Effects of the COVID-19 Pandemic on Grief: A Prospective Survey Among Community Hospital Next-of-Kin

Abstract

Background:

Over one million Americans have died from coronavirus disease 2019 (COVID-19). Increased isolation resulting from social distancing, public health restrictions, and hospital visit limitations may affect the ability to perform normal mourning practices. Grief experts expect higher rates of prolonged grief disorder (PGD) because of the pandemic, but empirical evidence is limited.

Objectives:

To compare grief severity using PG-13-R scores among those bereaved during the COVID-19 pandemic to prepandemic controls.

Design:

Prospective cohort trial.

Setting/Subjects:

One hundred twenty-three next-of-kin of deceased individuals at a secondary medical center in the United States.

Measurements:

We compared the PG-13-R score and demographic characteristics across three periods based on the time of their relative's death: prepandemic (n = 42, November 2019–March 19, 2020), early-pandemic (n = 42, March 20, 2020–June 2020), and mid-pandemic (n = 39, July 2020–October 2020). Linear regression investigated the association between the PG-13-R score and COVID-19-related death, hospice use, and grief support services while controlling for demographic characteristics.

Results:

There were no statistical differences in unadjusted PG-13-R score, individual PG-13-R questions, and demographic characteristics across three periods. In adjusted analyses, COVID-19-related death was associated with an increased PG-13-R score (coefficient: 6.17; p = 0.031), while hospice use was associated with a decline in the PG-13-R score (coefficient: −3.68; p = 0.049).

Conclusion:

Individuals have adapted to COVID-19 societal changes, including how they grieve. However, COVID-19-related deaths may lead to a higher risk for PGD, consistent with COVID-19 grief studies globally.

Introduction

While death and thus bereavement are universal parts of the human experience, most individuals grieve and complete the bereavement period without significant complications. However, ∼10% of bereaved individuals will experience persistent yearning for the deceased and pervasive preoccupation with the dead, characteristic of prolonged grief disorder (PGD), previously referred to as traumatic grief or complicated grief.^1,2

Coronavirus disease 2019 (COVID-19) has affected nearly every aspect of life worldwide. At the time of this writing, over 595 million global cases of COVID-19 had been reported, resulting in over six million deaths. Over 93 million COVID-19 patients in the United States have resulted in over one million deaths.³ Social distancing, public health restrictions in both travel and gatherings, and limitations in hospital visitation have all contributed to a sense of isolation. The inability to gather in person has affected performing customary mourning practices and rituals. Given that lack of preparedness for the death, the sense that the death was preventable or premature, perceived lack of social support after the death, and dying in a hospital/ICU are risk factors for PGD, COVID-19 deaths may lead to a higher risk for PGD.^4–9 Indeed, higher rates of PGD occur in other disasters that result in many fatalities.^10,11

Grief experts anticipate increased grief severity and PGD because of the pandemic.^12–15 A recent trial by Gang et al showed higher rates of probably PGD in the bereaved individuals who died from an accident during the COVID-19 pandemic.¹⁶ Preliminary studies outside the United States suggested that COVID-19 deaths resulted in higher grief severity and risk for PGD compared to natural deaths and pre-COVID-19 era deaths.^17–19

A recent self-reported online survey of grief severity in the United States showed a higher overall risk of PGD during the COVID-19 pandemic, at a rate of 66%.¹⁶ This is above the previous high PGD rate of 49% reported in Djelantik et al's study of bereaved individuals who lost a significant person due to unnatural causes, such as accidents, disasters, homicides, or suicides.¹¹ To our knowledge, this is the first study looking at the effects of the COVID-19 pandemic and grief in the United States with a prepandemic control cohort.

Identifying PGD in bereaved populations is vital as severe grief reactions have been associated with poorer health outcomes,^20–23 possibly mediated through immunomodulatory or hormonal mechanisms.^24–26 The adverse health effects include increased mortality.^22,27 PGD is distinct from other psychiatric diseases such as depression or PTSD, which helps explain why it does not respond to pharmacotherapy alone.^28–32

Our study seeks to identify if bereaved individuals during the COVID-19 pandemic have higher grief severity as a proxy measure of PGD than bereaved individuals who lost a significant person before the COVID-19 pandemic. Grief severity has previously been established to be predictive of PGD and thus used as a proxy measure in our study.^33–35

Methods

Setting

This study population was a cohort of patients who passed away in 2019–2020 at a 269-bed secondary care medical center in the Kaiser Permanente Southern California (KPSC) system. KPSC is an integrated health care system currently providing care to more than 4.7 million members throughout Southern California.

Data and sample

Recruitment and data collection started on April 14, 2021, and ended on August 1, 2021. Information was obtained from the electronic health record to get a list of deceased patients and extract the next-of-kin's contact information. To be included in the study, patients must have died during the following three periods (from any causes): November 2019–March 19, 2020 (pre-pandemic), March 20, 2020–June 2020 (early-pandemic), and July 2020–October 2020 (mid-pandemic). All participants lost a significant person to them at least six months before the interview in this study. A period of six months was chosen based on the International Classification of Diseases 11th (ICD-11) definition for PGD.¹ The list of deceased patients was extracted from the electronic health records, including the decedent affairs, palliative care, and hospice records (N = 493). Information extracted included their date of death, sex, age at the time of death, and their next-of-kin individual.

The next-of-kin group of the deceased was invited to complete the study's questionnaire, which included demographic questions (Table 1). Next-of-kin was defined as the primary contact listed in the deceased's electronic health record. Next-of-kin needed to be individuals aged 18 or older who spoke English. The deceased's next-of-kin were contacted by research staff (social workers) and coinvestigators (physicians) for a telephone interview (N = 493). Up to two additional attempts would be made if an individual was not reached on the first attempt. Non-English speaking next-of-kin were excluded.

Table 1.

Demographic Characteristics by Time of Death

	Prepandemic	Early-pandemic	Mid-pandemic	Total	p
	(N = 42)	(N = 42)	(N = 39)	(N = 123)	p
	n (column %) unless otherwise specified
Relationship					0.399^a
Spouse	13 (31%)	7 (16.7%)	11 (28.2%)	31 (25.2%)
Adult Child	26 (61.9%)	27 (64.3%)	23 (59%)	76 (61.8%)
Sibling	2 (4.8%)	2 (4.8%)	1 (2.6%)	5 (4.1%)
Other	1 (2.4%)	6 (14.3%)	4 (10.3%)	11 (8.9%)
Age (in years)					0.177^b
N	42	41	39	122
Mean (SD)	61.1 (12.56)	58.2 (15.29)	55.1 (15.84)	58.2 (14.68)
Median	61	63	54	59
Q1, Q3	51.0, 72.0	49.0, 69.0	46.0, 64.0	49.0, 69.0
Range	(38.0–84.0)	(30.0–79.0)	(24.0–90.0)	(24.0–90.0)
Age group					0.070^c
18–44	4 (9.5%)	10 (23.8%)	9 (23.1%)	23 (18.7%)
45–64	21 (50%)	13 (31%)	21 (53.8%)	55 (44.7%)
65+	17 (40.5%)	18 (42.9%)	9 (23.1%)	44 (35.8%)
Missing/unknown	0 (0%)	1 (2.4%)	0 (0%)	1 (0.8%)
Gender					0.845^c
Male	12 (28.6%)	12 (28.6%)	9 (23.1%)	33 (26.8%)
Female	30 (71.4%)	30 (71.4%)	29 (74.4%)	89 (72.4%)
Transgender male	0 (0%)	0 (0%)	1 (2.6%)	1 (0.8%)
Race/ethnicity					0.056^c
Non-Hispanic white	13 (31%)	10 (23.8%)	5 (12.8%)	28 (22.8%)
Non-Hispanic black	3 (7.1%)	1 (2.4%)	4 (10.3%)	8 (6.5%)
Non-Hispanic Asian/PI	4 (9.5%)	4 (9.5%)	0 (0%)	8 (6.5%)
Hispanic	22 (52.4%)	24 (57.1%)	26 (66.7%)	72 (58.5%)
Other	0 (0%)	3 (7.1%)	4 (10.3%)	7 (5.7%)
Education					0.749^c
Less than high school	1 (2.4%)	0 (0%)	2 (5.1%)	3 (2.4%)
High school graduate	5 (11.9%)	8 (19%)	7 (17.9%)	20 (16.3%)
Some college	19 (45.2%)	17 (40.5%)	17 (43.6%)	53 (43.1%)
College graduate or above	17 (40.5%)	17 (40.5%)	12 (30.8%)	46 (37.4%)
Missing/unknown	0 (0%)	0 (0%)	1 (2.6%)	1 (0.8%)
Lived in the same household					0.416^a
Yes	27 (64.3%)	21 (50%)	22 (56.4%)	70 (56.9%)
No	15 (35.7%)	21 (50%)	17 (43.6%)	53 (43.1%)
Death related to COVID					0.006^a
Yes	0 (0%)	7 (16.7%)	9 (23.1%)	16 (13%)
No	42 (100%)	35 (83.3%)	30 (76.9%)	107 (87%)
Death was expected					0.091^a
Yes	27 (64.3%)	17 (40.5%)	21 (53.8%)	65 (52.8%)
No	15 (35.7%)	25 (59.5%)	18 (46.2%)	58 (47.2%)
Used grief support					0.521^a
Yes	8 (19%)	8 (19%)	11 (28.2%)	27 (22%)
No	34 (81%)	34 (81%)	28 (71.8%)	96 (78%)
Used hospice					0.905^a
Yes	25 (59.5%)	23 (54.8%)	22 (56.4%)	70 (56.9%)
No	17 (40.5%)	19 (45.2%)	17 (43.6%)	53 (43.1%)

Chi-square test.

Kruskal-Wallis test.

Fisher's exact test.

COVID, coronavirus disease; SD, standard deviation.

KPSC's Institutional Review Board (IRB) approved this study. Participation in this study was voluntary. Interviewers informed respondents that agreeing to answer the survey questions implied their verbal consent to participate in this study. There was no compensation for study participation. A telephone recruitment script was used to inform the participants about the study, explain the study's procedures, and time commitment for completing the survey. On average, the interview lasted 30 minutes. The final sample included 123 participants whose significant person died during the prepandemic period (n = 42), the early-pandemic period (n = 42), and the mid-pandemic period (n = 39). The remaining 370 potential participants opted not to participate or were not reachable by phone after three attempts.

Following the interview, all participants were provided information to participate in a group bereavement class and offered a bereavement packet derived from KPSC's social services grief resources.

Outcomes

The primary outcome, grief severity, was evaluated using the PG-13-R instrument. The PG-13-R consisted of 13 questions. The first two questions asked whether participants had lost someone significant and the number of months since that person had died. The following 10 questions are listed in Table 2. They assessed cognitive, behavioral, and emotional symptoms associated with their significant person's death. Response categories included “not at all,” “slightly,” “somewhat,” “quite a bit,” and “overwhelmingly” and were given a point score from 1 to 5, respectively. Finally, participants were asked whether any reported symptoms caused significant impairment in social, occupational, or other important areas of functioning.

Table 2.

Prolonged Grief Disorder Score by Time of Death

	Prepandemic	Early-pandemic	Mid-pandemic	Total	p
	(N = 42)	(N = 42)	(N = 39)	(N = 123)	p
	n (column %) unless otherwise specified
PGD score					0.637^a
N	42	42	39	123
Mean (SD)	20.3 (9.62)	21.0 (8.40)	21.4 (8.73)	20.9 (8.87)
Median	17	18	21	18
Q1, Q3	15.0, 22.0	14.0, 26.0	15.0, 25.0	14.0, 25.0
Range	(10.0–47.0)	(10.0–42.0)	(10.0–44.0)	(10.0–47.0)
PGD_Q3: Longing					0.647^b
Not at all	5 (11.9%)	3 (7.1%)	5 (12.8%)	13 (10.6%)
Slightly	0 (0%)	4 (9.5%)	2 (5.1%)	6 (4.9%)
Somewhat	12 (28.6%)	8 (19%)	10 (25.6%)	30 (24.4%)
Quite a bit	10 (23.8%)	13 (31%)	9 (23.1%)	32 (26%)
Overwhelmingly	15 (35.7%)	14 (33.3%)	13 (33.3%)	42 (34.1%)
PGD_Q4: Trouble doing things					0.159^b
Not at all	21 (50%)	23 (54.8%)	17 (43.6%)	61 (49.6%)
Slightly	11 (26.2%)	4 (9.5%)	8 (20.5%)	23 (18.7%)
Somewhat	4 (9.5%)	10 (23.8%)	10 (25.6%)	24 (19.5%)
Quite a bit	3 (7.1%)	0 (0%)	2 (5.1%)	5 (4.1%)
Overwhelmingly	3 (7.1%)	5 (11.9%)	2 (5.1%)	10 (8.1%)
PGD_Q5: Confused					0.764^b
Not at all	26 (61.9%)	29 (69%)	26 (66.7%)	81 (65.9%)
Slightly	6 (14.3%)	2 (4.8%)	6 (15.4%)	14 (11.4%)
Somewhat	4 (9.5%)	4 (9.5%)	4 (10.3%)	12 (9.8%)
Quite a bit	2 (4.8%)	4 (9.5%)	1 (2.6%)	7 (5.7%)
Overwhelmingly	4 (9.5%)	3 (7.1%)	2 (5.1%)	9 (7.3%)
PGD_Q6: Trouble believing					0.634^b
Not at all	27 (64.3%)	23 (54.8%)	23 (59%)	73 (59.3%)
Slightly	6 (14.3%)	6 (14.3%)	2 (5.1%)	14 (11.4%)
Somewhat	3 (7.1%)	8 (19%)	8 (20.5%)	19 (15.4%)
Quite a bit	3 (7.1%)	3 (7.1%)	4 (10.3%)	10 (8.1%)
Overwhelmingly	3 (7.1%)	2 (4.8%)	2 (5.1%)	7 (5.7%)
PGD_Q7: Avoid reminders					0.989^b
Not at all	29 (69%)	29 (69%)	29 (74.4%)	87 (70.7%)
Slightly	4 (9.5%)	3 (7.1%)	3 (7.7%)	10 (8.1%)
Somewhat	5 (11.9%)	5 (11.9%)	4 (10.3%)	14 (11.4%)
Quite a bit	3 (7.1%)	3 (7.1%)	3 (7.7%)	9 (7.3%)
Overwhelmingly	1 (2.4%)	2 (4.8%)	0 (0%)	3 (2.4%)
PGD_Q8: Feel emotional pain					0.291^c
Not at all	16 (38.1%)	7 (16.7%)	6 (15.4%)	29 (23.6%)
Slightly	8 (19%)	10 (23.8%)	10 (25.6%)	28 (22.8%)
Somewhat	9 (21.4%)	13 (31%)	9 (23.1%)	31 (25.2%)
Quite a bit	3 (7.1%)	6 (14.3%)	8 (20.5%)	17 (13.8%)
Overwhelmingly	6 (14.3%)	6 (14.3%)	6 (15.4%)	18 (14.6%)
PGD_Q9: Trouble reengaging					0.975^b
Not at all	30 (71.4%)	27 (64.3%)	27 (69.2%)	84 (68.3%)
Slightly	4 (9.5%)	3 (7.1%)	4 (10.3%)	11 (8.9%)
Somewhat	4 (9.5%)	7 (16.7%)	3 (7.7%)	14 (11.4%)
Quite a bit	3 (7.1%)	3 (7.1%)	4 (10.3%)	10 (8.1%)
Overwhelmingly	1 (2.4%)	1 (2.4%)	1 (2.6%)	3 (2.4%)
Missing/unknown	0 (0%)	1 (2.4%)	0 (0%)	1 (0.8%)
PGD_Q10: feel emotionally numb/detached					0.771^b
Not at all	33 (78.6%)	30 (71.4%)	24 (61.5%)	87 (70.7%)
Slightly	3 (7.1%)	4 (9.5%)	5 (12.8%)	12 (9.8%)
Somewhat	3 (7.1%)	3 (7.1%)	6 (15.4%)	12 (9.8%)
Quite a bit	2 (4.8%)	3 (7.1%)	4 (10.3%)	9 (7.3%)
Overwhelmingly	1 (2.4%)	1 (2.4%)	0 (0%)	2 (1.6%)
Missing/unknown	0 (0%)	1 (2.4%)	0 (0%)	1 (0.8%)
PGD_Q11: feel life is meaningless					0.173^b
Not at all	32 (76.2%)	35 (83.3%)	24 (61.5%)	91 (74%)
Slightly	2 (4.8%)	2 (4.8%)	7 (17.9%)	11 (8.9%)
Somewhat	4 (9.5%)	2 (4.8%)	5 (12.8%)	11 (8.9%)
Quite a bit	3 (7.1%)	0 (0%)	1 (2.6%)	4 (3.3%)
Overwhelmingly	1 (2.4%)	2 (4.8%)	2 (5.1%)	5 (4.1%)
Missing/unknown	0 (0%)	1 (2.4%)	0 (0%)	1 (0.8%)
PGD_Q12: feel alone					0.966^c
Not at all	17 (40.5%)	16 (38.1%)	15 (38.5%)	48 (39%)
Slightly	7 (16.7%)	8 (19%)	6 (15.4%)	21 (17.1%)
Somewhat	9 (21.4%)	11 (26.2%)	7 (17.9%)	27 (22%)
Quite a bit	3 (7.1%)	3 (7.1%)	5 (12.8%)	11 (8.9%)
Overwhelmingly	6 (14.3%)	4 (9.5%)	6 (15.4%)	16 (13%)
PGD_Q13: have symptoms that caused significant impairment					0.059^c
Yes	6 (14.3%)	10 (23.8%)	2 (5.1%)	18 (14.6%)
No	36 (85.7%)	32 (76.2%)	37 (94.9%)	105 (85.4%)

Kruskal-Wallis test.

Fisher's exact test.

Chi-square test.

PGD, prolonged grief disorder.

The PG-13-R was then scored as a continuous measure by adding all symptom items (excluding the first two questions and the last question about the loss of a significant individual, duration, and impairment). The total score ranged from 11 to 55, with higher scores implying more intense grief severity. While there is currently no official recommendation for a cutoff point of the PG-13-R score, a PG-13-R score of 30 correlated with the DSM symptom criteria for PGD.³⁶

The PG-13 instrument, on which the PG-13-R is based, has been widely used in both the United States and other countries.^37–39 Both instruments have been proved to be a valid and reliable measure for the classification of PGD individuals, with a high internal consistency (Cronbach's alpha ranges from 0.83 to 0.93) and a statistically significant association with symptoms or diagnoses for major depressive disorder, post-traumatic stress disorder, and generalized anxiety disorder, suicidal ideation.³⁶

Other variables

During the interview, next-of-kin participants reported the following: their relationship to the deceased KPSC member (adult child, spouse, sibling, and other), age (categorized in this study as 18–44, 45–64, 65+, and missing/unknown), gender (female, male, and transgender male), race and ethnicity (non-Hispanic white, non-Hispanic black, non-Hispanic Asian or Pacific Islander, Hispanic, and other), educational attainment (less than high school, high school graduate, some college, college graduate or above, and missing/unknown), whether they lived in the same household as the deceased KPSC member, whether the death was related to COVID-19, whether the death was expected, whether the participant used grief support service, whether the deceased had died while on hospice, and the number of months since the deceased had passed away. Whether the death was COVID-19 related was based on the next-of-kin's interpretation, not defined by the Center for Disease Control's criteria.⁴⁰

Statistical analysis

Descriptive analysis was used to compare the PG-13-R scores, each of the 10 individual PGD symptoms, and the characteristics of participants who lost a significant person during the prepandemic, early-pandemic, and mid-pandemic periods. The authors used the Kruskal-Wallis test for continuous variables, the chi-squared test for categorical variables, and Fisher's exact test for categorical variables with small cell sizes (<5). Linear regression was used to examine the association between the PG-13-R score and the following variables (in separate models): whether the death was related to COVID-19, whether the deceased died in hospice, whether the participant used grief support services, and the participant's race and ethnicity. We controlled for the covariates listed in Table 1, including the relationship between the respondent and the deceased, age of respondent, gender (female, male, and transgender male), race/ethnicity, education, whether the respondent and the deceased lived in the same household, and the number of months since the deceased had passed away.

Following convention, a p-value of <0.05 was considered statistically significant. We did not correct for multiple testing due to the exploratory nature of this study. Therefore, the p-values should be interpreted with caution. All analysis was conducted in October 2021 using SAS version 9.4.

Results

Descriptive statistics for characteristics of participants who lost a significant individual during the prepandemic, early-pandemic, and mid-pandemic periods are presented in detail in Table 1. Most respondents were the deceased's adult child (62%) or spouse (25%). Many participants were female (72%), Hispanic (59%), and educated with high school degrees or above (97%). Thirteen percent were related to COVID-19 as defined by the respondent, not based on the electronic medical record diagnosis. While over half the patients reportedly died on hospice (59.6%), most participants did not use grief support services (78%). All characteristics did not significantly vary between cohort time of death.

On average, the PG-13-R score was 20.9 (median, 18; range, 10–47; standard deviation [SD] 8.87). There was no statistical difference found in the PG-13-R scores between the prepandemic (M = 20.3, SD = 9.62), early-pandemic (M = 21.0, SD = 8.4), and mid-pandemic (M = 21.4, SD = 8.73) cohorts (Table 2). Further investigation into each symptom and impairment suggested similar patterns, such that the prevalence of these symptoms did not significantly change before and during the pandemic.

A fully adjusted analysis controlling for the covariates listed in Table 1 suggested that COVID-19-related death was associated with an increase of 6.17 (95% confidence interval [CI]: 0.59 to 11.74) in the PG-13-R score (Table 3). Hospice use was associated with a decline of 3.68 (95% CI: −7.35 to −0.02) in the PG-13-R score. The use of grief support was not associated with the PG-13-R score (coefficient: 3.57; 95% CI: −1.39 to 8.53). Finally, there was no evidence that the PG-13-R score varied by race and ethnicity.

Table 3.

Unadjusted and Adjusted Linear Regression Analyses of Prolonged Grief Disorder Score on the Cause of Death, Use of Hospice, Use of Grief Support, and Race/Ethnicity

	Outcome: PGD score (continuous)
	N = 123 observations
	Unadjusted		Adjusted^a
	Coefficient (95% CI)	p	Coefficient (95% CI)	p
Model 1
Death related to COVID
No (reference)
Yes	5.73 (1.11 to 10.34)	0.015	6.17 (0.59 to 11.74)	0.031
Model 2
Used hospice
No (reference)
Yes	−2.84 (−6.01 to 0.33)	0.079	−3.68 (−7.35 to −0.02)	0.049
Model 3
Used grief support
No (reference)
Yes	5.26 (1.54 to 8.98)	0.006	3.57 (−1.39 to 8.53)	0.156
Model 4
Race/ethnicity
White (reference)
Black	3.48 (−3.64 to 10.6)	0.335	3.72 (−5.63 to 13.07)	0.431
Asian/PI	1.36 (−5.76 to 8.48)	0.707	−0.67 (−8.85 to 7.52)	0.872
Hispanic	0.84 (−3.11 to 4.80)	0.674	1.41 (−3.17 to 6.00)	0.542
Other	−1.00 (−8.51 to 6.51)	0.792	−4.02 (−13.23 to 5.20)	0.389

Adjusted analyses controlled for the relationship between respondent and the deceased family member (adult child [reference group], spouse, sibling, and other), age of respondent (18–44 [reference group], 45–64, 65+, and missing/unknown), gender (female [reference group], male, and transgender male), race/ethnicity (non-Hispanic white [reference group], non-Hispanic black, non-Hispanic Asian or Pacific Islander, Hispanic, and other), education (less than high school [reference group], high school graduate, some college, college graduate or above, and missing/unknown), whether the respondent and the deceased family member lived in the same household, and the number of months since the family member passed away.

CI, confidence interval.

Discussion

To our knowledge, this is the first study to look prospectively at the relationship between the COVID-19 pandemic and grief reactions in an American cohort, using a prepandemic control cohort. Similar to previously published literature on PGD and grief during the COVID-19 pandemic, individual circumstances may increase one's PGD risk. For instance, our study showed that grief severity was higher if the death was due to COVID-19.

Prigerson et al showed that a PG-13-R score of 30 or above correlated with the DSM 5 symptom criteria for PGD.³⁶ The clinical diagnosis of PGD also requires the loss to be more than 6 months (ICD-11) or 12 months (DSM 5) and impairment in function.^1,2 Our study used a six month cutoff consistent with the ICD-11 definition. It is possible that the scores for the mid-pandemic cohort could have changed before the deceased individual's first anniversary. While our mean scores did not exceed the 30 point cutoff, other studies such as Tang and Xiang have shown higher rates of PGD using the International Prolonged Grief Disorder Scale (IPGDS) and the Traumatic Grief Inventory Self Report (TGI-SR) scale in their Chinese cohort study of bereaved individuals who lost someone to COVID-19.⁴¹

Eisma and Tamminga's Dutch grief survey showed higher grief severity in those bereaved due to COVID-19 than other natural causes but not unnatural causes.^17,18 Our study showed an increase in grief severity for COVID-19-related deaths, but not different types of deaths when adjusted for other respondent demographic variables (Table 1). This contrasts with Gang et al's study, which showed a higher rate of probable PGD due to COVID-19 deaths only when compared to dementia deaths but not cancer, heart attacks, accidents, suicide, homicide, or natural/man-made disasters. We believe the difference between studies is based on the higher overall rate of probably PGD in Gang et al's cohort, which was based on self-selected online survey data from a cohort actively seeking information on PGD online.¹⁶

An unexpected death and perceived low social support have been associated with PGD. Often COVID-19-related deaths are perceived as premature. Compounded with the social restrictions during the COVID-19 pandemic, this perception may help explain why COVID-19 deaths have been associated with higher rates of PGD in other studies such as Eisma et al and Tang and Xiang.^18,41

Tang and Xiang found that close kinship to the deceased, for example, partner, child, or parent, was associated with higher rates of PGD, our study did not show such a relationship.⁴¹ Their study utilized the IPGDS, which includes a cultural screening item looking at whether the respondent's grief experience is beyond the normal grief response in their culture. Our study's respondents were majority Hispanic, and the differences between the two studies may be cultural; we did not see demographic differences in grief severity among gender or age lines. This contrasts with known risk factors for PGD in other settings, such as being female or younger age of the bereaved.⁴² However, given the small sample sizes, our study may have been underpowered to show cultural or demographic differences.

One protective feature against severe grief reactions was if the deceased individual was receiving hospice services. It is likely that the additional support, including the anticipatory grief guidance that hospice provides, may have assuaged the perceived trauma of the deceased's death. In addition, hospice services include bereavement support, which may have allowed the bereaved individuals to work through their grief in a structured environment. The differences in PG-13-R scores among those who received hospice services may partially be explained through a response bias, such as a courtesy bias in the next-of-kin respondents who received hospice services or care through Kaiser Permanente.

Median and mean PG-13-R scores did not differ significantly across the three cohorts. This finding was found in a Canadian matched cohort of pre- and postpandemic bereaved individuals using the Inventory for Complicated Grief revised. Of note, their cohorts showed a higher rate of severe grief reactions than our study demonstrated (28.9%).⁴³

The uniformity of the grief severity across cohorts may reflect an adaptive process to the COVID-19 pandemic cohort respondents' grief. Borghi and Menichetti's Italian study of family members of COVID-19-related decedents revealed several spontaneous coping strategies, including creating alternative goodbye rituals, for example, drive by or virtual funerals, placing special items in the coffin; turning inwardly toward faith and spirituality; or using time in isolation to allow additional processing of the death.⁴⁴ Likewise, Mitima-Verloop et al's study demonstrated similar grief severity between prepandemic and during the pandemic. While they specifically looked at the relationship between funeral rituals and bereaved individuals' perceptions, they found the bereaved to be resilient. This was qualitatively expressed in their questionnaire's comment section through comments about finding solidarity under challenging circumstances, meaning in small nontraditional rituals, and gratitude for the support they received.⁴⁵

Alternatively, given the disproportionately high death toll during the pandemic, the bereaved may have found comfort in the sense of communal grief.⁴⁶ The concurrent grief experience during the pandemic may also explain why the PG-13-R scores did not differ significantly between cohorts.

Our study had several limitations. First, it was limited to bereaved individuals of patients affiliated with a single community hospital in Southern California. Second, our study was underpowered to show differences between the three cohort groups. This limitation was due to the number of deaths in our community during the sample period being insufficient to power the analysis. Third, the study response rate was 25%, slightly lower than most telephone surveys.⁴⁷ It is possible that bereaved individuals suffering from PGD were not captured due to unwillingness to discuss a profoundly personal matter or that the death was too recent for them to be ready to participate. Given that many individuals are not identified as having PGD until years after the death, future studies may be needed to determine the true PGD frequency among the COVID-19 pandemic era bereaved.

As both a strength and a weakness, our study utilized a telephonic interview. This was possible due to the small sample size of our pilot study. While larger response numbers could have been obtained using an online survey, a phone interview was the most efficient way to ensure that the participant understood the questions. It also allowed the interviewer to assess whether the participant needed immediate bereavement support. The study's authors felt that using an online survey would be too impersonal, given the sensitive nature of grief. A telephonic interview also allowed us to capture respondents who did not have access to the internet. One of our study's strengths is that the majority of our respondent population was Hispanic, representing a historically understudied population.⁴⁸

Conclusion

Our study showed that, as grief experts predicted, deaths from COVID-19 appear to result in higher rates of grief severity, a key clinical component of PDG. In our study, hospice services may mitigate grief severity. As the COVID-19 pandemic continues to affect daily life and is entering an endemic phase, researchers and clinicians need to be aware of the higher rates of PGD that are likely to occur. Future multisite studies are required to see if higher grief severity after COVID-19-related deaths manifests in PGD clinical diagnoses and its associated poor health outcomes.

Footnotes

Acknowledgments

We sincerely thank our project manager, Stephanie Tovar, whose presence enabled this study to come to fruition.

Authors' Contributions

A.E.: conceptualization (equal), review (lead), and writing (lead). D.D.: conceptualization (equal), writing (equal), and formal analysis (lead). T.D.: Writing—review, and editing (equal).

Funding Information

No funding was received for this article.

Author Disclosure Statement

No competing financial interests exist.

References

World Health Organization. International classification of diseases for mortality and morbidity statistics (11th Revision). Available from: https://icd.who.int/browse11/l-m/en [Last accessed: December 22, 2020].

Lundorff

, Holmgren

, Zachariae

, et al. Prevalence of prolonged grief disorder in adult bereavement: A systematic review and meta-analysis. J Affect Disord, 2017; 212:138–149; doi: 10.1016/j.jad.2017.01.030

Dong

, Du

, Gardner

An interactive web-based dashboard to track COVID-19 in real time; 2020. Available from: https://coronavirus.jhu.edu/map.html [Accessed November 29, 2020].

Goveas

, Shear

. Grief and the COVID-19 pandemic in older adults. Am J Geriatr Psychiatry, 2020; 28(10):1119–1125; doi: 10.1016/j.jagp.2020.06.021

Barry

, Kasl

, Prigerson

. Psychiatric disorders among bereaved persons: The role of perceived circumstances of death and preparedness for death. Am J Geriatr Psychiatry, 2002; 10(4):447–457.

Lobb

, Kristjanson

, Aoun

, et al. Predictors of complicated grief: A systematic review of empirical studies. Death Stud, 2010; 34(8):673–698; doi: 10.1080/07481187.2010.496686

Wright

, Keating

, Balboni

, et al. Place of death: Correlations with quality of life of patients with cancer and predictors of bereaved caregivers' mental health. J Clin Oncol, 2010; 28(29):4457–4464; doi: 10.1200/JCO.2009.26.3863

Johns

, Blackburn

, McAuliffe

. COVID-19, Prolonged Grief Disorder and the role of social work. Int Soc Work, 2020; 63(5):660–664.

Hebert

, Dang

, Schulz

. Preparedness for the death of a loved one and mental health in bereaved caregivers of patients with dementia: Findings from the REACH study. J Palliat Med, 2006; 9(3):683–693; doi: 10.1089/jpm.2006.9.683

10.

Eisma

, Lenferink

LIM

, Chow

AYM

, et al. Complicated grief and post-traumatic stress symptom profiles in bereaved earthquake survivors: A latent class analysis. Eur J Psychotraumatol, 2019; 10(1):1558707; doi: 10.1080/20008198.2018.1558707

11.

Djelantik

AMJ

, Smid

, Mroz

, et al. The prevalence of prolonged grief disorder in bereaved individuals following unnatural losses: Systematic review and meta regression analysis. J Affect Disord, 2020; 265:146–156; doi: 10.1016/j.jad.2020.01.034

12.

Wallace

, Wladkowski

, Gibson

, et al. Grief during the COVID-19 pandemic: Considerations for palliative care providers. J Pain Symptom Manage, 2020; 60(1):e70–e76; doi: 10.1016/j.jpainsymman.2020.04.012

13.

Eisma

, Boelen

, Lenferink

LIM

. Prolonged grief disorder following the Coronavirus (COVID-19) pandemic. Psychiatry Res, 2020; 288:113031; doi: 10.1016/j.psychres.2020.113031

14.

Kokou-Kpolou

, Fernández-Alcántara

, Cénat

. Prolonged grief related to COVID-19 deaths: Do we have to fear a steep rise in traumatic and disenfranchised griefs?. Psychol Trauma, 2020; 12(S1):S94–S95; doi: 10.1037/tra0000798

15.

Lichtenthal

, Roberts

, Prigerson

. Bereavement care in the wake of COVID-19: Offering condolences and referrals. Ann Intern Med, 2020; 173(10):833–835; doi: 10.7326/M20-2526

16.

Gang

, Falzarano

, She

, et al. Are deaths from COVID-19 associated with higher rates of prolonged grief disorder (PGD) than deaths from other causes?. Death Stud, 2022; 46(6):1287–1296; doi: 10.1080/07481187.2022.2039326

17.

Eisma

, Tamminga

, Smid

, et al. Acute grief after deaths due to COVID-19, natural causes and unnatural causes: An empirical comparison. J Affect Disord, 2021; 278:54–56; doi: 10.1016/j.jad.2020.09.049

18.

Eisma

, Tamminga

. Grief before and during the COVID-19 pandemic: Multiple group comparisons. J Pain Symptom Manage, 2020; 60(6):e1–e4; doi: 10.1016/j.jpainsymman.2020.10.004

19.

Chen

, Tang

. Profiles of grief, post-traumatic stress, and post-traumatic growth among people bereaved due to COVID-19. Eur J Psychotraumatol, 2021; 12(1):1947563; doi: 10.1080/20008198.2021.1947563

20.

Aalbaek

, Graff

, Vestergaard

. Risk of stroke after bereavement-a systematic literature review. Acta Neurol Scand, 2017; 136(4):293–297; doi: 10.1111/ane.12736

21.

Ott

CH.

The impact of complicated grief on mental and physical health at various points in the bereavement process. Death Stud, 2003; 27(3):249–272; doi: 10.1080/07481180302887

22.

Stroebe

, Schut

, Stroebe

. Health outcomes of bereavement. Lancet, 2007; 370(9603):1960–1973; doi: 10.1016/S0140-6736(07)61816-9

23.

Utz

, Caserta

, Lund

. Grief, depressive symptoms, and physical health among recently bereaved spouses. Gerontologist, 2012; 52(4):460–471; doi: 10.1093/geront/gnr110

24.

Fagundes

, Brown

, Chen

, et al. Grief, depressive symptoms, and inflammation in the spousally bereaved. Psychoneuroendocrinology, 2019; 100:190–197; doi: 10.1016/j.psyneuen.2018.10.006

25.

Knowles

, Ruiz

, O'Connor

. A systematic review of the association between bereavement and biomarkers of immune function. Psychosom Med, 2019; 81(5):415–433; doi: 10.1097/PSY.0000000000000693

26.

O'Connor

MF.

Grief: A brief history of research on how body, mind, and brain adapt. Psychosom Med, 2019; 81(8):731–738; doi: 10.1097/PSY.0000000000000717

27.

Stahl

, Arnold

, Chen

, et al. Mortality after bereavement: The role of cardiovascular disease and depression. Psychosom Med, 2016; 78(6):697–703; doi: 10.1097/PSY.0000000000000317

28.

Zygmont

, Prigerson

, Houck

, et al. A post hoc comparison of paroxetine and nortriptyline for symptoms of traumatic grief. J Clin Psychiatry, 1998; 59(5):241–245; doi: 10.4088/jcp.v59n0507

29.

Jacobs

, Nelson

, Zisook

. Treating depressions of bereavement with antidepressants. A pilot study. Psychiatr Clin North Am, 1987; 10(3):501–510.

30.

Pasternak

, Reynolds 3rd

, Schlernitzauer

, et al. Acute open-trial nortriptyline therapy of bereavement-related depression in late life. J Clin Psychiatry, 1991; 52(7):307–310.

31.

Szuhany

, Malgaroli

, Miron

, et al. Prolonged grief disorder: Course, diagnosis, assessment, and treatment. Focus (Am Psychiatr Publ), 2021; 19(2):161–172; doi: 10.1176/appi.focus.20200052

32.

Bui

, Nadal-Vicens

, Simon

. Pharmacological approaches to the treatment of complicated grief: Rationale and a brief review of the literature. Dialogues Clin Neurosci, 2012; 14(2):149–157; doi: 10.31887/DCNS.2012.14.2/ebui

33.

Lenferink

LIM

, Nickerson

, de Keijser

, et al. Trajectories of grief, depression, and posttraumatic stress in disaster-bereaved people. Depress Anxiety, 2020; 37:35–44; doi: 10.1002/da.22850

34.

Boelen

, Smid

, Mitima-Verloop

, et al. Patterns, predictors, and prognostic validity of persistent complex bereavement disorder symptoms in recently bereaved adults: A latent class analysis. J Nerv Ment Dis, 2019; 207(11):913–920; doi: 10.1097/NMD.0000000000001053

35.

Boelen

, Lenferink

LIM

. Symptoms of prolonged grief, posttraumatic stress, and depression in recently bereaved people: Symptom profiles, predictive value, and cognitive behavioural correlates. Soc Psychiatry Psychiatr Epidemiol, 2020; 55(6):765–777; doi: 10.1007/s00127-019-01776-w

36.

Prigerson

, Boelen

, Xu

, et al. Validation of the new DSM-5-TR criteria for prolonged grief disorder and the PG-13-Revised (PG-13-R) scale. World Psychiatry, 2021; 20(1):96–106; doi: 10.1002/wps.20823

37.

, Tang

, Yu

, et al. The prevalence, comorbidity and risks of prolonged grief disorder among bereaved Chinese adults. Psychiatry Res, 2014; 219(2):347–352; doi: 10.1016/j.psychres.2014.05.022

38.

Lichtenthal

, Corner

, Sweeney

, et al. Mental health services for parents who lost a child to cancer: If we build them, will they come?. J Clin Oncol, 2015; 33(20):2246–2253; doi: 10.1200/JCO.2014.59.0406

39.

Tsai

, Prigerson

, Li

, et al. Longitudinal changes and predictors of prolonged grief for bereaved family caregivers over the first 2 years after the terminally ill cancer patient's death. Palliat Med, 2016; 30(5):495–503; doi: 10.1177/0269216315603261

40.

COVID-19 Mortality Overview; 2021. Available from: https://www.cdc.gov/nchs/covid19/mortality-overview.htm [Last accessed March 13, 2022].

41.

Tang

, Xiang

. Who suffered most after deaths due to COVID-19? Prevalence and correlates of prolonged grief disorder in COVID-19 related bereaved adults. Global Health, 2021; 17(1):19; doi: 10.1186/s12992-021-00669-5

42.

Kersting

, Brahler

, Glaesmer

, et al. Prevalence of complicated grief in a representative population-based sample. J Affect Disord, 2011; 131(1–3):339–343; doi: 10.1016/j.jad.2010.11.032

43.

Downar

, Parsons

, Cohen

, et al. Bereavement outcomes in family members of those who died in acute care hospitals before and during the first wave of COVID-19: A cohort study. Palliat Med, 2022; 36(8):1305–1312; doi: 10.1177/02692163221109711

44.

Borghi

, Menichetti

. Strategies to cope with the COVID-related deaths among family members. Front Psychiatry, 2021; 12:622850; doi: 10.3389/fpsyt.2021.622850. eCollection 2021

45.

Mitima-Verloop

, Mooren

TTM

, Kritikou

, et al. Restricted mourning: impact of the COVID-19 pandemic on funeral services, grief rituals, and prolonged grief symptoms. Front Psychiatry, 2022; 13:878818; doi: 10.3389/fpsyt.2022.878818. eCollection 2022

46.

Wang

, Paulson

, Pease

, et al. Estimating excess mortality due to the COVID-19 pandemic: A systematic analysis of COVID-19-related mortality, 2020–2021. Lancet, 2022; 399(10334):1513–1536; doi: 10.1016/S0140-6736(21)02796-3

47.

Sinclair

, O'Toole

, Malawaraarachchi

, et al. Comparison of response rates and cost-effectiveness for a community-based survey: Postal, internet and telephone modes with generic or personalised recruitment approaches. BMC Med Res Methodol, 2012; 12:132; doi: 10.1186/1471-2288-12-132

48.

, Galanter

, Thakur

, et al. Diversity in clinical and biomedical research: A promise yet to be fulfilled. PLoS Med, 2015; 12(12):e1001918; doi: 10.1371/journal.pmed.1001918