Research paperEvaluation of the factor structure, prevalence, and validity of disturbed grief in DSM-5 and ICD-11
Introduction
The American Psychiatric Association (APA) has made a significant change to the 5th edition of their Diagnostic and Statistical Manual of Mental Disorders (DSM-5, APA, 2013) concerning the classification of disturbed grief, by including Persistent Complex Bereavement Disorder (PCBD). PCBD is included as a condition for further study in Section 3, and can be classified as “Other Specified Trauma- and Stressor-Related Disorder.” PCBD comprises 16 symptoms, organized under two symptom-clusters, namely separation distress and a second symptom-cluster with additional symptoms, that is subdivided into signs of “reactive distress to the death” and “social/identity disruption”. A diagnosis of PCBD requires that the person has experienced the death of someone with whom s/he had a close relationship, and the endorsement of at least one separation distress symptom and six additional symptoms. Additionally, these symptoms must be associated with functional impairment, and have persisted for at least 12 months after the death. The World Health Organization has proposed a similar change to the forthcoming 11th edition of the International Classification of Diseases (ICD-11) by adding Prolonged Grief Disorder (PGD) to the category of “Disorders associated with stress” (WHO, 2018). PGD includes a description of 12 symptoms, categorized into separation distress and additional symptoms. A diagnosis of PGD requires having experienced loss, combined with at least one of two symptoms of separation distress and at least one of ten additional symptoms. These symptoms must be associated with functional impairment and have persisted for at least 6 months after the death (WHO, 2018).
Both conditions of disturbed grief in DSM-5 and ICD-11 are intended to capture the pervasive emotional distress that can occur following bereavement in a significant minority of people. Disturbed grief occurs in 5–10% of people confronted with natural losses (Lundorff et al., 2017) and a slightly higher percentage of people confronted with unnatural losses, including accidents, suicide, and homicide (cf. Kristensen et al., 2012). An apparent difference between both criteria-sets is that ICD-11 PGD includes a shorter list of symptoms. This reflects the aim of the ICD system to simplify the classification of disturbed grief and enhance clinical utility, which is particularly relevant in low-resourced contexts (Maercker et al., 2013). Differences in criteria for disturbed grief in DSM-5 and ICD-11 parallel the way that posttraumatic stress disorder (PTSD) is described in both systems, with DSM-5 distinguishing 20 symptoms and ICD-11 only 6 symptoms. Studies comparing PTSD in DSM-5 and ICD-11 have suggested that the predictive validity of both criteria-sets is similar but that ICD-11 PTSD generates lower prevalence rates (Hansen et al., 2015; Shevlin et al., 2018).
In contrast with the growing number of studies comparing psychometric properties of PTSD in DSM-5 and ICD-11, few studies have evaluated the psychometric properties of disturbed grief in DSM-5 and ICD-11. Five recent studies are pertinent to this issue. Using data from a community sample, Maciejewski et al. (2016) compared the prevalence and validity of four criteria-sets for disturbed grief, including DSM-5 PCBD, PGD as per ICD-11, a slightly different formulation of PGD put forth by Prigerson et al. (2009), and criteria for complicated grief (CG) proposed by Shear et al. (2011). Outcomes showed that the first three criteria-sets yielded similar prevalence rates and predictive validity, whereas CG criteria performed poorly as indicators of disturbed grief. Cozza et al. (2016) studied bereaved military family members, comparing DSM-5 PCBD, PGD as per Prigerson et al. (2009), and CG as per Shear et al. (2011). They found that CG criteria performed best in terms of distinguishing between people with disturbed and non-disturbed grief. However, this conclusion has been critiqued since they excluded participants with subthreshold grief symptoms (Maciejewski and Prigerson, 2017, Smid and Boelen, 2016). A further comparison between these three sets of criteria was conducted by Mauro et al. (2017); in a sample of treatment seeking bereaved individuals, they also found that CG performed better than DSM-5 PCBD and PGD as per Prigerson et al. (2009). Tay et al. (2016) compared different factor-models of disturbed grief in a sample of West Papuan refugees. They found that a six-factor structure (based on Simon et al. 2011) fit the data better compared to factor solutions reflecting DSM-5 PCBD and PGD as per Prigerson et al. Finally, Claycomb et al. (2016) evaluated the factor structure and clinical correlates of DSM-5 PCBD in a large sample of bereaved Bosnian adolescents and found preliminary evidence for a multidimensional structure of PCBD in this group. All these studies have advanced our understanding of criteria for disturbed grief. However, only the study by Maciejewski et al. (2016) evaluated PGD-criteria as proposed for ICD-11 and that evaluation is currently less relevant because criteria for ICD-11 PGD have changed since that study. Moreover, prior studies relied on different samples and grief measurement instruments which limits the possibility to compare findings between studies.
Altogether, there is a need to enhance knowledge on the psychometric properties of different criteria-sets for disturbed grief. Evaluation of the DSM-5 and ICD-11 criteria is particularly relevant because these are included in the most widely used diagnostic systems worldwide and—as such—are mostly used in clinical and research settings. Identifying a psychometrically sound conceptualization of disturbed grief is important for theoretical reasons (to inform research on the aetiology and maintaining mechanisms of disturbed grief) and clinical practice (to foster the identification of people in need of support). Comparing DSM-5 PCBD-criteria and ICD-11 PGD-criteria is particularly important in order to know whether research findings based on one of these criteria-sets can be generalized to people meeting criteria for the other set.
The present study represented a preliminary attempt to evaluate the psychometric properties of disturbed grief as per DSM-5 and ICD-11, using self-reported data from a large Dutch bereaved community sample. The first aim was to examine the factor structure of DSM-5 PCBD and ICD-11 PGD. Items from the Inventory of Complicated Grief Revised (ICG-R, Prigerson & Jacobs, 2001) and Beck Depression Inventory-II (BDI-II; Beck et al., 1996) were selected for inclusion in this study according to how closely they mapped onto both criteria-sets. We evaluated the fit of three PCBD-models (mirroring the clustering of PCBD-symptoms in DSM-5): (i) a one-factor model with all PCBD-items loading on a single factor, (ii) a two-factor model with PCBD-items forming distinct, but correlated clusters of separation distress (factor 1) and reactive distress and social/identity disruption (factor 2), (iii) a three-factor model with PCBD-items forming distinct, but correlated clusters of separation distress (factor 1), reactive distress (factor 2), and social/identity disruption (factor 3). In addition, we evaluated the fit of two PGD-models: (i) a one-factor model and (ii) a two-factor model with PGD-items clustering into correlated factors of separation distress and additional symptoms—resembling the ICD-11 proposal. Because of the scant evidence regarding the factor structure of PCBD and PGD no hypotheses were formulated. A second aim was to determine prevalence rates of disturbed grief as per DSM-5 and ICD-11, the diagnostic agreement between diagnostic criteria, and the number of ‘unique’ cases of PCBD and PGD (i.e., individuals meeting criteria for PCBD but not PGD, and vice versa). A third aim was to evaluate the performance of individual items of DSM-5 PCBD and ICD-11 PGD as indicators of PCBD and PGD. To this end, we determined the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of PCBD-items and PGD-items in relation to a diagnosis of probable PCBD and a diagnosis of probable PGD, respectively (cf. Nickerson et al., 2016). The fourth aim was to evaluate the concurrent validity of PCBD and PGD diagnoses. To this end, we compared symptom-levels of overall disturbed grief, depression, and PTSD between cases and non-cases of DSM-5 PCBD and between cases and non-cases of ICD-11 PGD. The fifth aim was to examine the predictive validity of PCBD and PGD diagnoses, using data from a subset of participants completing additional measures one year after baseline. We examined the degree to which meeting criteria for caseness according to both criteria-sets at T1 was associated with higher levels of overall disturbed grief, depression, and PTSD at T2 (one year later), while controlling for baseline symptom-levels.
Section snippets
Participants and procedure
Data were originally collected in the context of an IRB approved research project conducted in the Netherlands (see e.g., Boelen et al., 2015; Djelantik et al., 2017). Participants were recruited via professional and lay mental health care workers (e.g., grief counsellors, therapists) who distributed questionnaires among bereaved persons. Over 700 individuals entered the project. For this study, we selected data from 512 participants who were at least 18 years of age and bereaved more than 5
Participant characteristics
The mean age of participants was 53.83 (SD = 13.92) years. Most participants (n = 390; 76.2%) were women; 307 participants (60.0%) had followed primary/secondary education only, whereas 202 participants (39.5%) had been to college or university; 338 participants (66.0%) lost a spouse/partner, 55 (10.7%) a child, and 119 (23.2%) someone other than a partner or child (e.g., friend, parent, sibling). The mean time since loss was 28.64 (SD = 26.30, range 6–120) months; losses were due to a natural
Discussion
This study evaluated psychometric properties of disturbed grief as introduced in DSM-5, named PCBD (APA, 2013) and as proposed for ICD-11, named PGD (WHO, 2018). The first aim was to examine the factor structure of DSM-5 PCBD and ICD-11 PGD. Confirmatory factor analyses showed that the DSM-5 model, with PCBD-symptoms forming three distinguishable (but related) symptom-clusters of separation distress, reactive distress, and social/identity disruption fit the data well and fit better than the
Ethical approval
Ethical approval for conducting this study was obtained from an institutional review board (IRB).
Limitations of the study
Limitations include our reliance on self-reported data and symptoms of PCBD and PGD being derived from two scales.
Conflict of interest
The authors declare no conflicts of interest.
Author's contribution
PB, LL, and AN undertook the statistical analyses. PB and LL wrote the completed draft of the manuscript. GS assisted in study design, statistical analyses, and protocol. All authors contributed to and have approved the final manuscript.
Role of funding source
We did not receive funding for this research.
Acknowledgments
Participants are thanked for their participation in this study.
References (33)
- et al.
Symptoms of prolonged grief, post-traumatic stress, and depression after loss in a Dutch community sample: a latent class analysis
Psychiatry Res.
(2017) - et al.
Prevalence of prolonged grief disorder in adult bereavement: A systematic review and meta-analysis
J. Affect. Disord.
(2017) - et al.
Proposals for mental disorders specifically associated with stress in the International Classification of Diseases-11
Lancet
(2013) Diagnostic and Statistical Manual of Mental Disorders
(2000)Diagnostic and Statistical Manual of Mental Disorders
(2013)- et al.
Beck Depression Inventory-II
(1996) - et al.
Reliability and validity of the Dutch version of the Inventory of Traumatic Grief
Death Stud.
(2003) - Boelen, P.A., Djelantik, A.A.A.M.J., Keijser, J.de, Lenferink, L.I.M., Smid, G.E., in press. Further validation of the...
- et al.
Cognitive behavioural variables mediate the impact of violent loss on post-loss psychopathology
Psychol. Trauma Theory Res. Pract. Policy
(2015) - et al.
Commentary on the inclusion of persistent complex bereavement-related disorder in DSM-5
Death Stud.
(2012)
The traumatic grief inventory self report version (TGI-SR): introduction and preliminary psychometric evaluation
J. Loss Trauma
Persistent complex bereavement disorder symptom domains relate differentially to PTSD and depression: a study of war-exposed Bosnian adolescents
J. Abnorm. Child Psychol.
Performance of DSM-5 persistent complex bereavement disorder criteria in a community sample of bereaved military family members
Am. J. Psychiatry
Low specificity of symptoms on the post-traumatic stress disorder (PTSD) symptom scale
Br. J. Clin. Psychol.
Reliability and validity of a brief instrument for assessing post-traumatic stress disorder
J. Trauma. Stress
Less is more? Assessing the validity of the ICD-11 model of PTSD across multiple trauma samples
Eur. J. Psychotraumatology
Cited by (66)
What Is Prolonged Grief Disorder and How Can Its Likelihood Be Reduced?
2023, Evidence-Based Practice of Palliative Medicine, Second EditionPredictors of complicated grief in mourners of sudden cardiac death
2021, American Journal of Emergency Medicine