PSYCHO-ONCOLOGY

Psycho-Oncology 9: 113–126 (2000)

SOCIAL SKILLS AND PSYCHOLOGICAL

ADJUSTMENT OF CHILD AND ADOLESCENT
CANCER SURVIVORS
WENDY LEVIN NEWBYa, RONALD T. BROWNb,*, TERESA M. PAWLETKOc, STUART H. GOLDd and J.
KENNETH WHITTc
a
North Carolina State Uni6ersity at Raleigh, USA
b
Department of Pediatrics, Medical Uni6ersity of South Carolina, Charleston, SC, USA
c
Departments of Psychiatry and Pediatrics, Uni6ersity of North Carolina at Chapel Hill, NC, USA
d
Department of Pediatrics, Uni6ersity of North Carolina at Chapel Hill, NC, USA

SUMMARY
Social skills and psychological adjustment for survivors of childhood cancer were investigated. Cancer survivors
included 42 children and adolescents ranging in age at evaluation from 6 to 18 years with a mean age of 13.1 years.
Measures included teacher and parent ratings of social skills and adjustment and parent ratings of family
functioning. The findings showed that social skills and psychological adjustment as rated by both parents and
teachers were primarily associated with academic functioning. In addition, family cohesiveness was found to
account for nearly one third of the variance in survivors’ adjustment when rated by teachers, and length of time
off treatment accounted for a significant percentage of the variance in children’s adjustment when rated by
parents. The findings underscore the importance of a multi-informant approach to the assessment of psychological
adjustment of pediatric cancer survivors and demonstrate the role of learning difficulties and family functioning
in influencing social skills and adjustment for these children and adolescents. Copyright © 2000 John Wiley &
Sons, Ltd.

INTRODUCTION 1993). In this series of studies, Noll and his col-

leagues investigated peer relationships, interac-
tions, and perceptions during the critical period of
Research interest in children and adolescents who reintegration into school during the late-treatment
have survived cancer has increased over the past phase and immediately following termination of
several years. Cognitive impairments of cancer therapy. Noll et al. (1990) studied teacher ratings
survivors have been well investigated (Whitt et al., of children who were either in treatment or had
1984; Cousens et al., 1988, 1991; Fletcher and ceased therapy within the past year. Teachers
Copeland, 1988; Brown et al., 1992; Madan- rated the cancer survivors low in leadership and
Swain and Brown, 1992; Haupt et al., 1994; social skills and noted they were disengaged from
Brown et al., 1996), and attention recently has peers. Similarly, for children who had been off
focused on the long-term social outcomes of these treatment for at least 18 months, or for those who
children and adolescents. were in remission, Noll et al. (1991) found that
Although some recent investigations have fo- survivors were perceived by peers as more socially
cused on psychosocial adjustment and competen- isolated or withdrawn. However, these peer per-
cies of children and adolescents diagnosed with ceptions did not affect the popularity of survivors,
cancer, few investigators have examined social their number of mutual friends, or their ratings of
interactions between children with cancer and social competencies. In another study (Noll et al.,
their healthy peers (Noll et al., 1990, 1991, 1992, 1993), adolescent cancer survivors again were
rated by peers as more socially isolated, although
friendships and popularity were not affected. The
* Correspondence to: Department of Pediatrics, Medical Uni-
versity of South Carolina, Children’s Hospital, 171 Ashley
findings from the series of studies by Noll and
Avenue, PO Box 250561, Charleston, SC, 29425, USA; e-mail: associates (Noll et al., 1991, 1992, 1993) are
[email protected] consistent with the psychosocial literature

Copyright © 2000 John Wiley & Sons, Ltd. Recei6ed 12 June 1998
Accepted 14 October 1999
114 W. LEVIN NEWBY ET AL.

(Madan-Swain and Brown, 1992; Kupst, 1994), functional status of the child, maternal education,
suggesting generally adequate social adjustment and family functioning (Wallander et al., 1989;
for children with cancer, particularly during the Breitmeyer et al., 1992). For example, support
late-treatment or the early off-therapy period. within the family was the only correlate of psy-
On chronic illness in general, Spirito et al. chological adjustment for children with juvenile
(1991) noted that an illness frequently disrupts rheumatoid arthritis (Varni et al., 1988), and ade-
peer interactions because of observed physical quacy of family support was an important factor
differences and limitations. Studies specifically ex- for children who survived leukemia (Kupst and
amining the social competencies of children with Schulman, 1988). Rait et al. (1992) found that
chronic illnesses, however, have shown generally perceived family cohesion and adaptability were
adequate adjustment. Spirito et al. (1990) exam- strongly associated with post-treatment psycho-
ined the social adjustment of children aged 5 –12 logical adjustment. Similarly, some studies have
years who had completed treatment for cancer, found lower parent ratings of social competencies
with the range of time off treatment from 6 in survivors of childhood cancer to be associated
months to 5 years. In comparison with their with cognitive and physical impairments, the use
healthy peers, teachers rated the cancer survivors of cranial radiation therapy during treatment, and
as better socially adjusted. Specifically, the chil- being reared in a single-parent household (Mul-
dren were rated as being teased less and arguing hern et al., 1989; Greenberg and Meadows, 1991).
less frequently with classmates than were their Kupst et al. (1995) reported that adjustment in
peers. The survivors themselves, however, re- long-term cancer survivors was associated with
ported fewer friends of the same age and greater socio-economic status and academic difficulties.
loneliness and isolation from peers than did In an investigation from the Children’s Cancer
healthy controls. Madan-Swain et al. (1994), in Group, Noll et al. (1997) examined parent and
their study of adolescent cancer survivors, noted teacher reports of behavior and social competence
an enduring tendency for survivors to be overly for children who were survivors of leukemia. They
compliant in their social and interpersonal found that scores on the standardized measures
relationships. were similar to instrument norms, although par-
In a comprehensive review of the literature, ents reported heightened somatic concerns. In ad-
Van Dongen-Melman and Saunders-Woudstra dition, there was no effect of either radiation
(1986) also observed that adjustment difficulties in therapy or chemotherapy on behavioral adjust-
children and adolescents with cancer are typically ment during the late-effects period. The results
mild and transient. Consistent with these conclu- were interpreted to suggest minimal psychosocial
sions, Greenberg et al. (1989) investigated elemen- morbidity among long-term survivors of children
tary school children and adolescents following with leukemia given the stressors associated with
diagnosis and found generally adequate adjust- the cancer experience. The few studies published
ment and functioning, albeit with some variability on the psychosocial effects of cancer have primar-
on objective test measures. Other investigators ily focused on children’s adjustment during the
(Teta et al., 1986; Chang et al., 1988; Fritz and treatment period and shortly following termina-
Williams, 1988; Madan-Swain and Brown, 1992; tion of therapy. Less common are studies specifi-
Brown and Madan-Swain, 1993; Kupst, 1994; cally examining social functioning of cancer
Madan-Swain et al., 1994) reported findings simi- survivors longer into the period following termi-
lar to those of Greenberg and colleagues, who nation of therapy. In an investigation of long-
concluded that the cancer experience resulted in term survivors of childhood cancer, and
few long-term problems in adjustment. In an im- consistent with general chronic illness literature,
pressive longitudinal investigation of family cop- however, Greenberg and Meadows (1991) re-
ing with pediatric leukemia, Kupst and associates ported a positive association between academic
(Kupst et al., 1982, 1984, 1995; Kupst and Schul- functioning, physical impairments, and levels of
man, 1988) prospectively followed a group of distress in a group of cancer survivors who were 8
survivors for over one decade. These researchers years post-treatment. Similarly, Mulhern et al.
also found that long-term survivors and their (1989) found that lower parent ratings of social
parents continued to be well adjusted. competencies in survivors of childhood cancer
For predictors of adjustment, social compe- were associated with cognitive and physical im-
tence most often has been associated with the pairments, the use of cranial radiation therapy

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
 SOCIAL SKILLS AND PSYCHOLOGICAL ADJUSTMENT 115

during treatment, and being reared in a single- posite ratings, including frequency of treatment,
parent household. prognosis, extent of life-threatening complica-
A social–ecological and systems’ orientation is tions, and academic functioning) would account
also useful in the understanding of psychosocial for a significant portion of the variance in the
adjustment in children with a chronic illness social skills and psychological adjustment of sur-
(Kazak, 1994). In this model, the family repre- vivors during the late-effects period. Current liter-
sents the primary system that influences adjust- ature (Brown et al., 2000) suggests an association
ment and ‘shifts the focus away from only disease in children with cancer between social competen-
and treatment variables’ (p. 186). When the late cies and cognitive impairments (e.g. neurotoxic
effects of the cancer experience are considered, the late effects of cranioradiation therapy). In this
interactions of the child and family assume a article, academic functioning refers to the seque-
more prominent role than the disease process. In lae of learning problems that result from the
fact, some research suggests greater control within central nervous system prophylactic therapies (i.e.
families in response to a chronic illness (Kazak et cranial radiation and chemotherapy), many of
al., 1995). Excessive control within families has which have been found to result in neurocognitive
been associated with poorer adjustment in chil- toxicities (Mulhern, 1994). In fact, in one investi-
dren, particularly as they negotiate the disease gation, MacLean et al. (1995) provided com-
process (Wallander et al., 1989). However, the pelling data to suggest that for young children
influence of the family and children’s adjustment diagnosed with leukemia, the combined effects of
during the late-effects period has recently received cranial irradiation and intrathecal chemotherapy
considerable interest. For example, Kupst et al. resulted in neurocognitive toxicities less than 1
(1995) reported that coping and perceived adjust- year after therapy.
ment in long-term survivors were positively asso- Given the extant literature relating family func-
ciated with mothers’ coping, and mothers’ tioning to children’s disease adjustment, it was
adjustment was associated with survivors’ adjust- also posited that family environments character-
ment. Rait et al. (1992) reported that familial ized by cohesiveness, expressiveness, and organi-
adaptability and cohesion were associated with zation would account for a significant portion of
adjustment during the late-effects period. the variance in the psychological adjustment and
This investigation was designed to extend the social skills of cancer survivors. This investigation
literature on cancer survivors by examining the also compared teacher and parent ratings of sur-
social skills and the psychological adjustment of vivors’ social skills functioning and adjustment
cancer survivors two or more years following relative to normative data for healthy youth.
termination of treatment. It was hoped that this Given the aforementioned literature suggesting
would capture some of the complexity in the generally adequate adjustment, no differences
relationships of the variables. The few studies were predicted between survivors and the healthy
published on the psychosocial effects of cancer normative group in the areas of adjustment and
have primarily focused on children’s adjustment social skills. This study is unique because it exam-
during the treatment period and shortly following ined peer socialization skills and adjustment
termination of therapy. Fewer studies have spe- across situations (both at home and at school)
cifically examined social functioning of cancer and by multiple informants (teachers and par-
survivors longer into the period following termi- ents), whereas most studies in this area have
nation of therapy. This period following treat- examined adjustment and social skills rated pri-
ment is referred to as the late-effects period. For marily by parents and the survivors themselves
this investigation, the onset of the late-effects (Kupst et al., 1995).
period is at least 2 years off treatment.
Because previous studies primarily have exam-
ined psychosocial adjustment (Noll et al., 1990, METHOD
1991, 1992, 1993, 1997), additional information
on the social skills of cancer survivors occurring
during this late-effects period is needed. It is Participants
hypothesized that disease- and treatment-related
variables (measured by age at diagnosis, time off The subjects were 42 children and adolescent
therapy, physical late effects, and physician com- (21 females and 21 males) cancer survivors1 par-

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
116 W. LEVIN NEWBY ET AL.

ticipating in a late-effects follow-up clinic at a (SES) based on occupation codes from Holling-
university medical center. A comparison of those shead (1975) was 44.1 (SD= 11.5). Subjects at-
who did and did not participate in the study was tended high school (35.7%), middle school
made available, including current chronological (31.0%), elementary school (28.5%), and college
age, age at diagnosis, duration of treatment, time (4.8%).
since treatment, and type of cancer. No difference
was found between the two groups for any of the
variables. Subjects ranged in age at diagnosis Procedure
from birth to 12 years 4 months, with a mean age
Informed consent was obtained from the par-
at diagnosis of 4.8 years (SD = 3.5 years). The
ents and assent from the children and adolescents.
mean age at evaluation was 13.1 years (SD= 2.8
Cancer survivors, their parents, and their teachers
years), with a range from 6 to 18 years. The time
were requested to complete measures mailed to
since completion of treatment was at least 2 years them. Upon completion of the study, a compli-
(M=6.8 years, SD= 3.2 years) and ranged from mentary restaurant coupon was given to each of
2 years 6 months to 17 years 6 months. Survivors the child and adult participants.
of all types of pediatric cancers were considered
for participation, except those with brain tumors
because of the obvious cognitive impairments and
social late effects of this type of neoplasm and its MEASURES
associated therapies (Mulhern et al., 1993). Diag-
noses included leukemia (n =20), Hodgkin’s Demographics
lymphoma (n = 2), non-Hodgkin’s lymphoma
(n =5), Wilm’s tumor (n =5), neuroblastoma Parents were asked to complete a brief ques-
(n=2), Rhabdomyo sarcoma (n =2), osteogenic tionnaire identifying their marital status, educa-
sarcoma (n=2), Burkitt’s lymphoma (n = 1), and tion, occupation, and ethnicity.
other types of solid tumors (n =3).
The demographic characteristics of the cancer
survivors are presented in Table 1. Ethnicity was Disease and treatment characteristics
Caucasian (n= 35), African American (n = 6),
and American Indian (n = 1). Most of the sur- A disease severity index was computed based
vivors were from two-parent families (n =38; on information from medical records and an inde-
90.5%). Most of the subjects were from middle- pendent rating provided by the oncologist. Items
class families; the mean socio-economic status selected for inclusion on the disease severity index
were obtained from the literature (Kazak, 1994)
Table 1. Demographic characteristics of cancer survivors and were delineated for review by two pediatric
oncologists. Both oncologists reviewed all of the
Variable F % M SD items, and those retained were based on a consen-
sus of opinion.
Gender The disease severity index was based on the
Male 21 50.0 four components of treatment history (whether
Female 21 50.0 the child was treated with chemotherapy, radia-
Race tion, transplantation, or a combination), medical
Caucasian 35 83.3 and psychological late effects (growth failure, car-
African American 6 14.3 diac impairment, sterility, second malignancy,
American Indian 1 2.4 skeletal malformations, like scoliosis or asymme-
Family composition try, and demonstrated developmental late effects,
Single-parent household 4 9.5 like learning disabilities and attention deficit hy-
Two-parent household 38 90.5 peractivity disorder), the physician’s subjective as-
SES 44.5 11.5
sessment of the survivor’s disease (histology,
Age at evaluation 13.1 2.9 spread, and other biologic measures), and treat-
Number of siblings 2.4 0.9 ment variables (diagnosis, age at diagnosis, length
of treatment, type of cytotoxic therapies, com-
n= 42. plications, time since completion of medical

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
 SOCIAL SKILLS AND PSYCHOLOGICAL ADJUSTMENT 117

treatment, and number of relapses). Participants chopathology and pediatric literature and has
were rated only on the subjective component, good psychometric properties. This measure was
and scores ranged from 4 (mild) to 12 (severe). used to assess child adjustment as rated by par-
The reliability of the two raters was 0.90 (Pear- ents. The total problem–behavior score was
son correlation coefficient). When the frequency used in the analyses. Because none of the analy-
of late effects was calculated, a severity score ses yielded significant effects for either of the
was provided. The severity score allowed for the internalizing or the externalizing broadband
quantification of the various types of late ef- scales, only the total problem–behavior score
fects. This severity score was used in the final was used in the analyses. Parents also were
analyses. asked to complete the social competence factor
Information for children’s academic function- of the CBCL.
ing was derived from the School scale of the
Child Behavior Checklist (CBCL) (Achenbach Teacher Report Form (TRF) (Achenbach and
and Edelbrock, 1991) and the Teacher Rating Edelbrock, 1991). Teachers were asked to com-
Form (TRF) (Achenbach and Edelbrock, 1991). plete the TRF, a 113-item checklist that ad-
These scales are described more fully in the next dresses a broad range of internalizing and
section. Items on the CBCL included current externalizing behavioral symptoms for children
school performance in academic areas (reading and adolescents aged 5–16 years. The TRF has
or English, writing, arithmetic, and spelling), re- also been used extensively in child psychopathol-
ceipt of special education services, and whether ogy and pediatric literature and has good psy-
or not a child had repeated a grade. Items on chometric properties. Consistent with the CBCL,
the TRF included current performance across this measure was used to assess child adjust-
academic subjects. For each rating (CBCL and ment as rated by teachers. The total problem–
TRF), a separate composite academic compe- behavior score was used in the analyses.
tence measure is obtained. This measure was
used as an index of academic functioning.
Teacher ratings of academic competence on the Social skills
TRF have been found to be associated highly
with more objective measures of cognitive aca- Social Skills Rating Scale-Parent Form
demic performance, including achievement tests (SSRS-P) (Gresham and Elliott, 1990). The
(Singer, 1972). One-week test – retest reliability SSRS-P assesses social skills for children and
for the School scale was 0.92 and interrater adolescents at three developmental levels: pre-
agreement was 0.87. This latter correlation was school, elementary, and secondary. This measure
the highest among the three social competency was completed collaboratively by mothers and
scales of the CBCL and the TRF. Cronbach’s fathers. Specifically, the SSRS-P measures pro-
alpha for the School scale ranged from 0.59 to social skills, social competence, and adaptive
0.62, depending upon the age range (4 – 11 years functioning at home and at school. Subscales
versus 12–18 years) and gender of the subjects. include Cooperation, Assertion, Responsibility,
All School scale items significantly discriminated and Self-Control. Adequate reliability and valid-
between community and psychiatric-referred ity have been established for the SSRS-P, with a
youth (Achenbach and Edelbrock, 1991). test–retest reliability of 0.87. The summed social
skills rating score was used in the analysis.
Psychological adjustment
Social Skills Rating Scale-Teacher Form
(SSRS-T) (Gresham and Elliott, 1990). The
Child Beha6ior Checklist (CBCL) (Achenbach SSRS-T is comparable in format with the rating
and Edelbrock, 1991). Mothers and fathers of scale administered to parents. Subscales include
cancer survivors completed the CBCL, a 113- Cooperation, Assertion, and Self-Control. Ade-
item checklist that addresses a broad range of quate psychometric data are available for the
internalizing (e.g. anxiety and depression) and SSRS-T, including test–retest reliability (0.75–
externalizing (e.g. acting-out) symptoms for chil- 0.93) and validity (0.44–0.77). Consistent with
dren and adolescents 4 – 16 years of age. The the SSRS-P, a summed social skills rating was
CBCL has been used extensively in child psy- used in the analysis.

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
118 W. LEVIN NEWBY ET AL.

Family functioning as clinically elevated (T] 60), 14.3% (n= 6) of the

children met criteria for poor adjustment as rated
by parents, and 7.1% (n = 3) met criteria for poor
Family En6ironment Scale—third edition (FES)
adjustment as rated by teachers. Because no dif-
(Moos and Moos, 1994). Mothers and fathers were
ferences were found for any of the variables as a
asked to complete the FES collaboratively, which
function of gender, this variable was not con-
consists of 90 true – false items that form ten
trolled in subsequent analyses.
subscales assessing the social – environmental at- Results of the bivariate correlation coefficients
tributes of various families. Three scales from this are presented in Table 2. The correlations yielded
measure were chosen for analysis because of their significant associations between the SSRS-P and
relevance to social functioning: Family Cohesion, SES (r= 0.47, pB 0.001) and academic function-
Expressiveness, and Organization. The FES has ing as rated by parents (r= 0.42, pB 0.001) and
been demonstrated to have good psychometric teachers (r= 0.50, pB 0.001). Thus, parents’ rat-
properties, including test – retest reliability (0.86 ings of their children’s prosocial behaviors were
over a 2-month period) and internal consistency significantly associated with higher social class
(0.61–0.78), and has been used in research on and fewer academic difficulties as rated by parents
families with chronic illness. and teachers. For the SSRS-T, significant associa-
tions were found for academic functioning as
rated by parents (r= 0.54, pB 0.001) and teachers
RESULTS (r=0.56, pB 0.001), and the TRF (r= − 0.45,
pB0.001). Thus, teacher ratings of prosocial
skills were significantly associated with fewer
A series of Pearson correlation coefficients were symptoms of academic difficulty as rated by both
computed for measures of social skills as rated by parents and teachers, and with teacher ratings of
both parents (SSRS-P) and teachers (SSRS-T), adjustment; fewer behavioral problems were asso-
psychological adjustment as rated by both parents ciated with teacher ratings of prosocial skills. For
(CBCL) and teachers (TRF), and disease and the CBCL parent rating of adjustment, significant
treatment characteristics (age at diagnosis, time associations were found for time off therapy (r=
off therapy, number of physical late effects, aca- −0.48, pB 0.001) and academic functioning as
demic functioning as rated by teachers and par- rated by parents (r= − 0.55, p B 0.001). Thus, a
ents, and physicians’ ratings), demographic data higher frequency of academic difficulties and less
(SES, age at evaluation), and family functioning time off therapy were associated with a greater
(FES cohesiveness, expressiveness, and organiza- frequency of behavioral problems (i.e. poorer ad-
tion). One-way ANOVAs also were performed on justment) as rated by parents.
each of the dependent measures for gender. Fi- TRF scores were significantly associated with
nally, based on the bivariate correlations, a series time off therapy (r= − 0.51, p B 0.001) and aca-
of hierarchical regression equations were con- demic difficulties as rated by parents (r= − 0.53,
structed to predict social skills and psychological pB 0.001). In addition, as expected, TRF scores
adjustment as rated by both parents and teachers. were associated with parent ratings of adjustment
Because of the small sample size, the numerous as rated in the CBCL (r= − 0.53, pB 0.001).
analyses, and the possibility of Type I error, a Specifically, a higher frequency of behavioral
Bonferroni correction procedure was used for problems (poorer adjustment, as rated by teach-
each set of bivariate correlations for each of the ers, was associated with less time off therapy and
measures of interest (CBCL, TRF, SSRS-P, with a greater incidence of academic and adjust-
SSRS-T). Thus, only correlation coefficients at ment problems as endorsed by parents. With the
the 0.006 level of significance were considered as exception of academic functioning and time off
significant. therapy, none of the other disease and treatment
Survivors’ adjustment scores suggested signifi- characteristics, including treatment intensity (e.g.
cant variability, ranging from the 3rd to the 97th prophylactic chemotherapy or radiation), were
percentile as reported by parents and from the 4th significantly associated with either parent or
to the 97th percentile as reported by teachers. The teacher ratings of adjustment or social skills. In
mean adjustment score of both the CBCL and addition, treatment intensity was neither associ-
TRF corresponded to the 50th percentile of the ated with parent (r= − 0.37) nor teacher ratings
normative sample. Establishing the 80th percentile of academic difficulties (r= − 0.01).

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
 SOCIAL SKILLS AND PSYCHOLOGICAL ADJUSTMENT 119

Table 2. Correlations among variables

Adjustment and Adaptation Social Skills

CBCL TRF SSRS-P SSRS-T

Disease and treatment characteristics

Age at diagnosis −0.29* 0.07 0.11 0.07
Time off therapy −0.48**** −0.51**** 0.06 0.14
Academic functioning
Teacher report −0.36** −0.23 0.50**** 0.56****
Parent report −0.55**** −0.53**** 0.42**** 0.54****
Physician rating 0.27 0.31 −0.13 0.05
Demographics
SES −0.29* −0.26* 0.47**** 0.31*
Age at evaluation −0.04 −0.15 −0.01 0.07
Family environment
Cohesion −0.33** 0.38** 0.20 −0.09
Expressiveness 0.04 0.03 0.13 −0.12
Organization 0.14 0.23 0.14 −0.11
Adjustment and adaptation
CBCL — −0.53**** −0.32* −0.17
TRF 0.23 — −0.25 −0.45****

* pB0.05; ** pB0.02; *** pB0.01; **** pB0.001.

Bonferroni correction procedure for family-wise test (pB0.006) required for significance at 0.05
level.

Based on the bivariate correlation coefficients, Although the FES Cohesion measure was not
a series of hierarchical regression equations were significantly associated with the other measures
constructed for the SRSS-P, SSRS-T, CBCL, and after the Bonferroni correction procedure, it was
TRF. The results of these regression equations are entered as a last step on the CBCL and TRF
presented in Table 3. SES was included as a because the relationship between the familial co-
covariate for the SSRS-P because of the signifi- hesion score and teacher and parent ratings of
cant bivariate association between social class and adjustment approached significance and because
parent ratings of social skills. SES was included as of the theoretical importance of the FES. Parent
a covariate for the SSRS-P because the bivariate and teacher ratings of adjustment were not en-
correlation approached significance. Parental rat- tered into any of the regression analyses for the
ings of academic functioning were entered for SSRS-P or the SSRS-T because the association of
teacher ratings of social skills, and teacher ratings adjustment and social skills was not of interest in
of academic functioning were entered for parent this particular investigation.
ratings of social skills because of the significant The results of the regression equation for the
correlation coefficients. Ratings of academic diffi- SSRS-P yielded a significant main effect [R 2 =
culties as reported by either teachers or parents 0.31, F(2,25)=5.76, pB 0.009]. After controlling
for SES, only teacher ratings of academic func-
were entered on the subsequent step following
tioning were found to predict parents’ ratings of
SES. For the CBCL, time off therapy was entered
children’s social skills, with 21% of the variance in
as a covariate because of the strong bivariate
social skills being predicted by teacher ratings of
association between this variable and the CBCL.
academic functioning. Thus, poorer social skills as
For teacher ratings of adjustment (TRF), time off rated by parents were associated with a higher
therapy and parent ratings of academic function- frequency of academic difficulties.
ing were entered because of the strong bivariate The results of the regression equation for the
association between academic functioning and SSRS-T also yielded a significant effect [R 2 =
teacher ratings of adjustment. 0.30, F(2,24)=5.14, pB 0.01]. After controlling

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
120 W. LEVIN NEWBY ET AL.

Table 3. Summary of regression equations

Variable R2 R 2Change b F

SSRS-Parent
SES 0.10 0.02 0.01
Academic functioning, teacher 0.31 0.21 1.03 7.53***
CBCL
Time off therapy 0.23 −1.31 8.98***
FES Cohesion 0.28 0.05 −0.18 2.28a
SSRS-Teacher
SES 0.08 0.09 0.19
Academic functioning, parent 0.30 0.22 0.65 7.26***
TRF
Academic functioning, parent 0.29 −0.61 24.48****
FES Cohesion 0.59 0.30 0.44 19.48****

* pB0.05; ** pB0.02; *** pB0.01; **** pB0.001.

a
Represents a trend toward significance (pB0.10).

for SES, parent ratings of academic functioning ents. Table 4 presents the mean and SD for each
were found to predict teacher ratings of children’s measure.
social skills, with 22% of the variance accounted
for by academic functioning. Thus, poorer social
skills ratings by teachers were associated with a DISCUSSION
higher frequency of academic difficulties reported
by parents.
A regression equation computed for the TRF This study investigated social skills and psycho-
logical adjustment of child and adolescent cancer
problem–behaviors measure yielded a significant
survivors and is one of the few studies which
effect [R 2 =0.59, F(2,26) = 19.05, p B0.0001].
examines social skills well into the late-effects
Parent ratings of children’s academic functioning
period. Consistent with the findings of Noll et al.
and the FES Cohesion scale were entered into the
(1997), the current data indicate that, as a group,
equation. Ratings of academic functioning were these survivors generally evidence normal social
found to account for 29% of the variance in skills and few internalizing or externalizing behav-
teacher ratings of adjustment, and parent ratings ioral problems (i.e. adequate adjustment) when
of family cohesion were found to account for an compared with normative data of their healthy
additional 30% of the variance. Specifically, rat- peers. Nonetheless, 14% of the survivors met
ings of academic functioning and cohesion within criteria for poor adjustment as rated by parents, a
the family accounted for nearly 60% of the vari- finding consistent with the data provided by the
ance in teacher ratings of adjustment. A higher Isle of Wight study, where the rate of psychiatric
frequency of behavioral problems as reported by disorders in children with a chronic illness was
teachers was associated with greater family cohe- 17% (Thompson and Gustafson, 1996). However,
sion and survivors’ lower academic functioning as only 7% of the sample met criteria for poor
reported by parents. adjustment as rated by teachers. This percentage
Finally, for the CBCL, only time off therapy is quite low, even lower than the data reported by
was found to predict parent ratings of children’s Achenbach and Howell (1993). Nonetheless, these
adjustment [R 2 = 0.28, F(2,38) = 7.27, p B0.002]; findings support the observations of Thompson
there was a trend toward significance for the FES and Gustafson (1996), that different informants
Cohesion scale to predict CBCL scores. Time off may differentially rate the adjustment of children
therapy accounted for 23% of the variance in with chronic illness. Coupled with their observa-
parental ratings of survivors’ adjustment, with less tions, the current data underscore the importance
time off therapy associated with a higher fre- of engaging multiple informants (i.e. both parents
quency of behavioral problems endorsed by par- and teachers) to assess adjustment.

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
 SOCIAL SKILLS AND PSYCHOLOGICAL ADJUSTMENT 121

Table 4. Mean and standard deviation for each measure

Measure Cancer Survivors Normal Sample

M SD Range M SD

SSRS-Parenta 100.3 17.0 65–130 100.0 15.0

CBCL (adjustment/adaptation)b 49.1 9.7 28–69 50.0 10.0
SSRS-Teachera 104.5 13.5 80–128 100.0 15.0
TRF (adjustment/adaptation)b 46.4 9.7 31–69 50.0 10.0
FES
Cohesion 54.3 11.9 25–65
Expressiveness 51.0 10.7 28–71
Organization 54.7 10.9 32–69

a
SSRS n=922.
b
CBCL and TRF n= 2200.

These findings show that social skills as rated academic functioning as rated by both parents
by both parents and teachers are best predicted by and teachers significantly predicted children’s so-
academic functioning. In fact, for both parent and cial skills and adjustment lends support to the
teacher ratings of social skills, academic function- pediatric psychology literature, that social and
ing accounted for over 20% of the variance in emotional competencies of children with chronic
survivor social skills, with fewer school-related illnesses are primarily associated with academic
difficulties being associated with better psycholog- functioning (Wallander et al., 1989; Breitmeyer et
ical adjustment (fewer behavioral problems as en- al., 1992), or as indicated in the present study,
dorsed by parents and teachers). Similar to child challenges with learning at school. In fact, the
psychological adjustment as rated by teachers, current data might be interpreted to support the
parent ratings of academic functioning accounted findings of Mulhern et al. (1989), that for sur-
for nearly one third of the variance in child vivors of childhood cancer, lower parent ratings
adjustment, with a higher frequency of academic of social competencies were associated with a
difficulties associated with poorer adjustment. In higher frequency of neurocognitive toxicities.
addition, family cohesiveness was found to ac- That nearly one half of the sample received either
count for a significant percentage of the variance prophylactic chemotherapy or radiation lends
(nearly one third) for teacher ratings of adjust- some support to this notion. Other researchers
ment and was marginally associated with parent have noted that increased parent and teacher re-
ratings of behavioral problems (i.e. adjustment). porting of academic difficulties is associated with
Thus, reports by parents of greater cohesiveness cognitive neurotoxicities, which may be due to
within families were associated with a higher fre- intrathecal chemotherapy or cranial radiation
quency of behavioral problems in survivors as (Brown et al., 1992, 1996; Madan-Swain and
rated by teachers. Finally, length of time off Brown, 1992). This is unlikely in the current
treatment accounted for a significant percentage sample, however, because treatment intensity (i.e.
of the variance (nearly a quarter) in child adjust- cranial radiation, intrathecal chemotherapy) was
ment as rated by parents; children who were off not associated with either parent or teacher rat-
therapy longer were rated by their parents as ings of academic functioning. It is likely that the
having better psychological adjustment. observed academic weaknesses are at least par-
These findings are in accord with those studies tially the result of frequent absences from school.
indicating generally adequate adjustment for child For example, Williams et al. (1991) reported a
survivors of cancer (Teta et al., 1986; Van Don- significant association between poor ratings of
gen-Melman and Saunders-Woudstra, 1986; academic skills and number of days missed from
Chang et al., 1988; Fritz and Williams, 1988; school. Thus, the observed effect may be due to
Greenberg et al., 1989; Spirito et al., 1990; earlier periods of school absence. Unfortunately,
Madan-Swain and Brown, 1992; Kupst, 1994; information on school absences during the cancer
Madan-Swain et al., 1994; Noll et al., 1997). That experience would have been very difficult to

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
122 W. LEVIN NEWBY ET AL.

obtain because data were gathered many years literature relating to chronically ill children is that
after treatment had terminated (the mean time cohesiveness within families, which is usually con-
since completion of treatment was greater than 6 sidered to be a positive attribute of family life for
years). Regardless of the etiology of these lower healthy children, is generally associated with more
ratings of academic performance, they are associ- adjustment difficulties in those youth who have
ated with significant effects on the social skills survived cancer (Rait et al., 1992; Kazak et al.,
functioning and psychological adjustment of these 1995). One explanation for this is that children
survivors. Thus, consistent with the findings of who have survived cancer may have greater needs
Kupst et al. (1995) on survivors of leukemia, for autonomy than their healthy peers. Thus, any
children who experience diminished academic perceptions of intrusion from the family system
functioning seem to be at greatest risk for adjust- may be associated with symptoms of adjustment
ment difficulties and social skills problems. Future difficulties among survivors. Additional research
intervention studies should target such at-risk will be needed to determine the veracity of this
children with a focus on improving adjustment explanation. These data have implications for a
and social skills with peers. family systems approach to psychotherapy, partic-
Of particular interest in this investigation and ularly for children who are designated to be at
consistent with the pediatric chronic illness litera- risk for adjustment difficulties during either the
ture is the strong association between family func- treatment or late-effects period. The data suggest
tioning and child psychological adjustment. that less cohesiveness within the families of cancer
Cohesiveness within the family accounted for survivors may be associated with better adjust-
nearly one third of the variance in psychological ment during the late-effects period.
adjustment as rated by teachers. Specifically, for One final explanation for these findings may
survivors, a greater frequency of behavioral prob- relate to the impact of teacher perceptions of
lems as reported by teachers was associated with family functioning on their perceptions of child
greater reports of cohesiveness within their behavior. Teachers may view the increased cohe-
families. These data reflect the observations of sion that is typical of families who have experi-
Rolland (1984, 1987, 1990) and the findings of enced childhood chronic illness (Rolland, 1984,
others (Wertlieb et al., 1986; Hanson et al., 1989; 1987, 1990) as negative, and this in turn may
Wallander et al., 1989; Thompson et al., 1992), influence their perceptions of child adjustment.
who underscore the role of the family environ- Families who exert too much control during the
ment in predicting adjustment to the disease pro- late-effects period are likely to have children who
cess for both chronically ill children and their exhibit adjustment difficulties within the context
parents. The findings also are in accord with the of another system (school), where the decrease in
observations of Rait et al. (1992), who found that structure may lead to acting out.
both family cohesion and adaptability were asso- Childhood cancer differentially affects the fam-
ciated with adjustment during the late-effects ily system over time, from diagnosis to the end of
period. treatment to the late-effects period (Kazak, 1994).
Taken together with those studies that have Unfortunately, this study is unable to address
been conducted for children with cancer (Kazak how childhood cancer affects the family system
and Nachman, 1991; Madan-Swain and Brown, over time, from diagnosis to the late-effects pe-
1992; Kupst, 1994; Kazak et al., 1995), the cur- riod. However, when time off therapy was con-
rent findings support the importance of family trolled, family cohesion was still associated with
functioning in predicting overall adjustment dur- child adjustment during the late-effects period,
ing the late-effects period. For children who have suggesting significant influences of family func-
survived cancer, their families who report signifi- tioning during this period. Additional longitudi-
cant cohesiveness are likely to have survivors who nal studies are needed to evaluate the change of
exhibit poorer adjustment. These data might be family functioning over the course of the cancer
interpreted to indicate that enmeshment as re- experience.
ported by families is associated with teacher rat- Few of the disease and treatment characteris-
ings of adjustment difficulties. However, it should tics, with the exception of academic functioning
be noted that the mean FES scores are well within and time off therapy, predicted child adjustment
the average range. An alternative explanation for and social skills as rated by either parents or
the findings that has been echoed in the family teachers. In fact, physician ratings did not predict

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
 SOCIAL SKILLS AND PSYCHOLOGICAL ADJUSTMENT 123

outcome; only parent and teacher reports of aca- instrument may not detect more subtle adjustment
demic functioning were associated with child out- difficulties (Perrin et al., 1991). Thus, the data
come. One explanation for these findings is that may underreport the prevalence of adjustment
physician ratings did not include a category of difficulties in this sample of cancer survivors. In
cognitive or academic impairments, the primary addition, information on survivors’ adjustment
disease-related variable that predicted adjustment and social skills were obtained only from parents
and social skills. Interestingly, children who were and teachers; self-reports of adjustment were not
off therapy longer evidenced better adjustment as included. Although the use of self-report data
rated by parents. Perhaps children and adoles- from the survivors may have been useful, epidemi-
cents experience significant initial anxiety and ologic data generally suggest that teacher and
emotional turbulence following the cessation of parent reports are more reliable and valid in the
treatment because they are concerned that the identification of psychopathology, particularly for
disease may recur. Thus, the longer the period off externalizing difficulties for younger children
therapy, the less concern for possible relapse and (Thompson and Gustafson, 1996). Future re-
the better the psychological adjustment. Future search should validate the reliability of self-report
research should examine differential patterns of ratings with parent and teacher ratings, particu-
adjustment of cancer survivors during various larly for older adolescents. Moreover, because an
phases of the late-effects period, including imme- experimental design was not used, the data cannot
diately after termination of therapy. be considered causal. That parents and teachers
The contributions of these findings must be rated children as more psychologically adjusted
interpreted within the limitations of the investiga- and better skilled socially simply because they
tion. Unfortunately, specific data from individual performed better at school might explain the ob-
academic achievement or neuropsychologic tests served findings. Future studies will need to utilize
of cognitive impairments were not available, so longitudinal designs, following children prospec-
cognitive late effects had to be inferred from both tively through the various phases of the cancer
parent and teacher report measures of academic experience, including the late-effects period.
functioning. Parent and teacher ratings do not The data do not permit a definitive statement
provide as sensitive and specific a picture as that on these associations between academic impair-
obtained through direct academic or neuro- ments, social skills, and teacher and parent ratings
psychologic assessments. This shortcoming, how- of behavior and survivors of childhood cancer
ever, is mitigated by the choice of extensively specifically. In order to determine that the ob-
standardized checklists with strong psychometric tained associations are unique for survivors of
properties (i.e. the TRF and the School scale of childhood cancer, survivors must be compared
the CBCL). Although parents may inaccurately with other chronic illness groups. Intervention
report their children’s academic functioning, studies that examine ways to enhance cognitive
teacher ratings of academic achievement also were and academic skills in cancer survivors who are
administered. In this study, parent and teacher experiencing difficulties at school and then exam-
ratings of academic performance were signifi- ine adjustment and social skills will be useful.
cantly correlated (r= 0.54, p B 0.002). Moreover, Finally, the relatively small sample size, coupled
the reportedly high rates of interparent and in- with the heterogeneity of the sample with regard
terteacher agreement for the School scales of the to age, diagnosis, and time off treatment, may
CBCL and the TRF and the fact that all com- have obscured significant results; thus, future re-
posite items discriminated between community- search should include multicenter studies where
and psychiatric-referred samples minimized this greater numbers of subjects can be enrolled.
threat to the integrity of the study. Future investi- This investigation is one of the few studies to
gations using psychometric measures of neurocog- include teacher reports in addition to parent re-
nitive functioning are needed to examine more ports of behavioral adjustment. The finding that
intensively the associated relationships of neu- teacher ratings of adjustment were predicted pri-
rocognitive functioning and psychosocial late ef- marily by family cohesiveness underscores the im-
fects for childhood cancer survivors. portance of employing teacher ratings in
Another issue is that the utility of the CBCL in late-effect studies of children and adolescents with
the assessment of adjustment for children with cancer. In fact, some research has suggested that
chronic illness has been questioned because the teachers may be the most accurate reporters of

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
124 W. LEVIN NEWBY ET AL.

emotional adjustment in children (Loeber et al., differences between survivors and the healthy com-
1990). This also again emphasizes the importance parison group (all FB 1).
of a multi-informant approach, including teach-
ers, for the assessment of adjustment and social
skills in children with chronic illness. REFERENCES
Although the findings are in accord with much
of the chronic illness literature in suggesting gen- Achenbach, T. and Edelbrock, C. (1991) Manual for the
erally adequate adjustment and social skills for Child Beha6ior Checklist and Teacher Rating Form.
children having survived the cancer experience, University of Vermont, Burlington, VT.
the data highlight the role of academic function- Achenbach, T. and Howell, C.T. (1993) Are American
ing and familial functioning (cohesiveness) in in- children’s problems getting worse? A 13-year com-
fluencing psychological adjustment and social parison. J. Am. Acad. Child Adolesc. Psychiatry 32,
1145 – 1154.
skills for children who have survived cancer. The Breitmeyer, B.J., Gallo, A.M., Knafl, K.A. and Zoeller,
findings have important implications for interven- L.H. (1992) Social competence of school-aged chil-
tion programs for at-risk children, which can dren with chronic illness. J. Pediatr. Nursing 7, 181 –
include tutoring and additional time spent on 188.
school-related subjects. In addition, the data have Brown, R.T. and Madan-Swain, A. (1993) Cognitive,
implications for children evidencing social skills neuropsychological, and academic sequelae in chil-
deficits and adjustment difficulties, since remedia- dren with leukemia. J. Learn. Disabil. 26, 74 – 90.
tion of learning problems through special educa- Brown, R.T., Madan-Swain, A., Pais, R., Lambert,
tion programs and a family systems approach to R.G., Sexson, S.B. and Ragab, A. (1992) Chemother-
psychotherapy can likely enhance the quality of apy for acute lymphocytic leukemia: cognitive and
academic sequelae. J. Pediatr. 121, 885 – 889.
life for children and adolescents who have sur- Brown, R.T., Sawyer, M., Antoniou, G., Toogood, I.,
vived cancer. Rice, M., Thompson, N. and Madan-Swain, A.
(1996) A 3-year follow-up of the intellectual and
academic functioning of children receiving CNS-
ACKNOWLEDGEMENTS prophylactic chemotherapy for leukemia. J. De6elop.
Beha6. Pediatr. 17, 392 – 398.
Brown, R.T., Mulhern, R. and Simonian, S. (2000)
This research was completed by the senior author in Diseases of the blood and blood-forming organs, in
partial fullfilment for a PhD degree in psychology Health and Beha6ior (T. Boll Ed.), vol. 3. American
under the supervision of the second author. Psychological Association, Washington, DC (in
press).
Chang, P.N., Nesbit, M.E., Youngren, N. and
NOTES Robinson, L.L. (1988) Personality characteristics and
psychosocial adjustment of long-term survivors of
childhood cancer. J. Psychosoc. Oncol. 5, 43 – 58.
1. An effort to generate a healthy comparison cohort Cousens, P., Waters, B., Said, J. and Stevens, M. (1988)
resulted in a sample of 15 children and adolescents Cognitive effects of cranial radiation in leukemia: a
drawn from the same elementary, middle, and high survey and meta-analysis. J. Child Psychol. Psychi-
school classrooms as survivors that was matched to atry 29, 839 – 852.
the indexed subjects on gender and chronological Cousens, P., Ungerer, J.A., Crawford, J.A. and
age. Each teacher was instructed to select one Stevens, M.M. (1991) Cognitive effects of childhood
healthy child of the same gender and with the closest leukemia therapy: a case for four specific deficits. J.
birth date to the survivor participant. Of the 42 Pediatr. Psychol. 16, 475 – 488.
healthy controls that were recruited, only 14 (35.7%) Fletcher, J.M. and Copeland, D.R. (1988) Neurobehav-
returned consent forms for participation in the in- ioral effects of central nervous system prophylactic
vestigation. In addition, because the comparison treatment of cancer in children. J. Clin. Exp. Neuro-
control group was recruited from schools, only psychol. 10, 495 – 538.
teachers completed measures for these children. For Fritz, G.K. and Williams, J.R. (1988) Issues of adoles-
these reasons, the comparison control group was not cent development for survivors of childhood cancer.
included in this report. There were, however, no J. Am. Acad. Child Adolesc. Psychiatry 27, 712 – 715.
significant differences in SES, age, gender, or ethnic- Greenberg, H.S. and Meadows, A.T. (1991) Psycho-
ity between cancer survivors and available compari- social impact of cancer survival on school-age chil-
son subjects, and a series of one-way ANOVAs dren and their parents. J. Psychosoc. Oncol. 9, 43 –
performed on the TRF and SSRS-T yielded no 56.

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
 SOCIAL SKILLS AND PSYCHOLOGICAL ADJUSTMENT 125

Greenberg, H.S., Kazak, A. and Meadows, A. (1989) MacLean, W.E., Noll, R.B., Stehbens, J.A., Kaleita,
Psychological adjustment in 8–16-year-old cancer T.A., Schwartz, E., Whitt, J. K., Cantor, N.L.,
survivors and their parents. J. Pediatr. 114, 488–493. Waskerwitz, M., Ruymann, F., Novak, L.J.,
Gresham, F.M. and Elliott, S.N. (1990) Social Skills Woodard, A. and Denman, H.G. (1995) Neuropsy-
Rating System. American Guidance Service, Circle chological effects of cranial irradiation in young chil-
Pines, MN. dren with acute lymphoblastic leukemia 9 months
Hanson, C.L., Henggeler, S.W., Harris, M.A., Burgen, after diagnosis. Arch. Neurol. 52, 156 – 160.
G.A. and Moore, M. (1989) Family system variables Madan-Swain, A. and Brown, R.T. (1992) Cognitive
and the health status of adolescents with insulin-de- and psychosocial sequelae for children with acute
pendent diabetes mellitus. Health Psychol. 8, 239 – lymphocytic leukemia and their families. Clin. Psy-
253. chol. Re6. 11, 267 – 294.
Haupt, R., Fears, T.R. and Robinson, L.L. (1994) Madan-Swain, A., Brown, R.T., Sexson, S.B., Baldwin,
K., Pais, R. and Ragab, A. (1994) Adolescent cancer
Educational attainment in longterm survivors of
survivors: psychosocial and familial adjustment. Psy-
childhood acute lymphocytic leukemia. J. Am. Med.
chosomatics 35, 1–7.
Assoc. 272, 1427–1432.
Moos, R.H. and Moos, B.S. (1994) Family En6ironment
Hollingshead, A.B. (1975) Fi6e-Factor Index of Social Scale Manual De6elopment, Applications, Research.
Status. Department of Sociology, Yale University, Center for Health Care Evaluation, Palo Alto, CA.
Princeton, NJ. Mulhern, R.K. (1994) Neuropsychological late effects,
Kazak, A. (1994) Implications of survival: pediatric in Pediatric Psychooncology: Psychological Perspec-
oncology patients and their families, in Pediatric ti6es on Children with Cancer (D.J. Bearison and
Psycho-oncology: Psychological Perspecti6es on Chil- R.K. Mulhern Eds.). Oxford University Press, New
dren with Cancer (D.J. Bearison and R.K. Mulhern York, pp. 99 – 121.
Eds.). Oxford University Press, New York, pp. 171 – Mulhern, R.K., Wasserman, A.L., Friedman, A.G. and
192. Fairclough, D. (1989) Social competence and behav-
Kazak, A. and Nachman, G. (1991) Family research on ioral adjustment of children who are long-term sur-
childhood chronic illness: pediatric oncology as an vivors of cancer. Pediatrics 83, 18 – 25.
example. J. Family Psychol. 4, 462–483. Mulhern, R.K., Carpentieri, S., Shema, S., Stone, P.
Kazak, A., Segal-Andrews, A.M. and Johnson, K. and Fairclough, D. (1993) Factors associated with
(1995) Pediatric psychology research and practice: a social and behavioral problems among children re-
family systems approach, in Handbook of Pediatric cently diagnosed with brain tumor. J. Pediatr. Psy-
Psychology (M. Roberts Ed.). Guilford Press, New chol. 18, 339 – 350.
York. Noll, R., Bukowski, W., Rogosch, F., LeRoy, S. and
Kupst, M.J. (1994) Coping with pediatric cancer: theo- Kulkarni, R. (1990) Social interactions between chil-
retical and research perspectives, in Pediatric Psy- dren and their peers: teacher ratings. J. Pediatr.
chooncology: Psychological Perspecti6es on Children Psychol. 7, 75 – 84.
with Cancer (D.J. Bearison and R.K. Mulhern Eds.). Noll, R., LeRoy, S., Bukowski, W.M., Rogosch, F.A.
Oxford University Press, New York, pp. 35–60. and Kulkarni, R. (1991) Peer relationships and ad-
Kupst, M.J. and Schulman, J.L. (1988) Long-term cop- justment in children with cancer. J. Pediatr. Psychol.
ing with pediatric leukemia: a six-year follow-up 16, 307 – 326.
study. J. Pediatr. Psychol. 13, 7–22. Noll, R.B., Ris, M.D., Davies, W.H., Bukowski, W.M.
Kupst, M.J., Schulman, J.L., Honig, G., Morgan, E. and Koontz, K. (1992) Social interactions between
children with cancer or sickle cell disease and their
and Fochtman, D. (1982) Family coping with child-
peers: teacher ratings. J. De6elop. Beha6. Pediatr. 13,
hood leukemia: one year after diagnosis. J. Pediatr.
187 – 193.
Psychol. 7, 157–174.
Noll, R.B., Bukowski, W.M., Davies, W.H., Koontz,
Kupst, M.J., Schulman, J.L., Maurer, H., Morgan, E., K. and Kulkarni, R. (1993) Adjustment in the peer
Honig, G. and Fochtman, D. (1984) Coping with system of adolescents with cancer: a two-year study.
pediatric leukemia: a two-year follow-up. J. Pediatr. J. Pediatr. Psychol. 18, 351 – 364.
Psychol. 9, 149–163. Noll, R.B., MacLean, W.E., Whitt, J.K., Kaleita, T.A.,
Kupst, M.J., Natta, M.B., Richardson, C.C., Schul- Stehbens, J.A., Waskerwitz, M.J., Ruymann, F.B.
man, J.L., Lavigne, J.V. and Das, L. (1995) Family and Hammond, G.D. (1997) Behavioral adjustment
coping with pediatric leukemia: ten years after treat- and social functioning of long-term survivors of
ment. J. Pediatr. Psychol. 20, 601–617. childhood leukemia: parent and teacher reports. J.
Loeber, R., Green, M. and Lahey, B. (1990) Mental Pediatr. Psychol. 92, 827 – 841.
health professionals’ perception of the utility of chil- Perrin, E.C., Stein, R.K. and Drotar, D. (1991) Cau-
dren, mothers, and teachers as informants on child- tions in using the Child Behavior Checklist: observa-
hood psychopathology. J. Clin. Child Psychol. 19, tions based on research about children with a chronic
136–143. illness. J. Pediatr. Psychol. 16, 411 – 422.

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)
126 W. LEVIN NEWBY ET AL.

Rait, D.S., Ostroff, J.S., Smith, K. and Cella, D.F. Thompson, R., Gustafson, K., Hamlett, K.W. and
(1992) Lives in a balance: perceived family function- Spock, A. (1992) Stress, coping, and family function-
ing and the psychosocial adjustment of adolescent ing in the psychological adjustment of mothers and
cancer survivors. Family Process 31, 383–397. adolescents with cystic fibrosis. J. Pediatr. Psychol.
Rolland, J. (1984) Towards a psychosocial typology of 17, 741 – 755.
chronic and life-threatening illness. Family Syst. Van Dongen-Melman, J.E.W.M. and Saunders-
Med. 2, 245–262. Woudstra, J.A.R. (1986) Psychosocial aspects of
Rolland, J. (1987) Chronic illness and the life cycle: a childhood cancer: a review of the literature. J. Child
conceptual framework. Family Process 26, 203–211. Psychol. Psychiatry 27, 145 – 180.
Rolland, J. (1990) Anticipatory loss: a family systems Varni, J.W., Wilcox, K.T. and Hanson, V. (1988)
framework. Family Process 29, 229–244. Chronic musculoskeletal pain and functional status
Singer, H. (1972) Review of Metropolitan Readiness in juvenile rheumatoid arthritis: an empirical model.
Test, in The Se6enth Mental Measurements Yearbook Pain 39, 1–7.
(O.K. Buros Ed.). The Buros Institute of Mental Wallander, J., Varni, J., Babani, L., Banis, H. and
Measures, Lincoln, NB. Wilcox, K. (1989) Family resources as resistance
Spirito, A., Stark, L., Cobiella, C., Drigan, R., An- factors for psychological maladjustment in chroni-
drokites, A. and Hewitt, K. (1990) Social adjustment cally ill and handicapped children. J. Pediatr. Psy-
of children successfully treated for cancer. J. Pediatr. chol. 14, 157 – 174.
Psychol. 15, 359–371. Wertlieb, D., Hauser, S.T. and Jacobson, A.M. (1986)
Spirito, A., DeLawyer, D. and Stark, L. (1991) Peer Adaptation to diabetes: behavior symptoms and fam-
relations and social adjustment of chronically ill chil- ily context. J. Pediatr. Psychol. 11, 463 – 479.
dren and adolescents. Clin. Psychol. Re6. 11, 539 – Whitt, J.K., Wells, R.J., Lauria, M.M., Wilhelm, C.L.
564. and McMillan, C.W. (1984) Cranial radiation in
Teta, M., Po, M., Kasl, S., Meigs, J., Myers, M. and childhood acute lymphocytic leukemia: neuropsycho-
Mulvihill, J. (1986) Psychosocial consequences of logical sequelae. Am. J. Dis. Child. 138, 730 – 736.
childhood and adolescent cancer survival. J. Chron. Williams, K.S., Ochs, J., Williams, J.M. and Mulhern,
Dis. 39, 751–759. R.K. (1991) Parental report of everyday cognitive
Thompson, R.J. and Gustafson, K.E. (1996) Adaptation abilities among children treated for acute
to Chronic Childhood Illness. American Psychological lymphoblastic leukemia. J. Pediatr. Psychol. 16, 13 –
Association, Washington, DC. 26.

Copyright © 2000 John Wiley & Sons, Ltd. Psycho-Oncology 9: 113–126 (2000)