Abstract
Background: Although bereaved parents suffer greatly, some may experience positive change referred to as post-traumatic growth.
Objective: Explore the extent to which parents perceive post-traumatic growth after their child's death in a pediatric intensive care unit (PICU), and associated factors.
Design: Longitudinal parent survey conducted 6 and 13 months after a child's death. Surveys included the Post-traumatic Growth Inventory-Short Form (PTGI-SF), a 10-item measure with range of 0–50 where higher scores indicate more post-traumatic growth. Surveys also included the Inventory of Complicated Grief (ICG), the Patient Health Questionnaire-8 (PHQ-8) for depression, the Short Post-Traumatic Stress Disorder Rating Interview (SPRINT), a single item on perceived overall health, and sociodemographics.
Setting/Subjects: One hundred fifty-seven parents of 104 children who died in 1 of 8 PICUs affiliated with the U.S. Collaborative Pediatric Critical Care Research Network.
Results: Of participating parents, 62.4% were female, 71.6% White, 82.7% married, and 89.2% had at least a high school education. Mean PTGI-SF scores were 27.5 ± 12.52 (range 5–50) at 6 months and 28.6 ± 11.52 (range 2–49) at 13 months (p = 0.181). On multivariate modeling, higher education (compared with those not completing high school) and higher 6-month ICG scores (reflecting more complicated grief symptoms) were associated with lower 13-month PTGI-SF scores (p = 0.005 and 0.033, respectively).
Conclusion: Parents bereaved by their child's PICU death perceive a moderate degree of post-traumatic growth in the first 13 months after the death however variability is wide. Education level and complicated grief symptoms may influence parents' perception of post-traumatic growth.
Keywords: bereavement, child, infant, parent, pediatric intensive care unit, post-traumatic growth
Introduction
The death of a child is devastating and often leads to intense grief for parents. Bereaved parents are at high risk for adverse mental health, including anxiety, depression, post-traumatic stress, and complicated grief.1–3 Complicated grief, now referred to as Prolonged Grief Disorder in the upcoming Diagnostic and Statistical Manual of Mental Disorders, fifth edition, text revision (DSM-5-TR), is a maladaptive form of grief that is characterized by strong yearning for and preoccupation with the deceased, as well as difficulty accepting the reality of the death.4–6 Not all parents develop bereavement-related mental health challenges; some grieve intensely but do not develop full-blown disorders.
Despite the substantial challenges and pain they experience, some bereaved parents find positive ways in which their lives have changed following their child's death.7,8 Such positive changes have been referred to as stress-related growth, benefit finding, or post-traumatic growth.7–10
Post-traumatic growth is defined as positive change that occurs as a result of one's struggle with highly challenging life events. As posited by Tedeschi and Calhoun,10 when an individual experiences a traumatic event severe enough to challenge their assumptive world, the resulting distress stimulates cognitive and emotional processing, which in turn may contribute to making sense of the event, finding meaning, and post-traumatic growth. Post-traumatic growth is manifested as enhanced relationships, changed priorities, personal strength, spiritual or existential change, and increased appreciation of life.10,11 Post-traumatic growth has been identified in adults with a variety of traumatic experiences such as refugees,12 war veterans,13 and bereaved individuals,14,15 including bereaved parents.7,16,17
We have previously demonstrated that parents whose children die in pediatric intensive care units (PICUs) have high rates of adverse mental health symptoms during bereavement.18 The extent to which parents bereaved by their child's death in a PICU experience post-traumatic growth and the associated factors have not been well described. In parents whose children survived a PICU stay, Colville and Cream found post-traumatic growth to be more strongly associated with moderate levels of post-traumatic stress than high or low levels,19 while Rodríguez-Ray and Alonso-Tapia found higher levels of post-traumatic growth were associated with higher levels of post-traumatic stress, anxiety, and depression.20
We aimed to examine the extent to which parents perceive post-traumatic growth after their child's death in a PICU. We hypothesized that the extent of post-traumatic growth reported by parents is associated with parent characteristics, including mental health, child characteristics, and characteristics of the child's clinical course. Greater understanding of post-traumatic growth among parents bereaved by their child's PICU death may help guide supportive care.
Methods
Design and setting
This study was a longitudinal survey conducted between September 2016 and March 2019 across eight children's hospitals affiliated with the Eunice Kennedy Shriver National Institute of Child Health and Human Development Collaborative Pediatric Critical Care Research Network (CPCCRN).18 Participants were asked to complete surveys 6 and 13 months after their child's death in a PICU. The study was approved by the University of Utah Institutional Review Board (IRB), which serves as the central IRB for the CPCCRN. The need for written consent was waived.
Participants
Parents (i.e., biological and/or legal guardians) were eligible if their child died in a PICU, they were English or Spanish speaking, and were at least 18 years old. One or more parents of each deceased child were eligible. The children's medical records were reviewed to obtain parents' contact information and primary language for purpose of recruitment.
Study procedures
Eligible parents were mailed surveys in English or Spanish at 6 and 13 months after their child's death. If completed surveys were not returned within one month, telephone contact was attempted to offer parents the option of completing the survey by telephone. If the household was successfully contacted by telephone but the survey was not completed, parents were categorized as refusing to participate. If the household could not be contacted by telephone after three or more attempts, the parents were categorized as unable to contact. Survey packets included a study information sheet offering parents the assistance of the local PICU social worker or psychologist for bereavement support or mental health referrals. Social worker or psychologist assistance was available regardless of parental participation in this study.
Surveys administered at 6 and 13 months asked parents to complete the following measures: the Post-traumatic Growth Inventory-Short Form (PTGI-SF),21 the Inventory of Complicated Grief (ICG),22,23 the Patient Health Questionnaire-8 (PHQ-8),24,25 the Short Post-Traumatic Stress Disorder Rating Interview (SPRINT),26 and a single item on perceived overall health.27 Surveys administered at six months also requested parents to provide sociodemographic data. For parents who completed surveys, their deceased child's medical records underwent further review to collect characteristics of the child and their clinical course.
Outcome
The primary outcome for this study was the extent of post-traumatic growth perceived by parents 13 months after their child's death as assessed by the PTGI-SF.21 The 13-month rather than 6-month time point was chosen for the primary outcome to allow more time for post-traumatic growth to occur. Thirteen months rather than one-year was chosen to avoid the anniversary of the child's death.
The PTGI-SF is a 10-item measure of the extent of positive change resulting from adversity. Post-traumatic growth is assessed across five domains, including (1) relating to others, (2) new possibilities, (3) personal strength, (4) spiritual change, and (5) appreciation of life. Each domain has two items. Responses are rated from 0 (no change) to 5 (very great degree of change) and summed to obtain domain scores ranging from 0 to 10 and total scores from 0 to 50. Higher scores indicate more positive change. Cronbach's α is >0.8 for various samples.21
Independent variables
Independent variables included parent six-month responses to the ICG, PHQ-8, SPRINT, and overall health item; parent sociodemographics; and characteristics of the child and their clinical course.
The ICG is a 19-item measure that assesses the frequency of complicated grief symptoms.22 Responses are rated from 0 to 4 and summed to obtain total scores ranging from 0 to 76. Higher scores indicate more grief symptoms. Scores ≥30 at least six months after a death have been used as a positive screen for complicated grief.23 Cronbach's α has been reported as 0.94.22
The PHQ-8 is an 8-item measure that assesses the frequency of depression symptoms.24,25 Responses are rated from 0 to 3 and summed to obtain total scores ranging from 0 to 24. Scores of 5, 10, 15, and 20 have been used as positive screens for mild, moderate, moderately severe, and severe depression, respectively.25 Cronbach's α has been reported to range from 0.82 to 0.88.28,29
The SPRINT is an 8-item measure that assesses the frequency of post-traumatic stress symptoms.26 Responses are rated from 0 to 4 and summed to obtain total scores ranging from 0 to 32. Scores ≥14 have been used as a positive screen for post-traumatic stress disorder (PTSD).26 Cronbach's α has been reported to range from 0.77 to 0.87.26
Perceived overall health was assessed by one item,27 “In general, would you say your overall health is (1) excellent, (2) very good, (3) good, (4) fair, (5) poor.”
Parent sociodemographics included gender, age, race, ethnicity, marital status, education, relationship to the deceased child, and number of surviving children. Relationship to the deceased child was categorized as biological parent or other legal guardian.
Child characteristics included gender, age at time of death, cause of death, trajectory of death, mode of death, and PICU and hospital lengths of stay. Trajectory of death represents the trajectory of the illness that lead to death and included four categories: (1) sudden unexpected death (e.g., death from a car crash), (2) death from a potentially curable disease (e.g., brain cancer with an initial positive response to treatment followed by death due to disease progression), (3) death from a lethal congenital anomaly (e.g., nonrepairable cardiac defect with death in the first few weeks of life), and (4) death from a progressive condition with intermittent crises (e.g., muscular dystrophy with slowly deteriorating health interrupted by potentially fatal medical events).30
Mode of death was categorized as limitation or withdrawal of life support, brain death, or unsuccessful cardiopulmonary resuscitation (CPR).31 Limitation of life support was defined as not initiating or increasing life-sustaining interventions, and withdrawal of life support was defined as discontinuing a life-sustaining intervention that had already been started with the expectation that death would result. Brain death was defined as a formal determination of brain death consistent with institutional criteria. Unsuccessful CPR was defined as death despite resuscitation procedures and medications.31
Statistical analysis
Scores for each measure were calculated based on available data if at least 80% of the items in the measure were completed. Missing item responses were replaced by the mean of the completed items. If less than 80% of the items were completed for any measure, the score was considered missing. Only parents who responded to the 13-month survey with at least 80% of the PTGI-SF items completed were included in the current analysis.
Parent and child characteristics are summarized using frequencies and percentages for categorical variables, and means and standard deviations for continuous variables. Parents who responded to the 13-month survey with at least 80% completion of PTGI-SF items were compared with those that did not respond using t tests for continuous variables, Fisher's exact tests for categorical variables, and Cochran–Armitage trend test for discrete ordinal variables.
To account for the fact that more than one parent of the same deceased child could participate in the study and the characteristics and survey scores of parents of the same child are likely correlated, linear mixed models were used with child as the random effect to identify variables associated with post-traumatic growth. These models were considered appropriate after examination of distribution plots.
Post-traumatic growth was first analyzed by measuring the change in total PTGI-SF score and the five domains from 6 to 13 months, and assessing the significance of those changes using linear mixed models. Next, each parent and child characteristic was assessed as an independent predictor of the 13-month PTGI-SF score. Variables with p-value <0.2 in those univariable mixed models were then considered for the multivariate model. Multicollinearity was assessed for all candidate predictors, where two predictors had high correlation (>0.7), the most clinically relevant variable was included in the full model. The final multivariate model was constructed through manual backward selection, until all remaining predictors had a p-value <0.1.
Results
Parents of 710 deceased children were eligible for the study. Of these, 235 parents of 158 children responded to the 6-month survey (22% of eligible families). One hundred fifty-seven parents of 104 children responded to the 13-month survey with completion of at least 80% of the PTGI-SF items (65.8% of families responding at 6 months and 14.6% of all eligible families). Parents responding to 13-month surveys were more likely to be married and to report less complicated grief symptoms and better overall health at 6 months than those who did not respond (Table 1). Child characteristics were similar between parents responding to 13-month surveys and those who did not (Table 2).
Table 1.
Parent Characteristics by 13-Month Post-Traumatic Growth Inventory-Short Form Response
| Overall (N = 235), n (%) | Parent responded to 13-month PTGI-SF survey |
p | ||
|---|---|---|---|---|
| Yes (N = 157), n (%) | No (N = 78), n (%) | |||
| Parent gender | 1.000a | |||
| Male | 88 (37.4) | 59 (37.6) | 29 (37.2) | |
| Female | 147 (62.6) | 98 (62.4) | 49 (62.8) | |
| Parent age, years, mean (SD) | 38.7 (10.61) | 39.4 (10.81) | 37.1 (10.07) | 0.108b |
| Parent race | 0.118a | |||
| Black/African American | 46 (20.0) | 25 (16.1) | 21 (28.0) | |
| White/Caucasian | 157 (68.3) | 111 (71.6) | 46 (61.3) | |
| Other | 27 (11.7) | 19 (12.3) | 8 (10.7) | |
| Parent ethnicity | 0.244a | |||
| Hispanic or Latino | 36 (15.7) | 21 (13.5) | 15 (20.3) | |
| Non-Hispanic or Latino | 193 (84.3) | 134 (86.5) | 59 (79.7) | |
| Parent language (used for survey completion) | 0.764a | |||
| Spanish | 13 (5.5) | 8 (5.1) | 5 (6.4) | |
| English | 222 (94.5) | 149 (94.9) | 73 (93.6) | |
| Parent education | 0.156c | |||
| Some high school or less | 22 (9.4) | 17 (10.8) | 5 (6.6) | |
| High school graduate or GED | 62 (26.6) | 34 (21.7) | 28 (36.8) | |
| Vocational school or some college | 64 (27.5) | 42 (26.8) | 22 (28.9) | |
| College degree or higher | 85 (36.5) | 64 (40.8) | 21 (27.6) | |
| Parent marital status | 0.001a | |||
| Married | 177 (76.0) | 129 (82.7) | 48 (62.3) | |
| Single | 56 (24.0) | 27 (17.3) | 29 (37.7) | |
| Relation to child | 1.000a | |||
| Biological parent | 215 (91.5) | 143 (91.1) | 72 (92.3) | |
| Other | 20 (8.5) | 14 (8.9) | 6 (7.7) | |
| Surviving children | 0.470a | |||
| No | 42 (17.9) | 26 (16.6) | 16 (20.8) | |
| Yes | 192 (82.1) | 131 (83.4) | 61 (79.2) | |
| Six-month ICG score,d mean (SD) | 34.4 (14.94) | 33.8 (15.35) | 35.8 (14.10) | 0.317b |
| Six-month ICG scored ≥30 | 142 (63.4) | 86 (57.7) | 56 (74.7) | 0.013a |
| Six-month PHQ score,e mean (SD) | 9.1 (6.20) | 9.0 (6.36) | 9.3 (5.93) | 0.716b |
| Six-month PHQ scoree categories | 0.857c | |||
| None to mild depression (0 to <10) | 126 (55.3) | 85 (56.3) | 41 (53.2) | |
| Moderate to moderately severe depression (10 to <20) | 85 (37.3) | 54 (35.8) | 31 (40.3) | |
| Severe depression (≥20) | 17 (7.5) | 12 (7.9) | 5 (6.5) | |
| Six-month SPRINT score,f mean (SD) | 14.6 (8.19) | 14.0 (8.31) | 15.6 (7.90) | 0.156b |
| Six-month SPRINT scoref ≥14 | 122 (53.3) | 78 (51.7) | 44 (56.4) | 0.576a |
| Six-month PTGI-SF score,g mean (SD) | 27.3 (12.24) | 27.5 (12.52) | 26.8 (11.74) | 0.694b |
| Six-month overall health | 0.025c | |||
| Excellent or very good | 75 (33.2) | 57 (38.3) | 18 (23.4) | |
| Good | 94 (41.6) | 59 (39.6) | 35 (45.5) | |
| Fair or poor | 57 (25.2) | 33 (22.1) | 24 (31.2) | |
Parents of 710 children were eligible. Of these, 235 parents of 158 children responded at 6 months (22% of eligible families) and 157 parents of 104 children responded at 13 months (65.8% of families responding at 6 months and 14.6% of all eligible families).
The number of missing values are as follows: Parent age had 5 missing values, Parent race had 5 missing values, Parent Ethnicity had 6 missing values, Parent education had 2 missing values, Parent marital status had 2 missing values, Surviving children had 1 missing value, six-month ICG score had 11 missing values, six-month PHQ-8 score had 7 missing values, six-month SPRINT score had 6 missing values, six-month PTGI-SF score 10 missing values, and six-month overall health had 9 missing values.
Fisher's exact test (Monte Carlo approximation for tables larger than 2 × 2).
t Test with unpooled variance estimates.
Cochran–Armitage trend test.
ICG: Higher scores indicate more grief symptoms.
PHQ-8: Higher scores indicate more depression symptoms.
SPRINT: Higher scores indicate more PTSD symptoms.
PTGI-SF: Higher scores indicate more post-traumatic growth.
GED, General Educational Development test; ICG, Inventory of Complicated Grief; PHQ-8, Patient Health Questionnaire-8; PTGI-SF, Post-traumatic Growth Inventory-Short Form; PTSD, post-traumatic stress disorder; SD, standard deviation; SPRINT, Short Post-traumatic Stress Disorder Rating Interview.
Table 2.
Child Characteristics by 13-Month Post-Traumatic Growth Inventory-Short Form Response
| Overall (N = 158), n (%) | Parent responded to 13-month PTGI-SF survey |
p | ||
|---|---|---|---|---|
| Yes (N = 104), n (%) | No (N = 54), n (%) | |||
| Child gender | 0.867a | |||
| Male | 87 (55.1) | 58 (55.8) | 29 (53.7) | |
| Female | 71 (44.9) | 46 (44.2) | 25 (46.3) | |
| Child age at time of death, years | 0.347b | |||
| Mean (SD) | 7.9 (7.77) | 8.3 (8.19) | 7.1 (6.91) | |
| Trajectory of death | 0.061a | |||
| Sudden unexpected death | 49 (31.0) | 26 (25.0) | 23 (42.6) | |
| Death from a congenital anomaly | 29 (18.4) | 24 (23.1) | 5 (9.3) | |
| Death from potentially curable disease | 36 (22.8) | 25 (24.0) | 11 (20.4) | |
| Death from progressive condition with intermittent crises | 44 (27.8) | 29 (27.9) | 15 (27.8) | |
| Mode of death | 0.061a | |||
| Brain death | 25 (15.8) | 12 (11.5) | 13 (24.1) | |
| Failed CPR | 19 (12.0) | 11 (10.6) | 8 (14.8) | |
| Limitation or withdrawal of life support | 114 (72.2) | 81 (77.9) | 33 (61.1) | |
| Cause of death | 0.310a | |||
| Cardiac | 40 (25.3) | 26 (25.0) | 14 (25.9) | |
| Respiratory | 25 (15.8) | 16 (15.4) | 9 (16.7) | |
| Malignancy | 13 (8.2) | 9 (8.7) | 4 (7.4) | |
| Sepsis or multiple organ failure | 30 (19.0) | 24 (23.1) | 6 (11.1) | |
| Neurologic | 25 (15.8) | 16 (15.4) | 9 (16.7) | |
| Trauma | 20 (12.7) | 9 (8.7) | 11 (20.4) | |
| Other | 5 (3.2) | 4 (3.8) | 1 (1.9) | |
| PICU length of stay, days | 0.702b | |||
| Mean (SD) | 22.2 (38.93) | 23.1 (38.37) | 20.5 (40.30) | |
| Min, Max | 0.0, 209.2 | 0.0, 209.2 | 0.0, 191.8 | |
| Hospital length of stay, days | 0.674b | |||
| Mean (SD) | 27.2 (42.41) | 28.3 (42.77) | 25.3 (42.04) | |
| Min, Max | 0.1, 209.2 | 0.1, 209.2 | 0.2, 191.8 | |
Fisher's exact test (Monte Carlo approximation for tables larger than 2 × 2).
t Test with unpooled variance estimates.
CPR, cardiopulmonary resuscitation; PICU, pediatric intensive care unit.
Of 157 participating parents, 98 (62.4%) were female, 111 (71.6%) were White, 134 (86.5%) were non-Hispanic, 129 (82.7%) were married, and 140 (89.2%) had at least a high school education (Table 1). One-hundred forty-three (91.1%) were biological parents and 26 (16.6%) had lost their only child. Six-month ICG score was 33.8 ± 15.35 with 86 (57.7%) parents having scores ≥30; PHQ-8 score was 9.0 ± 6.36 with 66 (43.7%) having scores ≥10; and SPRINT score 14.0 ± 8.31 with 78 (51.70%) having scores ≥14. Six-month overall health was rated as fair or poor by 33 (22.1%) parents.
Of 104 deceased children, 58 (55.8%) were male and mean age at the time of death was 8.3 ± 8.19 years (Table 2). Trajectory of death was sudden unexpected for 26 (25.0%), and mode of death was limitation or withdrawal of life support for 81 (77.9%). Mean PICU and hospital length of stay was 23.1 ± 38.4 and 28.3 ± 42.8 days, respectively.
Parent PTGI-SF score was 27.5 ± 12.52 at 6 months and 28.6 ± 11.52 at 13 months (p = 0.181) (Table 3). The range of scores was wide; 5 to 50 at 6 months and 2 to 49 at 13 months. Domain scores were not significantly different between 6 and 13 months.
Table 3.
Post-Traumatic Growth Inventory-Short Form Total and Domain Scores at 6 and 13 Months after Child's Death and Change in Scores
| 6 Month | 13 Month | Change | p a | |
|---|---|---|---|---|
| PTGI-SFb total scores (N = 150) | ||||
| Mean (SD) | 27.5 (12.52) | 28.6 (11.52) | 0.1 (0.99) | 0.1810 |
| Min, Max | 5.0, 50.0 | 2.0, 49.0 | ||
| PTGI-SFb domain scores | ||||
| Relating to others | 5.7 (3.03) | 5.7 (2.92) | −0.1 (2.68) | 0.8110 |
| New possibilities | 4.4 (3.13) | 4.9 (3.00) | 0.4 (3.16) | 0.1161 |
| Personal strength | 6.2 (3.23) | 6.4 (3.14) | 0.1 (2.72) | 0.5490 |
| Spiritual change | 4.6 (3.87) | 5.0 (3.83) | 0.3 (2.73) | 0.1737 |
| Appreciation for life | 6.4 (2.73) | 6.7 (2.33) | 0.3 (2.60) | 0.1872 |
From linear mixed model accounting for between-parent correlation.
PTGI-SF: Higher scores indicate more post-traumatic growth.
In univariable analyses, the following variables were identified as candidate predictors of parental 13-month PTGI-SF score (Table 4): 6-month ICG, PHQ-8, SPRINT and overall health scores, race, ethnicity, education level, marital status, PICU length of stay, and hospital length of stay. Sets of variables with high correlation included PHQ-8 and SPRINT scores, as well as hospital and PICU length of stay. As both sets of variables were identified as candidate predictors for multivariable modeling, SPRINT and PICU length of stay were chosen for the full model.
Table 4.
Univariate Analyses for 13-Month Post-Traumatic Growth Inventory-Short Form Scores
| Effect (95% CI) | p a | |
|---|---|---|
| Parent gender | 0.246 | |
| Male | Reference | |
| Female | 1.96 (−1.39 to 5.32) | |
| Parent age | 0.01 (−0.17 to 0.2) | 0.898 |
| Parent race | 0.041b | |
| Black/African American | 6.95 (1.52 to 12.39) | |
| White/Caucasian | Reference | |
| Other | 2.66 (−3.45 to 8.77) | |
| Parent ethnicity | 0.153b | |
| Hispanic or Latino | 4.03 (−1.55 to 9.62) | |
| Non-Hispanic or Latino | Reference | |
| Parent education | 0.008b | |
| Some high school or less | Reference | |
| High school graduate or GED | −3.96 (−10.72 to 2.81) | |
| Vocational school or some college | −8.97 (−15.63 to −2.31) | |
| College degree or higher | −9.54 (−15.86 to −3.21) | |
| Parent marital status | 0.113b | |
| Married | −4.18 (−9.38 to 1.02) | |
| Single | Reference | |
| Relation to child | 0.384 | |
| Biological parent | −2.89 (−9.49 to 3.72) | |
| Other | Reference | |
| Surviving children | 0.211 | |
| No | Reference | |
| Yes | −3.31 (−8.55 to 1.93) | |
| Child gender | 0.793 | |
| Male | Reference | |
| Female | 0.54 (−3.55 to 4.62) | |
| Child age at time of death, years | −0.06 (−0.31 to 0.18) | 0.619 |
| Trajectory of death | 0.203 | |
| Sudden unexpected death | Reference | |
| Death from a congenital anomaly | 5.69 (−0.06 to 11.44) | |
| Death from potentially curable disease | 1.25 (−4.53 to 7.02) | |
| Death from progressive condition with intermittent crises | 3.75 (−1.86 to 9.36) | |
| Mode of death | 0.817 | |
| Brain death | 1.84 (−4.51 to 8.18) | |
| Failed cardiopulmonary resuscitation | 1.11 (−5.63 to 7.86) | |
| Limitation or withdrawal of life support | Reference | |
| Cause of death | 0.323 | |
| Cardiac | 4.99 (−3.34 to 13.32) | |
| Respiratory | 3.54 (−5.43 to 12.51) | |
| Malignancy | 2.67 (−7.46 to 12.79) | |
| Sepsis or multiple organ failure | 9.09 (0.61 to 17.57) | |
| Neurologic | 2.64 (−6.26 to 11.54) | |
| Trauma | Reference | |
| Other | 3.87 (−8.82 to 16.56) | |
| PICU length of stay, days | 0.05 (0 to 0.1) | 0.041b |
| Hospital length of stay, days | 0.05 (0 to 0.1) | 0.033b |
| Six-month ICG scorec | −0.16 (−0.28 to −0.03) | 0.017b |
| Sic-month PHQ-8 scored | −0.48 (−0.77 to −0.2) | 0.001b |
| Six-month SPRINT scoree | −0.26 (−0.49 to −0.04) | 0.023b |
| Six-month overall health | 0.028b | |
| Excellent or very good | Reference | |
| Good | 1.54 (−2.66 to 5.74) | |
| Fair or poor | −5.06 (−10 to −0.12) |
Each p-value is from a mixed model with the predictor listed and 13-month PTGI-SF score as the outcome.
All variables with p-values <0.2 are candidate predictors for the multivariable model.
ICG: Higher scores indicate more grief symptoms.
PHQ-8: Higher scores indicate more depression symptoms.
SPRINT: Higher scores indicate more PTSD symptoms.
CI, confidence interval.
The final multivariate model is shown in Table 5. While adjusting for other variables in the model, higher education (compared with those not completing high school) and higher 6-month ICG score were associated with lower 13-month PTGI-SF score (p = 0.005 and 0.033, respectively). PICU length of stay and 6-month overall health were also in the final model but not significant at the 0.05 level.
Table 5.
Post-Traumatic Growth Multivariable Model
| Characteristics (N = 151) | Effect (95% CI) | p |
|---|---|---|
| Parent education | 0.005 | |
| Some high school or less | Reference | |
| High school graduate or GED | −4.31 (−11.1 to 2.48) | |
| Vocational school or some college | −9.36 (−15.99 to −2.73) | |
| College degree or higher | −9.95 (−16.21 to −3.69) | |
| PICU length of stay, days | 0.05 (0 to 0.09) | 0.057 |
| Six-month ICG scorea | −0.14 (−0.27 to −0.01) | 0.033 |
| Six-month overall health | 0.089 | |
| Excellent or very good | Reference | |
| Good | 2.33 (−1.76 to 6.42) | |
| Fair or poor | −3.07 (−8.31 to 2.17) |
PTGI-SF: Higher scores indicate more growth. This multivariable mixed model is the result of manual backward selection.
ICG: Higher scores indicate more grief symptoms.
Discussion
Our study evaluates post-traumatic growth in bereaved parents after the death of their child in a PICU, and explores parent and child characteristics associated with post-traumatic growth. Bereaved parents in our study reported a moderate amount of post-traumatic growth, although variability was wide. Factors associated with less post-traumatic growth were greater parental complicated grief symptoms and higher parental education levels.
Parents whose children die in a PICU may represent a unique population of bereaved individuals. PICUs are characterized by technologically advanced care typically aimed at restoring a child's health.32 Death is often not expected by parents at the time of their child's admission to a PICU.
In our study, average PTGI-SF scores were 28.8 ± 11.52 at 13 months after the child's death. In a recent study by Albuquerque et al.,7 parents whose children died from a variety of circumstances, including fetal loss (27%) reported PTGI-SF scores of 40.4 ± 12.5 representing a greater degree of growth than observed in our parent cohort. Besides different circumstances and setting of death, time since death was also longer in Albuquerque et al.'s study, ranging from 6 months to 10 years with an average of 2.8 ± 2.3 years. One might expect that longer time from a child's death would allow for more growth; however, time since death was not a significant predictor of post-traumatic growth in Albuquerque et al.'s study. Importantly, Albuquerque et al.'s study was cross-sectional, and change in post-traumatic growth over time was not evaluated within the same parent.
The notion that personal growth can emerge as a result of the distress associated with losing a loved one has been suggested by bereavement researchers for decades.33 Yet, studies on the relationship between grief and post-traumatic growth have had heterogeneous results. When investigating grief symptoms and post-traumatic growth, some studies have shown a negative association,33 some a positive association,34 and others have shown that these outcomes are independent of each other.17 A curvilinear relationship has also been described in which bereaved individuals with moderate symptoms of grief, depression, anxiety, and post-traumatic stress have greater post-traumatic growth than those with low or high symptom levels.15,35,36
Findings from our study suggest a negative association between bereavement-related distress and growth. Greater symptoms of complicated grief, depression, and post-traumatic stress at 6 months after a child's death were associated with less post-traumatic growth at 13 months on univariate analyses in our parent cohort. Of these, complicated grief remained a significant predictor of less growth when adjusted for other variables. Notably, more than half of our bereaved parents had mental health symptoms of sufficient severity to screen positive for complicated grief, depression, and/or PTSD at six months after their child's death. Complicated grief is characterized by intense longing for the deceased and inability to accept the reality of the death. In contrast, depression is characterized by feelings of worthlessness and self-loathing, while PTSD is characterized by anxiety and fear.4,5,37 Post-traumatic growth is posited to occur as bereaved individuals attempt to make meaning from their loss and reconstruct their world in the absence of their loved one.14,17 Perhaps high levels of adverse mental health symptoms experienced by some bereaved parents interfere with the process of meaning reconstruction, and thereby delay or prohibit the attainment of growth after their child's death. It may also be that severely intense grief, known to occur more often after the death of one's child compared with other types of loss, is preventing meaning reconstruction and growth for parents.38
In our study, bereaved parents with some college/vocational school or higher education had less post-traumatic growth when compared with parents not completing high school. This result is surprising and difficult to explain. Parents with higher education likely possess greater health literacy, and therefore we expected them to more readily make sense of their child's death, at least in the medical realm, leading to greater post-traumatic growth.
Lichtenthal et al.8 questioned a group of bereaved parents about their ability to make sense of their loss, and some described concrete medical explanations for their child's death. Parents in Lichtenthal et al.'s study who were able to engage in sense making had less symptoms of complicated grief, which could potentially allow for more growth. Why parents in our study with higher education had less growth is not evident from our data. Parents with higher education may be able to intellectually make sense of how their child died but still be devastated and stuck in their grief. Higher education may bring parents to feel that more could have been done for their child and thus contribute to more distress, guilt, and challenges with meaning during bereavement.
PICU length of stay was associated with post-traumatic growth in our multivariable model although the association was not significant at the 0.05 level. Longer PICU stays may allow more trusting relationships to develop with PICU staff, potentially reducing anger and blame, and promoting personal growth. Perceived overall health was also a nonsignificant predictor of post-traumatic growth in our multivariable model suggesting that reduced physical health as well as mental health may contribute to less growth during bereavement.
Strengths of this study include the longitudinal design, which suggests a sequence of events in which adverse mental health symptoms interfere with the development of post-traumatic growth in bereaved parents. Strengths also include the racial, ethnic, and geographic diversity of participants. Limitations include our low overall response rate, which may be related to the distress of bereavement and added burden of research participation. Responders to 13-month surveys reported less complicated grief symptoms and better overall health at 6 months post-death than nonresponders. This difference may have biased our results toward healthier parents causing overestimation of the level of post-traumatic growth at 13 months. The PTGI-SF is a shortened version of the full 21-item PTGI.21 Although research comparing the PTGI-SF to the full version suggests the PTGI-SF is psychometrically sound and that it can be substituted for the full version to reduce respondent burden, in-depth exploration of the five domains of post-traumatic growth was not possible with the PTGI-SF due to the reduced number of items. The mental health surveys used in this study assess frequency of symptoms and are screening tools rather than diagnostic instruments. The ICG measure may soon be outdated given the DSM-5-TR criteria for Prolonged Grief Disorder. Several variables potentially associated with post-traumatic growth were not evaluated in our study, such as parent resiliency, social support, the nature and extent of parents' continuing bonds with their deceased child, and parents' use of spiritual and religious coping.7,16,39,40 Prior research suggests that parents' use of spiritual and religious activities early after the death of their child in a neonatal ICU or PICU is associated with greater personal growth, and that these associations are stronger for spiritual activities as compared with religious activities.40 Parents included in our study experienced their child's death in a PICU, a setting where death is often unexpected. Our findings may not apply to parents whose children die in other environments.
Conclusion
Parents bereaved by their child's PICU death perceive a moderate degree of post-traumatic growth in the first 13-months after the death, although variability is wide. Adverse mental health symptoms may hinder development of post-traumatic growth, particularly in parents with symptoms of complicated grief. Greater attention to interpersonal interactions and activities during a child's PICU stay that have been suggested to lessen complicated grief symptoms may help to promote personal growth in parents during bereavement. These include high-quality communication between parents and clinicians, as well as supporting parents' need to make meaning of their experiences.41 Parents with higher levels of education may in particular have difficulty achieving post-traumatic growth, the reasons for which remain unclear and warrant further investigation.
Future research should also investigate post-traumatic growth among bereaved parents for longer than 13 months after their child's death.
Acknowledgments
The authors thank Whit Coleman, MSRA, BSN, RN, CCRC, University of Utah; Stephanie Dorton, BSN, RN, CCRP, University of Utah; Nael Abdelsamad, MD, University of Utah; Kylee Arbogast, BS, RN, University of Utah; Kristi Flick, BS, MPH, University of Utah; Ann Pawluszka, BSN, RN, Children's Hospital of Michigan; Melanie Lulic, BS, Children's Hospital of Michigan; Carolann Twelves, RN, BSN, CCRC, Children's Hospital of Philadelphia; Mary Ann DiLiberto, BS, RN, CCRC, Children's Hospital of Philadelphia; Elyse Tomanio, BSN, RN, Children's National Medical Center; Katherine Stone, Children's National Medical Center; Kathryn Malone, Children's Hospital Colorado; Diane Ladell, MPH, CCRC, Children's Hospital Colorado; Ruth Grosskreuz, MD, CCRC, Children's Hospital Colorado; Lisa Steele, RN, BSN, CCRC, CCRN, Nationwide Children's Hospital; Maggie Flowers, BSN, Nationwide Children's Hospital; Anna Ratiu, MPH, University of California, Los Angeles; Tanaya Deshmukh, MS, University of California, Los Angeles; Anne McKenzie, BSN, CCRN, University of California, San Francisco; Yensy Zetino, University of California, San Francisco; and Leighann Koch, BS, BSN, RN, University of Pittsburgh Medical Center.
Contributor Information
for the Eunice Kennedy Shriver National Institute of Child Health and Human Development Collaborative Pediatric Critical Care Research Network (CPCCRN):
Whit Coleman, Stephanie Dorton, Nael Abdelsamad, Kylee Arbogast, Kristi Flick, Ann Pawluszka, Melanie Lulic, Carolann Twelves, Mary Ann DiLiberto, Elyse Tomanio, Katherine Stone, Kathryn Malone, Diane Ladell, Ruth Grosskreuz, Lisa Steele, Maggie Flowers, Anna Ratiu, Tanaya Deshmukh, Anne McKenzie, Yensy Zetino, and Leighann Koch
Collaborators: for the Eunice Kennedy Shriver National Institute of Child Health and Human Development Collaborative Pediatric Critical Care Research Network (CPCCRN)
Authors' Contributions
All authors have taken part in writing the article, reviewing it, and revising its intellectual and technical content. All authors assume responsibility and accountability for the results.
Funding Information
Supported in part, by the following cooperative agreements from the Eunice Kennedy Shriver National Institute of Child Health and Human Development: UG1HD083170, UG1HD049981, UG1HD63108, UG1HD083166, UG1HD083171, UG1HD049983, U01HD049934, and UG1HD050096.
Author Disclosure Statement
No competing financial interests exist.
References
- 1. Pohlkamp L, Kreicbergs U, Sveen J: Bereaved mothers' and fathers' prolonged grief and psychological health 1 to 5 years after loss—A nationwide study. Psychooncology 2019;29:1530–1536. [DOI] [PubMed] [Google Scholar]
- 2. Lykke C, Ekholm O, Schmiegelow K, et al. : Anxiety and depression in bereaved parents after losing a child due to life-limiting diagnoses: A Danish nationwide questionnaire survey. J Pain Symptom Manage 2019;58:596–604. [DOI] [PubMed] [Google Scholar]
- 3. Baumann I, Kϋnzel J, Goldbeck L, et al. : Prolonged grief, posttraumatic stress, and depression among bereaved parents: Prevalence and response to an intervention program. Omega (Westport) 2020. [Epub ahead of print]; DOI: 10.1177/0030222820918674. [DOI] [PubMed] [Google Scholar]
- 4. Shear KM: Complicated grief. N Engl J Med 2015;372:153–160. [DOI] [PubMed] [Google Scholar]
- 5. Zisook S, Reynolds CF: Complicated grief. Focus (Am Psychiatr Publ) 2017;15:12s–13s. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Prigerson HG, Kakarala S, Gang J, et al. : History and status of prolonged grief disorder as a psychiatric diagnosis. Annu Rev Clin Psychol 2021;17:109–126. [DOI] [PubMed] [Google Scholar]
- 7. Albuquerque S, Narciso I, Pereira M: Posttraumatic growth in bereaved parents: A multidimensional model of associated factors. Psychol Trauma 2018;10:199–207. [DOI] [PubMed] [Google Scholar]
- 8. Lichtenthal WG, Currier JM, Neimeyer RA, et al. : Sense and significance: A mixed methods examination of meaning making after the loss of one's child. J Clin Psychol 2010;66:791–812. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Helgeson VS, Reynolds KA, Tomich PL: A meta-analytic review of benefit finding and growth. J Consult Clin Psychol 2006;74:797–816. [DOI] [PubMed] [Google Scholar]
- 10. Tedeschi RG, Calhoun LG: Posttraumatic growth: Conceptual foundations and empirical evidence. Psychol Inq 2004;15:1–18. [Google Scholar]
- 11. Cadell S, Hemsworth D, Quosai TS, et al. : Posttraumatic growth in parents caring for a child with a life-limiting illness: A structural equation model. Am J Orthopsychiatry 2014;84:123–133. [DOI] [PubMed] [Google Scholar]
- 12. Teodorescu DS, Siqveland J, Heir T, et al. : Posttraumatic growth, depressive symptoms, posttraumatic stress symptoms, post-migration stressors and quality of life in multi-traumatized psychiatric outpatients with a refugee background in Norway. Health Qual Life Outcomes 2012;10:84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Mattson E, James L, Engdahl B: Personality factors and their impact on PTSD and post-traumatic growth is mediated by coping style among OIF/OEF veterans. Mil Med 2018;183:e475–e480. [DOI] [PubMed] [Google Scholar]
- 14. Currier JM, Mallot J, Martinez TE, et al. : Bereavement, religion, and posttraumatic growth: A matched control group investigation. Psycholog Relig Spiritual 2013;5:69–77. [Google Scholar]
- 15. Eisma MC, Lenferink LIM, Stroebe MS, et al. : No pain, no gain: Cross-lagged analysis of posttraumatic growth and anxiety, depression, posttraumatic stress and prolonged grief symptoms after loss. Anxiety Stress Coping 2019;32:231–243. [DOI] [PubMed] [Google Scholar]
- 16. Waugh A, Kiemle G, Slade P: What aspects of post-traumatic growth are experienced by bereaved parents? A systematic review. Eur J Psychotraumatol 2018;9:1506230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Bogensperger J, Lueger-Schuster B: Losing a child: Finding meaning in bereavement. Eur J Psychotraumatol 2014;5:22910. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Suttle M, Hall MW, Pollack MM, et al. : Complicated grief, depression and post-traumatic stress symptoms among bereaved parents following their child's death in the pediatric intensive care unit: A follow-up study. Am J Hosp Palliat Care 2021. [Epub ahead of print]; DOI: 10.1177/10499091211015913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Colville G, Cream P: Post-traumatic growth in parents after a child's admission to intensive care: Maybe Nietzsche was right? Intensive Care Med 2009;35:919–923. [DOI] [PubMed] [Google Scholar]
- 20. Rodríguez-Ray R, Alonso-Tapia J: Relation between parental psychopathology and posttraumatic growth after a child's admission to intensive care: Two faces of the same coin? Intensive Crit Care Nurs 2017;43:156–161. [DOI] [PubMed] [Google Scholar]
- 21. Cann A, Calhoun LG, Tedeschi RG, et al. : A short form of the Posttraumatic Growth Inventory. Anxiety Stress Coping 2010;23:127–137. [DOI] [PubMed] [Google Scholar]
- 22. Prigerson HG, Maciejewski PK, Reynolds CF, et al. : Inventory of complicated grief: A scale to measure maladaptive symptoms of loss. Psychiatry Res 1995;59:65–79. [DOI] [PubMed] [Google Scholar]
- 23. Shear K, Frank E, Houck PR, et al. : Treatment of complicated grief. A randomized controlled trial. JAMA 2005;293:2601–2608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Kroenke K, Strine TW, Spitzer RL, et al. : The PHQ-8 as a measure of current depression in the general population. J Affect Disord 2009;114:163–173. [DOI] [PubMed] [Google Scholar]
- 25. Kroenke K, Spitzer RL, Williams JBW, et al. : The Patient Health Questionnaire Somatic, Anxiety, and Depressive Symptoms Scales: A systematic review. Gen Hosp Psychiatry 2010;32:345–359. [DOI] [PubMed] [Google Scholar]
- 26. Connor KM, Davidson JR: SPRINT: A brief global assessment of post-traumatic stress disorder. Int Clin Psychopharmacol 2001;16:279–284. [DOI] [PubMed] [Google Scholar]
- 27. DeSalvo KB, Fisher WP, Tran K, et al. : Assessing measurement properties of two single-item general health measures. Qual Life Res 2006;15:191–201. [DOI] [PubMed] [Google Scholar]
- 28. Shin C, Lee SH, Han KM, et al. : Comparison of the usefulness of the PHQ-8 and PHQ-9 for screening for major depressive disorder: Analysis of psychiatric outpatient data. Psychiatry Investig 2019;16:300–305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Pressler SJ, Subramanian U, Perkins SM, et al. : Measuring depressive symptoms in heart failure: Validity and reliability of the patient health questionnaire-8. Am J Crit Care 2011;20:146–152. [DOI] [PubMed] [Google Scholar]
- 30. Institute of Medicine: When Children Die: Improving Palliative and End-of-Life Care for Children and Their Families. Washington, DC: The National Academies Press, 2003. [Google Scholar]
- 31. Burns, JP, Sellers DE, Meyers EC, et al. : Epidemiology of death in the PICU at five U.S. teaching hospitals. Crit Care Med 2014;42:2101–2108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Meert KL, Briller SH, Schim SM, et al. : Examining the needs of bereaved parents in the pediatric intensive care unit. Death Stud 2009;33:712–740. [DOI] [PubMed] [Google Scholar]
- 33. Hogan NS, Schmidt LA: Testing the grief to personal growth model using structural equation modeling. Death Stud 2002;26:615–634. [DOI] [PubMed] [Google Scholar]
- 34. Xu W, Zhongfang F, He L: Growing in times of grief: Attachment modulates bereaved adults' posttraumatic growth after losing a family member to cancer. Psychiatry Res 2015;230:108–115. [DOI] [PubMed] [Google Scholar]
- 35. Tian X, Soloman DH: Grief and post-traumatic growth following miscarriage: The role of meaning reconstruction and partner supportive communication. Death Stud 2018;44:237–247. [DOI] [PubMed] [Google Scholar]
- 36. Shakespeare-Finch J, Lurie-Beck J: A meta-analytic clarification of the relationship between posttraumatic growth and symptoms of posttraumatic distress disorder. J Anxiety Disord 2014;28:223–229. [DOI] [PubMed] [Google Scholar]
- 37. Zisook S, Iglewicz A, Avanzino J, et al. : Bereavement: Course, consequences, and care. Curr Psychiatry Rep 2014;16:482. [DOI] [PubMed] [Google Scholar]
- 38. Morris S, Fletcher K, Goldstein R: The grief of parents after the death of a young child. J Clin Psychol Med Settings 2019;26:321–338. [DOI] [PubMed] [Google Scholar]
- 39. Klass D, Silverman PR, Nickman SL, eds.: Continuing Bonds: New Understandings of Grief. Washington, DC: Taylor & Francis, 1996. [Google Scholar]
- 40. Hawthorne DM, Youngblut JM, Brooten D: Parent spirituality, grief, and mental health at 1 and 3 months after their infant's/child's death in an intensive care unit. J Pediatr Nurs 2016;31:73–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Meert KL, Templin TN, Michelson KN, et al. : The bereaved parent needs assessment: A new instrument to assess the needs of parents whose children died in the pediatric intensive care unit. Crit Care Med 2012;40:3050–3057. [DOI] [PMC free article] [PubMed] [Google Scholar]