MedicineLakex

 

Untitled

Treating Prolonged Grief DisorderA Randomized Clinical Trial Richard A. Bryant, PhD; Lucy Kenny, PhD; Amy Joscelyne, PhD; Natasha Rawson, MPsychol;Fiona Maccallum, PhD; Catherine Cahill, MPsychol; Sally Hopwood, MPsychol;Idan Aderka, PhD; Angela Nickerson, PhD IMPORTANCE Prolonged grief disorder (PGD) is a potentially disabling condition that affects
approximately 10% of bereaved people. Grief-focused cognitive behavior therapy (CBT) has
been shown to be effective in treating PGD. Although treatments for PGD have focused on
exposure therapy, much debate remains about whether exposure therapy is optimal for PGD.
OBJECTIVE To determine the relative efficacies of CBT with exposure therapy (CBT/exposure)
or CBT alone for PGD.
DESIGN, SETTING, AND PARTICIPANTS A randomized clinical trial of 80 patients with PGD
attending the outpatient University of New South Wales Traumatic Stress Clinic from
September 17, 2007, through June 7, 2010.
INTERVENTIONS All patients received 10 weekly 2-hour group therapy sessions that consisted
of CBT techniques. Patients also received 4 individual sessions, in which they were
randomized to receive exposure therapy for memories of the death or supportive counseling.
MAIN OUTCOMES AND MEASURES Measures of PGD by clinical interview and self-reported
measures of depression, cognitive appraisals, and functioning at the 6-month follow-up.
RESULTS Intention-to-treat analyses at follow-up indicated a significant quadratic
time × treatment condition interaction effect (B [SE], 0.49 [0.16]; t
0.18-0.81]; P = .003), indicating that CBT/exposure led to greater PGD reductions than CBTalone. At follow-up, CBT/exposure led to greater reductions in depression (B [SE], 0.35 [0.12];t = 2.83 [95% CI, 0.11-0.60]; P = .005), negative appraisals (B [SE], 0.68 [0.25]; = 2.66 [95% CI, 0.17-1.18]; P = .009), and functional impairment (B [SE], 0.24 [0.08]; = 3.01 [95% CI, 0.08-0.40]; P = .003) than CBT alone. In terms of treatment completers, fewer patients in the CBT/exposure condition at follow-up (14.8%) met criteriafor PGD than those in the CBT condition (37.9%) (odds ratio, 3.51; 95% CI, 0.96-12.89;χ2 = 3.81; P = .04).
CONCLUSIONS AND RELEVANCE Including exposure therapy that promotes emotional
processing of memories of the death is an important component to achieve optimal
reductions in PGD severity. Facilitating emotional responses to the death may promote
greater changes in appraisals about the loss, which are associated with symptom reduction.
Promotion of emotional processing techniques in therapies to treat patients with PGD is
needed.
Author Affiliations: School of
TRIAL REGISTRATION anzctr.org.au Identifier: ACTRN12609000229279
Psychology, University of New SouthWales, Sydney, Australia (Bryant,Kenny, Joscelyne, Rawson,Maccallum, Cahill, Hopwood,Nickerson); Department ofPsychology, University of Haifa, Haifa,Israel (Aderka).
Corresponding Author: Richard A.
Bryant, PhD, School of Psychology,
Mathews Bldg, Room 836, University
of New South Wales, Sydney, NSW,
Published online October 22, 2014.
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of New South Wales User on 10/22/2014
Research Original Investigation Treating Prolonged Grief Disorder have received considerable attention in recent for PGD by comparing CBT that contains exposure therapy (CBT/ years. The upcoming International Statistical Classi- exposure) and CBT alone.
fication of Diseases, 11th Revision, has proposed a new diag-nosis termed prolonged grief disorder (PGD) to describe thiscondition.1 This disorder involves persistent yearning for the deceased and associated emotional pain, difficulty inaccepting the death, a sense of meaninglessness, bitterness about the death, and difficulty in engaging in new activities.
Participants were bereaved patients who sought treatment at To diagnose PGD, these symptoms need to persist at least 6 the University of New South Wales Traumatic Stress Clinic from months after the death because evidence suggests that September 17, 2007, through June 7, 2010, after the death of bereaved people with these symptoms beyond this time are parents (23 patients [29%]), partners (24 [30%]), children (25 most likely to develop long-term impairment.2 Prolonged [31%]), or others (8 [10%]). The causes of death included sud- grief disorder has been associated with increased rates of den illness (16 [20%]), chronic illness (43 [54%]), accident (12 mental disorder and sleep disturbance,3,4 suicidality,5-7 poor [15%]), or suicide (9 [11%]). Inclusion criteria were that the pa- health behaviors,2,8 cardiovascular and cancer conditions, tient had experienced bereavement at least 12 months earlier and work and social impairment.9-11 Although most people and satisfied criteria for PGD. Exclusion criteria included a his- adapt to the loss by 6 months,12 studies indicate that tory of psychosis, current substance dependence, borderline approximately 10% of bereaved people experience PGD2,13; personality disorder, severe suicidal risk, an inability to con- this percentage equates to approximately 1 million new verse in English, and being younger than 17 years or older than cases of PGD each year in the United States alone, represent- 70 years. Participants underwent initial screening by tele- ing a major public health issue. Although PGD and depres- phone, and full assessments were conducted only for partici- sion can present with sadness, crying, suicidal ideation, and pants who did not report any exclusion criteria on telephone sleep disturbances, PGD is distinct from depression by the screening. The characteristics of patients are presented in preoccupation with yearning for the deceased. Accordingly, Table 1. This study was approved by the Human Research Eth- comorbidity between PGD and major depressive episodes ics Committee of New South Wales, and all participants com- ranges from only 50% to 70%.10,14-16 Factor analytic pleted written informed consent.
studies2,17,18 demonstrate that the impairment secondary toPGD extends beyond bereavement-related anxiety or Diagnostic Interview Reinforcing the proposal that PGD is distinct from de- The Complic ated Grief Assessment25 is a clinic ian- pression is evidence that P GD does not respond to administered semistructured interview for assessing PGD. The antidepressants.19-21 In contrast, emerging evidence suggests Complicated Grief Assessment interview is based on the In- that targeted cognitive behavior therapy (CBT) is an effective ventory of Complicated Grief26 and provides a diagnosis and strategy for reducing specific symptoms of PGD.22,23 In a semi- severity index of complicated grief. The interview assesses for nal study,22 patients with PGD were randomized to 16 sessions the presence of separation distress (criterion A) and difficulty of CBT (consisting of revisiting the death to facilitate emo- accepting the death, emotional numbness, bitterness, diffi- tional processing of the death, fostering positive memories of culty reengaging in life, and a sense of purposelessness and the relationship, and promoting new goals; in addition to usual meaninglessness (criterion B). A diagnosis of complicated grief CBT practice, this treatment also included imaginal dialogue is given if 6 months has passed since the death, criteria A and with the deceased (to address outstanding issues the patient felt B have been met for at least 6 months, and evidence of func- toward the deceased) or interpersonal psychotherapy (fo- tional impairment exists (criterion C).
cused on developing interpersonal skills, including understand- The Clinician-Administered PTSD Scale–127 is a struc- ing of the patient's relationship with the deceased). Cognitive tured clinical interview that indexes the 17 symptoms de- behavior therapy achieved markedly superior outcomes rela- scribed by the DSM-IV criteria for posttraumatic stress disor- tive to interpersonal psychotherapy. In a subsequent study,23 der (PTSD). Each symptom is rated on a 5-point scale in terms PGD patients were administered CBT with exposure therapy of severity and frequency of the symptoms in the past month.
(CBT/exposure) (involving repeated revisiting of the death) fol- This measure was used to assess for PTSD diagnosis before lowed by cognitive restructuring or the same components in re- treatment. The Mini-International Neuropsychiatric Inter- verse order. These studies highlight that exposure-based thera- view (version 5.5)28 was used to assess for comorbid Axis I de- pies are an effective means of alleviating the symptoms of PGD.
pression and anxiety disorders.
Because both of these studies integrated exposure therapy asa core component of the treatment, the outstanding question Self-report Measures is the extent to which exposure therapy is essential for treating Additional psychopathology measures included the Beck De- PGD. Despite their success, exposure-based approaches can elicit pression Inventory, Second Edition,29 to measure depression marked distress during the reliving. Although this distress is and the Posttraumatic Cognitions Inventory (PTCI)30 to in- transient, it can limit the appeal of exposure therapy for PGD.
dex cognitive responses to trauma. The PTCI is a 36-item self- Therefore, an understanding of whether exposure therapy is es- report scale that yields 3 factors, including negative cogni- JAMA Psychiatry Published online October 22, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of New South Wales User on 10/22/2014
Treating Prolonged Grief Disorder Original Investigation Research Table 1. Participant Characteristics Treatment Conditiona Age, mean (SD), y Time since death, mean (SD), y Educational level, mean (SD), y Relationship of deceased Abbreviations: CBT, cognitive behavior therapy; CBT/exposure, CBTwith exposure therapy; PTSD, Comorbid depression posttraumatic stress disorder.
a Unless otherwise indicated, data are Comorbid anxiety disorder expressed as number (percentage) Comorbid substance use disorder of patients.
Logic rating, mean (SD)b Calculated using the Expectancy rating, mean (SD)b tions about one's self (PTCI-Self), negative cognitions about severity total score. Participants were assigned according to a the world (PTCI-World), and self-blame.
random numbers system administered by an individual who The World Health Organization Quality of Life short-form was independent of the study and who worked at a site that assessment (WHOQOL-BREF)31 is a World Health Organiza- was distant from the treatment center. Every 6 months, allo- tion instrument that assesses quality of life across the follow- cation was amended by the independent assigner to ensure that ing 4 domains of functioning: physical health (daily living, pain, sex and grief severity were balanced across conditions. Pa- and work capacity), psychological health (mood, self- tients were considered dropouts if they commenced taking esteem, and concentration), social relationships (personal re- medication after starting treatment to ensure that observed lationships, social support, and sexual activity), and environ- changes could not be attributed to medication. Adverse reac- ment (financial resources, health care, and home environment).
tions were monitored by the therapist and recorded on the ba- The WHOQOL-BREF demonstrates good discriminant valid- sis of significant exacerbation of symptoms requiring re- ity, content validity, internal consistency, and test-retest moval or respite from the program. The Figure summarizes reliability.32 Functional impairment was defined by Austra- participant flow. Eighty patients were randomized into the lian norms for each domain of the WHOQOL-BREF.33 study and were randomized to CBT/exposure (n = 41) or CBT At the completion of session 1 and after the rationale had alone (n = 39). Sixty-one patients (76%) completed treat- been explained, patients completed the Credibility/ ment, and 56 patients (70%) completed the 6-month fol- Expectancy Questionnaire.24 Specifically, patients in the CBT/ low-up assessment.
exposure or CBT group rated their confidence in the treat- Initial pretreatment assessments were conducted before ment (1, not at all confident; 10, extremely confident) and the randomization. Posttreatment and 6-month follow-up assess- logic of the treatment (1, not at all logical; 10, extremely logi- ments were conducted by independent clinicians who were un- cal). All measures were administered at each assessment, with aware of the treatment condition of participants. Blindness was the exception of the therapy confidence and logic ratings, maintained by ensuring that clinicians who conducted assess- which were administered at baseline only.
ments did not have access to participant notes or condition al-location of participants.
ProcedureParticipants were informed that they would be randomized to Treatment Conditions one of 2 treatment conditions. Randomization was con- Therapy consisted of 10 weekly 2-hour group sessions and 4 ducted by a process of minimization stratified on sex and grief weekly 1-hour individual sessions. Group therapy was con- JAMA Psychiatry Published online October 22, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of New South Wales User on 10/22/2014
Research Original Investigation Treating Prolonged Grief Disorder sions. In the initial individual session, participants were in- Figure. Participant Flow in the Study structed in the rationale for exposure therapy and the need tominimize avoidance of grief memories to manage the dis- 100 Patients referred
tress and integrate the loss into one's memory. Participants then 1 Poor English
relived the time they experienced the death of the person for 2 Recent medication
40 minutes, including providing accounts in the first person 3 Suicidal/borderline risk
and present tense of their emotional, cognitive, sensory, and 4 Refused treatment
somatic reactions; if patients completed the narrative in less than 40 minutes, they were directed to repeat the narrative un-til 40 minutes had lapsed. Distress ratings were provided by 41 Randomized to CBT/exposure
39 Randomized to CBT alone
patients every 10 minutes to ensure they were engaging withemotional content. This process continued for 3 subsequent 31 Completed
30 Completed
sessions; in later exposure sessions, therapists focused pa- 5 Declined participation
6 Declined participation
tients' attention on critical aspects of the memories that elic- 2 Commenced medication use
2 Commenced medication use
2 Other deaths
3 Work commitments
ited the most distress to ensure that they were sufficiently en- gaged. Exposure sessions were not audiotaped, but patientswere instructed to conduct the exposure at least once be- 27 Completed follow-up
29 Completed follow-up
tween sessions for homework by following the same tech- 2 No contact
3 Dropouts contacted
nique used in the therapy session and to maintain the reliv-ing for 40 minutes.
CBT indicates cognitive behavior therapy; CBT/exposure, CBT with exposuretherapy; and PGD, prolonged grief disorder.
CBT AloneThe group therapy was identical to the treatment provided in ducted by 2 experienced master's-level clinical psychologists, the CBT/exposure condition. In each of the 4 weekly 1-hour in- and individual sessions were conducted by 1 clinical psy- dividual sessions that commenced after group therapy ses- chologist (from a pool of 6 clinical psychologists) who were sion 2, participants were invited to discuss anything they trained to use treatment manuals and who received weekly wished. The facilitators of the individual sessions (the same supervision from one of us (R.A.B.). All therapists provided therapists who conducted the exposure sessions) did not in- each type of treatment. Patients randomized to each condi- struct participants in any exposure-based approaches. Facili- tion participated in distinct group sessions dedicated to that tators responded to participants in a nondirective manner. To equate to the homework activity of those in the exposure con-dition, participants were asked to complete a diary of grief states between sessions.
Session 1 focused on clarification of group processes, educa-tion about grief, and an overview of treatment components.
Treatment Fidelity Session 2 included a detailed review of the rationales for the Audiotapes of 48 individual therapy sessions (20% of the treatment strategies (including the need to manage avoid- 240 completed individual therapy sessions) and 20 group ance, rumination, and excessively negative appraisals; to fa- therapy sessions (20% of the 100 therapy sessions) were cilitate positive memories; and to develop new goals). Ses- randomly selected and rated by 3 clinicians experienced in sions 3, 4, and 5 focused on teaching specific strategies in CBT who were independent of the study. Raters listened to cognitive restructuring to reframe common maladaptive grief- audiotapes and rated the presence or absence of each of related appraisals (eg, hopelessness and guilt). Session 6 ad- 44 treatment components without regard to treatment con- dressed rumination management, including identification of dition or treatment session. Raters also indicated the quality the merits and costs of repetitive thinking and distraction tech- of the therapy provided on a 7-point scale (1, unacceptable; niques. This session also included a letter-writing task in which 7, very good). All individual sessions of CBT/exposure participants expressed unresolved issues they wished to com- included adequate exposure sessions, and no individual municate to the deceased, which was intended to facilitate the CBT session included exposure. The mean (SD) quality cognitive restructuring. Session 7 continued with cognitive rating for treatment components across conditions was challenging and letter writing to the deceased and began fa- cilitation of positive memories, in which participants de-scribed memories of positive experiences with the deceased.
Statistical Analysis Session 8 continued letter writing and facilitation of positive Power analysis was derived with G*Power 3.34 We adopted a memories and initiated steps for new goals and activities. Ses- Hedges g effect size estimate in the power analysis to accom- sion 9 focused predominantly on identification of future goals modate smaller samples. Interpretation of the g statistic is cat- and steps to achieve them. This strategy was continued in ses- egorized as large (>0.80), medium (0.50-0.79), or small (0.20- sion 10, which also developed relapse prevention strategies for 0.49). To achieve a power of 80%, we based calculations on a high-risk times (eg, anniversaries). After group session 2, par- previous study that has used exposure-based therapy with ticipants commenced 4 weekly 1-hour individual therapy ses- PGD22 and determined that we required 80 participants to de- JAMA Psychiatry Published online October 22, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of New South Wales User on 10/22/2014
Treating Prolonged Grief Disorder Original Investigation Research tect an 8-point difference in Inventory of Complicated Grief significant linear time × treatment interaction from pretreat- scores at α = .05.
ment to follow-up (B [SE], −1.57 [0.39]; t = −4.00 [95% CI, The primary focus of analyses was intention to treat. Using −2.36 to −0.79]; P < .001), indicating that CBT/exposure led to commercially available software (SPSS, version 21; SPSS Inc), significantly greater decreases in prolonged grief symptoms we adopted hierarchical linear models to study treatment ef- relative to CBT. The level 2 model also indicated a significant fects because this method allows the number of observations quadratic time × treatment interaction (B [SE], 0.49 [0.16]; to vary between participants and effectively handles missing = 3.08 [95% CI, 0.18-0.81]; P = .003), indicating that par- data.35 Hierarchical linear models use a multilevel data struc- ticipants in the CBT/exposure condition evidenced a stronger ture that includes fixed and random intercepts and slopes. Time quadratic relationship between time and PGD symptoms com- (linear and quadratic), treatment condition, and their interac- pared with those in the CBT condition.
tion were included in the model. Fixed-effects parameters were Similarly, in relation to depressive symptoms, we found a tested with the Wald test (2-tailed t test) and 95% CIs.
significant linear time × treatment interaction (B [SE], −1.12 Levels 1 and 2 models were estimated; however, results fo- = −3.59 [95% CI, −1.74 to −0.50]; P < .01) and a sig- cus on level 2 models. Intraclass correlations were calculated nificant quadratic time × treatment interaction (B [SE], 0.35 to examine potential clustering effects of the data according = 2.83 [95% CI, 0.11-0.60]; P = .005), indicating that to treatment group. Results revealed no significant clustering patients in the CBT/exposure condition had a greater reduc- effects for any of the primary or secondary outcome mea- tion of depression over time relative to those in the CBT con- sures after treatment or at follow-up; therefore, a level 3 (ie, dition. In terms of the PTCI-Self, we found a significant linear group-level) model was not estimated. To determine the ef- time × treatment interaction (B [SE], −2.39 [0.64]; t fects of treatment on longer-term outcome, we focused analy- [95% CI, −3.65 to −1.12]; P < .001) and a significant quadratic ses on follow-up rather than posttreatment data. We calcu- time × treatment interaction (B [SE], 0.68 [0.25]; t lated effect sizes based on previous recommendations for [95% CI, 0.17-1.18]; P = .009), indicating that patients in the multilevel models using the following formula36: CBT/exposure condition had greater reduction in negative ap-praisals about the self than those in the CBT condition. We d = B × [Time /Raw Score of Pretreatment SD].
found a significant linear time × treatment interaction (B [SE], We also calculated PGD diagnostic rates and number −0.99 [0.21]; t = −4.77 [95% CI, −1.41 to −0.58]; P < .001) needed to treat, which were based on treatment completers.
and a significant quadratic time × treatment interaction (B [SE], In addition, we calculated the number of patients needed to 0.24 [0.08]; t = 3.01 [95% CI, 0.08-0.40]; P = .003) of PTCI- treat as 1 divided by the proportion responding in the CBT/ World scores, indicating that CBT/exposure resulted in better exposure group as an estimate of the number of patients who appraisals about the world than CBT. Regarding the PTCI–Self- would need to be given CBT/exposure for 1 of them to achieve blame score, we also found a significant linear time × treat- a response outcome (defined as no longer meeting criteria for ment interaction (B [SE], −0.50 [0.18]; t = −2.87 [95% CI, PGD) and who would not have achieved it with CBT alone. Ef- −0.85 to −0.16]; P = .005), indicating that CBT/exposure led to ficacious treatments typically have a number needed to treat greater decreases in negative cognitions relating to self- that ranges from 2 to 4.37 We also calculated the differential blame than CBT.
treatment gains made by patients according to antidepres- In terms of psychological functioning, we found a signifi- sant use and PGD secondary to sudden/traumatic death.
cant linear time × treatment interaction (B [SE], 2.24 [0.52];t = 4.31 [95% CI, 1.21-3.26]; P < .001) and a significant nega- tive quadratic time × treatment interaction (B [SE], −0.70 [0.20];t = −3.41 [95% CI, −1.10 to −0.29]; P = .001), indicating that CBT/exposure led to greater psychological functioning over Table 1 presents the patient characteristics. Planned com- time than CBT. In terms of social functioning, we found no sig- parisons of treatment completers and treatment dropouts nificant linear interaction, but a significant quadratic indicated no differences between conditions on any pre- time × treatment interaction (B [SE], −0.70 [0.25]; t treatment psychopathology, demographic, or bereavement- [95% CI, −1.18 to −1.96]; P = .007) suggested that CBT/ related factor. Similarly, we found no differences in those exposure also led to greater social functioning over time than who did and did not drop out of treatment on any of these variables. Patients in each condition did not differ in their Analyses of diagnostic status and number needed to treat expectations about therapy success. No adverse effects of were based on treatment completers. After treatment comple- the treatments were reported, although 2 patients dropped tion, marginally fewer participants in the CBT/exposure con- out of the CBT/exposure group owing to excessive distress, dition (6 participants [19%]) met criteria for PGD than partici- apparently as a result of other bereavements that occurred pants in the CBT condition (13 [43%]) (n = 61) (odds ratio [OR], in the course of treatment. Dropout rates during treatment 2.89 [95% CI, 0.92-9.07]; χ2 = 3.42; P = .06). At follow-up, fewer did not differ statistically between patients randomized to participants in the CBT/exposure condition (4 participants CBT/exposure and CBT.
[15%]) met criteria for PTSD than participants in the CBT con- Table 2 contains the least-squares mean scores for the pri- dition (11 [38%]) (n = 56) (OR, 3.51; 95% CI, 0.96-12.89; χ2 = 3.81; mary and secondary outcome measures by treatment condi- P = .04). In terms of follow-up diagnostic status, the number tion. In terms of PGD severity, the level 2 model indicated a needed to treat was 4.32.
JAMA Psychiatry Published online October 22, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of New South Wales User on 10/22/2014
Research Original Investigation Treating Prolonged Grief Disorder Table 2. Mean Psychopathology Scores Treatment Condition, Mean Score (95% CI) 47.49 (43.75-51.23) 46.04 (42.21-49.88) 27.52 (23.26-31.79) 38.16 (33.71-42.61) 28.13 (23.75-32.52) 36.44 (32.17-40.72) 31.07 (27.61-34.54) 28.08 (24.53-31.63) 16.76 (12.88-20.64) 22.41 (18.55-26.28) 16.98 (13.06-20.90) 20.93 (17.06-24.79) 89.66 (81.65-97.67) 82.08 (73.86-90.30) 58.94 (50.07-67.81) 68.68 (59.81-77.56) Abbreviations: BDI-2, Beck 61.75 (52.88-70.62) 69.55 (60.60-78.49) Depression Inventory, SecondEdition; CBT, cognitive behavior therapy; CBT/exposure, CBT with 31.63 (28.38-34.89) 29.49 (26.15-32.83) exposure therapy; ICG, Inventory of 22.99 (19.56-26.42) 27.43 (23.94-30.93) Complicated Grief; PTCI,Posttraumatic Cognitions Inventory; 22.73 (19.24-26.22) 27.35 (23.82-30.89) WHOQOL-BREF, World Health Organization Quality of Life 13.27 (11.27-15.27) 10.95 (8.90-13.00) 9.99 (7.76-12.22) 9.48 (7.26-11.70) Indicates negative cognitions aboutself.
8.47 (6.22-10.72) 10.51 (8.248-12.77) b Indicates negative cognitions about 33.65 (28.58-38.72 36.92 (31.72-42.12) c Indicates negative cognitions about 52.47 (46.64-58.29) 38.56 (32.79-44.33) blame of oneself.
52.58 (46.75-58.40) 42.00 (36.09-47.92) d Measures psychological health WHOQOL-BREF Sociale (mood, self-esteem, and 37.72 (31.48-43.95) 41.36 (34.97-47.75) 54.58 (47.41-61.75) 44.79 (37.38-52.20) Measures social relationships (personal relationships, social 46.13 (38.97-53.30) 46.69 (39.28-54.10) support, and sexual activity).
In secondary analyses, an analysis of antidepressant use toms of PGD. The convergent finding was that adding expo- found no significant difference between those randomized to sure to CBT led to greater reductions in grief and depressive CBT/exposure (8 participants [20%]) and CBT (10 [26%]) symptoms and increased psychological and social function- (n = 80) (χ2 = 0.43; P = .51). In terms of treatment response, we ing. This finding accords with those of previous studies stat- found no difference in meeting PGD criteria between those who ing that psychotherapy programs with some component of ex- used and did not use antidepressants after treatment (n = 61) posure-based treatment are effective.22,23 The novelty of this (6 participants [35%] vs 13 [30%]; χ2 = 0.19; P = .66; OR, 0.77 study is the finding that the exposure element was essential [95% CI, 0.23-2.52]) or at follow-up (n = 56) (3 [231%] vs 12 [28%]; if optimal treatment gains were to be achieved. This conclu- χ2 = 0.12; P = .73; OR, 1.29 [95% CI, 0.30-5.51]). In terms of type sion is consistent with evidence from treatment studies among of death, we found no difference in the rates of those be- patients with anxiety, in whom adding exposure therapy has reaved after sudden or violent death (n = 80) between those been found to augment the effects of other components of in the CBT/exposure condition (18 participants [44%]) and the CBT condition (18 [46%]) (χ2 = 0.43; P = .84). Regarding treat- Presuming that the added benefit of exposure therapy op- ment response, we found no difference in satisfying PGD cri- erated in the same way as shown in anxiety disorders may not teria between those bereaved after sudden/violent death and be correct. Exposure therapy is typically conceptualized in death after chronic illness after treatment (n = 61) (9 partici- anxiety disorders as a form of extinction learning, which in- pants [32%] vs 10 [30%]; χ2 = 0.02; P = .88; OR, 0.92 [95% CI, volves new learning that certain stimuli or thoughts do not sig- 0.31-2.72]) or at follow-up (n = 56) (7 [30%] vs 8 [24%]; χ2 = 0.26; nal threat or distress.40 Exposure techniques can involve a P = .61; OR, 0.73 [95% CI, 0.22-2.41]).
range of processes; however, including emotional processingof affective content and integration of corrective informationcan alleviate one's distress.41,42 The exposure element in thepresent study treatment may have facilitated emotional pro- cessing of the loss, and this facilitation had the additional ben- This study aimed to determine the additive benefit of includ- efit of augmenting more change in maladaptive cognitions or ing exposure therapy in psychotherapy targeted to symp- appraisals about themselves and their world. The finding that JAMA Psychiatry Published online October 22, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of New South Wales User on 10/22/2014
Treating Prolonged Grief Disorder Original Investigation Research CBT/exposure had a stronger beneficial effect on appraisals We note several limitations to this study. First, not all pa- than CBT underscores the suggestion that exposure may have tients were medication free during the course of the study, with facilitated patients' abilities to modify negative appraisals about 18 (23%) using antidepressants. This medication use may have their loss. Cognitive models of PGD posit that the condition is interacted with the psychotherapy; however, we lacked suf- maintained by entrenched negative interpretations about one's ficient numbers of these patients to compare treatment out- capacity to cope with the future or about one's role in the comes for CBT/exposure and CBT for those using and not using death.43,44 This supposition is supported by evidence of the medication. We included patients using antidepressants be- role of catastrophic appraisals in PGD.43 Although all patients cause (1) this approach allowed greater generalizability with received the same dosage of cognitive restructuring to ad- the PGD population, (2) patients maintained stability of medi- dress these thoughts, patients who were given the opportu- cation use during the course of the study (apart from 2 pa- nity to process the loss emotionally before the cognitive re- tients who changed medication), and (3) we found no differ- structuring may have been better able to achieve cognitive ence in medication use between treatment conditions. Second, change. This interpretation is consistent with a previous find- we included patients with PGD after a range of different types ing that providing exposure therapy before cognitive restruc- of death. Treatment response to exposure therapy may differ turing leads to greater improvement in grief symptoms than by the circumstances of the death; we have relatively few pa- when the components are delivered in the reverse order.23 tients who developed PGD after suicide or homicide, which Patients who received exposure therapy enjoyed better may have a distinct treatment response.
psychological and social functioning than those receiving CBTalone. Prolonged grief disorder leads to marked psychosocialimpairment, including suicidality, social withdrawal, and sub- stance abuse.7,9,10 The finding that exposure therapy resultedin a better quality of life 6 months after treatment under- The accumulating evidence points to the very significant pub- scores the conclusion that emotionally processing the loss, in lic health needs arising from PGD. Developing more effective combination with the other components of grief therapy, led treatments that can reduce the burden of this condition is im- to a marked shift in the PGD, which had a positive effect on portant. In the most valuable lesson from this study, optimal patients' abilities to function psychologically and socially.
gains with PGD patients are achieved when the emotions as- Exposure therapy did not result in adverse reactions. Ex- sociated with the memories of the death and the sequelae of posure therapy has been believed to be overly distressing for pa- the loss are fully accessed. Numerous studies attest to the tients and to lead to adverse outcomes or increased dropout from avoidant strategies in which PGD patients engage to mini- treatment.45,46 This distress is typically transient, and it abates mize their distress,51 and this reluctance to engage with their as exposure sessions continue. Consistent with this notion, the distressing emotions may be a major reason for not managing present study found that exposure did not result in negative re- the grief more effectively. Although focusing on appraisals and sponses in PGD patients. Controlled studies of exposure therapy developing future directions apparently was somewhat ben- in other anxiety conditions, such as PTSD, have also demon- eficial, directing patients to access their emotional memories strated that exposure therapy does not result in negative out- appears to achieve the most effective outcomes. Despite the comes relative to other interventions.47,48 Despite these re- distress elicited by engaging with memories of the death, this sults, much evidence exists that clinicians are often reluctant strategy does not lead to aversive responses. In light of evi- to use exposure therapy because of the perception that it can dence that many interventions provided to grieving people are exacerbate excessive distress.49,50 This evidence suggests that, not empirically supported,52 the challenge is to foster better despite the strength of the present finding, clinicians who treat education of clinicians through evidence-supported interven- PGD should be educated about the merits and safety of using tions to optimize adaptation to the loss as effectively as exposure techniques with PGD patients.
ARTICLE INFORMATION intellectual content: Bryant, Joscelyne, Rawson, interpretation of the data; preparation, review, or Submitted for Publication: February 27, 2014; final
Aderka, Nickerson.
approval of the manuscript; and decision to submit revision received May 20, 2014; accepted June 24, Statistical analysis: Bryant, Aderka, Nickerson.
the manuscript for publication.
Obtained funding: Bryant.
Administrative, technical, or material support: Published Online: October 22, 2014.
Bryant, Kenny, Joscelyne, Rawson, Maccallum, 1. Maercker A, Brewin CR, Bryant RA, et al.
Cahill, Hopwood, Nickerson.
Proposals for mental disorders specifically Author Contributions: Dr Bryant had full access to
Study supervision: Bryant, Rawson, Cahill.
associated with stress in the International all the data in the study and takes responsibility for Conflict of Interest Disclosures: None reported.
Classification of Diseases-11. Lancet. 2013;381 the integrity of the data and the accuracy of the data analysis.
Funding/Support: This study was supported by
Study concept and design: Bryant, Kenny, grant 568970 from the National Health and 2. Prigerson HG, Horowitz MJ, Jacobs SC, et al.
Joscelyne, Cahill, Hopwood.
Medical Research Council Program and grant Prolonged grief disorder: psychometric validation Acquisition, analysis, or interpretation of data: All 455341 from the National Health and Medical of criteria proposed for DSM-V and ICD-11 [published Research Council Project.
correction appears in PLoS Med. 2013;10(12)]. PLoS Drafting of the manuscript: Bryant, Kenny, Role of the Funder/Sponsor: The funding sources
Maccallum, Cahill, Hopwood, Nickerson.
had no role in the design and conduct of the study; Critical revision of the manuscript for important collection, management, analysis, and JAMA Psychiatry Published online October 22, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of New South Wales User on 10/22/2014
Research Original Investigation Treating Prolonged Grief Disorder 3. Germain A, Caroff K, Buysse DJ, Shear MK. Sleep
20. Zisook S, Shuchter SR, Pedrelli P, Sable J,
metric as for classical analysis. Psychol Methods.
quality in complicated grief. J Trauma Stress. 2005; Deaciuc SC. Bupropion sustained release for bereavement: results of an open trial. J Clin 37. Chatellier G, Zapletal E, Lemaitre D, Menard J,
4. Simon NM, Pollack MH, Fischmann D, et al.
Degoulet P. The number needed to treat: a clinically Complicated grief and its correlates in patients with 21. Reynolds CF III, Miller MD, Pasternak RE, et al.
useful nomogram in its proper context. BMJ. 1996; bipolar disorder. J Clin Psychiatry. 2005;66(9): Treatment of bereavement-related major depressive episodes in later life: a controlled study 38. Bryant RA, Moulds ML, Guthrie RM, et al. A
5. Mitchell AM, Kim Y, Prigerson HG, Mortimer MK.
of acute and continuation treatment with randomized controlled trial of exposure therapy Complicated grief and suicidal ideation in adult nortriptyline and interpersonal psychotherapy. Am and cognitive restructuring for posttraumatic stress survivors of suicide. Suicide Life Threat Behav.
J Psychiatry. 1999;156(2):202-208.
disorder. J Consult Clin Psychol. 2008;76(4):695-703.
22. Shear K, Frank E, Houck PR, Reynolds CF III.
39. Bryant RA, Mastrodomenico J, Felmingham KL,
6. Szanto K, Prigerson H, Houck P, Ehrenpreis L,
Treatment of complicated grief: a randomized et al. Treatment of acute stress disorder: Reynolds CF III. Suicidal ideation in elderly controlled trial. JAMA. 2005;293(21):2601-2608.
a randomized controlled trial. Arch Gen Psychiatry.
bereaved: the role of complicated grief. Suicide Life 23. Boelen PA, de Keijser J, van den Hout MA, van
Threat Behav. 1997;27(2):194-207.
den Bout J. Treatment of complicated grief: 40. Milad MR, Rauch SL, Pitman RK, Quirk GJ. Fear
7. Prigerson HG, Bridge J, Maciejewski PK, et al.
a comparison between cognitive-behavioral extinction in rats: implications for human brain Influence of traumatic grief on suicidal ideation therapy and supportive counseling. J Consult Clin imaging and anxiety disorders. Biol Psychol. 2006; among young adults. Am J Psychiatry. 1999;156(12): 24. Devilly GJ, Borkovec TD. Psychometric
41. Rothbaum BO, Mellman TA. Dreams and
8. Hardison HG, Neimeyer RA, Lichstein KL.
properties of the Credibility/Expectancy exposure therapy in PTSD. J Trauma Stress. 2001;14 Insomnia and complicated grief symptoms in Questionnaire. J Behav Ther Exp Psychiatry. 2000; bereaved college students. Behav Sleep Med. 2005; 42. Rothbaum BO, Schwartz AC. Exposure therapy
25. Zhang B, El-Jawahri A, Prigerson HG. Update on
for posttraumatic stress disorder. Am J Psychother.
9. Prigerson HG, Bierhals AJ, Kasl SV, et al.
bereavement research: evidence-based guidelines Traumatic grief as a risk factor for mental and for the diagnosis and treatment of complicated 43. Boelen PA, van den Bout J, van den Hout MA.
physical morbidity. Am J Psychiatry. 1997;154(5): bereavement. J Palliat Med. 2006;9(5):1188-1203.
The role of negative interpretations of grief 26. Prigerson HG, Maciejewski PK, Reynolds CF III,
reactions in emotional problems after 10. Silverman GK, Jacobs SC, Kasl SV, et al. Quality
et al. Inventory of Complicated Grief: a scale to bereavement. J Behav Ther Exp Psychiatry. 2003; of life impairments associated with diagnostic measure maladaptive symptoms of loss. Psychiatry criteria for traumatic grief. Psychol Med. 2000;30 44. Maccallum F, Bryant RA. A cognitive
27. Blake DD, Weathers FW, Nagy LM, et al. The
attachment model of prolonged grief: integrating 11. Boelen PA, Prigerson HG. The influence of
development of a Clinician-Administered PTSD attachments, memory, and identity. Clin Psychol Rev.
symptoms of prolonged grief disorder, depression, Scale. J Trauma Stress. 1995;8(1):75-90.
and anxiety on quality of life among bereaved 28. Sheehan DV, Lecrubier Y, Sheehan KH, et al.
45. Kilpatrick DG, Best C. Some cautionary remarks
adults: a prospective study. Eur Arch Psychiatry Clin The Mini-International Neuropsychiatric Interview on treating sexual assault victims with implosion.
(MINI): the development and validation of a Behav Ther. 1984;15:421-423.
12. Bonanno GA, Wortman CB, Lehman DR, et al.
structured diagnostic psychiatric interview for 46. Pitman RK, Altman B, Greenwald E, et al.
Resilience to loss and chronic grief: a prospective DSM-IV and ICD-10. J Clin Psychiatry. 1998;59(suppl Psychiatric complications during flooding therapy study from preloss to 18-months postloss. J Pers for posttraumatic stress disorder. J Clin Psychiatry.
Soc Psychol. 2002;83(5):1150-1164.
29. Beck AT, Steer RA, Brown GK. Beck Depression
13. Bonanno GA, Kaltman S. The varieties of grief
Inventory–Second Edition. San Antonio, TX: 47. Hembree EA, Foa EB, Dorfan NM, Street GP,
experience. Clin Psychol Rev. 2001;21(5):705-734.
Psychological Corp; 1996.
Kowalski J, Tu X. Do patients drop out prematurely 14. Neria Y, Gross R, Litz B, et al. Prevalence and
30. Foa EB, Ehlers A, Clark DM, Tolin DF, Orsillo SM.
from exposure therapy for PTSD? J Trauma Stress.
psychological correlates of complicated grief The Posttraumatic Cognitions Inventory (PTCI): among bereaved adults 2.5-3.5 years after development and validation. Psychol Assess. 1999; 48. Foa EB, Zoellner LA, Feeny NC, Hembree EA,
September 11th attacks. J Trauma Stress. 2007;20 Alvarez-Conrad J. Does imaginal exposure 31. WHOQOL Group. WHOQOL-BREF: Introduction,
exacerbate PTSD symptoms? J Consult Clin Psychol.
15. Newson RS, Boelen PA, Hek K, Hofman A,
Administration, Scoring and Generic Version of the Tiemeier H. The prevalence and characteristics of Assessment. Geneva, Switzerland: World Health 49. Becker CB, Zayfert C, Anderson E. A survey of
complicated grief in older adults. J Affect Disord.
Organization; 1996.
psychologists' attitudes towards and utilization of 32. Harper A, Power M; WHOQOL Group.
exposure therapy for PTSD. Behav Res Ther. 2004; 16. Simon NM, Shear KM, Thompson EH, et al. The
Development of the World Health Organization prevalence and correlates of psychiatric WHOQOL-BREF quality of life assessment. Psychol 50. Rosen CS, Chow HC, Finney JF, et al. VA
comorbidity in individuals with complicated grief.
practice patterns and practice guidelines for Compr Psychiatry. 2007;48(5):395-399.
33. Hawthorne G, Herrman H, Murphy B.
treating posttraumatic stress disorder. J Trauma 17. Prigerson HG, Bierhals AJ, Kasl SV, et al.
Interpreting the WHOQOL-BREF: preliminary Complicated grief as a disorder distinct from population norms and effect sizes. Soc Indic Res.
51. Schnider KR, Elhai JD, Gray MJ. Coping style use
bereavement-related depression and anxiety: predicts posttraumatic stress and complicated grief a replication study. Am J Psychiatry. 1996;153(11): 34. Faul F, Erdfelder E, Lang AG, Buchner A.
symptom severity among college students G*Power 3: a flexible statistical power analysis reporting a traumatic loss. J Couns Psychol. 2007; 18. Boelen PA, van den Bout J. Complicated grief,
program for the social, behavioral, and biomedical depression, and anxiety as distinct postloss sciences. Behav Res Methods. 2007;39(2):175-191.
52. Currier JM, Neimeyer RA, Berman JS. The
syndromes: a confirmatory factor analysis study.
35. Raudenbush SW. Comparing personal
effectiveness of psychotherapeutic interventions Am J Psychiatry. 2005;162(11):2175-2177.
trajectories and drawing causal inferences from for bereaved persons: a comprehensive 19. Pasternak RE, Reynolds CF III, Schlernitzauer M,
longitudinal data. Annu Rev Psychol. 2001;52:501- quantitative review. Psychol Bull. 2008;134(5):648- et al. Acute open-trial nortriptyline therapy of bereavement-related depression in late life. J Clin 36. Feingold A. Effect sizes for growth-modeling
analysis for controlled clinical trials in the same JAMA Psychiatry Published online October 22, 2014
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://archpsyc.jamanetwork.com/ by a University of New South Wales User on 10/22/2014

Source: https://www.psychologicalsociety.ie/page/file_dwn2/170/Treating%20Prolonged%20Grief%20Disorder%20-%20A%20Randomized%20Clinical%20Trial.pdf