As in many traumatic events [6,9], our study shows that psychological problems among Ebola survivors are common. Persons that have experienced severe trauma are at risk of increased psychological disorders [39]. The 1995 Kikwit, DR Congo Ebola outbreak, reported that 61% of the study participants presented issues related to mental/psychological consequences from the disease [4]. In Sierra Leone, a study of 74 survivors in Moyamba District, 2015 revealed that 48% of survivors show symptoms of psychological problems a few weeks after discharge from the ETC, placing them at risk of developing diverse levels of psychological disorders [40]. Similarly, 81 participants of another study in Kenema District, Sierra Leone, recorded around 35% had depression four months after discharge from the Ebola Treatment Unit [41]. Whereas in Liberia, the initial results of the ongoing research on neurological sequelae undertaken by Prevail III project showed a 49% prevalence of depression, for a cohort of 82 persons released from the treatment facility about six months earlier [36]. Another study performed by a Médecins Sans Frontières (MSF) team in Monrovia, reports a 40% prevalence of depression rate for a cohort of 136 persons discharged out of an Ebola Treatment Center for more than three months earlier and 12% for major PTSD-related depression [42]. The survivors of EVD are plagued by disease-related traumatic memories and associated death of relatives [43]. Our study also showed the survivors are struggling with extreme anxiety. This is due to fear of discrimination, as described by Reardon S. [43].
Contrary to assumptions, after three years of post-outbreak, psychological health levels did not subside, and depressed, nervous, suicidal, and posttraumatic symptoms are prevalent. It appears that instead of improving over time, EVD survivors' psychological health in terms of PTSD, depression, anxiety, and suicidal symptoms is elevated; this may have been attributed to civil wars, constant human rights violations, economic hardships and ght against deadly diseases such as malaria, HIV/AIDS and diarrhea [44]. Another logical rationale was that many of the volunteers recruited by aid programs were students, social workers, and others without specialized psychology experience or advanced psychological counseling expertise. The following years also saw the rise of several economic hardships, unemployment, social shifts, health inequality, education, making it di cult for survivors and supporters to engage in long-term psychotherapy. Therefore, survivors who only once received the support for mental health remembered painful thoughts and re-experienced horri c scenarios without adequate assistance to relieve and resolve the negative feelings that may have contributed to secondary psychological issues.
A signi cant contribution to the prevention and treatment of EVD survivors is the availability of longlasting and effective mental health support after an epidemic.
The study revealed a signi cant association between EVD and later diagnosis of PTSD, depression, anxiety, and suicidality. The other relationship studied included marital status, age, region, education, gender, being a health care worker. In the secondary analysis, the variables of signi cance for the association predicting PTSD were marital status (separated), ages 30-34, and 35-39, low educational level, and religion -Christian. Our ndings showed that the higher the level of education received, the less the psychological symptoms. It is necessary to combine local knowledge and understanding of the illness with a psychosocial approach, so, to help decrease the mental burdens surrounding the EVD epidemic. Training sessions can be structured classes regarding constructive mental health promotion or informal conversations between survivors and community members, peer support, education, and mental care services can be some methods shaping EVD survivors' policy initiatives. Training sessions can be structured classes regarding constructive mental health promotion or informal conversations between survivors and community members, peer support, education, and mental care services can be some methods shaping EVD survivors' policy initiatives [45]. The integration of local knowledge and understanding of the disease with psychosocial intervention is vital. Also, a marriage spilled is associated with poor mental health. A possible reason is due to the changes that occur after separation such as, poor sleep and appetite. The determinants of PTSD among Wenchuan earthquake survivors were also discussed by some recent articles [46][47][48]. During the post-earthquake period of the following the earthquake, Guo J. et al. [47] demonstrated that female gender, married, and low educational status were signi cantly associated with PTSD; and depression was signi cantly associated with PTSD 44 months after the earthquake. Additionally, Zhou X. et al. [48] further discovered that risk factors for PTSD were old age, female gender, and lonely living. While some of the causes as mentioned above are factors similar to our results, it is not fair to equate our ndings with those of the above articles is not rational since our study was conducted three years after the outbreak and this is an epidemic of Ebola rather than an earthquake; all the above papers are cross-sectional studies, whereas this study was a retrospective cohort study.
Our research ndings and that of other published reports have shown that many of the survivors have experienced psychological symptoms during and after the outbreak. It is, therefore, evident that a sensitive psychological trauma-informed care model is needed to meet the needs of the Ebola survivors, thus maximizing their mental and psychological health outcomes [49]. Such an approach enables health care providers, social workers, and policymakers to become aware of trauma, enhance screening and evaluation procedures, and implement evidence-based interventions that are tailored to the needs of EVD survivors [49].
We should also recognize existing efforts addressing the outbreak's psychological effects, as evidenced by the award of approximately US$3 million in psychological assistance from the World Bank and Japan to address the necessities of physical and psychological and social care of EVD survivors [43]. Nonetheless, signi cant gaps exist in resolving the survivors' psychological care, as shown by the ndings of our study. Mental Health care has been inadequate due to the absence of proper planning, lack of resources, staff shortages, and weak health systems, and insu cient knowledge of policymakers that support mental health needs [43].
A signi cant limitation of the research is exploring psychological symptoms linked to different types of Ebola virus disease-related event exposures. Some other limitations existed in the interview and data collection process before analysis. Most of the survivors (cases) were illiterate; therefore, we have to explain the questions into simple English/ dialect for him/her to understand. Thus, the responses were subject to information bias. The self-reports of the study subjects, describing the effects encountered over the three years, were also subject to recall bias. The nal major limitation in our study was the potential for participation bias. It is possible that, as with any cohort study, either cases or controls were differentially more likely to participate than the other if they somehow included in the results of the research.
Nevertheless, the results provide vital information about the psychological impact of an infectious disease epidemic for policymakers and mental health professionals globally, which can help brace the development of health systems and strategies to respond to possible future outbreaks of Ebola.

Conclusion
The present study show provides insight into the potential long-term adverse psychological effects of infectious diseases. As a group, the EVD survivors showed increased levels of psychological problems after three years, as evidenced by alarmingly high odds of depression, anxiety, suicidal and posttraumatic symptoms, as well as high prevalence of potential psychiatric morbidity cases.
The long-term psychological implications of infectious diseases should not be ignored. The results underscore the need for Mental Health services on a national scale for the EVD survivors.
An intervention in line with the IASC guidelines on mental health and psychosocial support for emergency settings and used with the unique characteristics of the Ebola virus outbreak is essential for the Ministry of Health, Liberia and its partners to implement and maintain rapidly throughout the recovery process. The IASC guidelines are structured around a 4-tiered pyramid of intervention: (1) restoring essential services and protection to the affected population; (2) strengthening family and community networks; (3) providing psychosocial assistance to those in distress; and (4) providing specialized mental health treatment to severely affected survivors.
Given the need for further research with a more extensive study to validate our ndings, we are optimistic that a potential intervention to reduce the rate of PTSD, depression, anxiety, and suicidality will help the survivors of EVD. Poor policy-making will not help; instead, ground-level implementation. Soon, the best solutions for minimizing the mental health problems associated with the EVD will be awareness, prevention, and therapeutic vaccination.