Assessing Quality of Life in Adult Cancer Survivors (QLACS)
Nancy E. Avis1, Kevin W. Smith2, Sarah McGraw2, Roselyn G. Smith3, Vida M. Petronis3& Charles S.
1Wake Forest University School of Medicine, Department of Public Health Sciences, Winston-Salem, NC,
USA (E-mail: firstname.lastname@example.org);2New England Research Institutes, Watertown, MA, USA;3University
of Miami, Coral Gables, FL, USA
Accepted in revised form 11 August 2004
This article describes development of a quality of life measure designed to assess issues relevant to long-
term cancer survivors. In-depth semi-structured interviews were conducted with 58 long-term cancer sur-
vivors to identify domains most relevant to long-term survivors (?5 years post-diagnosis). Self-report items
were developed from these interviews and administered to a second sample of 242 long-term survivors.
Domains and items were selected from the item pool by a combination of factor analysis and criterion-
based item selection. Five cancer-specific domains were identified (appearance concerns, financial problems,
distress over recurrence, family-related distress, and benefits of cancer) along with seven generic QOL
domains (negative feelings, positive feelings, cognitive problems, sexual problems, physical pain, fatigue,
and social avoidance). Cronbach’s a was 0.72 or greater for each domain. Correlations between domain
scores and criterion measures were 0.72 or higher in all but one generic domain (social avoidance), but
somewhat lower on cancer-specific domains. The new multidimensional measure has good internal con-
sistency and validity and is appropriate for comparisons between cancer and non-cancer populations, as
well as long-term follow-up of cancer patients.
Key words: Cancer survivors, Instrument development, Quality of life
The importance of quality of life issues for cancer
patients is well-recognized by both researchers and
clinicians [1–3]. Over the past several decades
numerous studies have addressed the physical,
emotional, social, and sexual well-being of cancer
patients [e.g., 4–8; for reviews see 1, 9–13]. The
focus of these studies, however, has been restricted
largely to the period of treatment following diag-
nosis. With improved early detection and treat-
ment, large numbers of cancer patients are now
surviving many years post-diagnosis with the
proportion of cancer patients surviving 5 or more
years now approximately one in two . As of
January 1997, there were an estimated 8.9 million
cancer survivors in the US . This large number
of people surviving many years post-cancer diag-
nosis has heightened interest in studying long-term
effects of cancer on quality of life (QOL) [16–18].
Adding to this interest is research suggesting that
cancer treatments can have long-term physical
effects that may influence quality of life [16, 17,
A number of cancer-specific QOL measures
have been developed, such as the Functional
Adjustment to Cancer Therapy (FACT) ,
European Organization for Research and Treat-
ment of Cancer (EORTC) [31–33], Functional
Living Index-Cancer (FLIC) , and Cancer
Rehabilitation Evaluation System (CaRES) ,
and its short form (CaRES-SF) . These mea-
sures, however, may not be appropriate for use
with long-term survivors. They were designed to
Quality of Life Research (2005) 14: 1007–1023
? Springer 2005
capture acute effects of being newly diagnosed
with cancer and the immediate effects of surgery
and treatment. Thus, they tend to focus on
symptoms and immediate problems.
The goal of the work described here was to
develop a QOL measure that captures issues rel-
evant to long-term (greater than 5 years) cancer
survivors. Studies suggest that long-term conse-
quences of cancer include issues present after
diagnosis and treatment that linger, but also new
concerns that develop over time . Conditions
that continue after treatment are pain and fatigue
[37–39], sexual problems [21, 22, 40–42], and
appearance and body-image concerns . Psy-
chological dysfunction can also be a problem [17,
29, 43]. Newer issues that may develop include
insurance concerns, worry about the health of
children, and worry about the family’s future in
the event of recurrence [18, 39, 41, 43]. Late
physical effects of cancer treatment, such as car-
diac toxicity or development of second malig-
nancies have also been identified [44–46]. It is
important to recognize that there are also positive
aspects of surviving cancer such as transforma-
tions in the survivor’s understanding of his or her
life and positive changes in relationships with
Despite the importance of these issues for long-
term survivors, currently there are only two QOL
measures designed specifically for long-term sur-
Researchers at the City of Hope National Medical
dimensions: physical, psychological, social, and
spiritual. They developed the Quality of Life–
Cancer Survivors scale (QOL-CS) based on these
dimensions . Although this scale reflects an
attempt to recognize QOL issues relevant to long-
term cancer survivors, it has a number of limita-
tions. Items were based on a small number of
cancer survivors. Validation of the scale was based
on survivors ranging from 4 months to 28 years
after diagnosis (thus including newly diagnosed
patients). Some items have problematic wording,
in that they ask about change but fail to indicate
its direction (e.g., ‘has your illness or treatment
caused changes in your self-concept?’). The do-
mains often measure multiple constructs at once
(e.g., social interaction includes appearance, sexual
functioning, and family distress). Further, several
of the other.
items ask about distress at the time of diagnosis
and treatment .
Wyatt and colleagues [51, 52] developed the
Long-term Quality of Life (LTQL) questionnaire
based on the same conceptual model. Wyatt and
colleagues developed an item set from focus
groups of female survivors. They then collected
data from a substantial sample of female cancer
survivors and conducted a factor analysis, an
internal consistency analysis, and determined
content validity. This yielded 34 items loading on
four factors that are slightly different from the
QOL-CS: somatic concerns, spiritual/philosophi-
cal view of life, fitness, and social support. While
the psychometric approach of the LTQL is an
improvement over that used for the QOL-CS,
some of the items themselves are still problematic
and the domains often encompass more than one
important aspect of QOL. For example, the so-
matic concerns domain includes both body-image
problems and pain. Although these are both so-
matic in character, they are different in focus. The
broad domains do not allow investigators to look
at more specific QOL domains. Further, this
measure is specific to female cancer survivors.
This article describes the development of a new
measure designed to assess issues relevant to long-
term cancer survivors. We call the measure the
Quality of Lifein Adult Cancer Survivors
(QLACS). We conceptualize cancer-related quality
of life based on the definition provided by Gotay
et al.  as the state of well-being that is a com-
posite of two components: the ability to perform
everyday activities that reflect physical, psycho-
logical, and social well-being; and patient satis-
faction with levels of functioning and control of
the disease. This conceptualization takes into ac-
count both functioning and patient satisfaction
with functioning. Because of its subjective nature,
QOL is best measured from the patient’s perspec-
tive [19, 53, 54]. As such, our approach also as-
sumes that it is critical to involve patient input into
determining relevant QOL domains and items .
This approach also views QOL as a multidi-
mensional construct. At a minimum, it includes
physical, social and psychological dimensions.
Other domains sometimes included are cognitive
functioning, intimacy or sexual functioning, role
limitations and productivity , pain , eco-
nomic factors [58, 59], and an existential/spiritual
domain [18, 60]. Our approach to instrument
development begins with a deliberately broad view
of possible domains that are then confirmed
The QLACS was developed and tested in two
phases, using two different samples of respondents.
During Phase I, qualitative semi-structured inter-
views with cancer survivors were conducted to
identify relevant domains and generate a pool of
questionnaire items. In Phase II, the preliminary
version of the instrument was tested to produce a
shortened final version and to evaluate its psy-
The first task in the development process was to
identify domains that were most relevant to long-
term cancer survivors. To do this, we conducted
in-depth semi-structured interviews with a sample
of 58 long-term survivors. Participants in these
semi-structured interviews were persons living in
southeast Florida who had been diagnosed at least
5 years earlier than the study date with breast
cancer, head/neck cancer, gynecological cancer,
bladder cancer, prostate cancer, or colorectal
cancer. These cancers were selected because they
have large numbers of long-term survivors. Our
goal was to obtain at least eight interviews for each
Participants were recruited through the Florida
Cancer Data System (FCDS), which is the state’s
record-keeping system for cancer diagnoses. Ev-
ery cancer diagnosis in the state of Florida is (by
law) entered into this data base. After approval
by IRBs from both the University of Miami and
the Florida Department of Health, FCDS staff
created a set of records containing the names and
addresses of all persons in Miami-Dade and
Broward counties who had been diagnosed with
the targeted cancers in 1994 or earlier. Miami-
Dade County contains the city of Miami; Bro-
ward County contains the city of Ft. Lauderdale.
Both counties are characterized by a wide range
of different socio-economic statuses, including
both very affluent and very poor areas. Both
counties also have substantial number of minority
residents, primarily Hispanic and African Amer-
The records generated by FCDS were delivered
by computer disk to one of the senior investigators
in Coral Gables. All subsequent handling of
information from FCDS was done by a core
member of the research group, to ensure confi-
dentiality. (All procedures concerning the acqui-
sition and handling of data from FCDS apply to
both phases of data collection.)
It should be noted that (by our request) the re-
cords provided were limited to the specific cancers
noted earlier, the geographic area of southeast
Florida, and diagnosis 5 years or more in the past.
There was no further restriction. In particular, we
wanted to ensure that the length of survivorship
would not be artificially constrained in any way.
Put differently, the procedure allowed us to recruit
persons who were as little as 5 years post-diagnosis
and also persons who were as much as 18 years
post-diagnosis (which was as far back as the reg-
istry went). In effect, we permitted ourselves the
full available range of length of survivorship.
The number of records FCDS supplied was
quite large. We began by extracting a random
sample of approximately 7000 from the full set.
We separated them into diagnostic categories, then
assigned each person (within category) a com-
puter-generated random number. Each diagnostic
category was then sorted by number, and the first
150 names in that category were sent letters
describing in general terms the nature of the pro-
ject and an invitation to participate. Subsequent
mailings (approximately another 100 in total) were
directed selectively toward categories where par-
ticipants were still needed (survivors of bladder
and head-and-neck cancers were harder to recruit
than survivors of the other cancers, for reasons
that are unclear).
The letter sent to these persons stated that the
project concerned the experiences of persons who
had been treated for cancer 5 years or more in the
past and that the research team was interested in
interviewing such persons. The interview was de-
scribed as approximately 2–2.5 h in length, for
which the interviewee would receive $50 compen-
sation for his or her time. Those willing to be
interviewed returned postcards on which they
wrote their phone numbers and convenient contact
hours. Research assistants contacted persons who
returned cards, described the process in more de-
tail, and made appointments to interview those
who remained interested in participating. Inter-
views took place in the respondent’s home or other
preferred meeting place.
The sample of 58 who were eventually inter-
viewed was 55% female. Forty-one of the respon-
dents were Caucasian, nine African-American, and
eight Hispanic. The mean age was 64.9 (SD ¼ 14.5,
range from 34 to 91). The sample was distributed
among cancer types as follows: breast (n ¼ 12),
prostate (n ¼ 10), colorectal (n ¼ 11), bladder
(n ¼ 6), head and neck (n ¼ 9), and gynecologic
(n ¼ 10).
Although it may appear that the sample we re-
cruited was relatively old (mean age of 65), they
actually were relatively young for the population
from which they were drawn. The average age at
diagnosis of the full set of records extracted for
this first phase of data collection was 66.54; aver-
age age at diagnosis of those sent letters was 63.68;
average age at diagnosis of those interviewed was
57.21. The diagnosis dates for the full set of re-
cords ranged from 5 to 18 years prior to recruit-
ment; the ranges of those contacted and those
interviewed were identical. Nor were the ages of
our participants particularly unusual on a national
basis. In the year of Phase I recruitment, 14% of
persons living in Miami-Dade County and 16% of
persons living in Broward County exceeded
65 years of age; the national figure was 12.7%.
Thus, the sample of Phase 1 appears representative
of the population we sought to recruit.
The interview itself included both very general and
more specific questions. Respondents first were
asked to describe in general terms how cancer had
affected their quality of life (‘‘free’’ responses).
They were then asked more structured questions
about how cancer affected their quality of life in 17
specific domains. These 17 domains were based on
domains frequently used in QOL instruments and
a review of the literature related to long-term
survivors. The domains were physical health, basic
appearance, sexual function, relationship with
partner, relationships with family, relationships
with friends, finances, health insurance, work,
other life responsibilities, relationships with health
care providers, philosophical/spiritual outlook,
feelings about self, and positive impact of cancer.
All interviews were tape-recorded and transcribed.
This approach corresponds to the systematic dis-
covery stage recommended for instrument design
by Kessler and Mroczek .
Study investigators reviewed the transcripts and
independently listed the domains that were men-
tioned most by respondents in the ‘free’ responses.
We focused primarily on the free responses for
identifying domains to pursue further, because
these areas were salient enough to respondents
that they identified them spontaneously. Re-
sponses to the more structured questions were
used to create specific items.
After the interviews were conducted, health care
providers and members of consumer advocacy
groups were contacted to complete mailed surveys.
Surveys were completed by five oncologists, two
surgeons, a clinical psychologist, a gynecologist,
and three advocacy group members (identified
through attendees at the 1999 NCI meeting on
survivorship and other contacts). The intent here
was to obtain information from people who had
experience with a large number of cancer survivors
and who had a broad view of what long-term
survivors reported as QOL issues. These interviews
were used to ensure that the survivors we had
interviewed had covered all major issues. The only
area mentioned by providers that had not been
mentioned by survivors was concerns about fer-
tility and premature menopause. We decided not
to include this as a domain because it would be
relevant to only a subset of survivors (i.e., younger
Twelve domains were identified from the free
responses in the interviews: positive feelings, neg-
ative feelings, cognitive problems, social avoid-
ance, fatigue, problems with sexual functioning,
physical pain, benefits from cancer, appearance
concerns, financial concerns, role limitations, and
concerns about recurrence. Our a priori domains
of relationships with partner, family, friends, and
health care providers were not prominently men-
tioned by the participants. Physical health was
mentioned largely in terms of fatigue and pain.
Basic activities, work, and other life responsibili-
ties seemed to come together as Role Limitations.
New domains of Social Avoidance and Cancer
Related Distress emerged.
Many interviewees commented that things had
changed in some of these areas, but could not
determine whether these changes were due to
cancer or aging. We therefore divided domains
into what we believed to be generic domains that
could not necessarily be attributed to cancer (e.g.,
Negative Feelings, Positive Feelings, and Fatigue)
and domains that were cancer-specific, such as
Benefits of Cancer and Cancer Related Distress.
The results of Phase I were used to generate a pool
of 83 items tapping the identified domains (both
generic and cancer-specific). Both positively and
negatively worded items were prepared. To go
beyond purely objective functioning and to get at
the subjective nature of QOL, many items were
worded to measure the person’s feelings about
aspects of his or her life. Many items incorporated
phrases used by Phase I participants to express
their concerns. Items were agreed upon by all
In many instruments, respondents are asked to
rate their satisfaction with various aspects of their
lives. However, satisfaction scales are subject to
ceiling effects because most people, even those with
life-threatening illnesses, tend to be satisfied with
most areas of their lives . For this reason we
decided to use a frequency response scale, in which
respondents indicate how often they felt a certain
way in the past 2 weeks.
To simplify administration of the instrument, a
single 7-category frequency scale (ranging from
never to always) was used for responses to all
items. The category anchors were selected on the
basis of the percentages people typically associate
with various frequency terms . The anchors
(never, seldom, sometimes, about as often as not,
frequently, very often, and always) were chosen to
produce approximately equal percentage intervals
from one category to the next. We chose seven
response categories rather than five to reduce the
tendency of respondents to choose extreme scale
After the item set was created, a second sample
was recruited through FCDS. The target sample
size for the planned statistical analyses was 200.
We extracted another random sample of approxi-
mately 45000 from the list of records obtained
earlier from FCDS. As in the first sample, we
placed no constraints on time since diagnosis. As
in the first sample, the records were segregated by
diagnostic category, assigned a random number,
and sorted (within category) by those numbers.
Letters were sent to potential participants in order
of these numbers.
Those who expressed interest were sent a copy
of the questionnaire set to complete and return.
Each person completing the questionnaire set re-
ceived $50. A total of approximately 4000 initial
contact letters were sent, and 372 questionnaire
sets were mailed to interested individuals. If the
questionnaire was not returned within 4 weeks, the
participant was contacted by phone. The research
remaining a participant and encouraged him or
her to complete the questionnaires in a timely
fashion. A total of 266 questionnaire sets were
returned, but 24 of them were incomplete and were
not included in analyses. The Phase II sample thus
consisted of 242 long-term cancer survivors.
The questionnaire set that was sent to Phase II
participants consisted of the initial item pool for
the new measure, a set of socio-demographic
items, several cancer treatment questions, and a set
of scales that were selected as criterion measures
for each domain. The inclusion of criterion mea-
sures was based on the following reasoning. A
number of measures exist that assess qualities that
are conceptually similar to qualities that are rep-
resented in the domains around which our items
are written. Although those measures are not de-
signed for long-term cancer survivors, they have
relevance to health-related aspects of QOL. We
included these measures in order to provide
converging information on the adequacy of a given
item from our item set as a reflection of the
domain it was intended to reflect. We chose these
subscales rather than scales from existing measures
of long-term QOL in order to maximize the extent
to which the content of the criterion scale focused
on the content domains of interest. Subscales from
instruments such as the WHOQOL , Sickness
Impact Profile , MOS Mental Health Index
, and Post-Traumatic Growth Inventory 
were chosen as criterion measures for this purpose,
separately for each domain. Respondents also
rated their overall quality of life by marking a
100 mm visual analog scale ranging from worst
possible to best possible QOL.
Domain and item selection
Our approach to item selection for the QLACS
used a two-pronged strategy. First, factor analyses
were performed separately for the 51 generic and
32 cancer-specific items to test the hypothesized
factor structure and determine which items loaded
best on the factors. Due to the large number of
items and factors, three separate factor analyses
were performed for conceptually similar sets of
items. Recent simulations have shown that sample
sizes of 200 are more than adequate to accurately
recover population factors and obtain unbiased
estimates of factor loadings when communalities
are high and there are at least three indicators per
factor . Second, we selected the four items that
explained the most variance in the criterion mea-
sure for each domain using stepwise regression.
This criterion-based procedure ensures that the
selected items measure different facets of each
domain while sacrificing only a small degree of
internal consistency [55, 69]. Finally, the factor
analysis and criterion-based results were compared
to select the final set of items.
The validity of individual domains was assessed
by correlations between domain scores and the
criterion instruments and visual analog ratings.
Internal consistency among the items constituting
each domain was assessed by Cronbach’s a. Scales
are generally considered to be reliable if a exceeds
Sample characteristics are shown in Table 1. The
mean age at interview was 71.4 years (SD ¼ 11.5,
Table 1. Characteristics of Phase II sample (N = 242)
Variablen % (Mean, SD)
Age (mean, SD)
71.4 (11.5) years
Doing something else
Living with partner
or spouse (%)
Years since diagnosis
235 12.6 (7.0)
Head and neck
aThese numbers added up to be greater than 242 because some
people reported multiple cancers.
range from 29 to 92 years), and respondents were
interviewed an average of 12.6 years after first
diagnosis (SD ¼ 7.0, range from 5 to 67 years).
Women comprised over half (58%) of the sample
and there were small percentages of minority
American, and 0.4% Other). 28% of the sample
reported multiple types of cancer. Several cases
occurred in which the cancer registry code for
cancer type differed from self-reported cancer type.
We re-contacted those persons to verify the
information. There were six cases, however, in
which the type of cancer could not be definitively
determined. These cases were excluded from any
subsequent analyses involving cancer type.
Nearly all subjects had had some form of can-
cer surgery (92%), with the lowest surgery rate
(64%) among the prostate group. Almost half of
the sample (48.7%) had undergone radiation
therapy and 23.7% had chemotherapy. Hormone
therapy, the least common treatment (12%), was
reported only for breast, gynecologic, and pros-
As in the Phase I sample, we examined how
representative the sample was of the population
from which it was drawn. The average age at
diagnosis was 65.2 years for the full set of records
extracted for this phase of data collection,
62.2 years for those sent letters, and 60.3 years for
the 242 subjects who formed our sample. Diag-
nosis dates for the full set of records ranged from 5
to 18 years prior to recruitment; the same was true
for persons sent letters and those in the final
sample. Thus, the Phase II sample also appears
representative of the population of long-term
cancer survivors. However, Phase II subjects, on
average, had survived somewhat longer than Phase
Domain measurement and item selection
Factor analyses reproduced the hypothesized
generic domains, with each item having its highest
loading on its intended factor (see Table 2). The
Sexual Problems domain split into correlated
interest and functioning factors, but the criterion-
based approach yielded a single domain combining
two interest and two functioning items.
Factor loadings for the cancer-specific items
differed in two ways from the expected structure.
First, the items reflecting Role Limitations sepa-
rated, loading across several domains. Because of
the lack of cohesion of the Role Limitations items,
this domain was eliminated from further consid-
eration. Second, the cancer distress items split into
two dimensions: family-related distress and fears
about recurrence. Because of this separation, sep-
arate domains were created for family distress and
distress about fears of recurrence.
Table 3 shows the correlations between domain
scores and the criterion measures. These correla-
tions were 0.72 or higher in all of the generic
domains but one (Social Avoidance). The corre-
lations were somewhat lower for the cancer-spe-
cific scales, which is not surprising, given that these
are the areas where criterion measures less closely
match the intended domains.
The final version of the QLACS consists of 47
items measuring 12 domains; seven are considered
Generic and five are Cancer-specific (the instru-
ment is in the Appendix). The Flesch Reading
Ease score for these items was 74.4 which is
equivalent to a Flesch–Kincaid reading level of 4.8
. Scores for each domain are the sum (after
appropriate reverse scoring) of the individual item
scores (1 ¼ ‘‘Never’’ through 7 ¼ ‘‘Always’’). Do-
main scores may therefore range from 4 to 28
points, with higher scores representing more
problems or poorer quality of life. Table 4 shows
the domain score means, standard deviations, al-
pha reliabilities, score ranges, and the percentage
of scores at the minimum (floor) or maximum
(ceiling) values. Cronbach’s a exceeded 0.71 for
each domain. This sample displayed substantial
floor effects only for appearance concerns and
QLACS summary scores
Correlations among domain scores are shown in
Table 5. Most domains were moderately corre-
lated with the others. An oblique factor analysis of
this matrix produced factors for the Generic
domains, Cancer Benefits, and the remaining
Cancer-Specific domains. Based on these results,
we created separate Generic and Cancer-Specific
summary scores by adding their constituent do-
main scores (omitting Cancer Benefits since it did
not load with the other domains). The summary
scales are scored so that higher scores represent
more problems or lower QOL. In our sample, the
mean generic summary score was 71.2 (SD=25.6)
Table 2. Factor loadings for (a) generic (N=204) and (b) cancer-specific (N=228) domains
Item Factor loadings
19 Bothered by mood swings
7 Felt blue or depressed
9 Worried about little things
24 Felt anxious
8 Enjoyed life
28 Content with life
6 Felt happy
22 Had a positive outlook on life
3 Bothered by having a short
4 Had trouble remembering things
2 Difficulty doing things requiring
23 Bothered by forgetting what
started to do
13 Bothered by pain preventing
17 Mood disrupted by pain or its
27 Pain interfered w/social activities
21 Had aches or pains
)0.187 0.4520.170 0.289
16 Lacked interest in sex
26 Avoided sexual activity
11 Lacked energy to do things wanted to
14 Felt tired a lot
1 Had energy to do things wanted to do
5 Felt fatigued
12 Dissatisfied w/sex life
10 Bothered by inability to function
18 Avoided social gatherings
20 Avoided friends
25 Reluctant to meet new people
15 Reluctant to start new
and the mean for the Cancer-Specific summary
was 38.0 (SD ¼ 17.4).
As an additional test of validity, correlations
were computed between the domain and summary
scores and respondent ratings of their overall QOL
using the visual analog scale (Table 6). The mean
visual analog score was 74.8 (SD ¼ 23.7). The
correlations were highest for the Positive Feelings,
Fatigue, and Social Avoidance domains. The
lowest correlations were for the Cancer-Specific
scales. As a result, the Generic Summary score was
more strongly associated with the analog ratings
(r ¼ )0.57) than was the Cancer-Specific summary
(r ¼ )0.24).
Domain scores by cancer type
Comparisons among the six types of cancer yiel-
ded statistically significant differences for five do-
Table 2. Continued
Item Factor loadings
43 Had money problems from cancer
45 Financial problems from loss of income
due to cancer
30 Financial problems from cost of cancer
surgery or tx
37 Problems with insurance because of cancer
0.8290.130 0.2690.360 0.411
40 Cancer helped recognize what important
41 Better able to deal w/stress because
32 Cancer helped cope better w/problems
29 Appreciated life more because of cancer
0.209 0.8050.2590.121 0.265
34 Worried whether family had cancer
31 Worried family members were at
risk for cancer
42 Worried family should have genetic
tests - cancer
0.326 0.052 0.8630.2120.450
35 Felt unattractive b/c of cancer or
33 Self-conscious about appearance
because of cancer
44 Felt treated differently b/c
of changes in appearance
38 Bothered by hair loss from cancer tx
0.532008 0.1900.678 0.278
0.3340.122 0.1500.435 0.367
39 Worried about cancer coming back
46 When felt pain, worried it was
36 Worried about dying from cancer
47 Preoccupied with concerns about cancer
mains (Table 7). To avoid confounding by multi-
ple cancers, these contrasts were made only among
participants reporting only one type of cancer (n
= 169). For most domains, respondents with
bladder or head and neck cancer had the least
favorable scores, while those with prostate cancer
had the most favorable scores. There were some
differences, however, across domains. For the
Sexual Problems domain, the scores were highest
for bladder and prostate cancer and lowest for
colorectal cancer. For Social Avoidance, scores
were highest for head and neck and bladder cancer
and lowest for colorectal cancer. There were highly
significant effects for cancer type on the appear-
ance and financial domains, with head and neck
cancer having the highest scores. Analysis of the
Generic summary score just failed to reach statis-
tical significance, with bladder and head/neck
cancer tending to have the highest (least favorable)
scores and prostate and colorectal cancer having
the lowest (most favorable) scores. For the Can-
cer-Specific summary score, the effect of cancer
type was highly significant. The study had 81%
power to detect effect sizes of 0.30 or greater
(a ¼ 0.05) for differences between the six cancer
types. These are slightly larger than what are
usually considered to be medium effect sizes.
Researchers and clinicians are increasingly aware
of the importance of quality of life issues among
long-term cancer survivors [16, 17, 19–29]. Quality
of life assessments can be an important indicator
of success of treatment, can help identify long-
term sequelae clinicians should monitor, and can
identify areas where services or interventions are
needed . However, few instruments are cur-
rently available that adequately capture a wide
range of QOL domains relevant to long-term
survivors. This article described development of
Table 3. Correlations between domain scores and criterion
Domain Correlation Criterion measure
WHOQOL Sex func-
SIP Social interaction
Social avoidance 0.62
)0.61 WHOQOL body im-
PTGI Total score
Benefits of cancer
Table 4. Domain means, standard deviations, ranges, percent at minimum and maximum, and a reliabilities
N Mean (SD)Range
Cancer specific domains
Distress – recurrence
Distress – familya
aThis domain consists of three items that are rescaled to make the metric comparable to the other domains.
the QLACS, a multidimensional measure designed
for long-term cancer survivors. The QLACS
measures 12 QOL domains, seven of which are
generic and five cancer-specific.
The generic item sets do not mention cancer and
therefore are also applicable to non-cancer popu-
lations. These domains lend themselves for use in
comparing cancer survivors with healthy people or
other disease populations, without having to use
multiple measures. The five cancer-specific do-
mains ask questions directly related to having
assessing the impact of interventions on long-term
QOL and for comparing QOL among cancer pa-
tients. Psychometric properties of the QLACS
suggest that the domains have good internal con-
sistency and good validity.
problems compared to other cancer types. Treat-
tohaveanegativeimpactonsexual functioning [40,
cancers and bladder cancer. Perhaps, because head
and neck cancer survivors have socially observable
in social situations. We should also note that for
gender-specific cancers (e.g., breast and prostate)
cancer type is confounded with gender. Larger
tease out the specific effects of gender.
In general, the most frequently reported problem
concerns, and fatigue. Long-term effects of fatigue
and sexual problems have been reported as impor-
tant issues in long-term survivors by others [21, 22,
37–42, 72]. Concerns about recurrence were also
Table 5. Correlations among domain scores (N = 224)
Table 6. Pearson correlations between domain scores and
visual analogue scale ratings of quality of life
Generic Summary score
Cancer-Specific Summary score
N = 216–240.
All correlation significantly different from zero (p < 0.05)
except for Family distress.
rated high in the present sample (as has also been
found within the first year after treatment .
Domains reported least frequently were Financial
Problems and problems with Appearance.
We found significant differences across cancer
types for the cancer-specific summary score, but
smaller differences for the generic summary score,
suggesting that the cancer-specific domains are
more sensitive to the type of cancer. This finding
highlights the advantage of including cancer-spe-
cific measures in assessing quality of life among
long-term survivors. Indeed, the different patterns
of scores across cancer types for different cancer-
specific domains support the desirability of multi-
dimensional measures, since summary scores do
not convey these differences . These compari-
sons should be viewed with some caution, however,
as the group sizes of each type were relatively small.
We also included a Benefits domain, which
received a high mean frequency score. This rein-
forces prior findings that people often find some
benefit from having had cancer [47–50, 73, 74].
Interestingly, Benefits scores had only small cor-
relations with other generic or cancer-specific
domains. For that reason, this domain was not
included in either summary score. These results
suggest that perceived benefits are a unique aspect
of survivorship that do not relate well to other
aspects of long-term physical and psychosocial
We should note some limitations to this work.
We intentionally recruited through a state registry
rather than through a cancer survivor support or
advocacy group, recognizing that such groups are
not likely to be representative of cancer survivors in
general. Nonetheless, participation in the research
involved some self-selection. Although recruitment
of participants began with a very wide solicitation,
for a person who received a letter from us to
actually become a participant required that person
to take the next step of contacting us. For the first
sample, there was an additional barrier to partici-
pation, because we asked respondents to indicate a
good time to call them by phone; anyone wishing to
be interviewed who did not have a phone number
to offer would have had greater difficulty in par-
ticipating than others. The study also required
potential participants to return a postcard, and
actual participants to spend approximately 2 h
reporting on their well-being. Given these consid-
erations, our sample may have been more cooper-
ative than a truly random sample of cancer
survivors. This limitation is common in research on
medical populations. Without recruiting people
who are willing to spend the time, it is difficult to
obtain large amounts of information.
Table 7. Mean (SD) domain scores by cancer type
(N = 44)
(N = 18)
(N = 20)
(N = 34)
(N = 31)
(N = 22)
Distress-recurrance 11.7 (5.8)
Potential issues of generalizability are also
raised by the fact that all participants came from
South Florida (i.e. their cancer diagnoses took
place in the state of Florida). It would be desirable
and valuable to have further information on sur-
vivors recruited from a wider geographical area
which is one goal for future investigation.
Another limitation is that the samples were
skewed towards older persons. This skew toward
the older also characterizes the population from
which we drew the samples. Comparison of par-
ticipant samples to the larger sets of records from
which the mailing lists were generated revealed
that those who participated were in fact a bit
younger than the two comparison groups (the full
set of records extracted, and those to whom letters
were sent). The age of the samples reflects the fact
that the cancers we studied are more common
among older than younger persons, and the fact
that we did not place any restriction on length of
survivorship at the time of data collection. To have
placed such a restriction would have created a bias
of a different type.
In conclusion, we offer researchers who are
interested in psychosocial well-being among long-
assessment tool for use in their investigations.
This instrument, the QLACS, covers five cancer-
specific areas that were suggested by long-term
survivors themselves as relevant to their lives,
along with seven additional areas that are rele-
vant to cancer but are not limited to cancer. The
QLACS therefore covers more areas relevant to
the cancer survivorship experience than do other
instruments such as the QOL-CS  and the
LTQL [51, 52]. It was developed in two phases,
using nearly equal numbers of men and women,
who were survivors of several different types of
cancers, and whose survivorship ranged from
recent (5 years) to prolonged (18 years). Its items
are clearly and unambiguously focused on spe-
cific effects in the domains that they represent.
We hope that this instrument will prove to be of
value to other researchers who are trying to
reach a better understanding of what factors
contribute to quality of life among long-term
cancer survivorsa multidimensional
This research was supported by grant no. R01
CA78995 from the National Cancer Institute.
Quality of Life in Adult Cancer Survivors Scale
Instructions: We’d like to ask you about some things that can affect the quality of people’s lives. Some of these questions may sound
similar, but please be sure to answer each one. Below is a scale ranging from never to always. Please indicate how often each of these
statements has been true for you in the past four weeks. [CIRCLE ONE ANSWER FOR EACH QUESTION]
NeverSeldom SometimesAbout as
often as not
Frequently Very often Always
In the past 4 weeks…
1. You had the energy to do the things
you wanted to do.
2. You had difficulty doing activities
that require concentrating.
3. You were bothered by having a
short attention span.
4. You had trouble remembering things.
5. You felt fatigued.
6. You felt happy.
7. You felt blue or depressed.
8. You enjoyed life.
9. You worried about little things.
Appendix A Continued
Never Seldom SometimesAbout as
often as not
10. You were bothered by being
unable to function sexually.
11. You didn’t have energy to do the
things you wanted to do.
12. You were dissatisfied with your sex life.
13. You were bothered by pain that kept
you from doing the things you
wanted to do.
14. You felt tired a lot.
15. You were reluctant to start new
16. You lacked interest in sex.
17. Your mood was disrupted by pain
or its treatment.
18. You avoided social gatherings.
19. You were bothered by mood swings.
20. You avoided your friends.
21. You had aches or pains.
22. You had a positive outlook on life.
23. You were bothered by forgetting
what you started to do.
24. You felt anxious.
25. You were reluctant to meet new people.
26. You avoided sexual activity.
27. Pain or its treatment interfered
with your social activities.
28. You were content with your life.
The next set of questions asks specifically about the effects of your cancer or its treatment. Again, for each statement, indicate how
often each of these statements has been true for you in the past four weeks.
29. You appreciated life more because
of having had cancer.
30. You had financial problems because
of the cost of cancer surgery or treatment.
31. You worried that your family members
were at risk of getting cancer.
32. You realized that having had cancer helps
you cope better with problems now.
33. You were self-conscious about the way
you look because of your cancer or its
34. You worried about whether your family
members might have cancer-causing genes.
35. You felt unattractive because of your
cancer or its treatment.
36. You worried about dying from cancer.12
37. You had problems with insurance because
38. You were bothered by hair loss from
39. You worried about cancer coming back.12
40. You felt that cancer helped you to
recognize what is important in life.
41. You felt better able to deal with stress
because of having had cancer.
1. Bottomley A. The cancer patient and quality of life. The
Oncologist 2002; 7: 120–125.
2. Gotay CC, Korn EL, McCabe MS, Moore TD, Cheson
BD. Quality of life assessment in cancer treatment proto-
cols: Research issues in protocol development. J Natl
Cancer Inst 1992; 84: 575–579.
3. Tanaka T, Gotay CC. Physicians’ and medical students’
perspectives on patients’ quality of life. Acad Med 1998; 73:
4. Vinokur AD, Threatt BA, Vinokur-Kaplan D, Satariano
WA. The process of recovery from breast cancer for
younger and older patients. Changes during the first year.
Cancer 1990; 65: 1242–1254.
5. McDonough EM, Varvares MA, Dunphy FR, Dunleavy
T, Dunphy CH, Boyd JH. Changes in quality-of-life
scores in a population of patients treated for squamous
cell carcinoma of the head and neck. Head Neck 1996; 18:
6. Ganz PA, Coscarelli A, Fred C, Kahn B, Polinsky ML, Pet-
ersen L. Breast cancer survivors: Psychosocial concerns and
quality of life. Breast Cancer Res Treat 1996; 38: 183–199.
7. Andrykowski MA, Bruehl S, Brady MJ, Henslee-Downey
PJ. Physical and psychosocial status of adults one year after
bone marrow transplantation: A prospective study. Bone
Marrow Transplant 1995; 15: 837–844.
8. Padilla GV, Grant MM, Lipsett J, Anderson PR, Rhiner
M, Bogen C. Health quality of life and colorectal cancer.
Cancer 1992; 70: 1450–1456.
9. Parker PA, Baile WF, deMoor C, Cohen L. Psychosocial
and demographic predictors of quality of life in a large
sample of cancer patients. Psycho-Oncology 2003; 12:
10. Osoba D. (ed.) The Effect of Cancer on Quality of Life.
Boca Raton: CRC press Inc., 1991.
11. Glanz K, Lerman C. Psychological impact of breast cancer:
A critical review. Ann Behav Med 1992; 14: 204–212.
12. de Haes JC, van Knippenberg FC. The quality of life of
cancer patients: A review of the literature. Soc Sci Med
1985; 20(8): 809–817.
13. Conroy T, Blazeby JM. Health-related quality of life in
colorectal patients. Expert Rev Anticancer Ther 2003; 3(4):
14. Parker SL, Tong T, Bolden S, Wingo PA. Cancer Statistics
1997. CA Cancer J Clin 1997; 47: 5–27.
15. Office of Cancer Survivorship website, http://cancercon-
16. American Society of Clinical Oncology. Outcomes of can-
cer treatment for technology assessment and cancer treat-
ment guidelines. J Clin Oncol 1996; 14: 671–679.
17. Gotay CC, Muraoka M. Quality of life in long-term sur-
vivors of cancer. J Natl Cancer Inst 1998; 90: 656–667.
18. Ferrell BR, Hassey Dow K, Grant M. Measurement of the
quality of life in cancer survivors. Qual Life Res 1995; 4:
19. Moinpour CM, Feigl P, Metch B, Hayden KA, Meyskens
FL Jr, Crowley J. Quality of life end points in cancer
clinical trials: Review and recommendations. J Natl Cancer
Inst 1989; 81: 485–495.
20. Nayfield SG, Ganz PA, Moinpour CM, Cella DF, Hailey
BJ. Report from a National Cancer Institute (USA)
workshop on quality of life assessment in cancer clinical
trials. Qual Life Res 1992; 1: 203–210.
21. Dorval M, Maunsell E, Deschenes L, Brisson J, Masse B.
Long-term quality of life after breast cancer: Comparison
of 8-year survivors with population controls. J Clin Oncol
1998; 16: 487–494.
22. Ganz P, Desmond KA, Leedham B, Rowland JH, Meye-
rowitz BE, Belin TR. Quality of life in long-term, disease-
free survivors of breast cancer: A follow-up study. J Natl
Cancer Inst 2002; 94: 39–49.
Appendix A Continued
NeverSeldomSometimes About as
often as not
FrequentlyVery often Always
42. You worried about whether your family
members should have genetic tests for
43. You had money problems that arose
because you had cancer.
44. You felt people treated you differently
because of changes to your appearance
due to your cancer or its treatment.
45. You had financial problems due to a loss
of income as a result of cancer.
46. Whenever you felt a pain, you worried
that it might be cancer again.
47. You were preoccupied with concerns
23. Mosconi P, Apolone G, Barni S, Secondino S, Sbanotto
A, Filiberti A. Quality of life in breast and colon cancer
long-term survivors: An assessment with the EORTC
QLQ-C30 and SF-36 questionnaires. Tumori 2002; 88:
24. Steer CB, Marx GM, Galani E, Harper PG, Khayat D.
Quality of life: It’s never too late. J Clin Oncol 2002; 20:
25. Wyatt G. Kurtz ME, Friedman LL, Given B, Given C.
Preliminary testing of the long-term quality of life (LTQL)
instrument for female cancer survivors. J Nurs Measure
1996; 4 (2): 153–169.
26. Ganz PA. Late effects of cancer and its treatment. Semin
Oncol Nurs 2001; 17(4): 241–248.
27. Bloom JR. Surviving and thriving? Psycho-Oncology 2002;
28. Tomich PL, Helgeson VS. Five years later: A cross-sec-
tional comparison of breast cancer survivors with healthy
women. Psycho-Oncology 2002; 11: 154–169.
29. Deimling GT, Kahana B, Bowman KF, Schaefer ML.
Cancer survivorship and psychological distress in later life.
Psycho-Oncology 2002; 11: 479–494.
30. Cella DF, Tulsky DS, Gray G, et al. The Functional
Assessment of Cancer Therapy (FACT) Scale: Develop-
ment and validation of the general measure. J Clin Oncol
1993; 11: 570.
31. Borghede G, Sullivan M. Measurement of quality of life in
localized prostatic cancer patients treated with radiother-
apy; development of a prostate cancer-specific module
supplementing the EORTC QLQ-C30. Qual Life Res 1996;
32. Osoba D, Aaronson N, Zee B, Sprangers M, te Velde A.
Modification of the EORTC QLQ-C30 (version 2.0) based
on content validity and reliability testing in large samples of
patients with cancer. Qual Life Res 1997; 6: 103–108.
33. Ringdal GI, Ringdal K. Testing the EORTC quality of life
questionnaire on cancer patients with heterogeneous diag-
noses. Qual Life Res 1993; 2: 129–140.
34. Schipper H, Clinch J, McMurray A, Levitt M. Measuring
the quality of life of cancer patients: The functional living
index-cancer: Development and validation. J Clin Oncol
984; 2: 472–783.
35. Schag CAC, Heinrich RL. Development of a comprehen-
sive quality of life measurement tool CARES. Oncology
1990; 4: 135–138.
36. Schag CAC, Ganz PA, Heinrich RL. Cancer Rehabilitation
Evaluation System-Short Form CARES-SF: A cancer
specific rehabilitation and quality of life instrument. Cancer
1991; 68: 1406–1413.
37. Berglund G, Bolund C, Fornander T, Rutqvist LE, Sjoden
PO. Late effects of adjuvant chemotherapy and postoper-
ative radiotherapy on quality of life among breast cancer
patients. Eur J Cancer 1991; 27: 1075–1081.
38. Bjordal K, Mastekaasa A, Kaasa S. Self-reported satisfac-
tion with life and physical health in long-term cancer sur-
vivors and a matched control group. Eur J Cancer 1995; 31:
39. Dow KH, Ferrell BR, Leigh S, Ly J, Gulasekaram P. An
evaluation of the quality of life among long-term survivors
of breast cancer. Breast Cancer Res Treat 1996; 39: 261–
40. Litwin MS, Hays RD, Fink A, et al. Quality-of-life out-
comes in men treated for localized prostate cancer. JAMA
1995; 273: 129–135.
41. Schag CAC, Ganz PA, Wing DS, Sim MS, Lee JJ. Quality
of life in adult survivors of lung, colon and prostate cancer.
Qual Life Res 1994; 3: 127–141.
42. Wei JT, Dunn RL, Sandler HM, et al. Comprehensive
comparison of health-related quality of life after contem-
porary therapies for localized prostate cancer. J Clin Oncol
2002; 20: 557–566.
43. Wyatt G, Friedman LL. Long-term female cancer survi-
vors: Quality of life issues and clinical implications. Cancer
Nurs 1996; 19: 1–7.
44. Tucker MA. Secondary cancers, Cancer Principles and
Practice of Oncology, 4th (ed.) 1993: 2407–2416.
45. Partridge AH, Burstein HJ, Winer EP. Side effects of che-
motherapy and combined chemohormonal therapy in wo-
men with early-stage breast cancer. J Natl Cancer Inst Mon
2001; 30: 135–142.
46. Burstein HJ, Winer EP. Primary care for survivors of breast
cancer. N E J Med 2000; 343: 1086–1094.
47. Wyatt G, Kurtz ME, Liken M. Breast cancer survivors: An
exploration of quality of life issues. Cancer Nurs 1993; 16:
MA. Posttraumatic growth following breast cancer: A con-
trolled comparison study. Health Psych 2001; 20: 176–185.
49. Antoni MH, Lehman JM, Kilbourn KM, et al. Cognitive-
behavioral stress management intervention decreases the
prevalence of depression and enhances benefit finding
among women under treatment for early-stage breast can-
cer. Health Psych 2001; 20: 20–32.
50. HelgesonVS, Snyder P, Seltman H. Psychological and phys-
ical adjustment to breast cancer over 4 years: Identifying
51. Wyatt G, Kurtz ME, Friedman LL, Given B, Given CW.
Preliminary testing of the long-term quality of life instru-
ment (LTQL) for female cancer survivors. J Nurs Meas
1996; 4: 153–170.
52. Wyatt G, Donze LF. The Long-term quality of life
instrument for female cancer survivors. In: O. Strickland
(ed.), Measurement of Nursing Outcomes. 2nd (ed.) New
York: Springer, 2003: 216–225.
53. Donovan K, Sanson-Fisher RW, Redman S. Measuring
quality of life in cancer patients. J Clin Oncol 1989; 7: 959–
54. Nayfield S, Hailey BJ, McCabe M. Quality of life assess-
ment in cancer clinical trials. Report of the Workshop on
Quality of Life Research in Cancer Clinical Trials, July 16–
17. Bethesda, MD: US Department of Health and Human
55. Kessler RC, Mroczek DK. Measuring the effects of medical
interventions. Med Care 1995; 33: AS109–AS119.
56. Shumaker SA, Anderson RT, Czajkowski SM. Psycholog-
ical tests and scales. In: Spilker B (ed.), Quality of Life
Assessment in Clinical Trials. New York: Raven Press,
57. Stewart AL, Hays RD, Ware JE Jr. The MOS short-form Download full-text
general health survey: Reliability and validity in a patient
population. Med Care 1988; 26: 724–735.
58. Ferrans C. Quality of life: Conceptual issues. Semin Oncol
Nurs. 1990 Nov 6(4): 248–254.
59. Padilla GV, Presant C, Grant MM, Metter G, Lipsett J,
Heide F. Quality of life index for patients with cancer. Res
Nurs Health 1983; 6(2): 117–126.
60. Peterman AH, Fitchett G, Brady MJ, Hernandez L,
Cella D. Measuring spiritual well-being in people with
cancer. The Functional assessment of chronic illness ther-
apy-spiritual well-being scale (FACIT-Sp). Ann Behav Med
2002; 24(1): 49–58.
61. Patrick DL, Erickson P. Health Status and Health Policy:
Allocating Resources to Health Care. New York: Oxford
Press, 1993: 100–103.
62. Budescu DV, Wallstein TS. Consistency in interpretation of
probability phases. Org Behav and Hum Dec Proc 1985; 36:
63. Alwin DF, Krosnick JA. The reliability of attitudinal sur-
vey measures: The role of question and respondent attri-
butes. Soc Meth and Res 1991; 20: 139.
64. Power M, Harper A, Bullinger M. The World Health
Organization WHOQOL-100: Tests of the universality of
quality of life in 15 different cultural groups worldwide.
Health Psychol 1999; 18: 495–505.
65. Bergner M, Bobbitt RA, Carter WB, Gilson BS. The
Sickness Impact Profile: Development and final revision of
a health status measure. Med Care 1981; 19: 787–805.
66. Stewart AL, Ware JE Jr, Sherbourne CD, Wells KB. Psy-
chological distress/well-being and cognitive functioning
measures. In Stewart AL, Ware JE Jr, (eds.), Measuring
Functioning and Well-being: The Medical Outcomes Study
Approach. Durham: Duke University Press, 1992: 102–142.
67. Tedeschi RG, Calhoun LG. The post-traumatic growth
inventory: Measuring the positive legacy of trauma. J
Traumatic Stress 1996; 9: 455–471.
68. MacCallum RC, Widaman KF, Preacher KJ, Hong S.
Sample size in factor analysis: The role of model error.
Multivariate Behav Res 2001; 36: 611–637.
69. Shrout PE, Yager TJ. Reliability and validity of screening
scales: Effect of reducing scale length. J Clin Epidemiol
1989; 43: 69–78.
70. Nunnally JC, Bernstein IH. Psychometric Theory. 3rd ed.
New York: McGraw-Hill, 1994.
71. Flesch R. A new readability yardstick. J Psychol Res 1948;
72. Holzner B, Kemmler G, Kopp M, et al. Quality of life in
breast cancer patients – not enough attention for long-term
survivors? Psychosomatics 2001; 42: 117–123.
73. Spencer SM, Lehman JM, Wynings C, Arena PE, Carver
CS, Antoni MH, Derhagopian RP, Ironson G, Love N.
Concerns about breast cancer and relation to psychosocial
well-being in a multiethnic sample of early stage patients.
Health Psychol 1999; 18: 159–168.
74. Ganz PA, Rowland JH, Desmond K, Meyerowitz BE,
Wyatt GE. Life after breast cancer: Understanding wo-
men’s health-related quality of life and sexual functioning.
J Clin Oncol 1998; 16: 501–514.
Address for correspondence: Nancy E. Avis, Wake Forest Uni-
versity School of Medicine, Department of Public Health Sci-
ences, Section on Social Sciences and Health Policy, Piedmont
Plaza II – 2nd floor, Winston-Salem, NC, 27157-1063, USA
Phone: +1-336-716-6974; Fax: +1-336-716-7554