You are on page 1of 9

Journal of Advanced Nursing, 1999, 29(2), 490±498 Issues and innovations in nursing practice

Quality of life in newly diagnosed cancer

T. Rustoen RN PhD
Associate Professor, Faculty of Nursing, Oslo College, Oslo

T. Moum PhD
Professor, Department of Behavioural Sciences in Medicine, University of Oslo,

I. Wiklund PhD
Department of Public Health and Primary Health Care, Division for Nursing
Sciences, University of Bergen, Bergen

and B.R. Hanestad RN PhD

Director of Educational Affairs, Bergen College, Bergen, Norway

Accepted for publication 11 February 1998

RUSTOEN T., MOUM T., WIKLUND I. & HANESTAD B.R. (1999) Journal of Advanced
Nursing 29(2), 490±498
Quality of life in newly diagnosed cancer patients
The aim of this study was to examine which domains of quality of life are most
strongly affected in patients with newly diagnosed cancer, and to examine if quality
of life was related to gender, age, educational level, cohabitation, time since
diagnosis, treatment or type of cancer in these patients. Ferrans and Powers
Quality of Life Index and the Cancer Rehabilitation Evaluation System, short form,
were completed by 131 newly diagnosed Norwegian adult cancer patients. The
patients had different cancer diagnoses, with breast cancer being the most
common type. None of them were terminally ill, and 80% had been diagnosed
within the last 6 months. The patients turned out to be satis®ed with their lives,
especially with the family and marital domains. They were least satis®ed with
health, functioning and sex life. Analyses of variance and multiple regression
analysis were conducted to examine the impact on quality of life of cohabiting,
age, gender, educational level, cancer diagnosis, treatment and time since
diagnosis. Those cohabiting had signi®cantly higher quality of life compared with
those living alone. In contrast the younger group
(19±39 years) living alone had signi®cantly lower quality of life than the older
groups living alone. Although age was only signi®cantly associated with quality of
life in one sub-scale, the elderly people reported their quality of life to be better in
almost all sub-scales. Time since diagnosis was not associated with quality of life,
while treatment was associated with quality of life as measured by Ferrans and
Powers Quality of Life Index. Gender and educational level were only associated
with one or two domains in quality of life, respectively.

Keywords: cancer, quality of life, nursing, demographic variables

Correspondence: T. Rustoen, Faculty of Nursing, Oslo College,
Bjerregaardsgt. 21, P.O. Box 86, 0172 Oslo, Norway.

490 Ó 1999 Blackwell Science Ltd

Issues and innovations in nursing practice Quality of life

INTRODUCTION subjective, includes both positive and negative facets of life

and is multidimensional (The WHOQOL group 1995). Even
There is a growing interest in quality of life (QOL) research
if objective factors such as vomiting in¯u-ence one's QOL,
in nursing, especially in relation to cancer patients (Padilla
the perceived importance or in what way it impacts on QOL
et al. 1992, Rustùen 1995). The main questions that have
are individual (Ferrans 1990). To get the correct picture of
been studied are what QOL entails, how to measure QOL
an individual's QOL, the individual's perception about both
and how cancer can threaten an individual's QOL. Since
positive and negative dimensions must be addressed (The
nurses are traditionally con-cerned about how to make
WHOQOL group 1995).
patients comfortable and enhance their well-being, another
focus of interest is how nurses can restore or increase an
After reviewing the literature on QOL (Rustùen 1995) we
individual's QOL in his or her daily work.
chose to base the study on Ferrans' (1990 p. 15) de®nition
of QOL: `a person's sense of well-being that stems from
There are very few studies of the impact of cancer on a
satisfaction or dissatisfaction with the areas of life that are
patient's life as a whole among Norwegian cancer patients,
important to him/her'. This de®nition em-phasizes the
and association between demographic variables and QOL
subjective perspective and takes into account both
among cancer patients has rarely been focused on. In the
satisfaction and dissatisfaction. This de®nition of QOL
general population demographic variables such as gender,
further comprises four underlying domains: a health/
age and educational level have not been shown to have a
functioning domain, a socio-economic domain, a psycho-
strong impact on QOL (Mastekaasa et al. 1988). Some
logical/spiritual domain and a family domain (Ferrans 1990).
studies have found that women experience lower QOL than
Thus QOL is de®ned as a multidimensional con-struct
men in primary care (Linzer et al. 1996) and among cardiac
covering all aspects of life. Cancer can affect all these
rehabilitation patients (Deshotels et al. 1995, Loose &
domains, and in order to maintain or improve QOL the
Fernhall 1995). There are few studies investigat-ing
nurse must bear all four in mind when dealing with cancer
whether this also applies to cancer patients. Ganz et al.
(1992) found a weak positive relationship between age and
As there is no `gold-standard' instrument to measure
QOL in women with breast cancer. The signi®-cance of
QOL, one should carefully examine the underlying theo-
social support for well-being and social func-tioning is often
retical framework that supports the questionnaire in order to
stressed in cancer patients (Krishnasamy 1996), and most
select an appropriate and relevant questionnaire (Padilla
studies conclude that satisfaction with family life and
1993). QOL being a subjective matter, the most appropriate
marriage is one of the most important precipitators of QOL
method for collecting QOL data relies on self-report from
(Mastekaasa et al. 1988). We there-fore decided to
the patients themselves, typically using self-administered
examine this area further, and since we are dealing with
scales (Aaronson 1990). Furthermore, it is recommended
cancer patients, we also decided to look for differences in
that QOL is addressed at a variety of levels in order to give
QOL related to time since diagnosis, differ-ent types of
information about as many relevant dimensions as possible.
cancer and treatment.
In spite of the value of global approaches, important
This study is an attempt to answer the following research
information about the patients can be masked if one only
uses a global approach. There is no established consensus
What are the speci®c domains of QOL most strongly as to what constitutes the key dimensions in QOL, but there
affected in patients with newly diagnosed cancer? seems to be a general agreement that physical/health
Is QOL related to gender, age, educational level, status, psychological sta-tus and social functioning have to
cohabitation, time since diagnosis, treatment or type of be included (The WHOQOL group 1995).
cancer in these patients?
Another debated issue in QOL research is the use of
generic versus more disease-speci®c instruments. Dis-
QUALITY OF LIFE ease-speci®c measures are designed for a single diagnos-
tic group or for a relatively broad range of related diagnostic
As a result of the increased interest in QOL, several categories. Examples of such instruments developed to
de®nitions exist. The existing de®nitions of QOL can be measure QOL in cancer patients are the Functional Living
categorized as related to normal life, happiness/satisfac- Index Cancer and the Cancer Rehabil-itation Evaluation
tion, achievement of personal goals, social utility or natural System (CARES). The Sickness Impact Pro®le and the
capacity (Ferrans 1990). The category related to Nottingham Health Pro®le are examples of generic
satisfaction/happiness seems best suited to serve as a measures. One shortcoming of generic measures might be
focus in nursing (Rustùen 1995). Even if there is no that they do not address issues of great relevance in
consensual de®nition of QOL, there is considerable evaluating the effect of a speci®c disease on QOL
agreement among QOL researchers today that QOL is (Aaronson 1990).

Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 29(2), 490±498 491
T. Rustoen et al.

MATERIALS AND METHODS of r ˆ 0á83 and Cronbach's alpha coef®cient was 0á88 at
test and 0á90 at retest.
The respondents were also asked about gender, age,
type of cancer diagnosis, length of time since diagnosis,
The patients were recruited from four different hospitals treat-ment, other diseases, level of education, marital status
with primary cancer treatment in the Oslo region, or from and whether they were living alone or together with another
outpatient clinics. The inclusion criteria were that the person.
subjects were adult men and women with a recent
diagnosis of cancer, a Norwegian cultural background,
Statistical analysis
could read and write Norwegian, lived in or close to Oslo
and were not terminally ill. All the patients were judged by a Analysis of variance (ANOVA) and multiple regression
physician to have a life expectancy of at least 1 to 2 years, analysis were used to examine the impact of the selected
in order to exclude terminally ill patients. In order to ensure variables on QOL. Statistical interactions between the
that standardized procedures were followed the independent variables were tested by multiple regression
questionnaires were sent by post to the patients' homes. analyses. The patients were divided into three age groups,
Hence, none of the patients completed the questionnaires younger patients (19±39 years), middle-aged (40±59 years)
in the hospital. The study was approved by the local and elderly (60±78 years), and for analytical purposes the
medical ethics committee. Among 465 patients initially sample was divided into two equally sized age groups
approached and asked by a nurse or a physician to (below and above 55 years).
participate in the study, 131 returned a signed written Normality in the dependent variables was checked by the
consent to participate. one-sample Kolmogorov-Smirnov test.

The Ferrans and Powers Quality of Life Index (QLI) and the
Cancer Rehabilitation and Evaluation Systems, short form The sample consisted of 131 newly diagnosed cancer
(CARES-SF) were used to measure QOL. The scales were patients, the majority of whom were women (76%) (Table
chosen because they both re¯ected Ferrans' de®ni-tion of 1). Ages ranged from 19 to 78 years, with a mean and
QOL. Furthermore, both scales consist of various sub- median age of 52 years. The subjects had different cancer
scales and also give a global score. In the present study diagnoses, but the majority had breast and gynae-cological
scores for various sub-scales and one overall score were cancer. The mean time since cancer diagnosis was 5
calculated. CARES-SF was also selected because it was months, and the majority did not receive any treatment
cancer speci®c, and because it assesses the patient's need when ®lling in the questionnaires. All the participants lived
for assistance (Schag et al. 1991). at home, the majority (74%) cohabited with a spouse,
The QLI consists of 34 items which measure satisfaction another adult, children or parents. Fifty-six participants
in the various domains on the one hand and the impor- (43%) had more than 12 years of education (Table 1).
tance of each domain to the subject on the other (Ferrans
1990). The items are scored on 6-point Likert scales with
the total score ranging from 0 to 30, with the high scores
Ferrans and Powers QLI
denoting better QOL. Psychometric tests have shown that it
is suitable for cancer patients (Ferrans 1990, Arzouman et The mean global score for the QLI was 21á5 ( SD ˆ 3á8,
al. 1991, Hughes 1993), and it has been adapted for deviation from normality not signi®cant). The patients
Norwegian cancer patients. The Norwegian version in the reported highest QOL in the family domain, and lowest in
present study showed a 3- to 4-week test±retest correlation the health and functioning domain (Table 1). Low scores on
(Pearson's) of r ˆ 0á78. Cronbach's alpha coef®cient was QLI were found with regard to sex life, anxiety and stress,
0á93 at test and 0á95 at retest. lack of energy, faith in God and health (Table 3). The best
The CARES-SF consists of 59 items measuring reha- QOL was found in connection with relationships with others,
bilitation needs and QOL (Schag et al. 1991). It com-prises friends, social support and home. The items considered
®ve different sub-scales (Table 2). It is scored on a ®ve- most important by the patients were relation-ships with
point scale, and the scoring range for the whole scale is children (x ˆ 5á8), family health (x ˆ 5á8), satisfaction with
4±0 with lower scores indicating better QOL. For each item
family (x ˆ 5á8) and relationships with family members (x ˆ
patients also indicate whether they would like help by
5á8) (scoring range 1Ð6). The items which were reported to
circling Y for yes and N for no. The Norwegian version in
be of less importance were faith in God (x ˆ 3á3),
the present study of CARES-SF showed a 3- to 4-week
neighbourhood (x ˆ 4á5), anxiety and stress (x ˆ 4á5) and
test±retest correlation (Pearson's)
sex life (x ˆ 4á6).
492 Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 29(2), 490±498
Issues and innovations in nursing practice Quality of life

Table 1 Demographic characteristics of the study participants (n = 131), and mean scores on QLI for the total scales and the different sub-

Health/ Psychological/ Socio-

functioning spiritual economic Family
QLI sub-scale sub-scale sub-scale sub-scale
(n = 126) (n = 130) (n = 128) (n = 130) (n = 129)

Mean score 21á5 20á0 20á8 23á1 25á0

(SD) (3á8) (4á5) (4á0) (4á3) (4á4)
Female 99 (76%) 21á2 19á8 20á6 22á9 25á0
Male 32 (24%) 22á2 20á8 21á6 24á0 24á8
19±39 23 (18%) 20á9 20á1 20á4 21á6 23á6
40±59 64 (49%) 21á2 19á5 20á4 22á9 25á2
60±78 44 (33%) 22á2 20á7 21á7 4 25á4
Living alone 34 (26%) 19á4 17á9 19á3 21á9 23á7
Cohabitation 97 (74%) 1 2 3 23á6 25á4
22á1 20á8 21á4
Educational level
<= 12 years 75 (57%) 21á5 20á2 20á9 22á7 25á7
> 12 years 56 (43%) 21á4 19á8 20á7 23á8 4
Time since diagnosis
1±4 months 71 (55%) 21á2 19á7 20á5 23á0 24á5
5±8 months 46 (36%) 21á8 20á4 21á1 23á1 25á7
9±12 months 7 (5%) 20á2 19á3 20á8 22á5 22á9
12 months+ 5 (4%) 23á9 22á6 23á2 25á7 26á4
Different cancer diagnoses
Breast 49 (37%) 21á5 20á0 20á6 23á2 25á5
Gynaecological 23 (18%) 20á7 19á1 20á5 22á3 23á9
Prostate 13 (10%) 22á6 21á3 21á5 24á4 25á0
Colon 17 (13%) 21á4 19á8 21á4 22á1 25á2
Current treatment
Radiotherapy 6 (5%) 24á8 23á9 24á2 25á7 27á1
Chemotherapy 24 (19%) 19á9 18á2 18á8 21á7 24á0
Hormone therapy 18 (15%) 22á4 20á6 21á3 25á1 26á3
No treatment 76 (61%) 4 4 3 4 24á8
21á5 20á2 21á1 23á0
1 2 3 4
Signi®cance levels by one-way ANOVA: P < 0á001, P < 0á005, P < 0á01, P < 0á05.

QOL, demographic and disease-related variables
The mean global score on CARES-SF was 0á76 ( SD ˆ
0á47, slightly skewed, P ˆ 0á04). The patients reported the An analysis of variance (ANOVA) showed that those living
highest QOL in the marital domain, and the lowest in the with another person had a signi®cantly higher QOL than the
sexual domain (Table 2). The items with the lowest QOL group living alone as measured by both QLI (P ˆ 0á0004)
scores were worry and anxiety, lack of energy and bodily (Table 1) and CARES-SF (P ˆ 0á0001) (Ta-ble 2). A search
changes (Table 3). The highest QOL scores were found for for interactions using multiple regression analysis, showed
the items connected with daily grooming, time off from work that the negative impact of living alone on QOL was
and transportation to treatment, talking to colleagues about particularly strong among younger patients in both QLI (P-
the cancer, and helping children to cope with the cancer. value of interaction ˆ 0á046, cf. Figure 1) and CARES-SF
(P-value of interaction ˆ 0á024, cf. Figure 2).
Thirty-nine patients did not ®ll in any of the questions The only statistical association between age and QOL
about needing assistance. The items that most patients was found in the socio-economic sub-scale of QLI where
wanted help with were in relation to: `I worry about whether the oldest of the three groups had a signi®cantly higher
the cancer is progressing' (n ˆ 34), `I frequently feel anxious' QOL (Table 1). For the groups above and below 55 years,
(n ˆ 29) and `I am uncomfortable with the changes in my the older group showed a higher QOL than the younger
body' (n ˆ 22). The correlation between QLI and CARES-SF group in all sub-scales of QLI (Figure 3). As measured by
was ±0á58. CARES-SF, the older group was more satis®ed than the

Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 29(2), 490±498 493
T. Rustoen et al.

Table 2 Mean scores on CARES-SF for the total scales and the different sub-scales by demographic characteristics (n = 131)

CARES-SF Physical Psychological Marital Sexual Medical interaction

(4-0) sub-scale sub-scale sub-scale sub-scale sub-scale
(n = 131) (n = 128) (n = 127) (n = 86) (n = 120) (n = 126)

Mean score 0á76 0á84 0á85 0á48 1á03 0á70

(SD) (0á47) (0á61) (0á56) (0á57) (0á91) (0á71)
Female 0á78 0á83 0á91 0á44 0á97 0á68
Male 0á68 0á86 4 0á56 1á21 0á80
19±39 0á75 0á77 0á81 0á26 1á12 0á72
40±59 0á79 0á89 0á91 0á57 0á95 0á68
60±78 0á71 0á80 0á78 0á44 1á10 0á74
Living alone 1á03 1á05 1á25 0á89 1á41 0á80
Cohabitation 1 4 1 0á47 3 0á67
0á66 0á77 0á71 0á91
Educational level
<= 12 years of education 0á80 0á87 0á90 0á54 1á01 0á85
> 12 years of education 0á70 0á80 0á78 0á41 1á06 3
Time since diagnosis
1±4 months 0á76 0á89 0á90 0á47 1á00 0á67
5±8 months 0á74 0á74 0á80 0á43 1á02 0á71
9±12 months 0á87 0á95 0á71 0á91 1á31 1á04
12 months+ 0á89 0á80 0á82 0á67 1á47 1á17
Different cancer diagnosis
Breast cancer 0á86 0á96 0á97 0á45 1á19 0á73
Gynaecological cancer 0á71 0á74 0á83 0á37 0á85 0á58
Prostrate cancer 0á61 0á72 0á46 0á46 1á71 0á56
Colon cancer 0á77 0á75 4 0á79 4 0á86
0á82 0á70
Current treatment
Radiotherapy 0á67 0á88 0á68 0á54 1á14 0á96
Chemotherapy 0á93 1á01 0á98 0á45 1á04 0á76
Hormone therapy 0á76 0á88 0á85 0á33 0á99 0á70
No treatment 0á71 0á74 0á81 0á52 1á03 0á67

1 2 3 4
Signi®cance levels by one-way ANOVA: P < 0á001, P < 0á005, P < 0á01, P < 0á05.
Table 3 The highest and lowest mean scores in CARES-SF and QLI

Items with the highest QOL (mean score) Items with the lowest QOL (mean score)

Dif®culty bathing, brushing teeth, or grooming Worry about whether the cancer is progressing (1á95)
myself (0á07)
Dif®culty asking for time off from work for medical Do not have the energy I used to (1á55)
treatments (0á13) Uncomfortable with changes in body (1á53)
Dif®culty talking to the people who work with me about the Worry about not being able to care for myself (1á41)
cancer (0á17)
Dif®culty with transportation to and from medical Cancer treatments interfere with my ability to
appointments (0á2) work (1á3)
Dif®culty helping my children cope with my illness (0á21) I frequently feel anxious (1á11)

Relationship with spouse (10á5) Sex life (1á03)
Friends (9á6) Amount of stress and worries (2á0)
The emotional support from others (9á6) Amount of energy for everyday activities (2á5)
Children (9á1)
Home (9á0) Faith in God (2á8)
Family's health (8á8) Health (2á8)

494 Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 29(2), 490±498
Issues and innovations in nursing practice Quality of life

Figure 1 Effect of age and living with another person on QOL Figure 3 QLI pro®les of patients' age. Mean scores. ˆ<55 years;
measured by QLI controlled for gender, educational level, treatment jˆ56+ years.
and time since diagnosis (nˆ131). bˆ0á24 for the group living alone,
using age as a linear covariate. bˆ0á097 for the group living with
another person, using age as a linear covariate. The interaction
between age and living with another person/living alone on QOL as
measured by QLI controlled for gender, educational level, treatment
and time since diagnosis, Pˆ0á046. Cohabiting: nˆ18<40,
nˆ50(40±59), nˆ25 60‡. Living alone: nˆ5 <40, nˆ13(40±59), nˆ13

Figure 4 CARES-SF pro®les by patients' age. Mean scores. ˆ>55

years; jˆ56‡ years.

men had a higher QOL than women. The group with the
lowest number of years of education had signi®cantly
greater satisfaction in the family domain than the better
educated group, and the latter had greater satisfaction with
medical interaction (Tables 1 and 2). Time since diagnosis
was not associated with QOL in our sample. Differences in
QOL associated with differences in the type of cancer were
Figure 2 Effect of age and living with another person on QOL only found in the psychological and sexual domain of the
measured by CARES-SF controlled for gender, educational level, CARES-SF. Treatment was associated with QOL only as
treatment and time since diagnosis (nˆ131). bˆ)0á13 for the group measured by QLI in the way that the patients receiving
living alone, using age as a linear covariate. bˆ0á083 for the group radiation therapy reported the best QOL, and the patients
living with another person, using age as a linear covariate. The receiving chemotherapy reported the worst QOL.
interaction between age and living with another person/ living alone
on QOL as measured by CARES-SF controlled for gender,
educational level, treatment and time since diagnosis, Pˆ0á024. DISCUSSION
*Reversed scoring (4-mean score). Cohabiting: nˆ18<40,
This study is one of the few which describes the impact of
nˆ51(40±59), nˆ26 60+. Living alone: nˆ5 <40, nˆ13(40±59), nˆ16
cancer on Norwegian cancer patients. Of the two different
scales used to measure QOL, CARES-SF is more speci®-
cally cancer-related, and measures problems in connec-tion
younger in the psychological and marital domains, while the with both cancer treatment and medical interaction. In spite
younger group was more satis®ed in the sexual domain of this, there was a fairly substantial correlation between the
(Figure 4). Gender was not associated with QOL except in scales, and both sets of scores gave the same overall
the psychological domain of CARES-SF, where picture of the cancer patient in the study sample. A

Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 29(2), 490±498 495
T. Rustoen et al.

comparison between the overall mean score on the QLI in Our ®nding of the highest level of QOL in the family
the present and in several previous studies is exhibited in domain and the lowest in the health and functioning domain
Table 4. The similarity of the mean score on QOL across is also supported by others (Hughes 1993, Bliley & Ferrans
different types of diseases seems to support the validity of 1993, Zacharias et al. 1994). The signi®cance of the family
the QLI (Table 4). However, the very small differences and marital domains for a high QOL in our sample was
observed between the healthy individuals and various supported by the fact that the patients themselves listed
patient groups (never more than one-sixth of a standard children and family as being of the greatest importance, and
deviation) may cast doubt on the sensitivity and/or by the signi®cantly higher QOL score for both scales for
relevance of the scale. those living together with another person compared with the
Given the range of scores for our QOL scales, the group living alone. For the group living alone, we found a
observed scores indicate a relatively high level of QOL in lower QOL in the younger patients compared with the older
our sample of patients. We would have expected QOL to be age groups. This could be because the cancer diagnosis
reduced because of the recent diagnosis of cancer, the may reduce their chances of establishing a family, which
attendant fear and uncertainty, and physical problems could be harder for the younger patients to adapt to.
related to the cancer treatment. This phenomenon might be
caused by the small proportion of those patients initially In spite of the importance of social support for cancer
approached who agreed to participate in the study. On the patients, social relationships can also be an additional
other hand, the level of QOL in our sample as measured by source of distress because the disease can provoke phys-
QLI was very similar to that of other samples (Table 4), and ical aversion and avoidance behaviour (Krishnasamy 1996).
the mean scores on QOL in our study as measured by This may be especially marked in the case of newly
CARES-SF were close to those of a sample of American diagnosed cancer patients because the experience is new
women with lung cancer (x ˆ 0á9) (Sarna 1993). There are for both patient and family. In our sample sex life was
still a number of unresolved questions regarding to what reported to be one of the least important domains and had
extent and in which way a cancer diagnosis impacts on the the lowest QOL score. During the initial period after the
individual's QOL. Courtens et al. (1996) found that al- diagnosis, closeness to a partner and seeing that close
though functioning and physical and psychological well- family members are coping with the new situation may be
being were negatively affected by cancer, people with a more important than a sex life. Some patients also probably
newly diagnosed cancer were satis®ed with life in general experience a less satisfactory sex life because of temporary
and had a high QOL. They suggested that this problems with recent surgery and bodily chan-ges,
phenomenon was due to the patient's ability to adapt to the especially given that 65% of our patients were diagnosed as
new situation. Graham & Longman (1987) found that the having gynaecological, breast and prostatic cancer.
level of QOL rose after a diagnosis of cancer. They pointed Nevertheless, our ®ndings emphasize the impor-tance of
out that the experience of cancer can lead to a reappraisal family relationships and social network for the QOL of
of values and priorities and a refocusing on different values cancer patients.
that actually increases QOL (Graham & Longman 1987). The lower score on QOL in the physical, health and
Other suggestions might be that in newly diagnosed, not functioning domains is not surprising, as the patients had
termi-nally ill cancer patients there is always a chance of been recently diagnosed and treated for cancer. For
cure, which may strengthen the psychological domain of instance, bodily changes such as climacteric and other
QOL. Denial also has a place here. Denial is a common hormonal problems, and problems stemming from loss of
reaction to a cancer diagnosis and acts as a defence breast and hair were reported. The breast cancer patients
mechanism for some patients in the ®rst phase after had the lowest QOL in both the psychological and the
receiving the diagno-sis (KuÈbler-Ross 1969). sexual domains compared with the other groups (Table 2).

Table 4 Mean score and standard deviation in overall QOL in studies using Ferrans and Powers QLI

Authors and year of publication Samples Mean score on QLI (SD)

Present study 131 cancer patients 21á46 (3á78)
Zacharias et al. 1994 40 gynaecological cancer patients 21á89 (3á58)
Arzouman et al. 1991 15 sarcoma patients 21á96 (2á97)
Hughes 1993 52 breast cancer patients 22á33 (4á33)
Ferrans & Powers 1992 349 haemodialysis patients 20á70 (4á77)
Hicks et al. 1992 35 liver transplant patients 22á20 (4á90)
Gustafsson & Hamrin 1996 90 disabled elderly patients 21á39 (3á59)
Ferrans et al. 1992 88 healthy persons 21á67 (3á67)

496 Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 29(2), 490±498
Issues and innovations in nursing practice Quality of life

The fact that the patients treated for prostate cancer were yielding higher scores for the well educated and another
the most satis®ed group in the psychological domain must (family) showing higher QOL among the less educated.
be seen in the context of the fact that the male patients in Thus there is no reason to conclude that the level of
general scored best in this domain. Lack of energy and education has any major impact on QOL in our sample.
ability to work had the lowest scores in both scales. Fatigue Another limitation of our study may be the heterogene-ity
is often described as a serious problem in cancer patients in cancer diagnoses and prognostic factors in our sample.
(Skalla & Lacasse 1992). However, the majority had been diagnosed within the
Emotional reactions such as stress, worry, anxiety, fear previous 6 months, and none was considered to be
and depression were also reported in our sample. The need terminally ill. Length of time since diagnosis had no impact
for assistance was most frequently reported in connection on QOL, and only small differences in overall QOL were
with dealing with anxiety and the fear that the cancer would found between the different types of cancer diagno-sis or
progress. Stress and worries, on the other hand, were between the time elapsed since diagnosis, so that this
reported to be of less importance for the patients. This may limitation does not seem to be a serious one.
mean that some emotional distress was expected. Another
®nding in this area was that men were more satis®ed in the CONCLUSION
psychological domain than women. In general, women are
found to have a higher frequency of psychological and The patients seem to be quite satis®ed with their lives in
psychosomatic complaints (Wiklund et al. 1993), to be more spite of the cancer. These ®ndings underline the impor-
depressed and tense (Tibblin et al. 1990) and to have more tance of measuring QOL in cancer patients. Data about the
mental disorders than men (Linzer et al. 1996). In our patients' own view of their life, about both global QOL and
sample 76% were women, and »70% were under 60 years their satisfaction in different domains in life provide a
old. Both care of children and climacteric problems among broader picture of how a cancer diagnosis might impact on
women may be an additional burden on top of a diagnosis patients' lives. The importance of the family or signi®cant
of cancer. others for QOL in the newly diagnosed cancer patient is
Age was only signi®cantly associated with QOL in one apparent. Thus nurses should help patients to avoid
sub-scale, the socio-economic one, in which older patients becoming isolated after a diagnosis of cancer, and regard
experienced a higher QOL than younger ones. One expla- family members and signi®cant others as a resource and if
nation for this might be that many of the elderly are retired possible include them in the care of the patients. The nurse
and that the economical consequences of the cancer in must be aware of the importance of QOL and, especially
terms of sick leave are easier to cope with for older people. that younger single patients are at risk here. Patients living
Overall, the older patients experienced slightly (but not alone may have greater dif®culty in coping with a cancer
signi®cantly) higher QOL in all domains except the sexual, diagnosis. Sexual matters should be given attention in order
physical and medical interaction domains. Pre-vious to prevent disease-related dif®culties from lasting so long
research has shown that older cancer patients manifest that they cause additional family problems.
fewer and less severe psychosocial problems than younger
ones (Mor et al. 1994). This may be because older people
are more tolerant of minor symptoms, dis-tress and
discomfort than younger age groups, and that the elderly We are grateful to the Faculty of Nursing, Oslo College for
tend to perceive their health as good even in the presence ®nancial support for this study.
of overt pathology (Hunt et al. 1984).

Study limitations
Aaronson N.K. (1990) Quality of life research in cancer clinical
Our sample may not be fully representative of the cancer trials: a need for common rules and language. Oncology
population because of the low response rate and that the (Huntingt) 4, 59±66.
sample contained a high proportion of patients that were Arzouman J.M., Dudas S., Ferrans C.A. & Holm K. (1991) Quality
cohabiting and well educated. The low response rate might of life of patients with sarcoma postchemotherapy. Onclogy
Nursing Forum 18, 889±894.
be caused by the fact that the patients were included in an
Bliley A.V. & Ferrans C.E. (1993) Quality of life after coronary
intervention study. It has been shown that »25% of cancer
angioplasty. Heart and Lung: Journal of Critical Care 22, 193±
patients do have interest in such interventions (Thiel de 199.
Bocanegra 1992). On the other hand, there were no signi-
Courtens A.M., Stevens F.C., Crebolder H.F. & Philipsen H. (1996)
®cant differences in the mean scores in overall QOL Longitudinal study on quality of life and social support in cancer
between the two educational groups. Two sub-scales patients. Cancer Nursing 19, 162±169.
showed statistically signi®cant differences between edu- Deshotels A., Planchock N., Dech Z. & Prevost S. (1995) Gender
cational groups, with one sub-scale (medical interaction) differences in perceptions of quality of life in cardiac rehabil-

Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 29(2), 490±498 497
T. Rustoen et al.

itation patients. Journal of Cardiopulmonary Rehabilitation 15, Mastekaasa A., Moum T., Nñss S. & Sùrensen T. (1988)
143±148. Livskvalitetsforskning. Institutt for Samfunnsforskning, Rapport
Ferrans C.E. (1990) Development of a quality of life index for nr.6, Oslo (Norwegian).
patients with cancer. Oncology Nursing Forum 17, 15±21. Mor V., Allen S. & Malin M. (1994) The psychosocial impact of
Ferrans C.E., Cohen F.L. & Smith K.M. (1992) The quality of life of cancer on older versus younger patients and their families.
persons with narcolepsy. Loss, Grief and Care 5, 23±32. Cancer 74, 2118±2127.
Ferrans C.E. & Powers M.J. (1992) Psychometric assessment of Padilla G.V. (1993) State of the art in quality of life research.
the Quality of Life Index. Research in Nursing and Health 15, Communicating Nursing Research 26, 71±80.
29±38. Padilla G., Grant M.M. & Ferrell B. (1992) Nursing research into
Ganz P.A., Lee J.J., Sim M.S., Polinsky M.L. & Schag C.A. (1992) quality of life. Quality of Life Research 1, 341±348.
Exploring the in¯uence of multiple variables on the relation-ship Rustùen T. (1995) Hope and quality of life - two central issues for
of age to quality of life in women with breast cancer. Journal of cancer patients - a theoretical analysis. Cancer Nursing 18, 355±
Clinical Epidemiology 45, 473±485. 361.
Graham K.Y. & Longman A.J. (1987) Quality of life and persons Sarna L. (1993) Women with lung cancer: impact on quality of life.
with melanoma. Preliminary model testing. Cancer Nursing 10, Quality of Life Research 2, 13±22.
338±346. Schag C.A., Ganz P.A. & Heinrich R.L. (1991) Cancer Rehabilita-
Gustafsson G. & Hamrin E.M. (1996) Assessing disabled elderly tion Evaluation System Ð short form (CARES Ð SF). A cancer
with the Swedish version of the Quality of Life Index. In Quality of speci®c rehabilitation and quality of life instrument. Cancer 68,
Life and Functional Capacity among Elderly with Locomotor 1406±1413.
Disability, Linkùping Studies in Health Sciences, thesis no. 23, Skalla K.A. & Lacasse C. (1992) Patient education for fatigue.
University of Linkùping, Linkùping, pp. 74±87. Oncology Nursing Forum 19, 1537±1541.
Hicks F.D., Larson J.L. & Ferrans C.E. (1992) Quality of life after The WHOQOL group (1995) The World Health Organization quality
liver transplant. Research in Nursing and Health 15, 111±119. of life assessment (WHOQOL): position paper from the World
Hughes K.K. (1993) Psychosocial and functional status of breast Health Organization. Social Science and Medicine 41,
cancer patients. Cancer Nursing 16, 222±229. 1403±1409.
Hunt S.M., McEven J. & McKenna S.P. (1984) Perceived health: Thiel de Bocanegra H. (1992) Cancer patient's interest in group
age and sex comparisons in a community. Journal of Epidemi- support programs. Cancer Nursing 15, 347±352.
ology and Community Health 38, 156±160. Tibblin G., Bengtsson C., Furunes B. & Lapidus L. (1990)
Krishnasamy M. (1996) Social support and the patient with cancer: Symptoms by age and sex. Scandinavian Journal of Primary
a consideration of the literature. Journal of Advanced Nursing 23, Health Care 8, 9±17.
757±762. Wiklund I., Herlitz J., Johansson S. et al. (1993) Subjective
KuÈbler-Ross E. (1969) On Death and Dying. Macmillan, New symptoms and well-being differ in woman and men after
York. Linzer M., Spitzer R., Kroenke K. et al. (1996) Gender, quality myocardial infarction. European Heart Journal 14, 1315± 1319.
of life and mental disorders in primary care: results from the
PRIME-MD 1000 study. American Journal of Medicine 101, Zacharias D.R., Gilg C.A. & Foxall M.J. (1994) Quality of life and
526±533. Loose M.S. & Fernhall B. (1995) Differences in quality of coping in patients with gynaecological cancer and their spous-es.
life among male and female cardiac rehabilitation participants. Onclogy Nursing Forum 21, 1699±1706.
Journal of Cardiopulmonary Rehabilitation 15, 225±231.

498 Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 29(2), 490±498