key: cord-0742402-gowj2pje authors: Sanderson, Emma A. M.; Humphreys, Sally; Walker, Fiona; Harris, Daniel; Carduff, Emma; McPeake, Joanne; Boyd, Kirsty; Pattison, Natalie; Lone, Nazir I. title: Risk factors for complicated grief among family members bereaved in intensive care unit settings: A systematic review date: 2022-03-10 journal: PLoS One DOI: 10.1371/journal.pone.0264971 sha: 0a362cda2c6e01fc90339b648f55ad93515b7f5e doc_id: 742402 cord_uid: gowj2pje BACKGROUND: Families of intensive care unit (ICU) decedents are at increased risk of experiencing complicated grief. However, factors associated with complicated grief in ICU and bereavement needs assessment are not available routinely. We aimed to conduct a systematic review identifying risk factors associated with complicated grief among family members of ICU decedents. MATERIALS AND METHODS: MEDLINE, EMBASE, CINAHL, PsycINFO, the Cochrane Library and Web of Science were searched to identify relevant articles. Observational studies and randomised and non-randomised controlled trials were included. Studies were screened and quality appraised in duplicate. Risk of bias was assessed using Newcastle-Ottawa Scale. A narrative synthesis was undertaken. RESULTS: Seven studies conducted across three continents were eligible. Four studies were of high quality. 61 risk factors were investigated across the studies. Factors associated with a decreased risk of complicated grief included age, patient declining treatment and involvement in decision-making. Factors associated with increased risk included living alone, partner, dying while intubated, problematic communication, and not having the opportunity to say goodbye. CONCLUSION: This systematic review has identified risk factors which may help identify family members at increased risk of complicated grief. Many of the studies has small sample sizes increasing the risk of erroneously reporting no effect due to type II error. Some factors are specific to the ICU setting and are potentially modifiable. Bereavement services tailored to the needs of bereaved family members in ICU settings are required. (PROSPERO registration ID 209503) Introduction Most people will experience bereavement during their lifetime. For many, the period following death can be disruptive and time is needed to adjust and accept the death of a loved one. Most individuals adapt to their loss and can integrate back into their life, often within a few months after the death [1] . However, a proportion of individuals experience prolonged, intense grieving that affects physical and mental health, and impacts on social and emotional wellbeing. This response is referred to as complicated grief [2] . Features may include guilt, anger, denial, emotional numbness or difficulty in engaging with activities [2] . Complicated grief is increasingly referred to as "prolonged grief disorder" and this term is now included in both the 11th edition of the International Classification of Diseases-11 and the 5th edition of the Diagnostic and Statistical Manual of Mental Disorders [3] . The proportion of individuals who experience complicated grief following a bereavement has been estimated to be between 5 and 10% in the general population [4] . In contrast, studies have shown that there is a much higher prevalence in family members bereaved in intensive care unit (ICU) settings, with studies estimating a prevalence of 46%-52% [5, 6] . Risk factors for complicated grief have been identified in non-ICU settings, which have led to the development of bereavement risk assessment screening tools for hospice and community settings [7] . However, given the highly technical environment and the frequency of sudden deaths in ICU compared to the wider hospital or community settings, it is possible that individual factors which are specific to the ICU context have not been considered in previous work. Furthermore, the COVID-19 pandemic has highlighted the unique challenges of supporting family members during and after end-of-life care in ICUs [8, 9] . Several reviews have been conducted previously to evaluate risk factors for complicated grief in the general population [10, 11] , and specifically in family caregivers [12, 13] . However, there are no published systematic reviews specifically investigating risk factors for complicated grief in the ICU context. Therefore, we aimed to systematically review published studies investigating factors associated with complicated grief among family members of patients dying in ICU (ICU decedents). The Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRIS-MA-P) [14] was used to develop the protocol for this systematic review and was then registered (PROSPERO ID: 209503). The PRISMA checklist was followed to conduct and report the review [15] . Studies included in the review were limited to those in which study participants were adult family members of patients admitted to adult ICUs. No limitations were in place for publication dates. Cohort studies, case-control studies, cross-sectional studies and randomised and non-randomised controlled trials were eligible to be included. Editorials, casereports, case-series, conference proceedings and qualitative research were excluded. Publications with only abstracts were also excluded. In addition, studies that did not provide a quantitative measure of association with the outcome were excluded. Furthermore, studies which investigated families of patients admitted to paediatric or neonatal critical settings were excluded. Due to a lack of translation resource, the review was limited to studies published in English. Search strategies were developed with the help of a medical librarian and the following databases were searched on 05/01/2022: MEDLINE, EMBASE, CINAHL, PsycINFO, the Cochrane Library and Web of Science. In addition, reference lists of the identified papers and review articles were searched to identify any further studies. We used a combination of controlled vocabulary and non-controlled terms to capture the main concepts in search strategies: "intensive care", "bereavement", "risk factors" and "complicated grief" (see S1 Table for complete search strategy). The results from the search were imported into a reference management software package and duplicates were removed. Abstracts were independently screened by two reviewers and full text of papers were retrieved and independently assessed in duplicate. Reviewers used eligibility criteria determined a priori. Disagreements were resolved by involving a third review author. The following data were extracted by a single reviewer and tabulated from eligible articles: publication information (study author, year and years conducted), study characteristics (including study design and statistical methodology), participant characteristics (population, sample size and relationship to deceased), risk factors (number reported or assessed, when they were ascertained and how they were ascertained) and outcome measures (definition, measurement tool, proportion of individuals with outcome and when outcome was assessed). The result of interest was the association between potential risk factors and complicated grief. Where possible odds ratios, risk ratios or hazard ratios were extracted for both univariable and multivariable analyses along with the relevant 95% confidence intervals. If this was not available, another relevant measure of association was extracted with its corresponding p-value. Data were extracted for all time points where results were reported. The Newcastle-Ottawa Scale (NOS) [16] was used to assess the risk of bias for observational studies. It was conducted independently by two reviewers, with the results synthesised together by a third reviewer. Any disagreements were resolved using the third reviewer. Results were tabulated and a narrative synthesis was conducted due to great variability in the risk factors assessed and measures of association, thus precluding a meta-analysis. Fig 1 shows the study selection process. Of the 574 articles identified, seven were eligible to be included in this review [6, [17] [18] [19] [20] [21] [22] The main characteristics of the studies are described in Tables 1 and 2 . The studies were conducted between 2005 and 2015, four were conducted in France and the remaining three in Australia, Canada and the USA. Almost all studies were observational in nature, apart from one interventional study [18] . Most relatives included were children or a spouse/partner of the patient. The Inventory of Complicated Grief (ICG) or ICG-Revised was used in six studies and the Core Bereavement Items questionnaire for one [17] . The ICG is a 19-item questionnaire with first-person statements scored using a Likert scale. A score greater than 25 suggests complicated grief is present [23] . The Core Bereavement Items questionnaire consists of 17 questions measuring the intensity of bereavement reactions by assessing 'core' bereavement experiences [24] . A total of 61 different risk factors were investigated across all the studies. The risk factors that each study investigated are presented in Table 3 . These risk factors were grouped into seven categories, discussed below. A small number of risk factors, mainly demographic ones, were assessed by multiple studies, but most by only a single study. The results of the quality appraisal are displayed in Table 4 . The overall quality of the studies varied. Four could be considered of higher quality, scoring 8 or 9 out of 9. The remaining three were of lower quality. Table 5 summarises the results by showing significant results at any time point for a risk factor associated with complicated grief and the direction of association, or no effect for univariable and multivariable analyses. Detailed results are available in the Supplement (S1 Table) . Demographics. Older patient age was associated with a decreased risk in two studies; one conducted univariable analysis (OR 0.98, 95%CI 0.96-0.99; p = 0.017) [6] and the other multivariable (OR 0.95, 95% CI 0.93-0.99, p<0.001) [18] . The gender of the family member had inconsistent associations with complicated grief; in univariable analyses, two studies showed no statistically significant results [17, 21] and the other two had opposing findings with one suggesting females were at an increased risk [6] and the other that men had higher scores on average [22] . Two studies indicated that if the family member was a spouse or partner, or if the individual was living alone, then they were at an increased risk. Two studies indicated that older age of the family member was associated with a decreased risk, although two showed no effect. No other demographic demonstrated an association. ICU characteristics/variables. The only ICU staff characteristic which showed an association was if the intensivist had been board certified before 2009 (OR 3.82, 95% CI 1.57-3.94; p = 0.003) [6] . Using vasopressors and dying while intubated were both associated with an increased risk. The patient declining ICU treatments was associated with a decreased risk of adverse outcomes among family members. Communication disagreements/conflict. Family disagreements over end-of life decisionmaking and problematic communication with the physician or nurse were associated with Table 1 . Publication data, study design information and details concerning the population for the 7 studies selected which assess potential risk factors for complicated grief. Year [17] and if the family was involved with the end-of-life decision-making (OR 0.57, 95% CI 0.35-0.92; p = 0.022) [6] were the only two factors associated with a decrease in complicated grief. Many of the other factors were associated with an increased risk during univariable analyses but not in multivariable analyses. There was some evidence suggesting not having the opportunity to say goodbye and being present at the time of death was associated with an increased risk. Other variables related to health and social needs. One study measured the concordance between three health-related questionnaires, measuring depression, post-traumatic stress disorder and social distress and risk of complicated grief. This demonstrated a slight concordance with complicated grief [25] , based on the kappa statistic. A previous history of depression (p = 0.09) or use of psychotropic medications (p = 0.74) appeared to have no bearing on the risk for an individual [21] . Coping mechanisms/determinants. One study assessed whether different coping mechanisms in bereavement affected the intensity of an individual's response [17] . This study indicated that three risk factors were significantly associated with increased bereavement intensity in both univariable and multivariable analysis: denial, self-blame and seeking emotional support. Acceptance of the death was the only coping determinant associated with a decreased risk, although this was no longer statistically significant in multivariable analyses. This systematic review identified seven studies that assessed 61 potential risk factors for complicated grief in family members bereaved in ICU settings. Despite the wide range of risk factors evaluated in the studies, most were investigated in a single study. Factors associated with a reduced risk of complicated grief included older age of the patient or family member, a patient declining ICU treatments, how prepared the family member was for the death and whether the family member was involved with end-of life care decision-making. Factors associated with an increased risk included living alone, being the spouse/partner of the deceased, the patient dying while intubated, problematic communication with doctors, and being present at time of death. Our review highlights risk factors that may help identify those at highest risk of poor bereavement outcomes. Several of these risk factors are amenable to interventions known to - - X - - - - Intensivist board certification<2009 - - X - - - - Nurses >2 years ICU experience - - X - - - - Patients Length of ICU stay - - X - - - - Need for vasopressors - - X - - - - Died will intubated - - X - - - - Immediate extubation vs terminal weaning - - - - - X - Refused treatment - - X - - - - Family disagreement EOL decision This study identified several ICU-specific factors not previously noted as risk factors for complicated grief. Where the patient had chosen to decline ICU treatments, family members How often at bedside - Medications for Depression - may well have felt treatments were aligned with patient preferences. In addition, they avoided the potential burdens associated with end-of-life care decision-making when patient wishes were uncertain. In contrast, family members experienced increased risk of complicated grief when a patient died while intubated and receiving mechanical ventilation or if receiving circulatory support with vasopressor therapy. These factors could reflect the impact on family members of the challenges in "negotiating a natural death" in patients receiving invasive life support therapies in a highly technical ICU setting [26] . However, in some cases they may signify family requests for escalating life support therapies when clinicians counsel comfort care, and therefore act as a marker for conflict between family members and clinicians or for family difficulties in accepting death [27] . Communication was another important identified factor. Communication between the patient and family member prior to death has been assessed previously [10, 13, 28] , but not communication with the medical staff in the context of bereavement outcomes. Sensitive and effective communication may be even more important in the ICU due to the distressing nature of the circumstances and environment. Problematic communication may make it harder for family members to understand what has happened and thus lead to difficulties during the grieving process. A previous study asking family members to make recommendations for improving end-of-life care identified communication as one of the key issues [29] . The global COVID-19 pandemic has amplified these factors and highlighted their importance [30] . Families who did not have the opportunity to say goodbye to their loved one had increased risk of complicated grief. However, being present at the time of death also increased the risk of complicated grief. This apparent paradox may be explained by the fact that these two factors describe different aspects of end of life care. Withdrawal of life sustaining therapies can lead to clinical signs or respiratory distress, which, if not adequately managed, families might find distressing if present at the time of death [31] . Ensuring that families are prepared for, and supported through, the process of withdrawal of life support may reduce the risk of complicated grief, offering potential targets to improve quality of end of life care. Evaluation of an individual's bereavement needs includes understanding coping mechanisms and responses to loss [32, 33] ; difficulties in these areas are associated with complicated grief in the non-ICU literature [10, 34] . However, these features were only assessed in a single study in this review, with limited significant associations in multivariable analyses possibly due to the small sample size [17, 35] . Previously published reviews, including those conducted systematically, reported factors associated with complicated grief in non-ICU populations. Two investigated risk factors in Yellow indicates the study showed no effect, green indicates the risk factor was associated with a decreased risk of complicated grief and red/orange indicates the study showed an increased risk for complicated grief. Lighter shades of either colour indicates that there is also a study which shows no effect for the association. Combined green and red/orange indicates conflicting evidence. Unless otherwise specified by the number of studies in brackets, the association was only investigated in a single study. Where applicable, further information is given in brackets relating to the risk factor. Abbreviations: family members with caregiving roles [12, 13] and the other two reviews did not specify the population of interest [10, 11] . Several risk factors were identified repeatedly, increasing the strength of the evidence-base. For example, being the decedent's spouse has been associated consistently with increased likelihood of complicated grief [10-13, 36, 37] and some studies also found parents of the decedent were at an increased risk [11, 13, 36] . Studies identified highly dependent relationships another risk factor [10, 11] . This review identified being a spouse as a risk factor in ICU decedent populations but without details of the spousal relationship. Consistent with previous reviews, we identified younger age of the deceased [10] [11] [12] and lack of preparedness for death as being risk factors for complicated grief [10, 12, 13] . A number of factors from previous reviews were non-significant or inconclusive in this review: female relative [10, 11] , low income [11] , prior psychiatric illness [10] [11] [12] , violent death [10, 11] and lack of education [11, 12] . This may be because some of these factors were not investigated in specific studies, the small sample size in most studies, or because these factors are of less relative important in the context of the ICU setting. There are several strengths to this review. Firstly, we searched multiple databases to help identify as many relevant studies as possible. Duplicate screening for study inclusion and quality appraisal aimed to reduce error and bias. In addition, we scrutinised studies to identify those which reported factors associated with complicated grief in secondary analyses, despite the primary aim and study design being unrelated, for example in a randomised controlled trial of a condolence letter [18] . This review had a number of limitations. Studies were restricted to those published in English, potentially leading to eligible studies being missed. Many of the studies has small sample sizes increasing the risk of erroneously reporting no effect due to type II error. Additionally, due to the lack of studies investigating the same risk factors, it was not possible to conduct a meta-analysis. Similar numbers of deaths occur in critical care and inpatient hospice settings in the UK [38] . Despite the high prevalence of complicated grief and mental illness experienced by families of ICU decedents, pro-active screening and bereavement support are uncommon in ICU settings [39] . Specialist palliative care involvement is and the impact of this on care around death in ICU is unclear. This is in contrast to hospice settings, in which bereavement care is an integral component of holistic care provided to patients and families, before, during and after death. Our findings highlight the need to develop and tailor bereavement follow-up services and screening for risk of complicated grief in the ICU context. Implementing such an approach brings its own challenges. Family member-specific demographics and characteristics are not included routinely in ICU clinical records. Specific risk factors related to perceived quality of communication, conflict and emotional burden may emerge during care but can be difficult to elicit reliably as part of routine care. Despite these challenges, possible models of ICU bereavement support services have been reported [40, 41] . ICU services can learn from other services such as hospices that have already integrated bereavement follow-up care pathways into standard practice, and implement similar models to ensure that families bereaved in ICU are provided with the support that they require. The COVID-19 pandemic with its high ICU death rates has been a catalyst for better palliative care and ICU integration in some areas, with proactive palliative care reviews in ICUs becoming routine [42] . Our review identified risk factors for that are modifiable within the ICU. Improving communication between clinicians and family members, involving families in end-of-life discussions and facilitating opportunities to say goodbye may be targets for interventions [6] . Similarly, patient involvement in decision-making about treatment escalation was associated with a markedly reduced risk of family members developing complicated grief. Whilst this is often impossible at the time of ICU admission due to impaired decision-making capacity, early engagement in advance (anticipatory) care planning with patients and families before a health crisis may have a positive impact on family members, particularly if it helps prepare them for proxy decision-making. Some of these factors have been the subject of clinical trials, with mixed results [43, 44] . Death is common in ICU settings and bereaved family members are at risk of poor outcomes. Yet, proactive screening and bereavement support are uncommon. This systematic review has identified potentially modifiable risk factors, some of which are specific to the ICU setting, which may help identify family members at highest risk of complicated grief. Our findings highlight the need to develop and tailor bereavement screening and follow-up services for family members bereaved in ICU settings. Supporting information S1 Table. Table of Grief and bereavement: what psychiatrists need to know Complicated Grief Prolonged grief disorder in DSM-5-TR: Early predictors and longitudinal measurement invariance. Australian & New Zealand Journal of Psychiatry Disturbed grief: prolonged grief disorder and persistent complex bereavement disorder Posttraumatic Stress and Complicated Grief in Family Members of Patients in the Intensive Care Unit Complicated grief after death of a relative in the intensive care unit A scoping review of bereavement risk assessment measures: Implications for palliative care Complicated Grief: What to Expect After the Coronavirus Pandemic Bereavement Support on the Frontline of COVID-19: Recommendations for Hospital Clinicians Predictors of Complicated Grief: A Systematic Review of Empirical Studies Prospective risk factors for complicated grief: A review of the empirical literature Complicated Grief: Risk Factors, Protective Factors, and Interventions Risks for Complicated Grief in Family Caregivers Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses The nature of death, coping response and intensity of bereavement following death in the critical care environment. Australian Critical Care Effect of a condolence letter on grief symptoms among relatives of patients who died in the ICU: a randomized clinical trial CAESAR: a new tool to assess relatives' experience of dying and death in the ICU Terminal weaning or immediate extubation for withdrawing mechanical ventilation in critically ill patients (the ARREVE observational study) Prevalence and predictors of severe grief reactions and desire for support following a death in the intensive care unit: a multicentre observational study Post-traumatic Stress Disorder and Complicated Grief are Common in Caregivers of Neuro-ICU Patients Inventory of Complicated Grief: A scale to measure maladaptive symptoms of loss Measuring core bereavement phenomena Medical Statistics at a Glance Negotiating natural death in intensive care Prevalence of Disagreement About Appropriateness of Treatment Between ICU Patients/Surrogates and Clinicians Predictors of Complicated Grief and Depression in Bereaved Caregivers: A Nationwide Prospective Cohort Study What Is Wrong With End-of-Life Care? Opinions of Bereaved Family Members End-of-life decisions and care in the midst of a global coronavirus (COVID-19) pandemic. Intensive and Critical Care Nursing Guidelines for the withdrawal of life-sustaining measures Guidance for bereavement needs assessment in palliative care. Help the Hospices Coping and wellbeing in bereavement: two core outcomes for evaluating bereavement support in palliative care Health outcomes of bereavement. The Lancet Long-term prevalence and predictors of prolonged grief disorder amongst bereaved cancer caregivers: A cohort study. Palliative and Supportive Care Who Needs Bereavement Support? A Population Based Survey of Bereavement Risk and Support Need Prevalence and determinants of complicated grief in general population Deaths in critical care and hospice-prevalence, trends, influences: a national decedent cohort study Time for change? A national audit on bereavement care in intensive care units Bereavement Care in the Adult Intensive Care Unit: Directions for Practice. Critical Care Nursing Clinics of North America Bereavement in critical care: A narrative review and practice exploration of current provision of support services and future challenges Palliative care in critically ill COVID-19 patients: the early New York City experience A Randomized Trial of a Family-Support Intervention in Intensive Care Units Randomized Trial of Communication Facilitators to Reduce Family Distress and Intensity of End-of-Life Care All authors made substantial contributions to the conception or design of the work; NP/SH/ FW/DH/ES/NL undertook searches, screening, data extraction, full text review, quality appraisal; NL/ES drafted the work; all authors revised the work critically for important intellectual content; all authors grant final approval of the version to be published. NL gives agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Template data collection forms and data extracted from included studies are available from the authors on request. PROSPERO reference: 209503.