Posttraumatic Stress Disorder in Mothers of Pediatric Cancer Survivors DAVID

PH.D.,

PELCOVITZ,

SANDRA

M.D.,

KAPLAN,

FRANCINE

BARBARA

VINCENT

Prevalence

of posttraumatic

survivors

was

stress

compared

Significantly more PTSD. Significant

with

mothers differences

M.D.

WEINBLATI’,

MARK

PH.D.,

MANDEL,

B.A.

GOLDENBERG,

BARBARA

M.A.

MEYERS,

VINCIGUERRA

disorder

(PTSD)

its prevalence

in 24 mothers

among

of pediatric cancer were also found

23 mothers

survivors in lifetime

of pediatric of healthy

were diagnosed with lifetime arousal, as well as current

and lifetime reexperience and avoidance symptom clusters. Significant isted in the distribution of the number of prediagnosis high-magnitude enced by the mothers diagnosed with current PTSD as compared with experience

of the mothers

who

level of perceived family Revised global severity and

T

PTSD-negative

his study

and index

were

not diagnosed

reports

the results

illness. The mothers vors were interviewed field

field trials which only persons

trials.

of an investiga-

study

who

are

the

exposed

illness

investigates

to an event

for

whether with for

would have implications treatment, and understanding

criterion a sizable

serious PTSD. for

A, in

“outside

illness

This

number

of

illness meet all Such a finding the diagnosis, of the effects

1994;

1996;

family

chiatry,

Division

Shore

University

of

College;

the

Cornell tistics,

Division

Hospital-Cornell

University Cornell

North

ill

adjustment

of

revised

February

the Department Psychiatry,

North Medical

Hematology/Oncology, Shore

University

Medical College; the Division of Research, North Shore Medical

College, Medical

18,

of Psy-

University

of Pediatric

quests to Dr. Pelcovitz, Department of Psychiatry, Psychiatry,

1993; From

chronically

survivors report some investiga-

and Adolescent

University

Medical University

the

Hospital-Cornell

of Pediatrics,

University Department

3,

4, 1994.

of Child

of

cancer Although

December May

37:116-126)

members studies

accepted

severity,

Symptom Checklist-90in the PTSD-positive

mothers of childhood inconsistent results.

Department

DSMillness as A.

on

Received

of the

criterion of PTSD

human experience.” rules out chronic

definition

mothers of children of the other criteria

116

question

was on how to define allows for a diagnosis

the range of usual III-R specifically meeting

central

PTSD.

and differ

children. Empirical

of pediatric cancer survias part of the DSM-IV One

current

difference exevents experithe prediagnosis

(Psychosomatics

tion of the prevalence of posttraumatic stress disorder (PTSD) in a sample of 24 mothers of survivors of childhood cancer, and in a community-recruited comparison group of 23 mothers of children who never had a chronic

PTSD

with

extrafamilial social support, scores did not significantly

groups.

cancer

children.

College;

Hospital-

of BiostaUniversity and

Cornell

the School of Public Health, College. Address reprint re-

North Shore University Hospital, Division of Child and Adolescent

300 Community Dr., Manhasset, NY 11030. © 1996 The Academy of Psychosomatic

Copyright

of

Medicine.

PSYCHOSOMATICS

Pelcovitz

tors report that parents cope treme stress that accompanies with

a serious

that

mothers

risk

for

illness,2

the exa child

researchers

find

survivors

are at

other

of pediatric

a variety

including discord.6

well with having

cancer

of psychosocial

depression,3’4

difficulties,

anxiety,5

and

marital

Perhaps one of the reasons for these contradictory findings is that systematic investigation of parental functioning of mothers of pediatric cancer measures

survivors rely almost that do not specifically

exclusively on focus on stress

related to the illness.7’8Researchers use self-report

measures

cal symptomatology, anxiety,9 medical clinical may

of general

such as staff ratings,2

interviews.’0 fall short

All

in their

reactions

child. logical

Failure to explore the reactions that typically

posure result

to an in false

from

insensitivity with

advantage

to having

extremely negative

posttraumatic of

specific

questions

of the illness and tions characteristically

for

examines seen

by extreme It has

been

stress. often

about

the

noted

that

atically evaluated 145 parents dren and found that, compared

significantly loss

more child.

higher

of confidence

levels

ability

parent. A better understanding tli)s chronic stress on mothers

VOLUME

ill children’s

37.

NUMBER

2

#{149} MARCH

-

APRIL

1996

with

A number ated

with

of risk

poor

cancer.9

There

adjustment for PTSD.

studies have demonstrated level of parental anxiety likely it is that the child

that the and distress, will develop factors

adjustment

is also

is associated A number of

have

higher the the more PTSD.’3”4 been

to persons

associexposed

of stressful

life

events, social isolation, and severity of stressor.’5 These factors have also been found to be to psychological with

outcome

a chronic

in parents

of

illness.

First, in a review of the literature on familiesof survivors of childhood cancer, Barbarin’6 highlights

the

impact

in parents

or reacaffected

al.’7 report of stressful impairment

of

risk for system-

and

to be a good

and

that parental child’s risk

need

to systematically

evaluate

dysfunction. Although we found no studies that systematically examined psychiatric symptoms

of the effects of also has implica-

adjustment

in children

evidence with the

the

tions for the ill child’s adjustment. The relationship between parental adjustment and the adjustment of the chronically ill child is well documented. Frank et al.’2 found an association between

justment

is that, method

of fatally ill chilwith the fathers,

of helplessness

high

anxiety,

tology. Barbarin’s review concludes that a cluster of stressful life events may be an important variable in predicting a family’s psychological

The

involved in the day-toThe mothers reported

in their

that

and

asso-

mothers

seriously ill children are at particular psychological maladjustment. Cook”

the mothers were day care of their

ill

assessing

symptoms in persons

reveal

the relationship between stressful and the development of psychiatric

reactions.

in this population approaches, this

studies depression

may stem

disturbances

systematically

prevalence of PTSD unlike more general allows

a seriously

stress

Such

maternal

and/or interference of strong emotions with the parents’ ability to appropriately cope and solve problems, are associated with problematic ad-

children

types of psychoresult from ex-

to specific

of

related

specific

traumatic event findings; this may

of distress.

levels

to severe stress,such as a history

approaches

to tap into

stress

ciated

psychologidepression or or structured

of these

ability

related

generally

level

et a!.

mothers’

of chronically

life events symptoma-

ill children,

Myers

et

an association between the number life events and degree of psychiatric in a community sample.

Social support can act as an important buffer against the stress of dealing with a seriously illchild.’5

In a study

that

included

survivors of childhood cancer, reported that perceived support

parents

of

Morrow et al.’8 from relatives

and friends was significantly correlated with more effective psychosocial adjustment to the child’s

illness. Severity

of

illness

threat of the illness predictors of PTSD.

and

the

extent

of the

to one’s life are important Kilpatrick et al.’9 demon-

strated that events that threaten one’s life, or the life of a close family member, are one of the major predictive factors for PTSD. There is also evidence that, when a traumatic event is particularly unexpected, the victim is more likely to

17

PTSD

in Mothers

develop

PTSD.

of Pediatric

The

“surprise”

Cancer

inherent

Survivors

in hav-

ing to deal with an illness more typically associated with adults may make serious chronic childhood

illness

posttraumatic

particularly

likely

symptomatology

to elicit

in both

child

and parents. In a study of adult cancer patients, Cella, Mahon, and Donavan2#{176} found that, when recurrence prise, the oping

of cancer came as a complete surpatients were more at risk for devel-

intrusive

and

avoidant

stress-response

symptoms. Our study, the prevalence

conducted of PTSD

in 1991, in mothers

cancer survivors, comparing prevalence in mothers of children. We

investigated of pediatric

treatment,

healthy

or about

if 1) the child’s diagnosis 18; 2) the child was off to enter

the

maintenance

40 patients who met these criteria, 16 did not participate. Reasons for nonparticipation were topic too painful to discuss (31%), moved out of state (25%), did not speak English too time-consuming (19%). A community control sample, neighborhood

(25%), living

and in the

as the illsubjects, was re-

cruited from a larger pool of adolescents who were interviewed as part of a study of physically abused subjects

also investigated certain potential

age

phase of treatment; and 3) the subject’s illchild was at least age 13 at the time of the study. Of

same

PTSD

it against physically

asked to participate was made before

adolescents. The subsample who matched the mothers

consisted of of cancer

risk factors for PTSD, including stressful events, severity of stressor (i.e., illness severity), and inadequate social support. Our hypothesis was that there would be significantly more PTSD in

survivors in age, race, gender, and socioeconomic status.2’ The comparison mothers were administered structured PTSD interviews as well as interviewed for an extra 2 hours (ap-

the mothers of pediatric in the comparison group.

proximately) vant to this

that who more

the

mothers

cancer survivors than We also hypothesized

of pediatric

cancer

survivors

with

more

severe

illness.

dents

Chronic were

history cidents

of pediatric cancer surviof adolescents with no

of chronic illness were administered

terviews,

as well

exposure presence

to traumatic of PTSD

or life-threatening structured PTSD

as interviews

about

acin-

history

of

events, to determine the symptoms. All interviews

were done at the subjects’ homes at their venience. However, because of scheduling ficulties, survivors

two of the mothers were interviewed

of pediatric by telephone.

condifcancer

Subjects Mothers of pediatric cancer survivors were recruited from a list of consecutively referred patients with childhood cancer at a teaching hospital’s Division of Hematology/Oncology. Subjects who met the following criteria were

118

were $250

not relefor their

illness and life-threatening acciruled out in the children of mothers

ministered mothers 23 mothers

that paid

in the comparison group by use of a structured interview, the Kiddie Schedule for Affective Disorders and Schizophrenia,22 which was ad-

METHODS Twenty-four vors and

data were

participation.

had PTSD would also have a history of stressful life events, lower social support,

and a child

to gather study and

as part

abuse study. This hensive section

of the interview asking

adolescent contains parents

physical a compreabout their

child’s medical history and history of threatening accidents. Two of the mothers were originally selected from the larger sample they

were had

dropped

a child

with

from

the

a history

study

lifewho sub-

because

of a chronic

illness. Demographic ers in both groups the demographic dren with cancer

characteristics are presented characteristics are presented

of the in Table

moth1, and

of their chilin Table 2. Our

assumption in designing the study even though the types and age at onset varied, the stress on the mothers was

was that, of cancer similar in

that all subjects were exposed to the trauma of having a child with a life-threatening illness that required painful treatments with significant side effects.

PSYCHOSOMATICS

Pelcovitz

1.

TABLE

Demograph

ic data

for the mothers

TABLE

2.

Type

Mothers

of Cancer

Acute

Sample 24

Interquartile

range

Socioeconomic

40-48.5

Interquartile

2

2

range

1-3

2-3

Squamous

American

4.2

1

4.2

carcinoma

Radiation

only

88% 8%

Chemotherapy

4%

4%

Radiation

only

and chemotherapy

Degree

Potential

Stressful

measure

is a structured

interview

systematically been exposed

assess to any

would

as an event

qualify

DSM-III-R

criterion

interview.23

whether traumatic

This

designed

to

the subject incidents

that

would

has that

meet

A guidelines

for

the

4.2

1

4.2

1

4.2

10

41.6

13

54.2

3

12.5

13

54.2

High

8

33.3

Rate of relapse

3

Intermediate

Events

4.2

I

of risk

Low

Measures

1

of treatment

83%

Other

4.2

1

cell carcinoma

12.5%

Caucasian

25.0

1

carcinoma

Nasopharyngeal Type

50.0

leukemia

embryonal

Percentage

6

lymphoma

Ovarian

Race

African

myeloblastic

Histiocytic

41-47

status

Median

with

12

tumor

Acute

42

leukemia

lymphoma

Wilm’s 44

Number

lymphoblastic

Hodgkin’s

25

Age Median

fo r the adol escents

Demographics cancer

Control

Cancer Mothers

Size

et a!.

12.5

Ages

Mean

Range

Currentage

16.13

14-23

Age at onset

10.50

2-18

Age of treatment

12.50

5-19

3.28

0-11

Years

since active

treatment

PTSD

(i.e., an event that is “outside the range of usual human experience”).’ Any event that met the conditions for criterion A was operationalized as a “high-magnitude stressor” (e.g., natural dis-

training manual that clearly delineated the standards needed to meet the criteria for a DSMIII-R diagnosis of PTSD. In addition, training

asters, sexual In addition, sessed other

jointly to audiotapes interviews. As noted

clusively

abuse, and extrafamilial assaults). this measure systematically asstressful events that occurred ex-

within

the past

criterion A guidelines, or economic hardship. tionalized

year

such These

as “low-magnitude

Structured

Clinical

that

did not meet

as marital discord events were opera-

for

ing

symptoms

prior

graduate

to 6 months. students

The

in doctoral

Diagnosis

interviewers programs

clinical psychology. Training ers consisted of having them

of the interviewread a SCID-PTSD

VOLUME

- APRIL

37.

NUMBER

2 #{149} MARCH

1996

was

held

the

for the interviewers

interviewed

reliability

as part

in

Module.25

of the DSM-IV

that recruited subjects from kappa coefficients measurof

Modified-Diagnostic PTSD

to listen

of previous SCID-PTSD earlier, the chronic illness

authors’ site as compared was 0.72 for current and

(SCiD).24 The PTSD module of the SCID was administered to all subjects. The interview assessed the presence of current and lifetime PTSD based on DSM-III-R criteria. Current PTSD is defined as presence of symptoms within the last 6 months, and lifetime refers to were

sample

were

PTSD field trials five sites. Interrater

stressors.”

interview

sessions

interviewers

interview The

PTSD

from

the

with the other 4 sites lifetime PTSD.23

design

field trials included assessing the connection PTSD symptoms and high-

Schedule of the

(DiS), DSM-IV

a mechanism for between positive or low-magnitude

stressors to which the symptoms were related. This was done by administering a modified version of the Diagnostic Interview Schedule, PTSD

Module,

which

contained

follow-up

119

PTSD

in Mothers

probes

to elicit

of Pediatric

event-specific

whether

positive

for

stressor. In the was used to deter-

PTSD

linked to the mothers’ child with cancer. The modified DIS was used

Survivors

information

each high- and low-magnitude present study, this measure mine

Cancer

symptoms

were

exposure to having a SCID rather than the to diagnose PTSD, be-

cause in the field trials the SCID-PTSD was viewed as the “gold standard”

module PTSD

for

The

distribution

P

>

0.05).

The

following measures were only adminto the mothers of childhood cancer surso as to gain a better understanding on

factors

for PTSD

in this

Data

were

21),

Checklist-90-R

measure

is a 90-item

assesses For this index,

This

instrument

the presence of psychological study, we examined the global which

emotional symptom ment

(SCL-90-R).26 self-report

is a global

distress. dimensions

ranges

from

measure

that distress. severity

of intensity

of

0.77

to 0.90.

ability ranges from 0.8 to 0.9. numerous studies that support

Test-retest

reli-

Derogatis cites the concurrent

pared

needed

to talk,

and

whom

they

de-

pended on when they needed help. The subjects were then asked how satisfied they were, on a scale of 1-6, with each type of support, and the number of friends and family members who offered such support.

Severity. the patients

A pediatric into three

(mild, moderate, nosis for survival

and severe) at the time

levels

of illness

of severity

lines established erature for each

120

oncologist dirisk categories

according of diagnosis. were

in the pediatric of the pediatric

based

to progThese on guide-

oncology malignancies.

lit-

of pediatric a PTSD

received

<

of the mothers

met

criteria

the control

with only 4% of the 13.9, df= 1, P = 0.001).

mothers,

none

for of the were

current and lifetime 17 PTSD symptoms. found

and lifetime symptom

between

of PTSD,

present

within

It is important the modified

DIS

7 were

present

in the mothers lifetime, and the past to note

6 months that,

the

for the reexclusters. The

cluster was only significant Moreover, of the 17 possible greater degree survivors for

of whom

PTSD (Fisher’s Exact Table 3 presents the num-

differences

symptoms

of

for life-

diagnosed in 25% of the cancer survivors as com-

manifested 0.02, NS).

Significant

for lifetime DSM-III-R to a signifiof pediatric 5 symptoms (current).

as indicated

interview, the diagnosis

by of

PTSD

in all of the mothers of pediatric cancer survivors resulted specifically from the stress associated with parenting a child with cancer. Even in the mothers who had PTSD symptoms related

Illness vided

was of pediatric

bers and percentages PTSD and for each

were

they

support

survivors

PTSD

with

when

social

cancer

currently Test, P

cantly cancer

of

who

(x2=

measured by two global interview questions. The mothers were asked to identify those persons whom they depended on to listen to them

Level

done

N for these analyses was

those

time PTSD, compared control mothers

arousal PTSD.

Support.

shown test,

were

on the mothers

of significance at 0.01. On the SCID-PTSD, 54%

was

Social

socioeconomic

of the SCID

(total

two groups for current perience and avoidance

and construct validity of the instrument.26

and

with those who did not. To control for number of chi-square tests, we set the

Current mothers

Internal consistency of the 9 measured by this instru-

done

comparing

pediatric Symptom

analyses

survivors

diagnosis the large level

population.

ages

groups of mothers was equal (Mann-Whitney

on the 24 mothers of pediatric cancer survivors and the 23 comparison subjects. Risk-factor cancer

istered vivors

of

status for the two to be approximately

analyses

diagnosis.

risk

RESULTS

to previous

diagnosis of PTSD their cancer-related magnitude stressors

high-magnitude

trauma,

the

was based exclusively on symptoms. The other highwere comparable for both

groups. High-magnitude events experienced by the mothers in both groups included (in descending order of frequency) sexual assault, physical assault, accidents, and natural disasters.

PSYCHOSOMATICS

Pelcovitz

TABLE

3.

Number and Posttraumatic

percentage of symptoms Stress Disorder (PTSD)

et a!.

on the Structured Clinical Interview for DiagnosisModule for mothers of cancer survivors vs. control

mothers

Number(%) CurrentLifetime

Cancer PTSD:

Summary

Control

Cancer

Control

6

(25)

0

(0)

13

(54)

1

(4)”

14

(58)

4

(17)”

19

(79)

6

(26)”

recollections

10

(42)

3

(13)

14

(58)

4

(17f’

dreams

4

(17)

3

(1 3)a

7

(29)

4

(17)

the experience

4

(17)

0

(#{216})a

5

(21)

0

(#{216})a

Distressing reminders

12

(50)

1

(4)”

15

(63)

2

(9)”

Criterion

10

(42)

0

(0)”

17

(71)

1

3

(13)

0

(#{216})a

9

(38)

3

4

(17)

1

(4)a

6

(25)

1

1

(4)a

1

(4)

1

(4)a

B: Reexperience

Criterion

Intrusive Distressing Reliving

C: Avoidance

Avoid

thoughts

Avoid

situations

of event

Diminished

amnesia interest

Feeling

are reminders

that

Psychogenic

1 in activities

of detachment

Restricted

range

from

Criterion Sleep

future

(0)”

9

(38)

(9)”

(71)

2

6

(25)

2

(9)’

14

(58)

0

(0)”

16

(67)

0

(0)”

8

(33)

1

(4)a

16

(67)

6

(26)’

8

(33)

1

(4)a

12

(50)

4

(17)

4

(17)

1

(4)a

10

(42)

4

(17)

response

“P
median

the time

0

7

(29)

1

(4)a

14

(58)

2

(9)”

9

(38)

0

(#{216})a**

10

(42)

2

(9)’

4

(17)

(4)a

6

(25)

3

(1 3)a

3

(13)

1

(4)a

5

(21)

3

(l3)a

For the mothers of cancer survivors (N = 24), for mothers Exact Test was used instead of chi-square.

‘P<0.0l;

The

(54)

17

reactivity

aFishers

13

(4)a

concentrating

Note:

(#{216})a**

(0)”

Hypervigilance startle

0 1

Irritability

Exaggerated

(38)

0

D: Arousal

Physiologic

9

(8)

disturbances

Difficulty

(4)’

2

others

of affect

Sense of foreshortened

(4)

(4)” (13)

in the comparison

group

(N

=

23).

“P<0.00l.

age

of the cancer

of the mothers’

survivors

interview

was

at

17 (in-

sent:

median

=

tion of the time

13, IQR since

The

3-15).

=

diagnosis

distribu-

for mothers

with

terquartile range [IQR] = 15-18.5). The median age of the control subjects was 14 (IQR = 1316, P = 0.003). The finding that the children

lifetime PTSD as compared with the mothers without lifetime PTSD was also not statistically significant (lifetime PTSD present: median

with

the

years

since

children did not affect the develop-

PTSD

absent: median

cancer

were

comparison

significantly

older

than

ment of PTSD in their mothers. The median of the children of mothers who had PTSD either (IQR

distribution larly,

of ages

without of child

cancer the

There

13in the

distribution magnitude

groups.

Simi-

diagnosed

of the age at onset

of their

two

was not statistically mothers with PTSD

(lifetime PTSD =

14, IQR

=

12-17;

37.

NUMBER

2

#{149} MARCH

significant and those

present: median lifetime - APRIL

PTSD 1996

2, IQR

=

1-5;

=6, IQR

=

2-10).

=

Risk-Factors

=

in these

the distributions

child’s between

VOLUME

age (in

the cancer or comparison group) was 17 = 15-19). The mothers without PTSD had

children with a median age of 15 (IQR 18). There was no significant difference

diagnosis

was

Analyses

a significant

difference

of the number of prediagnosis events experienced by the with

current

PTSD

the prediagnosis

experience

did not receive (Mann-Whitney

a diagnosis = 6.25, df

of pediatric

lifetime

=

age

mothers

cancer

ab-

not receive a current PTSD

in the highmothers

as compared

with

of the mothers

who

of current PTSD 1, P = 0.01). The survivors diagnosis

who had

did a 121

PTSD

in Mothers

median

of no high-magnitude

I), whereas vivors who a median years

events

of one high-magnitude their there

mothers

events with

lifetime PTSD. The mothers of pediatric who did not receive a lifetime of

no

whereas

cancer survivors who experienced a median event (IQR: time

in the years 1-1). In comparing and

current)

0-

event in the

lifetime

experience a diagnosis

cancer PTSD

their

factors

in the

for

had

the mothers

with

lifetime

PTSD

had fewer

low-magnitude events. The median SCL-90-R global T-Score for the mothers of childhood cancer survivors who

the distribution raw scores. On

of SCL-90-R

the

illness

severity

global

T-Scores

variable

that

(life-

of pediatric

or indi-

cated physician prognosis for survival at the time of diagnosis, 38% of the mothers of childhood cancer survivors with lifetime PTSD had children with the most “severe”

illness

PTSD

mothers

fact,

of in

had a median global T-Score of 55.5 (IQR: 49.5-60). There was no significant difference in

of pediatric

child’s

was no significant

of of

survivors diagnosis

had lifetime PTSD of one high-magnitude

There

0.5-3.5).

difference in the distribution of the number low-magnitude events for these two groups;

had lifetime PTSD was 54 (IQR: 44.5-61), whereas those mothers without lifetime PTSD

events

the mothers

of 1.5 (IQR:

as

by

PTSD

high-magnitude

preceding risk

of predi-

experienced

with the prediagnosis who did not receive

(IQR: 0-0.5),

(IQR:

of the number

diagnosed

a median

Survivors

child’s illness (IQR: 1-2). were no significant differ-

high-magnitude

compared the mothers

had

Cancer

of pediatric cancer surPTSD had experienced

in the distribution

agnosis the

the mothers had current

preceding In contrast,

ences

of Pediatric

46% 16%

had children had children

of risk.

Illness

severity

ers of childhood

levels of risk,

with “moderate” risk, and who were at the lowest level in the children

cancer

survivors

of mothwho

did

not

cancer survivors who received PTSD diagnoses as compared with those who did not, there were no other variables that were significantly corre-

receive a lifetime PTSD diagnosis was as follows: 25% “severe,” 62% “moderate,” and 13% “mild.” There was no significant difference in

lated with this diagnosis. Frequency distributions of the risk factors were evaluated to determine how much variability there was in their distribution.

the proportion of mothers of cancer survivors who had lifetime PTSD and those who did not

On the social support variables, mothers of childhood cancer survivors

Comparable variability tribution of social support,

16% of the with life-

time PTSD reported not being satisfied with the level of social support, 38% were “fairly satisfied,” and 46% were “very satisfied.” Similar levels of social support were reported mothers of childhood cancer survivors

by the who did

not receive a lifetime PTSD diagnosis (24% “not satisfied,” 50% “fairly satisfied,” 25% “very satisfied”). There was no significant difference in the proportion of the mothers of cancer survivors who had lifetime PTSD and those who did not on the amount of support received and

how

satisfied

they

vors

For the mothers of childhood cancer who had lifetime PTSD, the median

ber of low-magnitude whereas those mothers had a median number

l 22

were

with

that

support. survinum-

events was 1 (IQR: 1-2), without lifetime PTSD of low-magnitude events

on the severity

of their

child’s

illness

at time

of

diagnosis.

magnitude and illness

events, severity

was evident in disnumber of low-

SCL-90-R global T-Scores, scores in the current PTSD

groups as well. On all of these variables, there were no significant differences between the PTSD-positive and PTSD-negative groups for current

diagnosis.

VIGNETTES

In an attempt to convey the PTSD in mothers of pediatric the

following

case

vignettes

manifestations of cancer survivors, of mothers

with

PTSD as a result of their presented. All of the names

child’s cancer are and minor identify-

ing background information to ensure confidentiality.

have

been

changed

PSYCHOSOMATICS

Pelcovitz

Case

Vignette

1.

sales clerk,

only allow jor hospital

Reports

T., a 40-year-old, married, white after her first husband, father, had committed suicide. Her Mrs.

was remarried

the patient’s daughter, Nora, was 15 when she was diagnosed with Hodgkin’s lymphoma. Treatment consisted of radiation therapy and chemotherapy. Nora’s physician classified her as “high risk” at the time of diagnosis. Nora had just finished treatment, and the cancer was considered to be in remission when Mrs. T. was interviewed. In addition to the suicide by her first husband, Mrs. T. experienced a number of highmagnitude stressors before Nora’s illness: she was robbed at gunpoint, and she had a lengthy hospitalization following a serious car accident. Low-magnitude stressors that she experienced in the year that her daughter was treated included the loss of her second husband’s job and serious financial and marital difficulties. Mrs. T. reported having particularlyintrusive recollections about her first conversation with the doctor at the time of Nora’s diagnosis. She had nightmares where she experiences attending Nora’s funeral. Powerful reminders of her daughter’s treatment are evoked every time Mrs. T. sees children in the park wearing hats, because this reminds her of the period when it was necessary for her daughter to wear a hat because of treatment-related hair loss. Similarly,news reports of a celebrityhaving a bout with cancer also elicitedintrusivememories of Nora’s treatment. Avoidance symptoms include her feeling cut off from others and a belief that she has no future. Mrs. T. was also much less interested in activities that used to be important to her. Signs of physiologic arousal were manifested by sleep disturbance and hypervigilance. In the weeks preceding visits to the hospital for routine follow-up checkups, Mrs. T. would fantasize that she will be told that Nora has a recurrence of cancer. Sleep issignificantlyimpaired during this period untilshe is reassured by the oncologist that Nora ishealthy.Mrs. 1. ishypervigilant to any physical symptoms in Nora. For example, low fevers or diarrhea elicit such anxiety that even extensive reassurance from the oncologist that these symptoms are not related to cancer fail to calm her. These PTSD symptoms have led to frequent conflict

between

Mrs.

T. and

Nora,

who

feels

stifled

and overprotected by her mother’s anxieties. For cxample, Mrs. T. would refuse to allow Nora to apply to the college of her choice, insistingthatshe will

VOLUME

37

#{149} NUMBER

2

#{149} MARCH

-

APRIL

996

her to attend nearby.

a university

et

that has a ma-

Vignette 2. Mrs. R., a married, white, 41-year-old teacher, met the criteria for current and lifetime PTSD. Her son, Tim, age 17, was diagnosed as having acute myeloblastic leukemia 1 year prior to the interview.The physician’s severityratingat diagnosis placed her son in the highest risk category. Treatment included an extensive course of chemotherapy. Although Mrs. R. had experienced one high-magnitude

stressor

an attempted sexual symptoms

before

assault),

exclusively

and treatment. been plagued and prognosis.

Since

Tim’s

illness

she reports

in relation Tim’s

to Tim’s

diagnosis,

by intrusive

thoughts

(i.e.,

PTSD diagnosis

Mrs. about

R. has his illness

Although Tim has been off treatment for 4 months, she “continually fears for the future and his health.”

Mrs.

R. endorsed

symptom. For example, thing can trigger feelings

virtually

every

she complained of being back

PTSD

that “anythere” (i.e.,

in the hospital). Recurring dreams are dominated by themes related to Tim’s hospitalization and are often accompanied by feelings of terror and helplessness at not “being able to save him.” Avoidance symptoms include Mrs. R.’s diminished interest in areas that used to excite her and a feeling that she is cut off from others. Mrs. R. reported avoiding so-

cial events fore Tim’s included sponse,

and having fewer friends than she did beillness. Evidence of physiologic arousal hypervigilance,

and

exaggerated

startle

re-

irritability.

Mrs. R. explained that images about Tim’s illso dominate her life that she has found it necessary to postpone plans to become a psychotherapist, and she has dropped out of graduate school. ness

Vignette 3. Mrs. C., a 43-year-old, married,

white

housewife, met criteria for current and lifetime PTSD. Her daughter, Tiffany, was diagnosed with Hodgkin’s lymphoma at age 18, 1 year prior to the interview. Tiffany was off treatment for 4 months at

the time of the interview. Her physician classified her as “low risk” at the time of diagnosis. Treatment consisted 2

extremely

Tiffany’s

of radiation

stressful diagnosis.

therapy.

events Two

Mrs.

C. experienced

in the 2 years prior to

years

earlier

a close

friend’s daughter was killed in a hit-and-run accident, and 1 year later her sister was killed by a drunk driver. Although she described both traumas as “devastating,”Mrs. C. says thather emotional reaction to those events paled in comparison to the

123

in Mothers

rauma

of Tiffany’s

of Pediatric

diagnosis

Cancer

and treatment.

Survivors

Mrs.

interest

C. complained of several intrusive symptoms. She stated: “The word cancer keeps popping into my head, and whenever I hear something about cancer I start to cry.” Avoidance symptoms include a fear

nesia.

were also irritability,

In addition,

with

reported, jumpiness,

PTSD,

pervigilance,

magnitude finding

that

er prevalence atric

there

rates

cancer

were

significantly

high-

of PTSD

in mothers

of pedi-

survivors

than

group is not surprising research has revealed stressors

that

are

in the

chiatric

have

comparison

PTSD

with

symp-

There

are few life events the

coping

a life-threatening

with

range

as horrifying

of routine

and

experience

illness

as

in one’s

child. The process of parenting a child with cancer includes a number of factors that have been associated with high risk for PTSD, including life threat, physical injury, hospitalization,

and

factors studies

by

PTSD for

surprise.

In

associated

with

Kilpatrick

prevalence

crime

ened

who

review

PTSD,

Davidson27

and

rates

victims

and

a recent Resnick

whose

lives no

were

not

injuries.

15%

and

that a current with a lifetime

PTSD history

threat-

Smith28

found

that

has

risk.

also

(nonpsychiatric)

hospitaliza-

of subjective factors

associated

are

ers

of

with

PTSD

pediatric

unpreparedness,

risk

cancer

a risk

factor

relevant

fear,

for

to moth-

survivors,

extreme

including

and

lack

findings

suggest

that

there

may

be

a

specific symptom pattern that predominates in relatives of children with a life-threatening illness. For example, the majority

124

within the avoidance

of subjects

experienced

cluster,

diminished

who

high-magnitude

cancer and

subsequent

events.

regarding

prior

adjustment.

ing the diagnosis of childhood find no studies that investigated

life

Jacobs

and

cancer. We could lifetime history

of high-magnitude trauma in parents of chronically ill children. Our results suggest that this may be an important area to consider in future of the relatively to exercise that

perceived

support, and

family

current number

that

small

caution

sample

size,

in interpreting and

the

extrafamilial

psychiatric

symptoma-

of low-magnitude

stressors

with PTSD. The variability of each of the risk factors

a larger

sample

size

might

shed

severity was in the highest risk category at the time of diagnosis. In addition, there might have been

of

control.27 Our

psy-

further light on the role of these variables. It is of note that half of the mothers who met criteria for current PTSD had children whose illness

and

tion following a trauma PTSD. Finally, a number

is

for

families

Charles29 reported a clustering of stressful family life events in the period immediately preced-

social

been

needs

among

patients

tology,

Davidson

of high-

childhood

suggests

treatment

PTSD

diagnosis

of data in the research literafunctioning in families of

be threatened.

Medical

hydifficulty

There is a paucity ture on psychosocial

were not correlated in the distribution

with

increased

prior

rates rise to between 59% and 66% when the victim is injured and perceives his or her life to associated

experienced

suggests that clinicians need

experienced

finding

Prevalence

relatively

subjects

services

one needs

report

9% and

were

the physiologic often associated

disturbances,

to possible

research. In light

risk cites

that

of between

suffered

of

there

sleep

support

events

tomatology. as far outside

while

trauma

to be alert

in light of what previous about the nature of

associated

of detachment,

future. However, of psychogenic am-

many

concentrating. The finding was associated

DISCUSSION The

feelings

few subjects who experienced reactivity and startle response

the future,feeling “cut off” from others,and a general sense that lifehas lostits meaning. Symptoms of physiologic arousal marked by sleep difficulties, and trouble concentrating.

in activities,

and a sense of a foreshortened there was scarce evidence

PTSD

a bias in

toward this

study,

the as

underrepresentation

of

it is conceivable

those who refused to participate risk for this disorder. As noted

were earlier,

that more 31%

at of

those who declined to participate in the study did so because of what might be a component of undiagnosed PTSD (i.e., refusal to talk about the trauma).

PSYCHOSOMATICS

et a!.

Pelcovitz

Our

findings

of pediatric that is similar treme stress search.

suggest

cancer

that

Viewing

survivors

to other groups has implications

A number

of

mothers

studies

sault

such

as natural

are at particular

that stem pediatric

of

the case

PTSD

disasters risk

have

or sexual

for a first

as-

depressive

in mothers

of pediatric

cancer

survivors

toms vices

survivors

would

are a number

vignettes,

can directly via greatly

qual-

implications

PTSD in mothers of As is evident from

PTSD

impair

clearly

of clinical

from diagnosing cancer survivors.

significantly

PTSD extreme

episode, generalized anxiety disorder, and alcohol or substance abuse.30’3’ These studies suggest that using structured interviews to study the prevalence of these disorders

cancer

ify. There

exposed to exfor further re-

shown that, in addition to developing symptoms, persons exposed to other stressors

pediatric

as a population

in this

population

functioning.

can

Arousal

symp-

affect the use of medical increased phone calls

serand

office visits for reassurance and for “false alarms” where noncancer-related physical symptoms in the child are misinterpreted as a relapse.

Avoidance

symptoms

child’s well-being ply with follow-up

may

affect

the

via possible failure to comcare and recommendations.

should enhance our understanding of the needs of families of childhood cancer patients. Further research needs to be done with fa-

Future research aimed at systematic study these issues might document such effects.

thers would

that have been found to be effective in treating PTSD. We have recently begun support groups for mothers who have a PTSD diagnosis as a

of children with cancer. expect that mothers are

merely

because

for PTSD,3032 men to deal

being

female

it is possible with chronic

through avoidance” may cant risk for this disorder be important to investigate lence

of PTSD

from

for developing

is a risk

a longitudinal

in PTSD clearly A modifications criteria.

PTSD. mothers of meet criteria does not meet finding ill child

The proposed

DSM-IV

criterion

#{149} NUMBER

2

death or integrity response

helplessness, or horror.”33 criterion, the mothers of

#{149} MARCH

-

APRIL

1996

Early

recognition

result of their our treatment to support

can

children’s includes

mothers

lead

to treatments

cancer.34 assigning of newly

An aspect of these women diagnosed

chil-

dren. based

Initial findings suggest that this exposuretreatment has proven to be very helpful to

both

the

nosed

“veteran”

mothers

families. These findings

highlight

considering

the possibility

in psychosocial plications

and

newly

the

oncology

of

in mothers

Such an approach for future research and

in identifying

diag-

advantage

of PTSD

of pediatric cancer survivors. has important implications

has

clinical

a population

im-

in need

of

psychiatric services. Early identification of PTSD symptoms in this population can enhance the quality of cancer survivors.

A

person has been exposed to a which both of the following 1) the person has experienced, confronted with an event or

involved intense fear, Under this new

37

that can

lends support to the criproposed in the DSM-IV

events that involve actual or threatened serious injury, or a threat to the physical of oneself or others; and 2) the person’s

VOLUME

prevaSuch

in identifying which mothers survivors are at greatest risk

chronic

is as follows: “The traumatic event in have been present: witnessed, or been

perspective.

that PTSD of treatment.

the definition in criterion A. Our mothering a seriously chronically

draft

factor

place them at signifias well. It would also the relative preva-

Under DSM-III-R guidelines, pediatric cancer survivors cannot for PTSD because chronic illness

result terion

one risk

that the tendency of illness in their child

Clearly, one would expect lence would vary with stage studies might help of pediatric cancer

Although at greater

of

This through (NiMH),

life

for

research

was

the National Grant

families

childhood

supported

institute

of Mental

in

part

Health

1R01MH43772-04,

and

the “Psychopathology, Adolescent Abuse” DSM-iV PTSD field

Suicidal Behavior and the NiMH-funded trials. The opinions

and

pressed

are

in this

No.

of

article

and do not necessarily the American Psychiatric force

those

ex-

of the authors

represent the position of Association or its task

on DSM-IV.

125

PTSD

in Mothers

of Pediatric

Cancer

Survivors

References I. American

Psychiatric

tistical

Manual

Revised.

Association:

of

Mental

Washington,

DC,

ciation, 1987 2. Kupst MJ, Schulman aspects

of pediatric

the first

6 months

Diagnostic

Disorders,

3rd

American

and Sta-

and

parental

adjustment

Edition,

Clin

Psychol

1981;

Psychiatric

19. Kilpatrick

Asso-

JL, Maurer

H, et al: Psychosocial

leukemia:

from

of treatment.

diagnosis

Med

Pediatr

ment

through

Oncol

1983;

KN,

J Pediatr

and

parents’

Psychol G,

distress

Acta

5. Koocher

search

about

Psychol

1991;

of children

Journal 10. Brown

Childhood

staging

of pediatric

Cautions

D:

in

observations

with

a chronic

U,

Cobiella

successfully

can-

illness.

the

on

J Pediatr

for

cancer.

C: Predictors

adjustment

of children

in mothers

with

functioning

in children

J Am

Acad

31:495-502 11. Cook JA: Influence of fatally 1984;

with

Child

leukemia

Adolesc

and their 1992;

of gender Journal

parenting

E, Connell

Three

Mile

Island

1984;

14. Green

B, Korol

matic

tive

reactor.

Journal

effects

Adoles

Factors

Stress

1989;

OA:

psychiatric

2:489-5

of Psychosocial JK, Lindenthal impairment.

1991;

University,

1977 J: Issues

Davidson

disorder,

Journal

for

Oncology JJ, Pepper J Nerv

Ment

bases

survivors

a

of prevenof

cancer.

1987; 5:25-41 MP: Life events Dis 1971;

and

Psychiatry

1986;

152:149-

Hoagland

A, Carnrike

CL:

Social

support

the

Med

in psychiatric

1980;

42:11-24

Psychiatric

Helen’s

Swartz

reactions

experience.

MS:

among

Sexual

women

Med

community:

1990;

Cancer

J

assault

and

in a community

epidemiological

L,

Kochen

Symposium New

Psychiatric J, et al:

program

of cancer

Center,

DSM-IV

American

outreach

in parents

an

Association:

DC,

Gluck

Am

stress study.

21:713-721

Psychiatric

parent-to-parent

tering

DC,

Am J Psychiatry 1990; 147:335-341 J, Hughes DL, Blazer DO: Posttraumatic

Psycho-oncology GR,

12, edited

Washington

pp. 14 1-155

WM:

Saint

LK,

disorders

ria. Washington,

tology

posttraumatic

143:590-595

I, George

N,

A.

and

Hopkins

Vol

experience

Psychosom E, Vollmer

the Mount

in

(in

of Traumatic Stress l990, 3:459-476 E: Life events and the occurrence of

to disaster:

1993 34. Fried

Johns of

1993,

HA,

Press

Scoring,

MD,

diagnosis

Press,

in children.

33. American

Sourcebook,

Psychiatric

Tasman

post-

AJ, Pincus

of Psychiatry,

M,

State

on criterion

in DSM-IV Frances

R: Traumatic

JH, Tatum

Psychol and invulnerability:

and empirical services

30. Shore

disorder

14 risks

in the

York

JR. et al: The

Administration,

in Review J, Riba

New

emphasis

I. Baltimore,

J, Smith

32. Davidson

of Trau-

Washington,

M: Structured Clinical

American

Psychiatric

population.

between

review

trial:

TA,

SCL-90:

Manual.

psychiatric

30:945-951

157

18. Morrow

LR:

31. Winfield

symptoms.

that mediate

a review.

Psychosocial

family-focused

17. Myers

Psychi-

and disaster:

on PTSD

Psychiatry

in the victim

of the theoretical

Journal

near the

of Preventive

M, et al: Children

and psychopathology:

review

J Pediatr

field

DC,

Procedures

cancer

M, Grace

Child

MS:

16. Barbarin

stress.

health of children

M: Mental

and parental

J Am Acad disaster

parenting

Posttraumatic Meeting

York,

diagnosis,

outpatients. Journal 29. Jacobs TJ, Charles

2:275-301

gender,

15. Gibbs

and

JR:

at the Annual

New

by Widiger

American

16:361-371

Psy-

a comprehensive

JB, Gibbon

PTSD

Oncology illness,

depression Adolesc

Association,

DSM-III-R.

3, edited

of Psychosocial

et al: Child

Freedy

presented

Vol

by Oldham

AE,

Paper

stress disorder

of parents

Wagner

alliance,

1991;

13. Bromet

age,

CL,

W, et al: Retro-

Child

Institute, 1987 DG, Resnick HS, Freedy

28. Davidson

SJ, Olmsted

HS,

Psychiatric

for

on the problems

2:71-91

Acad

report:

A and overall

stress

ill children.

Psychol

126

27.

and Mental

major

J Am

as

1992

press) 26. Derogatis

and

Psychiatry

trial

et al. Washington,

cancer.

Class

1958 P. Chambers

Resnick field

results.

traumatic

of psycho-

recurrence

21:392-397

DG,

Psychiatric 25. Kilpatrick

J Pediatr

Be-

16:15-22

prepubertal

SADS.

disorder

Interview

adjustment

develop-

disorder.

Cancer

1990;

FC: Social

of

24. Spitzer RL, Williams

re-

M:

Med

Redlich

of the American DC,

using

based

C, et al: Social treated

1982;

stress

1986; 58:449-553

Cancer

the Kiddie

stress

Donavan

Behav

AB,

the

20:199-214

assessment

of initial

1990; 15:359-371 C, Forehand R, Saylor

parents.

12. Frank

SM,

event.

23. Kilpatrick

1981

of Psychosocial Oncology 1986; 3:43-54 R, Kaslow N, Hazzard A, et al: Psychiatric

family

atry

Syndrome:

Surviving

16:411-421

A, Stark

logical

D, Mahon

chiatry

Damocles

Drotar

children

with

leu-

of

families.

Checklist:

lymphatic

68:297-300

McGraw-Hill,

Stein RE,

EC,

Behavior

Psychol 9. Baskin

1989;

spective

J Consult A, et al:

with

posttraumatic

Therapy

21. Hollingshead

associated

Illness. New York, Wiley, 22. Orvaschel H, Puig-Antich

The

M: Psychological

Child

so-

Psychological

acute

1983;

JF:

and their

cer patients

the

Scand

York,

6. Adams-Greenly

8. Spirito

with

Consequences

New

7. Perrin

D, et al:

of children

JP, O’Malley

Cancer.

cancer, adjustment.

cancer.

BE, Amick-McMullan

factors

of crime-related

20. CelIa

17:15-31

C, DeLeo

Psychiatr

Psychosocial

childhood

psychological

1992;

Messina

in parents

kemia.

S: Surviving

Noh

support,

Saunders

crime

a traumatic

3. Speechley

4. Magni

and

havior

11:269-278 cial

DG,

Victim

to pediatric

49:763-765

survivors.

The

on PTSD:

Crite-

effects

September

of a

symptoma-

Paper presented

V. Memorial York,

Draft

Association,

Sloan 1993

PSYCHOSOMATICS

at Ket-