exploring parental grief: combining quantitative and qualitative measures

Exploring Parental Grief:Combining Quantitative and

Qualitative Measures

Joan Arnold, Penelope Buschman Gemma, and Linda F. Cushman

Archives of Psyc

Purpose. This study explores parental grief on the death of a child of any age.Study design. The sampling frame for the study consisted of 74respondents reporting that they experienced the death of a living child.With the exception of standard demographic measures, the quantitativeand qualitative items in the instrument were designed specifically for thisproject. The content validity of the instrument was assessed by a panel ofexperts on grief. Key quantitative items related to child loss and grief wereformatted with between three- and five-point ordinal answer categories.Two types of qualitative items were developed by the authors. Thereliability of the instrument designed for this study was not determined.Results. This survey of parents who experienced the death of a child duringtheir lifetime explores and extends current understanding of the complexemotional response of grief. The study offers empirical support for thenotion of grief as ongoing in the life of a parent whose child had died. Thefindings have significant implications for further clinical research support-ing studies to explore commonalities in the experience of grieving familiesregardless of the cause of and time since the death of their child. Inaddition, the findings may serve to inform the development and provisionof services for bereaved parents.D 2005 Elsevier Inc. All rights reserved.

From The College of New Rochelle School of Nursing,New Rochelle, NY 10805, Columbia University Schoolof Nursing; and Columbia University Mailman School ofPublic Health.

Address reprint requests to Joan Arnold, PhD, RN,The College of New Rochelle School of Nursing, 29Castle Place, New Rochelle, NY 10805.

E-mail address: [email protected] 2005 Elsevier Inc. All rights reserved.0883-9417/1801-0005$30.00/0doi:10.1016/j.apnu.2005.07.008

GRIEVING IS A universal human experience.

Historically, grief has been defined as a

response to loss and death that is organized by

sequential steps, stages, or phases and bound by

the dimension of time, requiring closure for

resolution (Kubler-Ross, 1969; Lindemann, 1944;

Worden, 1982). In this perspective, grief is a

quantifiable, episodic event related to the crises

of loss and death, described as a human response to

these events (Parad, 1965; Worden, 1982) and

characterized as a temporary condition, requiring

crisis intervention, treatment for resolution, and

emotional disengagement from the lost object to

overcome it (Bowlby, 1961; Engel, 1961, 1964;

Freud, 1957, 1963; Parkes, 1965, 1975).

The experience of clinicians providing care for

the bereaved has not been consistent with this view

of grief as an episode that must come to an end

through resolution (Arnold, 1995, 1996; Benoliel,

hiatric Nursing, Vol. 19, No. 6 (December),

1983; Cody, 1991; Cowles & Rodgers, 1991,

2000; Davies, 2004; Klass, Silverman, & Nickman,

1996; Lindgren, Burke, Hainsworth, & Eakes,

1992; Martocchio, 1985; McClowry, Davies, May,

Kulenkamp, & Martinson, 1987; Miles & Crandall,

1983; Miles & Demi, 1986; Murphy, 1983; Pilking-

ton, 1993). Grief is described as paramount in

the lives of bereaved parents; grieving, as a process

of maintaining connectedness (Arnold & Gemma,

2005: pp 245–255 245

ARNOLD ET AL.246

1983, 1994; Davies, 2004; Hutti, 1984; Martinson,

1992; Miles, 1985; Miles & Demy, 1986; Uren &

Wastell, 2002). Child death is viewed by parents as

an empty space within (McClowry et al., 1987). In

a landmark study comparing adult bereavement

following the death of a parent, spouse, and child,

Sanders (1980) found significantly higher inten-

sities of grief among those surviving the death of a

child and subsequently developed an integrated

theory detailing phases of grief (Sanders, 1989).

Following the lead of such research, the authors

seek to expand and support the understanding of

parental grief.

PURPOSE

This study explores the ongoing nature of

parental grief. Few empirical studies have exam-

ined this evolving view of the grieving process with

a combination of quantitative and qualitative mea-

sures. Moreover, this study includes parents whose

child had died regardless of the years that had

passed since the death, the age of the child at the

time of death, and the cause of the child’s death.

STUDY DESIGN AND METHODS

This is a cross-sectional, retrospective, self-

administered survey by mail. An adaptation of

the Dillman method was used to maximize the

respondents’ understanding of and participation in

the project (Dillman, 1978). The primary eligibility

requirement was having experienced the death of a

natural-born or adopted child.

Data Collection

An introductory letter with a self-addressed

stamped postcard was sent to the entire sampling

frame, which was composed of 5,211 nursing

alumni (described in the Sample section). Among

them, 74 respondents (72%) reported that they

experienced the death of a living child and 29

(28%) reported that they experienced a stillbirth.

The 74 respondents constitute the sample popula-

tion of this study.

Instruments

The study used a new survey instrument

designed to explore the premises that child death

has lasting effects on parents and that parental grief

is an ongoing process (Arnold, Gemma, & Cush-

man, 2000). In addition to standard demographic

items (17 items), the questionnaire included

quantitative (52 items) as well as qualitative

(11 items) measures of grief, loss, and related

variables. Key quantitative items related to child

loss and grief were formatted with between three-

and five-point ordinal answer categories (e.g.,

intense, moderate, mild, and no feelings of loss).

Qualitative items allowed respondents to provide

details in their own words after answering a

precoded item and to describe how specific images

of grief related to their experiences of loss. A panel

of experts on grief reviewed the quantitative and

qualitative items for content and completeness.

Two rounds of pretesting were conducted. Finally,

the construct validity of the key outcome variable

(whether respondents felt that their grief had ended

or continues) received preliminary support in the

bivariate analysis. The reliability of the instrument

was not tested.

Analysis

All quantitative data were entered into an SPSS

data file, cleaned for wild codes and inconsisten-

cies, and analyzed by the authors using standard

bivariate techniques. All associations should be

interpreted as preliminary, owing to the sample

size. Qualitative data were analyzed by the clini-

cal authors, who independently read all open an-

swers and developed a coding scheme. Once this

was accomplished, coded domains were compared

and reconciled.

Sample

The sampling frame for the study consisted of

the comprehensive list of members of the Colum-

bia University–Presbyterian Hospital School of

Nursing Alumni Association (1926–2000). By

design, all parents in the sample had experienced

the death of a child at some point during their adult

lives. Sociodemographic characteristics are sum-

marized in Table 1.

Causes of Child Death

Specific causes of death were wide ranging in

this sample, reflecting all the common as well as a

number of rare reasons for death in infancy,

childhood, adolescence, and adulthood. For ana-

lytic purposes, causes of death were assigned to one

of four categories: congenital (12%), prematurity/

birth-related complications (26%), unexpected

(40%), and illness (22%). In the congenital

Table 1. Sociodemographic Characteristics of the Sample

Race

White 99%

Sex

Female 97%

Age

Mean age 66 years

Age range 42–90 years

Education

Baccalaureate degree or higher 84%

Completed some graduate study

or graduate degree

54%

Marital status

Legally married 68.5%

Widowed 36%

Divorced 35%

Religion

Protestant 59%

Very/Somewhat religious 66%

Births

Live births 1–10

Two–four live births 74%

Nature of work in the past year

Employed 40%

Volunteered 84%

Household composition

Two individuals 51%

Alone 31%

Between three and five individuals 18%

Income

z$75,000 45%

$35,000–74,999 30%

b$35,000 25%

EXPLORING PARENTAL GRIEF: COMBINING QUANTITATIVE AND QUALITATIVE MEASURES 247

category, causes of death included congenital heart

diseases, Down syndrome, metabolic renal dis-

ease, and multiple birth defects. The prematurity/

birth-related complications category encompassed

hyaline membrane disease, RH incompatibility/

erythroblastosis, vasa previa, anencephaly, and

atelectasis. Unexpected deaths were attributed to

murder, suicide, drug overdose, drowning, horse-

back riding accident, car accident, being hit by a

car, farm equipment accident, heat stroke during

football practice, and Sudden Infant Death Syn-

drome. Illnesses causing death ranged from septic

shock, medulloblastoma, non-Hodgkins lym-

phoma, brain tumor, leukemia, breast cancer,

esophageal cancer, seizures, diabetes mellitus,

and HIV/AIDS to complications related to cardiac

surgery and cardiomyopathy.

Characteristics Related to Child Death

Six respondents had experienced the death of

two children and were asked to discuss experiences

regarding the most recent death. Approximately

40% of the sample reported that their child had

died at the age of 1 year or younger. The age at the

time of child death ranged from 0 to 48 years

(M = 13 years). The mean age of children older

than 1 year who died was 22 years. When asked if

their child’s death was expected or not, 15% of the

parents said that it was. The range of years since

the child’s death was 1 year or less to 62 years. The

mean number of years since the child’s death at the

time of the survey was 24.4 years. The parents’

ages ranged from 22 to 80 years at the time of

death (M = 42 years).

RESULTS

Intensity of Loss

Virtually, all respondents reported feeling in-

tense loss in the weeks that followed their child’s

death (90.5%); most reported that, over time, those

feelings became less intense. However, half the

sample reported that the intensity level of their

current feelings of loss varies, thus defying one

descriptor of intensity.

Social Support

With regard to social support, most of the parents

(86%) reported that most people who knew them

and knew about the death reached out to them, with

only 10% reporting that they felt shunned. How-

ever, when asked if people tended to ignore them

after the death, one quarter gave an affirmation.

Respondents were also asked about the extent to

which several resources helped them during the first

difficult year after the death of their child. Parents

most frequently stated that their own inner strength

and the support of family helped ba lot Q (73% and

62%, respectively). The support of friends, spiritu-

ality or religion, and working were reported to have

helped ba lot Q by smaller but still notable propor-

tions of the sample (51%, 44%, and 40%, respec-

tively). During the year after death, support groups

and individual psychological counseling and ther-

apy were viewed by few parents as having helped

ba lot Q (5% and 7% respectively).

Perceptions Regarding the Continuationof Grief

Parents were asked directly about the continuing

nature of their grief using the following item

developed and pretested specifically for this study:

bPeople have different ideas about grief after the

ARNOLD ET AL.248

death of a child. Some people say it is a process

that has an ending, and others believe that grief

never ends. How about you? Would you say your

grief has come to an end, or that your grief

continues? Q A key finding is shown in Table 2:

63.5% of parents, the majority, believe that their

grief continues and 36.5% report that their grief

had come to an end.

Other Feelings of Loss and the Continuationof Grief

To more fully understand the perception and

experience of lasting grief, the authors compared

parents whose grief had ended with those whose

grief continues on a series of other characteristics.

The first variable set consists of other feelings of

loss, thus providing a preliminary test of the

construct validity of the item. As shown in Table 2,

the two groups of parents are compared with re-

gard to bletting goQ of their deceased child, the

level of connection they currently feel with their

child, the frequency with which they currently

think about him or her, the proportion of

themselves that they feel is currently bmissingQ,and the degree (intensity) of loss they currently

feel. On every measure, the group of parents who

believe that their grief continues is significantly

different from the group of those whose grief had

ended, in the expected direction. For example,

Table 2. Comparing Those Whose Grief had Ended With Those

Would you say your grief has

come to an end or that it

continues?

Some people say parents must Have already let go [n (%)]

eventually let go of a child Will eventually let go [n (%)]

who died (n = 73) Will never let go [n (%)]

How much of yourself continues All/Most [n (%)]

to feel connected to your Some/A little [n (%)]

child? (n = 73) None [n (%)]

These days, how often do you Constantly [n (%)]

think of your child? Often/Sometimes [n (%)]

(n = 74) Rarely/Never [n (%)]

Some people feel that part of All/Most is missing [n (%)]

them is missing because Some/A little is missing [n (%)]

their child died (n = 60) None is missing [n (%)]

What is the degree of loss you Intense/Moderate [n (%)]

currently feel regarding your Mild [n (%)]

child’s death? None [n (%)]

(n = 73) Varies [n (%)]

Note. P values were calculated with v2.

Percentages are column percentages.

whereas 34% of those whose grief continues state

that they have let go of their child, most (81%) of

those whose grief had ended feel that they have

done so. Similarly, 43% of those whose grief

continues but only 8% of those whose grief had

ended report that all or most of them still feel

connected to their child. It is interesting to note

that most (81%) of those whose grief had ended

still feel some connection to their child. Similarly,

parents whose grief continues are more likely than

those whose grief had ended to think about their

child, to feel that part of them is missing at

present, and to experience strong feelings of loss

at the present time. In contrast, it should be noted

that overall life satisfaction does not distinguish

the two groups of parents. Indeed, more than 90%

of grief enders and grief continuers report that

they are very or somewhat satisfied with their life

these days. Although the former subjects are more

likely to report that they are very satisfied, the

difference is not statistically significant (P = .378,

not shown).

Characteristics of Child Death and theContinuation of Grief

The authors examined whether circumstances

surrounding a child’s death, as well as time since

the death, were associated with parents’ current

feeling regarding the continuation or end of their

Whose Grief Continues With Other Grief-Related Variables

All responses Grief had ended Grief continues P

36.5% 63.5%

37 (50.7) 21 (80.8) 16 (34.0) .000

4 (5.5) 1 (3.8) 3 (6.4)

32 (43.8) 4 (15.4) 28 (59.6)

22 (30.1) 2 (7.7) 20 (42.6) .001

47 (64.4) 21 (80.8) 26 (55.3)

4 (5.5) 3 (11.5) 1 (2.1)

2 (2.7) 0 (0.0) 2 (4.3) .010

58 (78.4) 18 (66.7) 40 (85.1)

14 (18.9) 9 (33.3) 5 (10.6)

4 (6.7) 0 (0.0) 4 (9.8) .089

48 (80.0) 15 (78.9) 33 (80.5)

8 (13.3) 4 (21.1) 4 (9.8)

13 (18) 1 (4) 12 (26) .002

19 (26) 11 (42) 8 (17)

4 (5) 4 (15) 0 (0.0)

37 (51) 10 (37) 27 (57)

Table 3. Comparing Those Whose Grief had Ended With Those Whose Grief Continues on Circumstances of Child Death

All responses Grief had ended Grief continues P

Cause of death (n = 73)

Congenital [n (%)] 9 (12.3) 3 (23.1) 3 (6.4) .006

Prematurity/Birth complications [n (%)] 19 (26.0) 9 (34.6) 10 (21.3)

Unexpected [n (%)] 29 (39.7) 8 (30.8) 21 (44.7)

Illness [n (%)] 16 (21.9) 3 (11.5) 13 (27.7)

Average time since death among those whose child was older than 1 year

Mean years since death 18.79 (n = 43) 20.09 (n = 11) 18.34 (n = 32) .650

Age of child (n = 74)

V1 year [n (%)] 31 (41.9) 16 (59.3) 15 (31.9) .085

N1 yearV21 years [n (%)] 20 (27.0) 4 (14.8) 16 (34.0)

z21 years [n (%)] 23 (31.1) 7 (25.9) 16 (34.0)

Note. P values for expected versus unexpected death were calculated with v2.

P values for (a) child’s age at death and (2) years since death were calculated with t test.



grief. Table 3 presents these data with mixed

results. A higher proportion of those whose grief

had ended had children die of congenital problems

and prematurity/birth complications (primarily

Table 4. Comparing Those Whose Grief had Ended With Those Who

Death of C

All responses

How has your own health changed since the death of your child? (n =

Better due to death [n (%)] 0 (0.0)

About the same [n (%)] 37 (86.0)

Worse due to death [n (%)] 6 (14.0)

Total [n (%)] 43 (100.0)

How has your sex life changed since the death of your child? (n = 30




Total [n (%)] 30 (100.0)

How has your relationship with your spouse/partner changed since th




Total [n (%)] 38 (100.0)

How has your relationship with surviving children changed since the d




Total [n (%)] 42 (100.0)

How has your involvement with friends and the community changed s




Total [n (%)] 46 (100.0)

How has your employment productivity changed since the death of yo




Total [n (%)] 23 (100.0)

Note. P values were calculated with v2.


babies). Indeed, more than half of this group

(59%), compared with approximately one third of

those whose grief continues (32%), had a baby die

at an age younger than 1 year. However, this

se Grief Continues on Life and Personality Changes Owing to

hild

Grief had ended Grief continues P

43)

0 (0.0) 0 (0.0) .226

16 (94.1) 21 (80.8)

1 (5.9) 5 (19.2)

17 (100.0) 26 (100.0)

)

0 (0.0) 1 (5.6) .107

12 (100.0) 11 (61.1)

0 (0.0) 6 (33.3)

12 (100.0) 18 (100.0)

e death of your child? (n = 38)

1 (6.7) 4 (17.4) .365

13 (86.7) 18 (78.3)

1 (6.7) 1 (4.3)

15 (100.0) 23 (100.0)

eath of your child? (n = 42)

2 (13.3) 14 (51.9) .007

12 (80.0) 13 (48.1)

1 (6.7) 0 (0.0)

15 (100.0) 27 (100.0)

ince the death of your child? (n = 46)

0 (0.0) 5 (15.6) .647

14 (100.0) 24 (75.0)

0 (0.0) 3 (9.4)

14 (100.0) 32 (100.0)

ur child? (n = 23)

0 (0.0) 4 (26.7) .794

8 (100.0) 8 (53.3)

0 (0.0) 3 (20.0)

8 (100.0) 15 (100.0)

ARNOLD ET AL.250

difference only approaches statistical significance

and should be interpreted cautiously. Finally, it

should be noted there is no significant relationship

between the grief continuation/end variable and

time since the death among those who lost a child

who was older than 1 year.

Insight from the open-ended items reinforces

the notion that circumstances surrounding the

death do shape the experience of parental grief.

For example, death by suicide was found to add

guilt and anger to parental grief, making it

complicated for parents to discuss their grief.

Other examples include the following: death by

homicide had prevented a parent from forgiving

the murderer who had shown no remorse; death

from profound chronic illness had afforded the

parent, while missing her child, a measure of

relief by sparing the child prolonged suffering;

Table 5. Comparing Those Whose Grief had Ended With Those W

All responses Grief h

Long-lasting effects on respondents include their feeling stronger (n =

True [n (%)] 58 (80.6) 23 (

False [n (%)] 14 (19.4) 4 (

Long-lasting effects on respondents include their feeling more vulnera

True [n (%)] 41 (57.7) 12 (

False [n (%)] 30 (42.3) 14 (

Long-lasting effects on respondents include their feeling more depress

True [n (%)] 19 (27.1) 2 (

False [n (%)] 51 (72.9) 24 (

Long-lasting effects on respondents include their feeling more sensitiv

True [n (%)] 60 (82.2) 21 (

False [n (%)] 13 (17.8) 6 (

Long-lasting effects on respondents include their feeling more angry (

True [n (%)] 11 (15.7) 0 (

False [n (%)] 59 (84.3) 26 (

Long-lasting effects on respondents include their feeling more hopeles

True [n (%)] 8 (11.4) 0 (

False [n (%)] 62 (88.6) 26 (

Long-lasting effects on respondents include their feeling more courage

True [n (%)] 43 (61.4) 17 (

False [n (%)] 27 (38.6) 9 (

Long-lasting effects on respondents include their feeling more fearful

True [n (%)] 22 (31.0) 5 (

False [n (%)] 49 (69.0) 21 (

Long-lasting effects on respondents include their feeling more detache

True [n (%)] 11 (15.9) 2 (

False [n (%)] 58 (84.1) 24 (

Long-lasting effects on respondents include their feeling more connec

True [n (%)] 2 (2.8) 0 (

False [n (%)] 69 (97.2) 26 (

Long-lasting effects on respondents include their feeling more creative

True [n (%)] 21 (30.9) 6 (

False [n (%)] 47 (69.1) 17 (

Note. P values were calculated with v2 (two tailed).


death of an adult child suffering from mental

illness left a parent feeling that if she had lived

closer to him she could have secured care that

might have saved his life; a parent whose young

child died a painful death from cancer expressed

relief in releasing him at death as the end to his

suffering. Finally, particularly strong regret was

expressed by parents who were unable to see or

touch their infants. Another parent stated that she

was encouraged not to see her deceased child by

her spouse, physician, and pastor and now

wonders why she ever agreed.

Continuation of Grief and Other ParentalCharacteristics

Although it is widely recognized that a child’s

death has a profound impact on parents, this impact

has not been well operationalized. Parents in the

hose Grief Continues on Perceived Characteristics of Self

ad ended Grief continues P

72)

85.2) 35 (77.8) .442

14.8) 10 (22.2)

ble (n = 71)

46.2) 29 (64.4) .133

53.8) 16 (35.6)

ed (n = 70)

7.7) 17 (38.6) .005

92.3) 27 (61.4)

e (n = 73)

77.8) 39 (84.8) .450

22.2) 7 (15.2)

n = 70)

0.0) 11 (25.0) .005

100.0) 33 (75.0)

s (n = 70)

0.0) 8 (18.2) .021

100.0) 36 (81.8)

ous (n = 70)

65.4) 26 (59.1) .601

34.6) 18 (40.9)

(n = 71)

19.2) 17 (37.8) .103

80.8) 28 (62.2)

d (n = 69)

7.7) 9 (20.9) .145

92.3) 34 (79.1)

ted to the past than the present (n = 71)

0.0) 2 (4.4) .276

100.0) 43 (95.6)

(n = 68)

26.1) 15 (33.3) .541

73.9) 30 (66.7)


study were asked whether they felt their child’s

death had impacted several specific areas of their

lives, including their other roles and relationships

as well as internal characteristics. These data are

presented in Tables 4 and 5 for the entire sample

and separately for parents whose grief had ended

and those whose grief continues.

As shown in Table 4, the modal response of the

sample overall was that their other relationships,

health, and community/work involvement stayed

the same after the death of their child. This pattern

was most pronounced among those whose grief

had ended: 80% or more of that group gave the

bstayed the sameQ response for every area dis-

cussed. Although the bstayed the sameQ responsewas also the modal one among grief continuers,

they displayed more variability in their answers.

For example, grief continuers were somewhat more

likely than enders to report changes in their own

health and sex life (worse), relationship with

spouse and living children (better), and involve-

ment with community/work productivity (both

better and worse). The only statistically significant

difference between the two groups was in reported

changes in their relationships with surviving

children. This area was described as changed by

more than half of the grief continuers, and all of

them rated that relationship as better. This pattern

was reinforced in the qualitative data. When asked

if their child’s death affected their relationships

with other children in the family, 29 respondents

described, in an open-ended question, a continuum

of reactions including difficulty loving their other

children, being anxious and guarded in their

relationships, and experiencing a heightened sense

of love and attachment with the need to protect

them from harm.

Qualitative findings from the open-ended items

are illustrative and increase understanding of other

life changes related to child death. Nearly half the

sample (n = 34) reported a wide range of changes,

including strained marital relationships, reduced

communication with spouses/partners, and signifi-

cant distancing resulting in divorce. Some reported

the strong desire to replace their deceased child.

Others reported feeling less competent and less in

control whereas some felt stronger, bolder, more

sensitive, and more spiritual. Thirteen respondents

described the transforming nature of their grief,

reflected in changed careers, volunteer efforts, and

demonstrations of greater compassion.

With regard to perceived characteristics of self,

as shown in Table 5, most of the sample overall

report that the experience of child death has made

them stronger (81%), more sensitive (82%), and

more courageous (61%). Of note, grief enders

and continuers are quite similar along these

dimensions, with most of them reporting these

lasting feelings. At the same time, a significantly

higher proportion of those whose grief continues

report lasting feelings of depression and anger

(and, to a lesser extent, vulnerability, hopeless-

ness, and fear) as compared with their counter-

parts whose grief had ended. Once again, however,

it is important to note that most continuers—

between 61% and 82%—do not report such

feelings. Grief continuers are more likely to

report that they have become more spiritual than

the grief enders (not shown).

Other factors in respondents’ culture and back-

grounds were explored qualitatively. Important

influences noted by 55 respondents that shaped or

gave perspective to their experience of grief were

clustered into three broad categories: family of

origin experience with death, religion and spiritu-

ality, and support from others. Many respondents

referred to the significance of their nursing educa-

tion and clinical experience/expertise in helping

them with the loss and the associated grief.

Images as Measures of Grief

Based on clinical experience caring for grieving

families subsequent to child death, the authors

developed innovative qualitative measures in an

attempt to describe with more texture and depth the

emotions of parental grief. These measures were

conceived as images representing the experience of

grief, including an erupting volcano; a well into

which one descends; a tree that has lost a limb;

and, finally, a hollow or empty space. For each

qualitative measure, there was a dichotomous item

that asked, bHas this image of grief applied to you

since the death of your child?Q In addition, a

number of open-ended questions provided an

opportunity for respondents to describe freely the

varying ways their lives were changed by their

child’s death.

Three of the images (a volcano that could erupt,

descending into a well, and feeling like their tree of

life has lost a limb) had meaning only for a

minority of parents, although nearly one quarter

responded in the affirmative to the latter construct.

ARNOLD ET AL.252

For these images, the only difference between

parents whose grief continues and those whose

grief had ended was observed for the construct of

descending into a well, with the former group of

parents more likely to feel that the image applied to

their experience of grief (21% vs. 4%, table not

shown). It should be noted, however, that the

relationship only approaches statistical significance

(P = .058). In contrast, the fourth image, a hollow

or empty space inside, was described by most

(~75%) with regard to their child’s death. There

was no difference observed between grief contin-

uers and enders related to this last image. More-

over, multiple expressions of emptiness were

reported by 45 respondents as a hollow or empty

space, specifically describing a hole in the fabric of

life, a void, a broken place in the heart forever,

empty arms and belly, and an empty chair.

Associated emotions included loneliness, never

being able to feel pure joy again, hopelessness,

an emptiness that can never be filled or recovered,

and despair. This critical qualitative finding illus-

trates the meaning of the quantitative item of a

hollow or empty space inside, regardless of

whether grief continues or ends, the age of the

child at death, and the cause of death.

Parents were asked if other images had meaning;

41 responded with a plethora of their own images

that seemed to enhance connectedness to their

child in a highly individualized fashion. In this

way, images are consoling and represent emotions

of grief. Some of the images offered included a

beautiful child happy in heaven or waiting in

heaven to be reunited, angels and cherubs playing

in the clouds, an empty space at the table, the

absence of light, a soul joining other souls, a

broken heart, and being connected to other women

past and present who had experienced child death.

Parents were asked to share any other informa-

tion about the death of their child or their grief.

Their views of grief are like snapshots taken at a

point in time. The survey captures their emotions

and reactions at a particular point moment in their

lives. Some took the opportunity to express the

ways they keep connected to their child by creat-

ing memorials, saying goodnight to a photograph,

writing poetry, recognizing the need to nurture

living things such as taking care of sick birds and

animals, smelling the child’s hair as an olfactory

memory, and reliving the day the child died by

wishing to undo what happened and hoping the

child could live. There was also recognition that

certain songs, hymns, and places intensify the pain

of loss. Parents described ways they found to

believe that their child was all right in death.

A final qualitative survey item asked parents

about the word acceptance and its meaning in

their grief for their deceased children. The word

acceptance triggered two types of responses: one

that death must be accepted because it cannot be

changed and another that acceptance is not possible

because the loss is intolerable. In the first instance,

acceptance refers to the inevitability of death

because the child’s condition was incompatible

with life. Acceptance means resignation to the fact

of death, an acknowledgement of the finality of

death. Implied is the knowledge that the child

cannot come back. Acceptance was viewed as

active, in contrast to passive resignation. In the

second instance, the refusal to accept the death was

described as bliving next to itQ and as not being

able to accept but to grieve endlessly. Acceptance

has many facets: dealing with the reality and

inevitability of death, the necessity of going on

without one’s child, a measure of God’s will, and

the inability to accept the death of one’s child.

DISCUSSION

This survey of nurses who experienced the death

of a child during their lifetime explores and extends

our current understanding of the complex emotional

response of grief. Specifically, the study offers

empirical support for the notion of grief as ongoing

in the life of a parent whose child had died.

Tantamount to documenting that most parents

believe that their grief continues at present are

several noteworthy comparisons between this

group and those parents who state that their grief

had come to an end. First, the significant correla-

tions between continuing grief and other feelings

of loss are methodologically and substantively

important. These associations are perhaps to be

expected: it is not surprising that those who report

that they are still grieving are more likely than

others to feel they will never let go of their child,

that part of them is missing, and that they are still

strongly connected to their child. Similarly,

although only a minority of current grievers report

feeling that the death of their child made them

more depressed, hopeless, or angry, the proportions

were nonetheless higher than in the group of

parents whose grief had ended. Thus, the suggested


construct validity of the newly devised bend/continueQ is promising and should be explored

with larger and more diverse samples of griev-

ing parents.

Cause of death as an influence on grief proved

an interesting variable suggesting that parents

whose child died in infancy may be more likely

than others—in later years—to view their grief for

that child to have ended rather than to be

continuing. In contextualizing this finding, it is

likely that most infant deaths in this sample

occurred within a social and health care context

that did not recognize the grief of parents,

particularly parents grieving the death of an infant.

Indeed, prior to the 1970s, when perinatal grief

programs were developed, parents of a deceased

baby were often isolated and discharged without

any support service whatsoever. Parents were

encouraged by health care professionals to sup-

press their grief or at least end it quickly. Thus, the

authors interpret this association between infant

death and grief ending as a potential methodolog-

ical artifact. These data overall support the position

that grief on the death of a child, regardless of the

cause of death, the age of the child at death, and the

time since the death, is universal and ongoing.

These findings have implications for care of

bereaved parents and can serve to guide evi-

dence-based clinical practice.

Overall, other comparisons between parents

whose grief continues and those whose grief had

ended reveal similarities that are striking and

suggest that continuing grief can be and is

combined with positive, life-affirming feelings.

For example, parents whose grief continues are

no more likely than their counterparts to feel that

important areas of their lives (e.g., their health and

involvement with friends and community) are

worse due to their child’s death. Moreover, those

who report that their grief continues are equally

likely to feel that their child’s death made them

stronger and more courageous and sensitive over-

all. Those who continue to feel grief are signifi-

cantly more likely than others to report that the

experience of child death has improved their

relationship with their remaining children and

made them more spiritual overall. In addition,

parents who say their grief had ended are just as

likely to respond to the image of a hollow or empty

space inside and have similar levels of overall life

satisfaction as those whose grief continues. These

data yield a picture of continuing grief that

includes more intense feelings of loss and related

emotions than are reported by other parents, yet

similar levels of courage, strength, sensitivity, and

increased spirituality and improved relationships

with surviving children. Those parents who said

that their grief had ended also described ongoing

feelings of loss and emptiness when responding to

the open-ended questions. The authors view this

disparity as a revealing paradox.

LIMITATIONS

This study has several limitations that should be

considered in interpreting the results. These

limitations lie in three areas: sample, time frame,

and omitted questions. First, the size of the

sample limits generalizability. The sample is

composed exclusively of nurses, predominately

female, who graduated from one school of

nursing, with very similar demographic character-

istics. Second, the extended time frame for the

study spans the entire 20th century. This was a

time during which there were dramatic changes in

services, policies, and procedures for childbirth

and for hospitalized children and their families.

Despite high infant and child mortality rates,

perinatal and child death were devalued and

parents were not given support for the expression

of their grief. Data presented here are undoubtedly

influenced and shaped by this milieu. A study

with a shorter and more well-defined time frame

would have standardized the effect of the milieu.

Another limitation addresses a small number of

respondents who stated that they wished there had

been attention given to the somatic reactions of

grief, surviving sibling reactions, parental dreams

following the death of a child, and peer contact

with other bereaved parents.

It should be noted that this study did not collect

in-depth data from parents who experienced

miscarriage and stillbirth. Clearly, these profound

losses need to be studied; however, at this

juncture, the authors cannot generalize the find-

ings to these losses. Furthermore, the impact of

multiple child death was not explored. This also

requires further investigation.

SUMMARY

Overall, the findings of this study support an

understanding of parental grief as complex, non-

linear, and ongoing. Correlates of continued

Suggested Clinical and Research

Implications

Recognize the importance of including child

death as part of family history;

Provide support services and resources on an

ongoing basis over the life course of the family in

all health care settings;

Treat child death, regardless of age, cause, or time

since the death, not only as a woman’s health

concern in obstetric and gynecological care but

also in pediatrics, men’s health, and geriatrics;

Explore grief as an ongoing process with diverse

sample populations of bereaved parents;

Support studies involving loss of a child in

pregnancy, at birth, in infancy, in childhood, in

adolescence, and in adulthood to explore com-

monalities in the experience of grieving parents,

regardless of the cause of death, the time since the

ARNOLD ET AL.254

parental grieving were identified and include feel-

ing connected to one’s child, feeling they will not

let go of their child, feeling part of them is missing,

and experiencing a hollow empty space inside

them. These emotions form a composite of parental

grief. Findings suggest that grieving continues and

can be associated with a myriad of other emotions,

including those that are positive and life trans-

forming. Continued grieving is highly individual-

istic and is associated with heightened and variable

emotionality, life satisfaction, and improved rela-

tionships with surviving children. Furthermore,

parental grief on the death of a child is profound,

regardless of the years since the death, the age of

the child at the time of death, and the cause of the

child’s death.

CALLOUTS

death, or the age of the child
bThis research supports parental grief as a non-linear,
individualized, ongoing process.Q

bContinued grieving can be associated with a wide range of

emotions, many of which are positive and life trans-

forming.Q

bParental grief is profound regardless of the years that have

passed since death, the age of the child at the time of death,

and cause of the childTs death.Q

bThe death of a child is a significant loss with implications

for further clinical research and practice considerations.Q

ACKNOWLEDGMENTS

We are profoundly grateful to the following

organizations for their support in funding the

development of the Grief Over a Lifetime Instru-

ment: Lucie S. Kelly Research Award Grant, Alpha

Zeta Chapter of Columbia University, Sigma Theta

Tau International (2000); Frances G. Crane

Research Grant, Foundation of the New York State

Nurses Association (2000); Nursing Research

Grant, Columbia University–Presbyterian Hospital

School of Nursing Alumnae Association (1998);

Faculty Fund, The College of New Rochelle

(1996–1997; 1999–2000). We also thank the

Columbia University–Presbyterian Hospital School

of Nursing Alumnae Association for providing

access to the alumni who so graciously participated

in this study. Our gratitude goes to Kimary Kulig,

PhD, and Curt Malloy, MPH, for assisting with the

computer-based analysis.

REFERENCES

Arnold, J. (1995). A reconceptualization of the concept of grief

for nursing: A philosophical analysis. Unpublished doc-

toral dissertation. New York University, New York.

Arnold, J. (1996). Rethinking grief: Nursing implications for

health promotion. Home Health Care Nurse, 14(10),

777–783.

Arnold, J., & Gemma, P. B. (1983). A child dies. A portrait of

family grief. Rockville, MD7 Aspen.

Arnold, J., & Gemma, P. B. (1994). A child dies. A portrait of

family grief. (2nd ed.). Philadelphia, PA7 Charles Press.

Arnold, J., Gemma P. B., & Cushman, L. F. (2000). Grief over a

lifetime: A study of the lifelong effects of child death on

parents. Unpublished research instrument.

Benoliel, J. Q. (1983). Nursing research on death, dying, and

terminal illness: Development, present state, and pros-

pects. In H. Werley, & J. Fitzpatrick (Eds.), Annual

review of nursing research, I. New York7 Springer

Publishing Company 101–130.

Bowlby, J. (1961). Processes of mourning. International

Journal of Psychoanalysis, 42, 317–340.

Cody, W. K. (1991). Grieving a personal loss. Nursing Science

Quarterly, 4(2), 61–68.

Cowles, K. V., & Rodgers, B. L. (1991). The concept of grief: A

foundation for nursing research and practice. Research

in Nursing & Health, 14, 119–127.

Cowles, K. V., & Rodgers, B. L. (2000). The concept of grief: An

evolutionary perspective. In B. L. Rodgers, &K. A. Knafl

(Eds.), Concept development in nursing: Foundations,

techniques, and applications (pp. 103-118). Philadelphia7

Saunders.

Davies, R. (2004). New understandings of parental grief:

Literature review. Journal of Advanced Nursing, 46(5),

506–513.

Dillman, D. A. (1978). Mail and telephone surveys: The total

design method. New York7 Wiley.


Engel, G. (1961). Is grief a disease? Psychosomatic Medicine,

23, 18–22.

Engel, G. (1964). Grief and grieving. American Journal of

Nursing, 64(9), 93–98.

Freud, S. (1957). Mourning and melancholia. In J. Strachey

(Ed.). The standard edition of the complete psycho-

logical works of Sigmund Freud, Vol. XIV, 243–258.

London7 Hogarth Press.

Freud, S. (1963). General psychological theory. New York7

Collier Books.

Hutti, M. H. (1984). An examination of perinatal death

literature: Implications for nursing practice and

research. Health Care for Women International, 5,

387–400.

Klass, D., Silverman, P., & Nickman, S. (1996). Continuing

bonds: New understandings of grief. Washington, DC7

Taylor & Francis.

Kubler-Ross, E. (1969). On death and dying. New York7

Macmillan.

Lindemann, E. (1944). Symptomatology and management

of acute grief. American Journal of Psychiatry, 101,

141–148.

Lindgren, C. L., Burke, M. L., Hainsworth, M. A., & Eakes,

G. G. (1992). Chronic sorrow: A lifespan concept.

Scholarly Inquiry for Nursing Practice: An International

Journal, 6(1), 27–40.

Martinson, I. M. (1992). Response to bChronic sorrow: A life-

span conceptQ. Scholarly Inquiry for Nursing Practice:

An International Journal, 6(1), 41–42.

Martocchio, B. C. (1985). Grief and bereavement: Healing

through hurt. Nursing Clinics of North America, 20(2),

327–341.

McClowry, S. G., Davies, E. B., May, K. A., Kulenkamp, E. J.,

& Martinson, E. M. (1987). The empty space phenom-

enon: The process of grief in the bereaved family. Death

Studies, 11, 361–374.

Miles, M. S. (1985). Emotional symptoms and physical health

in bereaved parents. Nursing Research, 34(2), 76–81.

Miles, M. S., & Crandall, E. K. B. (1983). The search for

meaning and its potential for affecting growth in

bereaved parents. Health Values: Achieving High Level

Wellness, 7(1), 19–23.

Miles, M. S., & Demi, A. S. (1986). Guilt in bereaved parents.

In T. A. Rando (Ed.), Parental loss of a child. 97–118.

Champaign, IL7 Research Press.

Murphy, S. (1983). Theoretical perspectives on bereavement. In

(Ed.). Advances in nursing theory development. 191–206.

Maryland7 Aspen Systems Corporation.

Parad H. J. (Ed.), (1965). Crisis intervention: Selected readings.

New York7 Family Services Association of America.

Parkes, C. M. (1965). Bereavement and mental illness. British

Journal of Medical Psychology, 38, 1–26.

Parkes, C. M. (1975). Determinants of outcome following

bereavement. Omega, 6, 303–323.

Pilkington, F. B. (1993). The lived experience of grieving the

loss of an important other. Nursing Science Quarterly,

6(3), 130–139.

Sanders, C. M. (1980). A comparison of adult bereavement in

the death of a spouse, child, and parent. Omega, 10,

303–322.

Sanders, C. M. (1989). Grief: The mourning after: Dealing with

adult bereavement. New York7 Wiley-Interscience.

Uren, T. H., & Wastell, C. A. (2002). Attachment and meaning-

making in perinatal bereavement. Death Studies, 26,

279–308.

Worden, J. W. (1982). Grief counseling and grief therapy: A

handbook for the mental health practitioner. New York7

Spring Publishing Co.

exploring parental grief: combining quantitative and qualitative measures

Documents