Partially dead, partially separated - establishing the mechanism between ambiguous loss and grief reaction among caregivers of patients with prolonged disorders of consciousness Clinical Rehabilitatio

Prévia do material em texto

https://doi.org/10.1177/0269215518802339
Clinical Rehabilitation
 1 –12
© The Author(s) 2018
Article reuse guidelines: 
sagepub.com/journals-permissions
DOI: 10.1177/0269215518802339
journals.sagepub.com/home/cre
CLINICAL
REHABILITATION
Partially dead, partially separated: 
establishing the mechanism 
between ambiguous loss and 
grief reaction among caregivers 
of patients with prolonged 
disorders of consciousness
Yael Zaksh1, Einat Yehene1 , Maya Elyashiv2 
and Arie Altman2
Abstract
Objectives: To trace the psychological mechanism underlying caregivers’ emotional experience in 
prolonged disorders of consciousness, by examining the mediating role of boundary ambiguity in the 
relationship between ambiguous loss and grief.
Design: Cross-sectional design.
Setting: The Respiratory Rehabilitation Division of a long-term medical and rehabilitation institute.
Subjects: A total of 64 primary caregivers (69% female) of patients in a vegetative state (n = 49) or 
minimally conscious state (n = 15), with a mean age of 55.5 (SD = 12.3) years. Participants were mostly the 
patient’s children (62%) or partners (27%). The mean caregiving duration was 4.9 (SD = 5.1) years.
Main measures: The Boundary Ambiguity Scale, the Revised Need for Closure Scale, an adapted version of 
the multifactor Two-Track Bereavement Questionnaire, and a sociodemographic questionnaire, which included 
items regarding caregiving: frequency of visits and perception of the patient’s psychological presence.
Results: (1) Caregivers’ grief scores (total TTBQ: mean = 2.97; SD = 0.55) did not significantly differ 
from those exhibited by a normative bereavement sample. (2) Time since injury did not affect caregivers’ 
grief scores (three-year cut-point; P > .05). (3) Mediation analyses revealed that boundary ambiguity 
(mean = 34.03; SD = 7.55) significantly mediates the relationship between need for closure (B = .11; 
confidence interval (CI) = .04–.23) and grief; frequency of visits (B = .05, CI = .02–.10) and grief and 
perception of psychological presence (B = .26, CI = .00–.61) and grief in three separate models.
Conclusion: Ambiguity concerning relational boundaries hinders caregivers’ ongoing grief reaction and 
impedes their ability to integrate the loss. High frequency of visits, belief in the patient’s psychological 
1 School of Behavioral Sciences, The Academic College of Tel 
Aviv-Yaffo, Tel Aviv-Yaffo, Israel
2 Respiratory Rehabilitation Division, Reut Medical-
Rehabilitation Center, Tel Aviv-Yaffo, Israel
802339 CRE0010.1177/0269215518802339Clinical RehabilitationZaksh et al.
research-article2018
Original Article
Corresponding author:
Einat Yehene, School of Behavioral Sciences, The Academic 
College of Tel Aviv-Yaffo, P.O.B 8401, Tel Aviv-Yaffo 61161, 
Israel. 
Email: einat.yehene@gmail.com
https://uk.sagepub.com/en-gb/journals-permissions
https://journals.sagepub.com/home/cre
mailto:einat.yehene@gmail.com
http://crossmark.crossref.org/dialog/?doi=10.1177%2F0269215518802339&domain=pdf&date_stamp=2018-09-26
2 Clinical Rehabilitation 00(0)
presence and difficulties in tolerating uncertainty are all clinical manifestations of caregivers’ entangled 
experience.
Keywords
Ambiguous loss, boundary ambiguity, caregivers, disorder of consciousness, grief, minimally conscious 
state, persistent vegetative state
Received: 9 May 2018; accepted: 30 August 2018
Introduction
Loss of consciousness is a key feature associated 
with severe brain injury. While most patients regain 
consciousness completely, others may continue to 
exhibit a Prolonged Disorder of Consciousness, 
which encompasses two major conditions. Despite 
intermittent wakefulness, the Vegetative State is 
characterized by complete absence of any self or 
environmental awareness, while the minimally 
conscious state (MCS) is characterized by limited 
conscious awareness and reproducible goal-
directed behaviors.1 Patients emerging from a veg-
etative state might enter minimal conscious state 
with or without further recovery,2 which indicates 
the long-lasting potential of these sequelae. Given 
that the prevalence of these patients has been rising 
due to advances in medical technology,3 caregiv-
ers’ persistent state of caring for a loved one who’s 
neither fully dead nor fully alive has become a 
point of interest.
Over the last decade, much of the literature has 
focused on caregivers’ psychological symptoms 
and continuous demands. Studies indicated high 
levels of anxiety and depression,4–6 a poor quality 
of life,7 and a drastic reduction in their interests and 
self-dedicated time.8 Taken together, these studies 
represent a dominant approach that conceptualizes 
caregivers’ emotional reactions as derivatives of 
the multifactored burden they are forced to bear 
due to the patient’s chronic condition.9–11
Recently, due to the lingering nature of their 
loss experience, grief has become to gather interest 
as a construct to understand caregivers’ psycho-
logical symptoms. Studies indicated that 15%–
60% of caregivers meet the diagnostic criteria for 
prolonged grief disorder12–14 for years afterwards,15 
regardless of depression and post-traumatic stress 
disorder (PTSD) symptoms.14
While quantitative studies have measured the fre-
quency and impact of caregivers’ loss, using differ-
ent conceptual models, qualitative explorations 
further elucidated caregivers’ multifaceted loss 
experience.16,17 Studies described a form of unend-
ing loss, resonating with Boss’s (1999) concept of 
Ambiguous Loss, in which a loved one is physically 
present but psychologically absent. Boss18 states 
that the grieving process is inherently impeded since 
the clarity needed for emotional boundary mainte-
nance or closure is unattainable. In-depth reports 
concur with this premise by depicting how caregiv-
ers’ perceptions and relationships with their loved 
one are fragmented by ambiguity, thus hindering 
resolution and re-engagement in ongoing life.19–21
Despite growing research, little is known about 
the contributing factors to caregivers’ emotional 
ordeal, indicating the need to deepen our under-
standing of the psychological mechanism under-
lying this loss experience.22Therefore, this 
study sought to (1) comprehensively explore 
caregivers’ grief reaction by endorsing the bifocal 
perspective of the adapted “Two-Track Model of 
Bereavement”,23 which assesses the general func-
tioning of caregivers, alongside their continuing 
bond with the patient, and (2) study the interplay 
between grief reaction and psychological factors 
that are relevant to ambiguous loss experience. To 
the best of our knowledge, this study represents the 
first attempt to operationalize caregivers’ experi-
ence of psychological/emotional limbo.
Zaksh et al. 3
Methods
This study has been conducted on family caregiv-
ers of adult patients with prolonged disorders of 
consciousness, admitted to the Respiratory 
Rehabilitation Division of a Long-term Medical 
and Rehabilitation Institute in central Israel. All the 
patients met the standards of the Royal College of 
Physicians1 for the diagnosis of vegetative or mini-
mal conscious state.
Only primary caregivers were invited to partici-
pate in the study on a voluntary basis. This meant 
that it was limited to those who spent more time 
caring for the patient informally and who assumed 
the greatest responsibility, following the approach 
used by Giovannetti et al.10
Data collection was performed from May 2017 
to August 2017. The study was approved by the 
Institutional Review Board (IRB) at the Medical-
Rehabilitation center (protocol number: 2017-2) 
and by the ethics committee of the Academic 
Institute (approval number: 2017086).
The participants provided written informed con-
sent, and confidentiality was preserved by record-
ing the data in the de-identified form. The 
assessment protocol was composed of self-report 
questionnaires and was applied by psychologists in 
a single session of approximately 60 minutes.Sociodemographic questionnaire
In addition to sociodemographic and clinical data, 
the questionnaire included items regarding the car-
egiver’s ongoing relationship with the patient, 
Frequency of Visits and caregivers’ perception of 
psychological presence, in which the caregivers 
were asked whether they perceived that the patient 
was aware of their presence.
Boundary Ambiguity Scale
Boundary Ambiguity Scale (BAS) is a question-
naire used to assess the way family members per-
ceive the situation of ambiguous loss and how the 
injured relative now fits into their family.24 It has 
several versions for measuring family members’ 
perceptions of incongruent presence–absence (e.g. 
missing in action (MIA), divorce, chronic illness). 
This study used the Boundary Ambiguity scale for 
Dementia (BAS-6),24 with two items modified so 
that they would apply to the current research popu-
lation. It includes 14 items, each evaluated on a 
1–5 Likert-type scale. Each statement measured 
lingering feelings of loyalty, guilt and dependence 
regarding their loved one and the extent to which 
they were still affected by his condition. Total score 
ranges 14–70, with higher scores reflective of 
greater ambiguity and confusion regarding the 
relationship with the patient. Boss et al.25 report a 
Cronbach’s alpha of .80 for the BAS-6. In this 
study, the standard “forward-backward” procedure 
was applied in order to translate the scale to 
Hebrew, and the Cronbach’s alpha for the modified 
scale used was .76.
Revised Need for Closure Scale
Revised Need for Closure (NFC) Scale is a ques-
tionnaire used to measure caregivers’ tolerance of 
uncertainty.26 It is an abbreviated 15-item index, in 
which each item is evaluated on 1–6 Likert-type 
scale and relates to a situation of cognitive uncer-
tainty. Total score ranges 1–6, with higher scores 
signifying a stronger discomfort with ambiguity 
and greater aversion when facing situations that are 
devoid of closure. Roets and Van Hiel27 report a 
Cronbach’s alpha of .87 for the abbreviated scale. 
The standard “forward–backward” procedure was 
also applied in this case, and the Cronbach’s alpha 
of the translated scale used in this study was .86.
Two-Track Bereavement Questionnaire
Two-Track Bereavement Questionnaire (TTBQ) is 
a multifactored, 70-item questionnaire28 con-
structed in accordance with the Two-Track Model 
of Bereavement.23 It assumes a bifocal perspective 
on bereavement, namely, the biopsychosocial func-
tioning of the bereaved and the nature of the ongo-
ing relational bond to the deceased. Each item is 
rated on a 1–5 Likert-type scale. Total score for 
each factor ranges 1–5, with higher scores repre-
senting greater levels of distress, mourning and 
intense involvement with the object of loss. Due to 
4 Clinical Rehabilitation 00(0)
the scale’s substantial length, this study used the 
abbreviated 40-item version (an unpublished meas-
ure designed for clinical use). It was found to be a 
reliable tool, as indicated by its Cronbach’s scale 
and its strong correlations with the factors of the 
full 70-item version: TTBQ-total score (α = .89, 
r = .973, P < .01), Relational Active Grief (RAG; 
α = .90; r = .964, P < .01), Close and Positive 
Relationship with the Deceased (PC; α = .77; 
r = .962, P < .01), Conflictual Relationship with the 
Deceased (CN; α = .77; r = .870, P < .01), General 
Biopsychosocial Functioning (F; α = .82; r = .973, 
P < .01) and Traumatic Perception of the Loss (D; 
α = .75; r = .918, P < .01). Adaptation to this study 
involved rephrasing 14 items in order to apply it to 
the current research population. The scale’s adapted 
version yielded a satisfactory Cronbach’s reliabil-
ity of α = .89 for the TTBQ-total score and was also 
satisfactory for each of the factors (.73 ⩽ α ⩽ .84).
Data Analysis
All the statistical analyses were performed using 
SPSS v.23 (SPSS Inc., Chicago, IL, USA). A 
descriptive analysis was performed of the sociode-
mographic characteristics of the sample and of the 
scores on all the questionnaires. Cronbach’s alpha 
was calculated, and values higher than .70 indi-
cated a good consistency for each of the measures 
(all modified and translated).
Preliminary analyses (independent samples 
t-test and chi-square) were conducted to investigate 
subgroup differences based on the patient’s diagno-
sis (i.e. vegetative or minimal conscious state).
To assess caregivers’ grief in comparison to 
bereavement, national norms were obtained from 
the International Center for the Study of Loss, 
Bereavement and Human Resilience. Given that 
the data were based on the full 70-item scale, we 
extracted new means and SDs for each of the fac-
tors of the abbreviated scale used, according to the 
scale’s scoring key.28 Finally, one sample t-test was 
used to compare caregivers’ grief scores to the 
respective normative bereavement sample. Pearson 
correlations were calculated among all the varia-
bles, and an independent samples t-test was used to 
determine whether “Time since Injury” affected 
caregivers’ grief reaction, using a median cut-off 
point of three years.
Subsequently, regression analyses were per-
formed, entering only those variables that were sig-
nificantly associated with caregivers’ grief scores 
(i.e. NFC, frequency of visits, perception of 
psychological presence and boundary ambiguity). 
The mediating effects of boundary ambiguity 
were tested based on the four-step procedure 
described by Baron and Kenny.29 The bootstrapping30 
(PROCESS Macro) test indicated that the mediat-
ing effect was significantly greater than zero with 
reference to the 95% confidence interval (CI).
Results
Clinical and sociodemographic data relating to car-
egivers and their respective relatives (patients in a 
vegetative or minimal conscious state) are pre-
sented in Table 1. The majority of the sample were 
female, with a mean age of 55.5 (SD = 12.3) years, 
and the patient was their parent. Most of the cases 
involved a vegetative state and their main etiology 
was cerebrovascular accident (CVA).
Caregivers’ frequency of visits, NFC, levels of 
boundary ambiguity and grief scores were not 
related to the patient’s diagnosis (all P values >.05). 
Moreover, the percentage of caregivers who per-
ceived the patient as psychologically present did not 
differ by diagnosis, χ2(1, N = 64) = .89, P = .67. 
Following this conformational analysis, the data 
were addressed as a single sample of caregivers of 
patients with prolonged disorders of consciousness.
In order to assess caregivers’ scores on the 
TTBQ, caregivers’ grief scores were compared 
with those we yielded from the national reference 
norms of bereavement28 (see Table 2). As can be 
seen from Table 2, scores did not significantly dif-
fer for the full scale or for any individual factors, 
indicating that caregivers’ grief reaction was simi-
lar to that exhibited in bereavement samples, fol-
lowing the death of a loved one.
The correlations between time-related variables 
(i.e. Age at Injury and Time since Injury) and car-
egivers’ grief reaction are depicted in Table 3.
A significant negative correlation was found 
between the patient’s “Age at Injury” and the factor 
Zaksh et al. 5
“Perception of Trauma” (r = –.349, P < .01). “Time 
since Injury” did not significantly correlate to any 
of the factors, and the Shapiro–Wilk test revealed 
that the variable is not normally distributed 
(P < .01). However, no significant differences 
were found among the caregivers’ grief scores 
using a cut-off point of three years, indicating that 
the time elapsed did not affect caregivers’ grief 
reaction (P values >.05).
The full correlation matrix of the other inde-
pendent variables associated with caregivers’ 
grief scores on the TTBQ (i.e. Boundary 
Ambiguity; NFC; Perception of Psychological 
Presence; Frequency of Visits) is presented in 
Table 4.
The results indicate that all the independent var-
iables correlated positively with the “TotalTTBQ” 
score and the “RAG” factor. This demonstrates that 
Table 1. Sample characteristics.
Characteristics (n = 64) n % Range M SD
Caregivers
Gender Male 20 31.3 – – –
Female 44 68.8 – – –
Age (in years) – – 34–80 55.5 12.35
Employment status Employed 44 68.8 – – –
Unemployed 7 10.9 – – –
Retired 13 20.3 – – –
Marital status Married 49 76.6 – – –
Single 5 7.8 – – –
Divorced 7 10.9 – – –
Widowed 1 1.6 – – –
Life partner 2 3.1 – – –
Educational level ⩽12 years 38 59.4 – – –
>12 years 26 40.6 – – –
Relation to patient (the 
caregiver is)
Spouse/partner 17 26.6 – – –
Son/daughter 40 62.5 – – –
Parent 6 9.4 – – –
Sibling 1 1.6 – – –
Frequency of visits (per week) – – 0.2–7 3.4 2.6
Is the patient aware of your 
presence?
Yes 55 85.9 – – –
No 9 14.1 – – –
Patients
Age (in years) – – 19–94 70.8 18.73
Age at injury (in years) – – 18–93.7 65.9 20.61
Time since injury (in years) – – 0.2–22 4.9 5.1
Diagnosis PVS 49 76.7 – – –
MCS 15 23.4 – – –
Type of event Motor vehicle 
accident
11 17.2 – – –
Fall 3 4.7 – – –
CVA 24 37.5 – – –
Neurodegenerative 
disease
10 15.6 – – –
Other 16 25 – – –
PVS: persistent vegetative state; MCS: minimally conscious state; CVA: cerebrovascular accident.
6 Clinical Rehabilitation 00(0)
Table 2. Mean, SD and results of one-sample t-test comparing TTBQ following adaptation with national reference 
norms of bereaved samples.
TTBQ factors TTBQ
PVS/MCS
Sample (N = 64)
TTBQ
Bereavement
Sample (N = 430)
P value
M (SD) M (SD)
Relational Active Grief 3.21 (.81) 3.23 (.86) .81 (NS)
Close relationship 3.69 (.88) 3.79 (.71) .38 (NS)
Conflictual relationship 2.07 (1.13) 1.86 (.79) .15 (NS)
General function 1.93 (.54) 2.06 (.62) .06 (NS)
Perception of trauma 4.29 (1.81) 4.18 (.80) .26 (NS)
Total TTBQ 2.97 (.55) 2.97 (.51) .99 (NS)
PVS: persistent vegetative state; MCS: minimally conscious state; TTBQ: Two-Track Bereavement Questionnaire.
Normative sample scores of the abbreviated TTBQ version were yielded based on previously tested bereaved samples of the full-
scale version;28 NS: non-significant for P value >.05.
Table 3. Correlations of TTBQ factors with time-related variables.
Variables TTBQ factors
Relational 
Active Grief
Close 
relationship
Conflictual 
relationship
General 
function
Perception 
of trauma
Total 
TTBQ
Age at injury (patient) –.133 .007 –.190 –.107 –.349** –.215
Time since Injury –.033 –.018 –.049 –.018 .088 –.017
TTBQ: Two-Track Bereavement Questionnaire.
**P < .01.
Table 4. Correlations between TTBQ factors and the independent variables.
1 2 3 4 5 6 7 8 9
1. TTBQ–RAG – 
2. TTBQ–PC .562** – 
3. TTBQ–CN .145 –.129 – 
4. TTBQ–F .315* .057 .092 – 
5. TTBQ–D .701** .521** .065 .309* – 
6. TTBQ Total .894** .605** .266* .589** .788** – 
7. BA .625** .337** .082 .301* .489** .574** – 
8. NFC .363** .109 .260* .148 .188 .335** .347** – 
9. PPP .444** .379** –.067 .108 .282* .389** .320* .000 –
10. FV .474** .448** –.039 .124 .259* .405** .470** .352** .331**
TTBQ: Two-Track Bereavement Questionnaire; RAG: Relational Active Grief; PC: Close and Positive Relationship; CN: conflic-
tual relationship; F: general function; D: traumatic perception of the loss; BA: boundary ambiguity; NFC: need for closure; PPP: 
perception of psychological presence; FV: frequency of visits.
PPP is a dichotomous variable, which was coded 0 if the caregiver responded, “the patient is not aware of my presence,” and 1 if 
the caregiver responded, “the patient is aware of my presence.”
*P < .05; **P < .01.
Zaksh et al. 7
caregivers’ high grief reaction was associated with 
a greater NFC, greater frequency of visits, higher 
boundary ambiguity and the caregivers’ perception 
that the patient was aware of their presence. 
Furthermore, boundary ambiguity was positively 
correlated to most of the scale’s factors and to all 
the other independent variables, indicating that 
greater levels of ambiguity concerning the new 
relationship with the patient were associated with 
greater NFC, greater frequency of visits and a 
greater perception that the patient was psychologi-
cally present.
Table 5 provides an overview of the analysis 
conducted in order to determine whether the four 
conditions required to establish mediation were 
met.29 Age, Gender and Relation to Patient were 
entered as control variables, as the effect of kinship 
and gender on grief scores had previously been 
demonstrated in bereavement samples.28
In Model 1, greater NFC was significantly asso-
ciated with increased boundary ambiguity (path a), 
which, in turn, was positively associated with 
higher levels of grief (path b). Greater NFC was 
also significantly associated with higher levels of 
grief (path c), and this correlation was no longer 
significant when boundary ambiguity was entered 
as a mediator (path c′ in Figure 1). The model is 
significant (F(5,58) = 7.58, P < .01) and it explains 
about 40% of the variance. The bootstrapping anal-
ysis revealed that boundary ambiguity was a sig-
nificant intervening variable between NFC and 
caregivers’ grief reaction (95% CI = .040–.232).
In Model 2, perception of psychological pres-
ence was used as a predictor variable. Similarly to 
Model 1, paths a, b and c were all significant and 
path c′ was no longer significant when boundary 
ambiguity was entered as a mediator (P > .05). The 
model is significant (F(5,58) = 8.25, P < .01) and 
explains about 42% of the variance. The bootstrap-
ping analysis revealed that boundary ambiguity 
was a significant intervening variable between per-
ception of psychological presence and caregivers’ 
grief reaction (95% CI = .001–.616).
Table 5. Mediation models, path coefficients and bootstrapping results to test the significance of the mediation 
effect.
Model B SE t P value 95% CI
Model 1 (NFC-BA-GR)
 NFC-BA (path a) 2.83 1.01 2.81 .007** –
 BA-GR (path b) .04 .01 4.76 .000** –
 NFC-GR (path c) .18 .07 2.53 .014* –
 NFC-GR via BA (path c′) .08 .07 1.15 .256 –
 Indirect effect (path a∙b) .11 .04 – – .040–.232
Model 2 (PPP-BA-GR)
 PPP-BA (path a) 6.95 2.72 2.55 .013* –
 BA-GR (path b) .04 .01 4.72 .000** –
 PPP-GR (path c) .57 .19 3.01 .004** –
 PPP-GR via BA (path c’) .32 .17 1.84 .071 –
 Indirect effect (path a∙b) .26 .16 – – .001–.616
Model 3 (FV-BA-GR)
 FV-BA (path a) 1.54 .37 4.19 .000** –
 BA-GR (path b) .03 .01 4.06 .000** –
 FV-GR (path c) .10 .03 3.73 .000** –
 FV-GR via BA (path c′) .05 .03 1.74 .088 –
 Indirect effect (path a∙b) .05 .02 – – .021–.109
NFC: Need for Closure; BA: Boundary Ambiguity; GR: Grief Reaction; PPP: Perception of Psychological Presence; FV: Frequency 
of Visits.
*P < .05; **P < .01.
8 Clinical Rehabilitation 00(0)
In Model 3, frequency of visits was used as a 
predictor variable. As in the previous models, paths 
a, b and c were all significant and path c′ was 
no longer significant when boundary ambiguity 
was entered as a mediator (P > .05). The model is 
significant (F(5,58) = 8.13, P < .01) and explains 
about 41% of the variance. The bootstrapping 
analysis revealed that boundary ambiguity was a 
significant intervening variable between frequency 
of visits and caregivers’ grief reaction (95% 
CI = .021–.109).
Although the relationships between the varia-
bles in Table 5 are referred to as “paths,” they are 
not the same as “paths” calculated in Path Analysis. 
In this study, the term “path” is used to refer to a 
regression coefficient that quantifies the associa-
tion between two variables.
Discussion
This study reveals the adverse impact of vague rela-
tional boundaries on grief reactions among caregiv-
ers of patients with prolonged disorders of 
consciousness. Our findings indicate that elevated 
levels of boundary ambiguity, stemming from the 
incongruity between physical and psychological 
presence, accentuates grief and hinders caregivers’ 
resolution of the loss. Furthermore, caregivers’ 
grief level was found similar to that exhibited by a 
normative bereavement sample, following the death 
of a loved one. Grief reaction was not affected bythe time elapsed, hence demonstrating the unre-
solved nature of caregivers’ loss. To our knowledge, 
this is the first study to use a multifactorial rela-
tional grief measure, thus enabling to fully grasp the 
multidimensional nature of this inimitable form of 
loss. To date, the very few previous studies assessed 
caregivers’ grief similarly, by using the same meas-
ure (i.e. Prolonged Grief-12).22
Our findings further elucidate the notion that 
unlike a clear-cut loss (i.e. death), an ambiguous 
loss can often be more difficult to cope with due to 
the frequent uncertainty as to what has been lost.31,32 
Patients with disorders of consciousness represent 
the end-point of a continuum that measures altera-
tions in cognition and awareness, in which the 
patient as a psychological entity is immensely dis-
solved. Not surprisingly, we found caregivers’ grief 
reaction closely akin to death, with a prominent 
Perception of the Event as Traumatic (Track I), 
accompanied by an acute Active Relational Grief 
(Track II). The latter indicates an intense yearning 
and preoccupation with the loss years after the 
event. These important clinical findings accord 
with those observed in bereaved people whose loss 
remains at the forefront of their lives.33 In a typical 
bereavement process, the grief reaction tends to 
recede over time, especially its aspects relating to 
the psychological bond with the deceased.28 
However, since a final resolution of grief cannot 
typically occur until physical death, caregivers 
find the patient’s concurrent presence–absence 
Figure 1. A conceptual model underlying the present mediation analysis of model 1, in which boundary ambiguity 
mediates the relationship between need for closure and grief reaction. Same structure was applied to models 2 and 
3, using other predictors.
NFC: Need for Closure; BA: Boundary Ambiguity; GR: Grief Reaction.
Zaksh et al. 9
particularly confounding and struggle to elaborate a 
strategy of mourning.18,32,34 Interestingly, despite 
their emotional state, caregivers’ General Function-
ing (Track I) was found to be adequate. This finding 
is consistent with a previous study which showed 
that caregivers’ perceived burden was not affected 
by anxiety, depression and quality of life,7 hence 
potentially dissociating grief from burden.
Caregivers’ emotional reaction in response to 
such an ambiguous loss was primarily conducted in 
the field of dementia.35,36 While the loss in demen-
tia is theoretically similar in form (psychological 
absence yet continued physical presence), it differs 
in several substantial ways. “Dementia grief” is a 
specific type of anticipatory grief,37 in response to 
compounded serial losses of varying magnitude.38 
In dementia, the disease trajectory is progressive 
and deterioration is gradual. There is a growing 
sense of detachment and receding of the known 
self, preceding physical death. In contrast, the 
onset of disorders of consciousness is often sudden 
and dramatic. There is an immediate loss of reci-
procity, an abrupt relational change19 and uncer-
tainty regarding the patient’s minimal, if any, 
responsiveness. The condition continues to allow 
for the possibility of change, even if unlikely as 
time progresses.39,40 Contributing factors to car-
egivers’ ongoing hope for recovery are recent evi-
dence that suggests some patients may retain a 
degree of consciousness41,42 and unhelpful portray-
als in the media.43,44 These may fulfill caregivers’ 
hope and belief in the patient’s presence and pre-
serve ambiguity.
Accordingly, our results indicated an increased 
perception of the patient’s psychological presence, 
regardless of diagnosis (i.e. MCS/persistent vege-
tative state (PVS)), alongside a high frequency of 
visits. These findings validate that caregivers’ sub-
jective perceptions regarding the patient’s condi-
tion affect their level of physical and emotional 
commitment.45 During this liminal state, caregivers 
mourn their relational bond with the patient. 
However, given his tangible presence, physical 
proximity prevails46 and the irrevocability of the 
loss is not acknowledged. Boss18,47 affirms that in 
such cases, of “Goodbye without leaving,”32 
caregivers may perceive and act on the patients’ 
presence when it is no longer a reality. As a result, 
they will have difficulties shifting their psycholog-
ical bond and recognizing their loss. 
Correspondingly, previous studies on disorders of 
consciousness depicted how physical contact 
became a means of interaction that compensated 
for the void created by the patient’s psychological 
absence. Furthermore, caregivers’ attempts to 
“give voice” to the living body of the patient 
became a life’s mission.19,20 Caregivers’ belief in 
the patient’s presence could be also understood as 
their way of resolving the cognitive dissonance of 
caring for a body that is psychologically unattain-
able. Given his presumed presence, the inherent 
self-blame and guilt arising from the thought over 
his potential neglect impede caregivers’ re-engage-
ment in ongoing life.
Within an overarching context of grief, our results 
corroborate earlier research exploring the role of 
boundary ambiguity and relationships.48–50 As seen 
in both the present and previous studies, the pro-
tracted state of an unclear relationship with a loved 
one creates confused perceptions about who is in or 
out of the family.47 The mediating effect of boundary 
ambiguity suggests that difficulties in relocating the 
relational bond with their “present-absent” loved 
one, by redefining it cognitively, behaviorally and 
affectively across time, can be a complicating imped-
iment to caregivers’ integration of the loss.
Despite these important findings, our study 
bears certain limitations. First, caregivers’ NFC 
levels, which may represent a derivative of their 
continuous state of uncertainty, rather than a per-
sonality trait. Second, only inpatients were studied. 
It is unfortunate that the study did not include car-
egivers who provide care for patients at home, as 
the constant physical proximity might facilitate 
boundary ambiguity. Furthermore, our participant 
sample involved only caregivers who maintain 
active relations with the patient, which appears to 
involve great emotional cost. Therefore, future 
research should examine the emotional experience 
of caregivers whose coping strategy involves an 
“emotional cut-off”51 from the situation, accompa-
nied by a concrete disengagement from the patient.
In conclusion, this study shows that the unre-
solved nature of the loss exerts a tremendous 
10 Clinical Rehabilitation 00(0)
impact on caregivers’ grief reaction, by forming 
confused perceptions of the new nature of the rela-
tionship with the injured person. It provides pre-
liminary support for the application of ambiguous 
loss theory and boundary ambiguity to clinical 
work with this population. This core feature sug-
gests that interventions should focus on forging 
clearer emotional boundaries between caregivers 
and their present–absent loved one. It is possible 
that by doing so, caregivers will gradually separate 
themselves psychologically from the patient, where 
separateness implies both withdrawal and continu-
ation. It appears that reconstructing their life while 
revising their continuous bond with him will ena-
ble their grieving process to unfold into clarity as 
to what has been lost. This will allow the possibil-
ity of adaptation to the new, emergent situation, 
without being haunted by lingering pain, guilt and 
doubts. Incorporating the new reality into their 
ongoing life will contribute to a sense of closure, 
albeit imposed.
Finally, the present investigation highlighted 
caregivers’ enduring distress, which is rooted in 
their inability to attain resolution. As such, it bears 
relevance to the medical–ethical–legal controversy52 
regarding life-sustaining procedures associated 
with chronic disorders of consciousness.53 With much 
sensitivityto caregivers’ condition, our results sup-
port the idea that a key policy priority should be 
given to a change in the medical legislation54 to 
help caregivers reach emotional salvation and 
finality, regarding their “partially dead” loved one.
Clinical Messages
•• Caregivers of patients with prolonged 
disorders of consciousness experience a 
prolonged grief reaction, similar to the 
aftermath of death.
•• Patient’s continuous physical presence 
creates great ambiguity regarding the 
nature of the loss, hence impeding re- 
solution.
•• Tolerating uncertainty and redefining the 
relational bond are crucial features in 
mitigating caregivers’ grief.
Acknowledgements
Yael Zaksh and Einat Yehene are shared first co-authors. 
We would like to thank all the caregivers for their par-
ticipation in the study; to Meital Davidian, MA from the 
Academic College of Tel-Aviv Yaffo, Israel, for her pro-
ficient and sensitive data collection and Ofri Bar-Nadav, 
PhD from the International Center for the Study of Loss, 
Bereavement and Human Resilience, University of 
Haifa, Israel, for his insightful comments and for provid-
ing the comparative data of the national normative 
samples
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest 
with respect to the research, authorship and/or publica-
tion of this article.
Funding
The author(s) received no financial support for the 
research, authorship and/or publication of this article.
ORCID iD
Einat Yehene https://orcid.org/0000-0001-9780-3974
References
 1. Royal College of Physicians. Prolonged disorders of con-
sciousness: national clinical guidelines. London: RCP, 
2013.
 2. Monti MM, Laureys S and Owen AM. The vegetative 
state. BMJ 2010; 341: c3765.
 3. Leonardi M, Sattin D and Raggi A. An Italian population 
study on 600 persons in vegetative state and minimally 
conscious state. Brain Inj 2013; 27: 473–484.
 4. Corallo F, Bonanno L, Buono VL, et al. Psychological 
distress of family members of vegetative and minimally 
conscious state patients. Acta Medica Mediterr 2015; 
31(2): 297–302.
 5. Moretta P, Estraneo A, De Lucia L, et al. A study of the 
psychological distress in family caregivers of patients 
with prolonged disorders of consciousness during in-hos-
pital rehabilitation. Clin Rehabil 2014; 28(7): 717–725.
 6. Pagani M, Giovannetti AM, Covelli V, et al. Physical 
and mental health, anxiety and depressive symptoms in 
caregivers of patients in vegetative state and minimally 
conscious state. Clin Psychol Psychother 2014; 21(5): 
420–426.
 7. Giovannetti AM, Covelli V, Sattin D, et al. Caregivers of 
patients with disorder of consciousness: burden, quality of 
life and social support. Acta Neurol Scand 2015; 132(4): 
259–269.
https://orcid.org/0000-0001-9780-3974
Zaksh et al. 11
 8. Chiambretto P, Rossi Ferrario S and Zotti AM. Patients in 
a persistent vegetative state: caregiver attitudes and reac-
tions. Acta Neurol Scand 2001; 104(6): 364–368.
 9. Covelli V, Sattin D, Giovannetti AM, et al. Caregiver’s 
burden in disorders of consciousness: a longitudinal study. 
Acta Neurol Scand 2016; 134(5): 352–359.
 10. Giovannetti AM, Leonardi M, Pagani M, et al. Burden 
of caregivers of patients in vegetative state and mini-
mally conscious state. Acta Neurol Scand 2013; 127(1): 
10–18.
 11. Leonardi M, Giovannetti AM, Pagani M, et al. Burden 
and needs of 487 caregivers of patients in vegetative state 
and in minimally conscious state: results from a national 
study. Brain Inj 2012; 26(10): 1201–1210.
 12. Chiambretto P, Moroni L, Guarnerio C, et al. Prolonged 
grief and depression in caregivers of patients in vegetative 
state. Brain Inj 2010; 24(4): 581–588.
 13. De la Morena MJE and Cruzado JA. Caregivers of patients 
with disorders of consciousness: coping and prolonged 
grief. Acta Neurol Scand 2013; 127(6): 413–418.
 14. Guarnerio C, Prunas A, Della Fontana I, et al. Prevalence 
and comorbidity of prolonged grief disorder in a sample 
of caregivers of patients in a vegetative state. Psychiatr Q 
2012; 83(1): 65–73.
 15. Bastianelli A, Gius E and Cipolletta S. Changes over time 
in the quality of life, prolonged grief and family strain of 
family caregivers of patients in vegetative state: a pilot 
study. J Health Psychol 2014; 21: 844–852.
 16. Hamama-Raz Y, Zabari Y and Buchbinder E. From hope 
to despair, and back: being the wife of a patient in a per-
sistent vegetative state. Qual Health Res 2013; 23(2): 
231–240.
 17. Kitzinger C and Kitzinger J. Grief, anger and despair in 
relatives of severely brain injured patients: responding 
without pathologising (Editorial). Clin Rehabil 2014; 28: 
627–631.
 18. Boss P. Ambiguous loss: learning to live with unresolved 
grief. Cambridge, MA: Harvard Press, 1999.
 19. Cipolletta S, Pasi M and Avesani R. Vita tua, mors mea: 
the experience of family caregivers of patients in a vegeta-
tive state. J Health Psychol 2016; 21(7): 1197–1206.
 20. Giovannetti AM, Černiauskaite M, Leonardi M, et al. 
Informal caregivers of patients with disorders of con-
sciousness: experience of ambiguous loss. Brain Inj 2015; 
29(4): 473–480.
 21. Kitzinger C and Kitzinger J. “This in-between”: how 
families talk about death in relation to disorders of con-
sciousness. In: Van Brussels L and Carpentier N (eds) 
The social construction of death. London: Palgrave, 2014, 
pp.239–258.
 22. Soeterik SM, Connolly S, Playford ED, et al. The psycho-
logical impact of prolonged disorders of consciousness 
on caregivers: a systematic review of quantitative studies. 
Clin Rehabil 2017; 31(10): 1374–1385.
 23. Rubin SS. The two-track model of bereavement: over-
view, retrospect, and prospect. Death Stud 1999; 23(8): 
681–714.
 24. Boss P, Greenberg J and Pearce-McCall D. Measurement 
of boundary ambiguity in families. Saint Paul, MN: 
Minnesota Agricultural Experiment Station, University of 
Minnesota, 1990.
 25. Boss P, Caron W, Horbal J, et al. Predictors of depression 
in caregivers of dementia patients: boundary ambiguity 
and mastery. Fam Process 1990; 29: 245–254.
 26. Webster DM and Kruglanski AW. Individual differences 
in need for cognitive closure. J Pers Soc Psychol 1994; 
67(6): 1049–1062.
 27. Roets A and Van Hiel A. Item selection and validation of 
a brief, 15-item version of the need for closure scale. Pers 
Indiv Differ 2011; 50(1): 90–94.
 28. Rubin SS, Nadav OB, Malkinson R, et al. The two-track 
model of bereavement questionnaire (TTBQ): develop-
ment and validation of a relational measure. Death Stud 
2009; 33(4): 305–333.
 29. Baron RM and Kenny DA. The moderator-mediator vari-
able distinction in social psychological research: concep-
tual, strategic, and statistical considerations. J Pers Soc 
Psychol 1986; 51(6): 1173–1182.
 30. Preacher KJ and Hayes AF. Asymptotic and resampling 
strategies for assessing and comparing indirect effects in 
multiple mediator models. Behav Res Methods 2008; 40: 
879–891.
 31. Boss P. The trauma and complicated grief of ambiguous 
loss. Pastoral Psychol 2010; 59(2): 137–145.
 32. Boss P. Ambiguous loss research, theory, and practice: 
reflections after 9/11. J Marriage Fam 2004; 66(3): 551–
566.
 33. Rubin SS, Malkinson E and Witztum E. Working with 
the bereaved: multiple lenses on loss and mourning. New 
York: Routledge, 2012.
 34. Stern JM, Sazbon L, Becker E, et al. Severe behavioural 
disturbance in families of patients with prolonged coma. 
Brain Inj 1988; 2(3): 259–262.
 35. Dupuis SL. Understanding ambiguous loss in the context 
of dementia care: adult children’s perspectives. J Gerontol 
Soc Work 2002; 37(2): 93–115.
 36. Boss P. Loving someone who has dementia. How to find 
hope while coping with stress and grief. San Francisco, 
CA: Jossey-Bass, 2011.
 37. Rando TA. A comprehensive analysis of anticipatory 
grief: perspectives, processes, promises, and problems. In: 
Rando TA (ed.) Loss and anticipatory grief. New York: 
Lexington Books, 1986, pp.1–36.
 38. Blandin K and PepinR. Dementia grief: a theoretical 
model of a unique grief experience. Dementia 2017; 
16(1): 67–78.
 39. Crawford S and Beaumont JG. Psychological needs of 
patients in low awareness states, their families, and health 
professionals. Neuropsychol Rehabil 2015; 15(3–4): 548–
555.
 40. Jox RJ, Kuehlmeyer K, Klein AM, et al. Diagnosis and 
decision making for patients with disorders of conscious-
ness: a survey among family members. Arch Phys Med 
Rehabil 2015; 96(2): 323–330.
12 Clinical Rehabilitation 00(0)
 41. Monti MM, Vanhaudenhuyse A, Coleman MR, et al. 
Willful modulation of brain activity in disorders of con-
sciousness. N Engl J Med 2010; 362(7): 579–589.
 42. Fernández-Espejo D and Owen AM. Detecting awareness 
after severe brain injury. Nat Rev Neurosci 2013; 14(11): 
801–809.
 43. Rodrigue C, Riopelle R, Bernat JL, et al. How contextual 
and relational aspects shape the perspective of healthcare 
providers on decision making for patients with disorders 
of consciousness: a qualitative interview study. Narrat Inq 
Bioeth 2013; 3(3): 261–273.
 44. Samuel G and Kitzinger J. Reporting consciousness in 
coma: media framing of neuro-scientific research, hope, 
and the response of families with relatives in vegetative 
and minimally conscious states. JOMEC J 2013; 3: 1–15.
 45. Tresch DD, Sims FH, Duthie EH, et al. Patients in a per-
sistent vegetative state attitudes and reactions of family 
members. J Am Geriatr Soc 1991; 39(1): 17–21.
 46. Bowlby J. Attachment and loss: retrospect and prospect. 
Am J Orthopsychiatry 1982; 52(4): 664–678.
 47. Boss P. A clarification of the concept of psychological 
father presence in families experiencing ambiguity of 
boundary. J Marriage Fam 1977; 39: 141–151.
 48. Kaplan L and Boss P. Depressive symptoms among 
spousal caregivers of institutionalized mates with 
Alzheimer’s: boundary ambiguity and mastery as predic-
tors. Fam Process 1999; 38(1): 85–103.
 49. Berge JM and Holm KE. Boundary ambiguity in par-
ents with chronically ill children: integrating theory and 
research. Fam Relat 2007; 56(2): 123–134.
 50. Landau J and Hissett J. Mild traumatic brain injury: 
impact on identity and ambiguous loss in the family. Fam 
Syst Health 2008; 26(1): 69–85.
 51. Bowen M. Family therapy in clinical practice. New York: 
Jason Aronson, 1978.
 52. Holland S, Kitzinger C and Kitzinger J. Death, treatment 
decisions and the permanent vegetative state: evidence 
from families and experts. Med Health Care Philos 2014; 
17(3): 413–423.
 53. Constable C. Withdrawal of artificial nutrition and hydra-
tion for patients in a permanent vegetative state: changing 
tack. Bioethics 2012; 26(3): 157–163.
 54. Kitzinger J, Kitzinger C and Cowley J. When “sanctity of 
life” and “self-determination” clash: Briggs versus Briggs 
[2016] EWCOP 53—implications for policy and practice. 
J Med Ethics 2017; 43: 446–449.