This is the first study to estimate the prevalence of PGD in ICD-11 and PCBD in DSM-5 and to identify demographic and loss-related factors associated with PGD and PCBD symptoms in a large sample of individuals bereaved due to COVID-19. Using the latest diagnostic criteria of pathological grief in ICD-11 and DSM-5 and the most updated measures for PGD and PCBD, this study demonstrates that over one-third of COVID-19 related bereaved individuals suffered from PGD or PCBD. Factors associated with more severe grief symptoms were losing a close person by COVID-19 rather than related complications, losing a partner, child, parent, grandparent, feeling more traumatic about the loss, being closer to the deceased, and having more conflicts with the deceased.
Prevalence of PGD found in our COVID-19 related bereaved sample (37.8%) is tripled than that in Chinese bereaved people (12.7%) and five times more than that in a German-speaking sample; both studies used the same measure and diagnostic algorithm (27). The number is also higher than the prevalence of PGD in a sample of bereaved earthquake survivors in China (8.47%)(16). Notably, the number is similar to the prevalence in Chinese Shidu parents (i.e., parents who lost their only child; 35.5%)(31), who are undergoing the most painful and traumatic experience than other types of loss and facing physical and psychological challenges (32). Except for the prevalence, the symptom levels of PGD (41.85 ± 9.61) were higher than those reported in the German-speaking sample (29.22 ± 10.83) and the Chinese sample (36.29 ± 11.35).
Rate of PCBD in our study (39.9%) is doubled or tripled than that in two Dutch bereaved patient samples, namely 17.7% (29) and 12.5% (30). The number is even six times more than that in Dutch citizens bereaved due to the Ukrainian plane disaster (6.6%)(30). While the prevalence of PCBD in our sample is higher than in previous studies, the symptom levels measured by the 18-item version of TGI-SR (57.39 ± 13.10) seems equivalent to that reported in a small group of Dutch people bereaved due to COVID-19 (57.37 ± 9.60)(26).
While the time frame of the survey (approximately 9 months since the first COVID-19 case reported in China) could be partially accounted for the relatively high prevalence of PGD and PCBD in the COVID-19 bereaved population, our findings support concerns from researchers that there would be a rise in PGD (8, 9), because acute grief is among the strongest predictors of the development of PGD (33). The finding that no difference of grief symptoms was detected between participants bereaved 6 months ago or less than 6 months echoes in part that acute grief might greatly correlate with prolonged grief. Another interesting point is that over one-third of participants believed their grief reactions went beyond cultural norms, yet only half of them met the symptom and functioning criteria. These findings add to the discussion on the PGD and PCBD criteria in terms of (a) whether the symptom criteria should be more strict for unnatural death as grief reactions for unnatural death might be more severe in nature; (b) whether the time criterion should be shortened for mass bereavement so that those in need could be helped in time; and (c) whether the consensus could be made between professionals from the same culture and context and between the professional and the bereaved regarding the consideration for social, cultural or religious norms (34).
Most of the associated factors identified in this study are consistent with previous research. First, those who grieve the most were the ones who lost a first-degree family member (i.e., partner, parent, child), followed by grandparents, friends, relatives, and other relationships. This pattern strengthens the importance of considering the kinship between the bereaved and the deceased across causes of death (13, 16, 35–42). Noticeably, no difference was found in grief severity among participants who lost a grandparent, relative, friend, and someone with another relationship, suggesting that bereavement care for COVID-19 bereaved individuals should not be restricted to family members, but extending to friends, colleagues, or even acquaintances that are affected by the death. Second, although all participants lost their close ones due to COVID-19, a handful of them endorsed COVID-19 induced complications as the causes of death and they experienced less severe grief symptoms than their peers. Despite the fact that deaths during the pandemic were generally unexpected, since those who died from complications was older [59.14 ± 16.79 versus 47.42 ± 21.61, t(420) = 2.008, p = 0.045] and usually had a history of diseases, the bereaved might be more mentally prepared for the death than others, and preparedness for death served as a protective factor for grief symptoms (43, 44).
We found that in the grieving process during the pandemic, subjective traumatic level of the loss was more crucial than unexpectedness of the death in association to grief symptoms. Compared to the objective traumatic deaths (e.g., violent deaths), subjective experience of death as traumatic was a more significant factor that contributing to grief symptom (18). During the COVID-19 pandemic, because of the nationwide lockdown measures taken to prevent the spread of the virus in China, people were unable to celebrate the Spring Festival by gathering together in late January, hold funerals and farewell rituals for the deceased, offer condolences to bereaved families by physical company, and memorialize decedents by sweeping the tomb on the Tsing Ming Festival in April. Additionally, owning to the contagious nature of the pandemic, family members were not allowed to say goodbye to the dying patient, let alone keeping some meaningful personal belongings of the deceased (45). All these created obstacles to the grieving process, which may result in prolonged grief.
Quality of relationship between the bereaved and the deceased also plays a part concerning grief severity and our findings suggest that assessing the ambivalence in the relationship is necessary as both closeness to and conflicts with the deceased were positively correlated with grief severity. The closer to the deceased before the death, the stronger attachment was established with the deceased. The bond of attachment would continue even when they are separated by death, yet it might imply disbelief that the person is dead and thus lead to unresolved grief (46). Death of an attachment figure presents a “temporarily irreconcilable mismatch between an unrevised mental representation of a loved one and a dramatic change in the ongoing relationship with that person” (p. 454), resulting in acute grief symptoms such as yearning for the deceased, preoccupation of the deceased, and loss of interest in the world (47). Once the mental representation of the deceased was revised by incorporating the reality of death, acute grief symptoms would be resolved (48). However, for individuals who are deeply attached to the deceased, the revision of the mental representation could take a longer time, and thus manifested as prolonged grief symptoms. Another finding we would like to highlight is that more conflicts with the deceased before the death, although at a smaller magnitude than closeness, is related to more severe grief symptom. Clinical observations showed that conflicts existed at the time of death and earlier may lead to pathological mourning as it is accompanied by a strong psychological dependence on the deceased (49), and greater dependency is a risk factor for prolonged grief symptoms (50).
This study has several limitations. First, the sample is recruited by convenience sampling, whose representativeness may be affected by the self-selection bias. The current sample might experience less severe grief symptoms than a random sample, as it may be more taxing for those who suffered from more severe grief symptoms to fill out all questions. Thus, the study may underestimate the severity of PGD and PCBD symptoms among COVID-19 bereaved individuals. Second, although strictly following the ICD-11 and DSM-5 diagnostic guidelines, this study adopted self-reported measures rather than structured clinical interviews to determine the prevalence rates. Third, due to the cross-sectional design, effects of demographic and loss-related variables in predicting the development of PGD and PCBD could not be examined. Longitudinal study is needed to determine the causality. Last, loss-related characteristics specifically affected by the COVID-19 pandemic was not included in explaining the variance of grief symptoms. Future research can investigate the relationship between grieving experience altered by COVID-19 and the development of PGD and PCBD so that bereavement support could be tuned for the COVID-19- or pandemic-related bereaved population.
Notwithstanding these considerations, the study contributes to the field by providing the first evidence of the prevalence, symptom severity, and associated factors of PGD and PCBD in a sample of COVID-19 bereaved individuals. It compared the prevalence of pathological grief between ICD-11 and DSM-5 in a Chinese population, adding cross-cultural data of adopting these two diagnostic systems in the midst of utilizing and revising the diagnostic criteria (7, 51).