Discussion

We found that bereaved families suffered from more psychological distress than non-bereaved survivors during the acute period of epidemic disasters. About 60% – 80% of bereaved family members experienced sadness/depression and anger within a month of the MERS epidemic; this was higher than bereaved families who had lost a family member in a chemical accident. ^{Reference Yoo, Sim and Choi8} There have been some evidence that funeral ceremonies are helpful in the mourning process. ^{Reference Irion9} During the MERS outbreak, many bereaved families couldn’t meet their deceased family member who was in the isolation room prior to death, and they were unable to hold a formal funeral process to control the MERS spread. Bereaved family members often talked about negative feelings regarding this situation during counseling. As is specific to epidemics, patients and families are not able to prepare for death, nor conduct traditional death-rituals and funerals; this disruption could have interfered with the normal grief processing. ^{Reference Leong, Lee and Ng10}

It is noteworthy that sadness/depression and anger were most common in bereaved-only family members, who did not contract the MERS infection themselves. Since bereaved MERS survivors showed additional concern about their physical recovery and long-term sequelae, we considered the possibility that their grief reactions, such as sadness and anger, might be disturbed or delayed. Instead, bereaved MERS survivors showed prominent avoidance/isolation, especially when their family member was a MERS spreader. Epidemic patients frequently suffer from negative attention from the public and rejection by neighbors and society. ^{Reference De Roo, Ado, Rose, Guimard, Fonck and Colebunders2,Reference Robertson, Hershenfield, Grace and Stewart11,Reference Verma, Mythily, Chan, Deslypere, Teo and Chong12} Considering that bereaved MERS survivors had a greater MERS-related burden within family, their avoidance/isolation might be related to stigmatization.

Over 50% of MERS survivors showed somatic responses, which were 3.29 times higher in women and those with an underlying medical disease. Common somatic responses were pain, especially headache, fatigue, muscle tension, chest pressure/epigastric mass, changed appetite, respiratory difficulty, and gastrointestinal problems. When approaching the somatic response of MERS survivors, it is necessary to consider both psychological problems and physical effects of the disease and its treatment. Interestingly, 42.2% of the bereaved-only group also showed somatic responses. Since they did not have MERS, the effects of contracting the disease and its treatment could be ruled out. Previous studies have demonstrated actual deterioration of health, as well as the experience of sadness and anger, following bereavement. ^{Reference Maddison and Viola13,Reference Marris14} These problems include pain, digestive issues, respiratory problems, cardiovascular problems, and immunological dysfunction, such as skin rashes or infections, which appear during the first 6 months and can last for more than a year after bereavement, sometimes causing chronic health problems. ^{Reference Ott15,Reference Parkes16} Thus, our results implicate that physical health care also is needed among the bereaved family members.

During outbreaks of epidemics, the increased workload on healthcare workers and their constant exposure to the risk of infection can lead to high levels of stress. Previous studies have found that healthcare workers who treated patients during the SARS epidemic exhibited high rates of anxiety, depression, PTSD, and other psychiatric morbidities. ^{Reference Chan and Huak17-Reference Tam, Pang, Lam and Chiu19} However, the current study found that healthcare workers were at a lower risk of anxiety/fear and insomnia compared to non-healthcare workers in the MERS-survivors group. Healthcare workers have accurate medical knowledge of infectious diseases, which could prevent them from experiencing psychological distress. This is consistent with previous findings showing severe PTSD symptoms following disasters in non-professional rescuers with less expertise and experience. ^{Reference Dyregrov, Kristoffersen and Gjestad20-Reference Guo, Chen, Lu, Tan, Lee and Wang22} However, we should not interpret this to mean that healthcare workers are safer from disaster or epidemic-related mental health issues, than general victims. In studies about mental health of healthcare workers in Korean MERS disaster, healthcare workers at hospitals with MERS cases showed higher levels of posttraumatic stress or depressive symptoms, compared to those at hospital without MERS cases. ^{Reference Um, Kim and Lee23,Reference Lee, Kang, Cho, Kim and Park24} As such, many healthcare workers in epidemic situations can suffer from infection anxiety or burnout. In such a situation, clear guidelines for infection prevention and adjusting strict protective measures for healthcare workers help to relieve stress. ^{Reference Khalid, Khalid, Qabajah, Barnard and Qushmaq25} These results highlight the importance of providing accurate information and education not only to healthcare workers, but also to the public for efficient disaster management.

In this study, psychological responses were not associated with the number of MERS patients in the family, living together with a deceased family member, or quarantine experience. Considering the possibility of a heroic and honeymoon phase of disaster, which is a short time during which survivors feel grateful and relief for their survival, a long-term monitoring process is necessary for evaluating the psychological responses of MERS survivors. A third of the bereaved survivors continued to have psychiatric problems 6 years after the Southeast Asian Tsunami ^{Reference Kristensen, Weisaeth, Hussain and Heir26} and 63.5% of MERS patients had more than 1 mental health problem such as post-traumatic stress symptoms, insomnia, or anxiety after 1 year of the outbreak. ^{Reference Shin, Park and Kim27} An additional limitation was that this study was based on a review of initial psychological counseling records, not research data, so several information were omitted and the psychological responses were not assessed by a structured measure. Nevertheless, the data were derived from ongoing statements made by the participants and not later recollections, thus our findings are likely to reflect a realistic representation of the situation at the time of the epidemic.