Journal of Psychiatric and Mental Health Nursing, 2009, 16, 546552

Quality of life for patients with schizophrenia living in

the community in Greece
P. D I M I T R I O U 1 , 2 r n m s c m b a p h d ,
D. ANTHONY3 rmn srn rn(canada) ba(hons) msc phd &
S. DYSON4 rn rm dip nursing bsc(hons) rnt msc edd
1

Lieutenant Colonel (N), Hellenic National Defense General Staff, 3Professor of Nursing, Mary Seacole Research
Centre, 4Head of Nursing and Midwifery Research, School of Nursing & Midwifery, De Montfort University,
Leicester, UK; and 2Lieutenant Colonel (N), Hellenic National Defense General Staff, Greek Air Force, Athens,
Greece

Correspondence:
D. Anthony
Mary Seacole Research Center
School of Nursing and Midwifery
Charles Frears Campus
266 London Road
Leicester LE2 1RQ
UK
E-mail: danthony@dmu.ac.uk
doi: 10.1111/j.1365-2850.2009.01413.x

DIMITRIOU P., ANTHONY D. & DYSON S. (2009) Journal of Psychiatric and Mental
Health Nursing 16, 546552
Quality of life for patients with schizophrenia living in the community in Greece
Several quality of life instruments were considered for use in a Greek mental health
environment. Subjective Quality of Life Profile was chosen as it was easy to complete and
covered the issues raised by patients with schizophrenia through interviews. Confirmatory
factor analysis gave credence to the four-dimensional structure identified by the original
authors. Patients with schizophrenia were generally satisfied with their quality of life,
found the items in the instrument important and were optimistic about expectations for
change. Age, gender, education, marital status and years of sickness were not statistically
significant in a general linear model with quality of life as the outcome for the 27 core
questions. There were some statistically significant results for the three disease-specific
questions; positive expectation was correlated positively with education and negatively
with years of sickness.
Keywords: community, Greece, quality of life, schizophrenia
Accepted for publication: 2 February 2009

Introduction
The trend in Greece (in common with other European
countries) has been to treat patients with schizophrenia in
the community. The Greek psychiatric reform started in
1984 after special funding by the European Community
and under the Regulation 815/84 (Stefanis et al. 1986).
Schizophrenia is one of the more important enduring
mental health problems and thus it would be useful to
measure the quality of life of patients with schizophrenia.
This is particularly pertinent as the focus of the Greek
psychiatric reform was not only on the mental health care
but also on the psychosocial and vocational rehabilitation
of the psychiatric patients (Bellali & Kalafati 2006).
546

Quality of life of people with schizophrenia has been

measured with a variety of instruments in many countries.
Examples include USA (Lehman et al. 1986), Germany &
Wales (Kaiser et al. 1997), the UK, Denmark, Holland,
Spain & Italy (Gaite et al. 2002), Canada, Cuba & the USA
(Vandiver 1998), Hong Kong (Chan & Yu 2004), Malaysia
(Mubarak et al. 2003), Morocco & USA (Green et al. 2001)
and Germany (Leibe & Kallert 2000). There is one paper
concerned with patients moving into the community from
hospital (Zissi et al. 1998), but otherwise no study exploring
the quality of life of patients with schizophrenia in the
community in Greece has been located (Dimitirou 2007).
Generic quality of life instruments include Short
Form-36 (SF-36) (Ware & Sherbourne 1992) and the
2009 Blackwell Publishing

Quality of life for patients with schizophrenia

Subjective Quality of Life Profile (SQLP) (Gerin et al. 1992,

Dazord 1997). There are several commonly employed
quality of life instruments designed for mental health
environments, e.g. the Lehman Quality of Life Interview
(QOLI) (Lehman 1983), the Lancashire Quality of Life
Profile (LQoLP) (Oliver et al. 1997), the Self Report
Quality of Life Measure (Wilkinson et al. 2000).
We considered several quality of life instruments for
their appropriateness in a Greek mental health setting. We
selected two of these on the basis they are commonly
employed, have been used in Greece and are available in
Greek.
Short Form-36 was used to assess the health status and
health-related quality of life of the personnel of the Hellenic Network of Health Promotion Hospitals (Tountas
et al. 2003), but this was not published at the time of our
study so we excluded SF-36. The LQoLP is available in
many languages, English, Dutch, Danish, Spanish and
Swedish but not Greek, so we excluded LQoLP. The QOLI
was used to measure quality of life in Greek hospital
patients with schizophrenia moving to a community hospital (Zissi et al. 1998). The SQLP has been translated in
Greek by Damigos and Siafaka in order to be used in renal
patients. Thus we had in Greek a specialized mental health
quality of life instrument (QOLI) and a very commonly
used generic instrument (SQLP) that has been used in many
populations including mental health (Dazord 2002).

Method
We interviewed eight patients with schizophrenia, who
attended a day centre in Athens, to determine what they
considered quality of life to mean using a simple interview
topic guide (see Table 1). We employed a semi-structured
interview approach to allow the interviewer to pick up and
further explore issues.

Table 1
Interview topic guide
Demographic
questions

Other questions
Quality of life
questions

How old are you?

Are you married (if yes)/ do you have children?
What is your educational background?
Do you work?
Whom do you live with?
Would you like to describe your daily life for
me?
Which of your life areas are the most
important to you?
Which of these do fulfil you the most?
What does quality of life mean to you? (In
cases not understood this question had to be
questioned in the way: What does satisfy
you? Or: What does fulfil you?

2009 Blackwell Publishing

We employed content analysis to form themes. We compared these themes with the constructs of the two quality of
life instruments.
Having selected the instrument that appeared most
suited to a Greek mental health population, we performed
confirmatory factor analysis to explore if the constructs
were the same in the Greek population. We then measured
the quality of life of a population of 100 patients with
schizophrenia. Finally we explored various demographic
variables to see if these were related to quality of life.

Ethics
The study was approved by the relevant ethics committee
of De Montfort University. Formal ethical approval was
obtained from the director of the mental health community
centre where the participants of the study attended as
outpatients.

Results
Sample
Three hundred outpatients met the selection criteria. Ten
of them were randomly selected to participate in the first
phase of the study and 90 more for the second phase of the
study. Out of the randomly selected 10 outpatients, two did
not agree to be interviewed and they expressed their wish
to participate in the second phase of the study, all selected
patients agreed to the second phase. Therefore, the final
number of the participants for the first phase was eight and
for the second phase 100, including the eight outpatients
who had been interviewed. However, in error 91 patients
were recruited for the second stage and thus a sample of
101 was available. All patients completed the survey
instrument.

Interviews
Initially, the data were read and reread by the one of us (P.
D.) to identify the quality of life indicators as expressed by
the interviewees. The indicators were then categorized into
themes and finally the clarification of the concept of quality
of life was extracted according to the themes.
In order to achieve data verification, the charge nurse of
the centre was asked to read the transcripts and to categorize the data. There were no big differences between the
categories between the two coders; both agreed upon the
final list of themes (see Table 2).
The analysis of the qualitative data identified five indicators or domains of quality of life, as experienced by the
eight participants, who had been interviewed. The five
547

P. Dimitriou et al.

Theme

Item

Work/money

To have a permanent job

To have a job
To be able to work
To be productive
To keep self busy
To create
To own a house
To live in a good house/apartment
To have money
To have money to spend for self and family
To buy things for self and son
To cover daily expenses
To buy clothes
To be married
To have family
To have many children
To take care of the child
To spend time with wife
Father
Son
Family, children
To have good relationships with sister
To visit son
To take care of the family
To help the family
To have a partner
To socialize
To go to the cafeteria
To make people laugh
To go out and enjoy self
To meet people
To be with other people
To talk with other people
To come to the community centre
Participate in the centres programme
Not to be alone
Boyfriend
Relationship with the other sex
To go out with wife
To go out with friends or cousins
Visit friends
To have friends
To have friends
Spend time with other people
Get out of the routine
To go out for walks
To go to the theatre
To paint
To attend concerts
Knit
Sew
To read books
Watching movies
To study computers and foreign languages
Travelling
Music
Hiking
Not to be sick
Not to have the side effects of the medication
To be active
No drinks
Not to take medications
To be healthy
To lose weight
To be loved/cared
Feeling shame for getting lunch for the poor people
The others to accept the sick people
To be perceived as normal person
To be respected by other people
The others not to avoid my company

Family

Social

Health

Psychological

548

domains of quality of life constitute the concept clarification of quality of life as it was given by the informants and
are work/money, family, social, health and psychological.
The Lehman QOLI has eight life domains, which are
living situation, daily activities and functioning, family
relations, social relations, finances, work and school, legal
and safety issues as well as health. The domains of life
that are explored through the SQLP are: (1) functional
life referring to motor, sexual, sleep and so on; (2) social life
referring to social roles, relationships; (3) material life
referring to financial issues, goods, house; and (4) spiritual
life referring to the aesthetics, religion, thinking. Thus
either has domains that map onto the quality of life issues
mentioned by the Greek community patients. The QOLI
consists of 143 items and takes 45 min to complete; SQLP
is composed of 27 core items and three disease-specific
items and takes 1520 min to complete. Pragmatically, we
therefore decided to employ the shorter, faster to complete
instrument as we considered this would give a better
response rate.

Analysis of the quantitative data

Demographic characteristics of participants
Fifty-one of the sample were female and fifty were male.
Seventeen had children and 84 did not. All received medication. Seventy-seven were single, 12 divorced and 12
married. The profiles for age, years of sickness and educational level are seen in Figs 13. It is seen that the sample is
almost exactly split by gender, most do not have children,
are typically middle-aged, have been sick for up to about

20

15
Frequency

Table 2
Themes from interviews of the concept quality of life

10

Mean = 41.92
Std. Dev. = 7.951
N = 101

0
20

40

60

Age
Figure 1
Age profile

2009 Blackwell Publishing

Quality of life for patients with schizophrenia

Table 3
Categorization of the core questions

12.5

Frequency

10.0

7.5

5.0

2.5
Mean = 16.64
Std. Dev. = 7.641
N = 101

0.0
0

20

40

Years of sickness
Figure 2
Years of sickness

40

Count

30

20

10

0
Elementary Gymnasium

Lyceum

University
student

University
graduate

Education
Figure 3
Level of education

Life domain

Question number

Functional
Social
Material
Spiritual

1,2,3,4,6,17,23
5,7,8,9,13,18,19,21
10,11,12,14,22
15,16,20,24,25,26,27

independently categorized the 27 core questions; the last

three are disease-specific, therefore they do not refer to any
of the four domains) into functional, social, material and
spiritual (Gerin et al. 1992).The classification (there was
minimal differences and agreement was reached easily) is
shown in Table 3.
Data were described and condensed using simple counts,
tables and graphs. Further exploration of the dimensionality of the SQLP was conducted with factor analysis. The
differences between the core and disease-specific questions
tested by a general linear model. General linear model is
multivariate and tests whether the several variables are
different in one or more groupings and with one or more
co-variants. By giving a single P-value for many simultaneous tests, it removes the type II errors that multiple
testing would make more likely. Only where an overall
significant P-value is found are individual differences even
looked at.
The 27 questions that form the core of SQLP were
summed to give three values: the sum of satisfaction for the
items, sum of perceived importance of items and sum of
expectation of change of items.
The means are centred on zero, so positive values show
positive attitudes and likewise negative ones show negative
attitudes. The findings show that the participants experience an overall satisfaction with the items addressed in
the core questions, with a mean score of between -0.5 and
+1.0; a minority are either unsatisfied or very satisfied (see
Fig. 4). Most of the participants perceive as important
(from 0.5 to 2.0) (see Fig. 5). Concerning their expectations
for change, most of the subjects seem to be optimistic and
very few see no change to the current situation referring to
the core questions (see Fig. 6).

Factor analysis
40 years, and typically 1520 years, and are mostly
not graduates, but typically have completed secondary
education.
To explore the domains in SQLP, it is necessary to identify each question with a domain. However, while SQLP
is stated to have four domains, in practice these are not
identified, and all 27 core and three disease-specific questions are considered as a profile. Two of us (P. D. and D. A.)
2009 Blackwell Publishing

In principal components analysis (factor analysis is a moregeneral method of principal components analysis, although
the terms are often used interchangeably), a correlation
matrix is computed for all the variables that form the
dataset. It can be shown that a set of new variables, called
eigenvectors, can be created that completely describe the
correlation matrix but these new variables are independent
of each other (i.e. they are not correlated with each other).
549

P. Dimitriou et al.

25

30

Frequency

Frequency

20

20

15

10

10
5

Mean = 0.136
Std. Dev. = 0.32077
N = 95

0
-0.50

0.00

0.50

20

Frequency

0.50

1.00

1.50

2.00

2.50

Figure 6
Mean expectation for change

25

15

10

5
Mean = 1.3242
Std. Dev. = 0.30191
N = 95

0
1.20

1.50

1.80

2.10

2.40

Mean perceived importance

Figure 5
Mean perceived importance

In a dataset where some of the variables are totally redundant, the eigenvectors will be fewer in number than the
original variables but will contain exactly the same information. In general, however, for a dataset with N variables,
there will be N eigenvectors, but some of them will contain
very little information. The amount of information in each
550

0.00

Mean expectation for change

Figure 4
Mean satisfaction

0.90

0
-0.50

1.00

Mean satisfaction

0.60

Mean = 0.434
Std. Dev. = 0.3494
N = 94

eigenvector is measured by the amount of variance in the

dataset it describes.
Each eigenvector has an associated eigenvalue, which is
the measure of the amount of variance in the dataset it
describes. If all the eigenvalues are summed, then the ratio
of each eigenvalue to this sum is the percentage of the
variance for which it accounts. Thus a sensible strategy
for reducing the number of the variables in the dataset
is to take the first few eigenvectors with the highest
eigenvalues.
There are two common ways of establishing the number
of factors in a dataset. Using Kaisers criterion, we would
use all eigenvalues above unity (1.0), but this has been
criticized as being arbitrary. A scree plot is a graph of
descending eigenvalues, and where the plot changes slope
most is assumed to be the point beyond which the eigenvectors are those describing noise (i.e. are random).
However, both Kaiser and scree plots are believed to retain
too many factors. A parallel analysis method generates
factors from random datasets. When eigenvectors from the
random dataset are larger than those found in the dataset
under examination, we assume the latter are random noise
(Pallant 2007).
Factor analysis was conducted separately on the three
types of question. For each there were four components
that had eigenvalues above one (the Kaiser criterion).
While this appeared to confirm the four factors proposed
for the tool, the loadings suggest this is not precisely the
case (see Table 4).
2009 Blackwell Publishing

Quality of life for patients with schizophrenia

Table 4
Variables loaded on the factors
Type of question

Factor 1 (all except

functional)

Factor 2 (functional
and social)

Factor 3 (functional
and social)

Factor 4 (social
and material)

A satisfaction for the items

B: perceived importance of items
C: expectation of change

527
527
527

1,2, 47
13 and 47
17

14 and 57
14
14 and 57

813
4 and 813
813

This shows a consistent loading on the first factor

largely for items not to do with functional life, a loading on
the second factor for functional life and social life, for
functional life and social on the third factor and for social
and material life on the fourth factor. Thus there are four
factors but not precisely those designed to be measured by
SQLP. However, they are similar: one factor is functional,
two and three are combinations of functional and social
and the fourth social and material. Spiritual does not load
on itself but only with all but functional on the first factor.
Thus factor analysis could justify four domains.
However, the more conservative parallel analysis shows
three factors for satisfaction, two for importance and
one for perceived change. The interpretation is similar, but
there is less evidence for material life being a factor in this
group, which may be because the group is homogeneous in
terms of income.

General linear model: analysis of variance

For each type (satisfaction, perceived importance and
expectation of change) of the core questions a multivariate analysis was conducted employing a general linear
model. These showed no significant differences at the
multivariate level for any of the covariates (continuous
variables) age, years of sickness, educational level or
factors (nominal variables), gender, having children or
being single. Thus there was no point considering individual items.
The three disease-specific questions also showed no significance for satisfaction or perceived importance but for
years of sickness and education in the expectation of
change questions. When individual items were considered
some were significant indicating that how long the subjects
had been ill and the level of education affected their perception of these items. For years of sickness, there were
differences in alcohol and drugs and for education drugs
and medication.
To explore these differences a correlational analysis was
conducted, which showed that higher educational levels
were associated with better expectation of change with
drugs and medication, and increased years of sickness with
more negative expectations of alcohol and drugs (see
Table 5).
2009 Blackwell Publishing

Table 5
Correlations for expectations

Education
Alcohol
Drugs
Medication

Alcohol

Drugs

Medication

0.075

0.221*
0.320**

0.326**
-0.031
0.207*

Years of
sickness
0.060
-0.300**
-0.306**
0.053

*Correlation is significant at the 0.05 level (2-tailed).

**Correlation is significant at the 0.01 level (2-tailed).

Conclusion
Subjective Quality of Life Profile is a valid quality of life
instrument to use in Greece in a mental health environment. There were three or four factors in the SQLP, and
these map onto the domains it is designed to measure, or
combinations of these.
At the multivariate level there are no differences in this
population with respect to age, gender, education, marital
status or years of sickness in the core questions. For the
disease-specific questions there are also no significant
differences except for the expectation of change where
education and years of sickness were significant. Years of
sickness was negatively correlated with positive expectation and education positively correlated. I.e. subjects who
were better educated and/or had less prolonged illness were
more positive about the future.
While there are these differences in disease-specific questions, these only relate to expectation of change. For the
core questions there are no differences noted for any of the
demographic variables. This is consistent with the entire
group having a homogeneous view of quality of life. One
might argue that people who were diagnosed with schizophrenia at least 2 years ago and who remain on medication
share a similar quality of life. Because the demographic
variables would be expected to impact on quality of life,
this might mean suffering from (or being diagnosed with)
schizophrenia puts people of different age, gender, having
children and marital status at the same social level. There is
no evidence that educational level or even years of sickness
make any difference in quality of life in this group.
Thus this study shows results in Greece are similar to
studies that show age, gender and marital status are not
related to quality of life for people with schizophrenia
551

P. Dimitriou et al.

(Skantze et al. 1992, Brown 1996, Vandiver 1998)

although other studies show, e.g. marital status is highly
relevant (Corrigan & Buican 1995a,b). The Greek public
image of the mentally ill was studied by Parashos (Parashos
1998) who concluded that 42% of the Greek general public
refuse to employ someone who is mentally ill and 36%
would not like to live in the same neighbourhood where
mental healthcare services are hosted. One explanation
thus for the homogeneity of the Greek population may be
that the stigma in Greek society remains very high for
mental illness, and people diagnosed with schizophrenia
find it very difficult to obtain employment, and this adds to
difficulties in finding a partner. Thus the population is
largely unmarried and unemployed, making differences in
these areas difficult to assess. However, stigma in mental
illness is not unique to Greece, with negative attitudes
noted in Japan (Yutaka et al. 1999) and a more recent
study in Germany (e.g.) showed negative attitudes to
people diagnosed with schizophrenia with a preference
to increase social distance (Angermeyer & Matschinger
2003). Thus these results are probably generalizable to
other countries in Europe, and probably worldwide.

References
Angermeyer M.C. & Matschinger H. (2003) The stigma of mental
illness: effects on labelling on public attitudes towards people
with mental disorder. Acta Psychiatrica Scandinavica 108, 304
309.
Bellali T. & Kalafati M. (2006) Greek psychiatric care reform:
new perspectives and challenges for community mental health
nursing. Journal of Psychiatric and Mental Health Nursing 13,
3339.
Brown C. (1996) A comparison of living situation and loneliness
for people with mental illness. Psychosocial Rehabilitation
Journal 20, 5963.
Chan S. & Yu I.W. (2004) Quality of life of clients with schizophrenia. Journal of Advanced Nursing 45, 7283.
Corrigan P. & Buican B. (1995a) The needs and resources assessment interview for severly mentally ill adults. Psychiatric
Services 46, 504505.
Corrigan P.W. & Buican B. (1995b) The construct validity of
subjective quality of life for the severely mentally ill. Journal of
Nervous and Mental Disease 183, 281285.
Dazord A. (1997) The Subjective Quality of Life Questionnaire
(SQLP): an updated review of validation and results. Quality of
Life Newsletter 18, 78.
Dazord A. (2002) Le concept de qualite de vie, resultats
denquetes effectuees aupres de 13,000 sujets. Recherche en
Soins Infirmiers 70, 2338.
Dimitirou P. (2007) Quality of life for patients with schizophrenia
living in the community in Greece. PhD thesis. Leicester, De
Montfort University.
Gaite L., Vzquez-Barquero J.L., Borra C., et al. (2002) Quality
of life in patients with schizophrenia in five European countries:

552

the EPSILON study. Acta Psychiatrica Scandinavica 105, 283

292.
Gerin P., Dazord A., Boissel J., et al. (1992) Quality of life assessment in therapeutic trials: rationale for the presentation of a
more appropriate instrument. Fundamental & Clinical Pharmacology 6, 263276.
Green C.A., Fenn D.S., Mousaoui D., et al. (2001) Quality of
life in treated and never treated schizophrenic patients. Acta
Psychiatrica Scandinavica 103, 131142.
Kaiser W., Priebe S., Barr W., et al. (1997) Profiles of subjective
quality of life in schizophrenic in- and out-patient samples.
Psychiatry Research 66, 153166.
Lehman A. (1983) The well-being of chronic mental patients.
Archives of General Psychiatry 40, 369274.
Lehman A.F., Possidente S. & Hawker F. (1986) The quality of life
of chronic patients in a state hospital and in community residences. Hospital & Community Psychiatry 37, 901907.
Leibe M. & Kallert T.W. (2000) Social integration and the quality
of life of schizophrenic patients in different types of complementary care. European Psychiatry 15, 450460.
Mubarak A.R., Baba I., Chin L.H., et al. (2003) Quality of life of
community based chronic schizophrenia patients in Penang,
Malaysia. Australian and New Zealand Journal of Psychiatry
37, 577585.
Oliver J., Huxley P. & Priebe S. (1997) Measuring the quality of
life of severely ill people using the Lancashire Quality of Life
Profile. Social Psychiatry and Psychiatric Epidemiology 32,
7683.
Pallant J. (2007) Title SPSS Survival Manual: A Step by Step Guide
to Data Analysis Using SPSS for Windows 3edn. Open University Press, Maidenhead.
Parashos J. (1998) Athenians views on mental and physical
illness. Lundbeck, Hellas, Athens.
Skantze K., Malm U., Dencker S.J., et al. (1992) Comparison of
quality of life with standard living of schizophrenic outpatients.
British Journal of Psychiatry 161, 797801.
Stefanis C., Madianos M.G. & Gittelman M. (1986) Recent developments in the care, treatment and rehabilitation of the chronic
mentally ill in Greece. Hospital & Community Psychiatry 37,
10411044.
Tountas Y., Demakakos P.T., Yfantopoulos Y., et al. (2003) The
health related quality of life of the employees in the Greek
hospitals: assessing how healthy are the health workers. Health
and Quality of Life Outcomes 1. Available at: http://
www.hqlo.com/content/1/1/61.
Vandiver V.L. (1998) Quality of life, gender and schizophrenia: a
cross-national survey in Canada, Cuba, and U.S.A. Community
Mental Health Journal 34, 501511.
Ware J. & Sherbourne C. (1992) The MOS, 36-item Short-Form
Health Servey (SF-36): conceptual framework and item selection. Medical Care 30, 473483.
Wilkinson G., Hesdon B., Wild D., et al. (2000) Self-report quality
of life measure for people with schizophrenia: the SQLS. British
Journal of Psychiatry 177, 4246.
Yutaka O., Satsumi Y., Kim Y., et al. (1999) Schizophrenia: is it
time to replace the term? Psychiatry and Clinical Neurosciences
53, 335341.
Zissi A., Barry M.M. & Cochrane R. (1998) A mediational model
of quality of life for individuals with severe mental health problems. Psychological Medicine 28, 12211230.

2009 Blackwell Publishing