A Delphi study to identify healthcare users’ priorities for cancer care in Greece

ARTICLE IN PRESS European Journal of Oncology Nursing (2008) 12, 362–371

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/ejon

A Delphi study to identify healthcare users’ priorities for cancer care in Greece Nikolaos Efstathioua,, Jamal Ameenb, Anne-Marie Collc a

School of Health Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK Applied Statistics, University of Glamorgan, Pontypridd, Wales, UK c Faculty of Health Sport and Science, Department of Health Sciences, University of Glamorgan, Pontypridd, Wales, UK b

KEYWORDS Cancer services; Greece; Delphi technique; User involvement; Healthcare users; Priorities

Summary User involvement is widely promoted within the European healthcare sector as an effective means of developing patient-centred health services. In addition, user involvement is recommended as an important aspect of developing effective cancer services. Despite the acknowledged benefits of user involvement, this concept has not been fully endorsed in Greece. Greek healthcare users may be increasingly involved in the decisions made for their health, but their involvement in cancer services planning remains very limited. The aim of this study was to identify healthcare users’ key areas of cancer care and services that needed to be developed or improved in Greece and prioritise them through a consensus building study. A Delphi technique was used to collect data from a sample of 30 healthcare users in two rounds. The response rate for each round was over 80%. The main consensual issues prioritised by the healthcare users were the improvement of education for health professionals and the development of effective treatments. Other areas that were also rated highly included the organisation of cancer services, the need for more resources and improvements in the attitude of healthcare professionals. Healthcare users’ priorities were focused on timely diagnosis and treatment by well-educated healthcare professionals, demonstrating a strong will for survival. Based on the findings from this study and the literature recommendations, it is expected that the reorganisation and fair distribution of the existing resources could improve the experience of healthcare users. In addition, the number of employed nurses should increase and improvements in healthcare professionals’ education should be implemented in order to develop and improve cancer services. Public health education needs to be emphasised and prioritised especially in areas such as prevention and screening. The Greek government needs to address and encourage user involvement in planning healthcare services with the introduction of relevant legislation. & 2008 Elsevier Ltd. All rights reserved.

Introduction Corresponding author. Tel.: +44 121 4158587.

E-mail address: [email protected] (N. Efstathiou). 1462-3889/$ - see front matter & 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ejon.2008.04.010

Over the recent years there has been an increasing call to promote user involvement in healthcare. The World Health Organisation has called for more user involvement in cancer

ARTICLE IN PRESS A Delphi study to identify healthcare users’ priorities care through many publications (WHO, 2004, 2007). It is greatly accepted that user involvement increases accountability, promotes the interests of service users and improves the effectiveness of services and the health of the public (Crowford et al., 2003). Although the concept of user involvement has been studied extensively, it requires further clarification as it is usually believed that user involvement denotes mainly patients and carers taking part in making decisions about their own healthcare, which is only one aspect, of user involvement (Evans et al., 2003). The other aspects such as user participation in health services planning, evaluation and research in healthcare are not often acknowledged. The reason could be that involving users in health services planning and delivery or research is not an easy process (Cormie, 1999). Furthermore, listening to users requires the development of different types of skills in order to communicate effectively with people who have been traditionally excluded from decision-making (Barnes and Bennet, 1998). In response to the increasing call for user involvement in health care, several countries have demonstrated a commitment to involving healthcare users in the planning of cancer services. For example, in the United Kingdom (UK), the Department of Health has produced a series of documents to clarify the concept of user involvement and promote user participation (Department of Health, 1997, 2001a, b). Furthermore, the UK’s National Institute for Clinical Excellence (2004) suggests that individuals whose lives are affected by cancer can be viewed as experts since they are living with its consequences and emphasises the importance and value of patient and carer led activities as an integral part of cancer services. Because of these efforts, patient and public involvement in the UK are more advanced in cancer services than any other services (Sitzia et al., 2006); however, it is also accepted that the concept of user involvement has proven difficult to implement (Taylor et al., 2004). Despite the recommendations for user involvement and the evidence of its benefits, in Greece, even the latest reform of the National Health System (NHS) (MOHAW, 2005) does not make any reference to user involvement. Most of the reforms tend to focus on the structure and administration of the Greek NHS (Davaki and Mossialos, 2005). Furthermore, in the context of decision making for the planning and delivering of healthcare services, decision makers are political parties, the healthcare professionals associations and the trade unions. Healthcare users are not usually involved or informed. According to Mossialos et al. (2005), policies that could improve the healthcare sector and consequently develop user involvement have been hindered due to the resistance of doctors and trade unions. Although user involvement in Greek healthcare has not been prioritised, this has not prevented the study of some aspects of user involvement such as the carers’ informational needs and satisfaction with communication with healthcare professionals (Iconomou et al., 2001) or satisfaction with the NHS services (Niakas et al., 2004; Gnardellis and Niakas, 2005), all of which give an indication of healthcare users’ views about the healthcare services offered. However, Edwards and Staniszewska (2000) suggest that satisfaction surveys give little indication of a user’s

363 experience of care and what users would like to see improved. It remains that no published research exists to identify the healthcare users’ views on service planning and delivery, in particular in identifying priorities in the area of cancer care services. It was expected that research in this field would give healthcare users an active role to directly influence healthcare decisions by making their priorities in cancer care known through the publication of this study.

The study Aim This study was undertaken to investigate healthcare users’ views on cancer care services in Greece and prioritise those areas that needed to be developed or improved. It is a part of a larger project, which also investigated healthcare providers’ views on this area for the same purpose (Efstathiou, 2004).

Methods The Delphi technique was considered for this study. The Delphi technique is a multi-staged approach involving a series of questionnaires that are distributed to an ‘expert panel’. The aim is to achieve a consensus of opinion amongst the expert panel (Hauck et al., 2007). The first questionnaire seeks to elicit a response relating to a broad subject area. The subsequent questionnaires are then developed from the returns gained from the initial questionnaire. The responses of each questionnaire are summarised and fed back to the panel who are asked to reconsider their judgement and respond in the light of this new information. Questionnaires are distributed until a consensus is reached (Hasson et al., 2000). Additionally, Delphi technique offers the combination of methods embedded both in the quantitative and in the qualitative paradigms (Bowles, 1999), which would not only help in identifying the participants’ personal views in this study, but reach a consensus on their priorities based on statistical means. It also meant that travelling for panel meetings was not necessary (Irvine, 2005). Furthermore, the use of a Delphi technique for this study would make comparisons possible based on the identical study which investigated healthcare providers’ views and priorities in cancer care (Efstathiou et al., 2007). Although Delphi technique has proved useful to researchers in health care (Zinn et al., 2001), there is considerable criticism on the adaptations to the Delphi method which have threatened the reliability and validity of findings (McKenna and Keeney, 2008). The most common areas of concern are the ‘expert’ panel composition, feedback and consensus (McKenna, 1994; Hasson et al., 2000). These areas are addressed and discussed in relevance to this study in the sections to follow. The Delphi ‘expert’ panel Green et al. (1999) indicate that the first stage of a Delphi technique is the formation of a ‘panel of experts’. Although it is generally acknowledged that there is little agreement on what an ‘expert’ is (Baker et al., 2006), a critical review

ARTICLE IN PRESS 364 of the Delphi technique by Keeney et al. (2001) cites a range of definitions of ‘expert’ including ‘informed individual’, ‘specialist in the field’ or ‘someone who has knowledge about a specific subject’. The participants in this study were selected and considered as ‘experts’ based on their knowledge and experience of living with cancer or caring for a cancer patient and using cancer services. Cancer patients and carers had to fulfil the criterion of having been diagnosed with cancer or caring for a cancer patient for at least a year. A year was considered a reasonable time that could provide adequate exposure to a range of cancer services within the healthcare system for cancer patients and their carers. The number of participants in Delphi studies is another area of consideration among researchers. According to Hasson et al. (2000), the scope of the problem and the resources available dictate the number of participants. However, a heterogeneous sample is necessary to ensure that the entire spectrum of opinion on a topic is determined (Keeney et al., 2001). Linstone and Turoff (2002) advise a panel size of anywhere from 10 to 50 participants. For this study, a purposive convenience sample was used to recruit a panel of 30 healthcare users from four hospitals in Athens, Greece (two oncology and two general hospitals). It was anticipated that 30 participants (15 cancer patients and 15 carers) would provide a wide spectrum of opinion for this study. Either a cancer patient or a carer was allowed to participate in the study from any eligible pair, but not both participating at the same time. The recruitment of the panel took place in the broader area of Athens, where 35% of the population of Greece resides (General Secretariat of National Statistical Service of Greece, 2004) with a caution to create a panel that would represent a wider geographical area.

Ethical considerations University ethical approval was obtained to conduct this study and approval was granted by the individual healthcare settings in which the participants were approached. Major ethical considerations had to be taken into account as there is a culture of not revealing a cancer diagnosis to cancer patients in Greece. Although Mystakidou et al. (2005) acknowledge that more physicians tend to inform patients of their cancer diagnosis, a study by Iconomou et al. (2002) revealed that of 100 cancer patients receiving chemotherapy, only 37% knew they had cancer. Only patients who were aware of their cancer diagnosis were recruited for this study. For this reason, the nurses in charge of the wards, where recruitment was taking place for the study, advised on which patients knew of their cancer diagnosis before being approached and invited to participate. Furthermore, in order to avoid possible unwanted revelations, only carers whose patients knew they had cancer were involved. The participants were approached directly by the researcher and were provided with realistic information about the requirements and time commitment involved in participating in the study. Reassurance was given that their identities would remain anonymous to the rest of the panel members and that they could withdraw from the study at any time. Participants were also assured that any suggestions provided would not influence their current or future

N. Efstathiou et al. treatment. A consent form was also signed by all the participants.

Questionnaires and development of consensus Round 1 Traditionally, round 1 is used to generate ideas, and the panel members are asked for their responses to or comments about an issue. Round 1 starts with an open-ended question, thus allowing panel members freedom in their responses (Keeney et al., 2001). Accordingly, the first round questionnaire asked the 30 participants to suggest the areas of cancer care and services, which needed development or improvement in Greece. The analysis of the first round data included the transcription of the responses in a data handling software programme (Approach by Lotus Smart Suite). A panel of seven experts in qualitative data analysis clustered the data into thematic groups. Some of the statements received were detailed and contained more than one variables; however, in line with a classical Delphi approach they were retained without editing, preserving their meaning. The unedited statements provided by the healthcare users’ panel within the thematic groups formed the questionnaire for the second round. Round 2 The second round questionnaire was posted to the same 30 healthcare users. This questionnaire formed the first feedback the participants received of the suggestions provided by the rest of the panel. The process of feedback in Delphi studies is paramount as this is the only communication between the participants (Murphy et al., 1998). It has been suggested that the feedback between rounds can widen knowledge and stimulate new ideas and be highly motivating and educational for the participants (Powell, 2003). Furthermore, it leads to perceptions of ownership and an increased chance of acceptance of the findings (McKenna, 1994). The second round questionnaire asked the panel to rate each statement on a seven Likert type scale from 1 ‘very low priority’ to 7 ‘very high priority’. The data from the second round were analysed using the Statistical Package for Social Sciences (SPSS Version 11). The means and standard deviations (SDs) were calculated for item scores (Norwood, 2000). Furthermore, correlation coefficients were computed for the mean scores between the groups of healthcare users (patients/carers from general/cancer hospital) in order to examine the association between raters. Consensus The purpose of a Delphi technique is to elicit a consensus of opinion about a subject area (Beretta, 1996); yet one of the most contentious issues that features in the literature is how consensus is established in a Delphi study. Most of the studies set consensus at different levels, using different measurements such as percentages (Orton, 1981), median or mean scores (Mead, 1993) and SDs (White, 1991). In this study, mean scores were used to measure the group opinion of the panel with regard to the priorities, and SDs were used to represent the extent of agreement within the panel. This

ARTICLE IN PRESS A Delphi study to identify healthcare users’ priorities strategy was implemented based on the experience gained from an identical Delphi study of Greek healthcare providers (Efstathiou et al., 2007) and the Delphi studies conducted by Broomfield and Humphris (2001) and Irvine (2005).

Results All the recruited sample members responded in round 1, reflecting the effective recruitment process, a clear commitment to the study and an interest in the topic. Fifty percent of the sample were cancer patients or carers from oncology hospitals and 50% cancer patients or carers from general hospitals. Among the 30 healthcare users of the first round, 27% stated an agricultural or semi-urban district as their permanent residence. Not surprisingly, 80% of the sample were female mainly due to the fact that in Greece, the carer’s role is often undertaken by a female member of the family. The mean age for cancer patients was 54 years (SD ¼ 9.1) and the average period they had been diagnosed with cancer was 5.9 years (SD ¼ 2.4). For the carers the mean age was 42.6 years (SD ¼ 7.4). It was difficult to identify how long carers were caring for their cancer patients as there were periods during the cancer trajectory where no form of caring was offered by the carers. The types of cancer reported by the participants were breast cancer (27%), bowel cancer (20%), lung cancer (13%), cancer of the uterus (10%), cancer of the larynx (10%), prostate cancer (10%), stomach cancer (7%) and kidney cancer (3%). The participants were asked to provide a maximum of five suggestions for the first round. In response to this, 123 statements were generated. The average number of statements generated by each participant was four. The cancer patients and carers from the two general hospitals provided 73 statements (60%), while the cancer patients and carers from the two oncology hospitals provided 50 statements (40%). After the research panel analysed the first round data, 27 themes emerged and the 123 statements were allocated under these themes. Forty-four statements were listed under the themes of ‘education/training’, ‘resources’, ‘psychological support’ and ‘information’. Other themes that emerged included ‘organisation’, ‘public health/health promotion’, ‘staff attitude’ and ‘pain management’. For the second round, 24 questionnaires were returned (80% response rate). In this round, the percentage of cancer patients treated in general hospitals fell from 27% initially to 21%. Half of the 123 statements had a mean score of six and above out of a maximum of seven. The main issues prioritised by the healthcare users were the education of health professionals and treatment. Other areas that were also rated highly included the organisation of cancer services, the need for more resources and the attitude of healthcare professionals. In general, healthcare users were focused on timely diagnosis and treatment by well-educated healthcare professionals, demonstrating a strong will for survival. The highest 10 rated statements were not uniformly highly rated by the panel (Table 1). The need for research to find more effective cancer treatments received high ratings by cancer patients and carers from cancer hospitals but not by their counterparts from general hospitals.

365 Although half of the 123 statements received a mean score of six and above by the participants, there were statements that received very low priority ratings (Table 2). Healthcare users were not entirely undiscriminating. It was not surprising that the lower rated statement by the participants expressed a positive view about the healthcare professionals and the setting where care was offered. A Pearson’s moment–product correlation was measured to demonstrate the correlation of the mean scores of the statements among the groups within the panel (Table 3). A strong correlation coefficient would mean that if one group had rated a statement highly, the other group would have also done the same, thus demonstrating an indication of consensus. A significant positive correlation was found between the scores obtained from all the groups within the panel of healthcare users. The highest correlation coefficient was observed between the patients in general and cancer hospitals, probably due to the similarities of personal experiences among those two groups. Based on the earlier described consensus strategy, to accord with the definitions ascribed to the numerical values on the Likert type scale, a mean score of 6.0 or above was selected to represent the panels’ priority with no rating below 5.0, thus a small SD, which would indicate agreement. Eighteen statements met the consensus criteria and are presented in Table 4. The statements are presented under the thematic group from which they originated and the number next to each category indicates the number of statements under this thematic group. It is generally difficult to judge the results obtained from a Delphi study as there is little evidence to suggest that the same results could be obtained if the same information were given to two or more panels (Hasson et al., 2000). However, the 30 participants in this study were regarded as a fairly heterogeneous group which is an essential criterion to increase the reliability of results in priorities setting Delphi studies (Duffield, 1993). In some Delphi studies, the concept of trustworthiness is normally addressed rather than that of reliability. With regard to the trustworthiness of the Delphi technique for this study, this was based on the clear formulation of the initial question, the careful transcribing of individual responses and documenting response rates over successive rounds as suggested by Crisp et al. (1997). Content validity in Delphi studies is usually assured when the participants have knowledge and an interest in the topic investigated (Goodman, 1987), which was the case in this study. In general, the validity of Delphi studies is judged by obtaining good response rates at every round (Lopez, 2003; Hasson et al., 2000) and evaluating the findings by comparison with the results of other existing studies (Powell, 2003). In this study, high response rates in both rounds were achieved and the findings are discussed and compared with relevant published evidence in the following section.

Discussion Using a Delphi technique yielded insights of healthcare users’ cancer care priorities. This study shed light on the services and care expected to be developed or improved by participants diagnosed with cancer or caring for a cancer

ARTICLE IN PRESS 366

Table 1

N. Efstathiou et al.

The 10 highest rated statements and their rank by each group.

Statements

All doctors should be aware of cancer symptoms in order to diagnose cancer early, at its beginning New methods of treatment should be discovered, more effective Better allocation of the existing CAT scans and all the special equipment and not concentration in Athens and Thessaloniki Doctors and nurses working in special cancer centres should be properly educated First of all, there should be sufficient space, because it is not fair for 16 patients to be in the same room, and it is depressing for somebody at the beginning of the disease to watch somebody in the last stages There should be a possibility for patients with aggressive cancer to be treated out of waiting lists The Greek state should give more attention to the healthcare sector; patients should not be nursed in the hospital’s corridors Free drug provision by all insurance companies Hospitals should grant the appropriate equipment and the appropriate specialised staff Doctors should give more attention and afford more time for the examination of the patients

All healthcare users

General hospital carers

Cancer hospital patients

Cancer hospital carers

General hospital patients

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

Mean

Rank

6.75

1

6.71

7

6.50

10

6.83

1

7.00

1

6.63

2

6.57

16

6.67

3

6.67

2

6.60

20

6.63

3

6.86

2

6.50

10

6.17

19

7.00

1

6.58

4

6.57

16

6.33

24

6.67

2

6.80

7

6.58

5

6.43

32

6.50

10

6.50

6

7.00

1

6.58

6

6.57

16

6.83

1

6.17

19

6.80

7

6.54

7

6.71

13

6.33

24

6.33

11

6.80

7

6.54

8

6.57

16

6.67

3

6.17

19

6.80

7

6.54

9

6.86

2

6.17

39

6.50

6

6.60

20

6.54

10

6.57

16

6.33

24

6.33

11

7.00

1

patient. Eighteen statements were consensual and despite some of them containing more than one variable, it was not difficult to identify the prevailing thematic groups. The improvement of medical training and education in order to develop skills to be able to diagnose cancer early was highly consensual among the panel of healthcare users. It is acknowledged that early detection of cancer, by either diagnosing it early in symptomatic populations or screening in asymptomatic populations, offers a greater chance for curative treatment to be effective (WHO, 2002). In Greece, there have been concerns over the undergraduate education regarding oncology and a recent study by Karamouzis et al. (2006) of 210 medical students revealed that the quality of education in oncology has improved, but it has not reached high standards. Although early diagnosis was highly prioritised, the panel did not provide any statement on screening. As no organised cancer screening programme is provided on

a national level, the efficacy of screening has not been demonstrated in Greece (Vivilaki et al., 2005). It is possible that the Greek population is not fully aware that early detection of many cancers through screening may lead to improved long-term survival rates. The notion of research to discover new and more effective treatment was highly consensual among the panel. This priority may demonstrate a lack of confidence in currently provided treatments and forms an issue that could be further investigated in Greece. A different attitude was demonstrated by Corner et al. (2007) in a study investigating cancer research priorities, where 105 cancer patients, representing broadly the UK population, ranked research on any type of treatment seventh. The organisation of diagnosis and treatment was also highly consensual among healthcare users. Three statements reached consensus in the second round, which indicated that patients should not

ARTICLE IN PRESS A Delphi study to identify healthcare users’ priorities have to complete tests in different places and that test results should be completed and communicated more efficiently. The panel also stressed the need for cancer treatment to be offered in one place. These findings were not surprising, as the absence of central planning and the lack of objective criteria and rules for the dissemination of biomedical technology in Greece have resulted in inequalities of healthcare provision from hospital to hospital, and the unnecessary movement of patients (Giokas, 2001). A number of consensual statements indicated the development of special cancer services providing care by specialist staff. Campbell et al. (1999) have indicated that patients cared for by specialists have been reported to receive more up to date treatment and have improved chances of survival. Furthermore, centralisation of cancer services may have advantages in terms of concentrating clinical expertise, enhancing the range of ancillary facilities and rationalising the provision of expensive specialist equipment, but it is not known to what extent patients may be affected by additional travel (Payne and Jarrett, 2000). Due to the geographical particularities of Greece, centralisation of services might pose problems for patients in accessing high-quality services. Further research would be required to explore the willingness of Greek cancer patients from remote areas to travel for treatment to cancer centres. Among the highest consensual statements, ‘resources’ were a repetitive theme with consensual statements suggesting increases in the number of beds and the space Table 2 The five lowest rated statements by healthcare users. Statements

Mean

SD

My views regarding the staff and the place of treatment are positive The nursing staff in general hospitals are uninterested and they are not properly educated to treat cancer patients National information net There should be a place where all the oncologists could meet and exchange views and information A lot of doctors do not inspire the feeling of trust to the patients and their relatives, so patients prefer to go abroad for better attention, diagnosis and treatment

3.96

1.63

4.63

1.44

4.88 4.88

1.33 1.73

5.00

1.35

Table 3

367 available for patients or increases in human resources. The prioritisation and consensus on these areas were not unexpected as there have been previous reports of dissatisfaction with accommodation facilities and the lack of nursing staff in hospitals (Niakas et al., 2004; Polyzos and Yfantopoulos, 2000). A recent large satisfaction survey in Greece revealed that 4209 patients were only fairly satisfied with accommodation facilities in hospitals (Gnardellis and Niakas, 2005). With regard to the lack of human resources and especially nursing shortages, this is an issue that has been well documented for a long time (Plati et al., 1998). Greece remains a country with one of the lowest ratios of nurses per 1000 habitants (3.9), less than half the European Union average (Mossialos et al., 2005). More recently, data from the General Secretariat of National Statistical Service of Greece (2008) for the number of nurses employed in hospitals in the year 2005 do not demonstrate a vast improvement. It has to be mentioned that the issue of human resources was a consensual priority, although it was not very highly prioritised. The possible reason could be the close bonds found in Greek families, especially towards severe problems such as cancer. Greek members of the family tend to provide basic care requirements to cancer patients either at hospital or in the community (Mystakidou et al., 2002; Iconomou et al., 2001). Merkouris et al. (1999), in a study of 103 patients in a Greek teaching hospital, revealed that a large percentage of patients (68.5%) were assisted by relatives or private nurses; however, even in their study 31.6% of the sample mentioned staff shortages as a major concern. Aspects of palliative care such as pain relief and psychological support, which are essential in the support of patients and their families facing the problems associated with life-threatening illness (WHO, 2002), were also consensual in this study. With regard to pain management, the consensual statement suggested improvements in the procedures of supplying opioid analgesics. This re-iterates a recommendation by WHO (2002) for all countries, regardless of their available resources for health. It was expected that further statements relevant to pain management would be highly rated, as cancer pain is reported by almost 50% of patients at all stages of the disease and by over 70% of patients with advanced neoplasms (Higginson and Hearn, 1997). Even the little evidence from Greece demonstrates that pain among cancer patients is evident and unrelieved (Panteli and Patistea, 2007; Mystakidou et al., 2007). It is anticipated that cancer patients and their carers may perceive pain as an inevitable part of the disease and treatment (Newell et al., 1998), and this is possibly why they did not consider pain management as a highest priority issue. Further research would be required in this area to

Pearson correlation (**po0.01) for the groups within the healthcare users’ panel.

Groups of participants

Patients cancer hospital

Carers general hospital

Carers cancer hospital

Patients general hospital Patients cancer hospital Carers general hospital

0.605** x x

0.581** 0.480** X

0.580** 0.580** 0.441**

ARTICLE IN PRESS 368

Table 4

N. Efstathiou et al.

The consensual statements by healthcare users.

Category (no. of statements under category)

Statements

Mean score

Education/training (11)

All doctors should be aware of cancer symptoms in order to diagnose cancer early, at its beginning Doctors and nurses working in special cancer centres should be properly educated New methods of treatment should be discovered, more effective Better allocation of the existing CAT scans and all the special equipment and not concentration in Athens and Thessaloniki Tests and, in particular, biopsies should be completed faster, giving priority to patients who need immediate treatment Organised provision of cancer care services, so that cancer patients are not transferred from hospital to hospital for examinations or special treatments First of all, there should be sufficient space, because it is not fair for 16 patients to be in the same room, and it is depressing for somebody at the beginning of the disease to watch somebody in the last stages Improvement of the wards. Less beds in the rooms Free drug provision by all insurance companies Hospitals should grant the appropriate equipment and the appropriate specialised staff They should not leave the patients in the corridors for hours or even days, having as a result their physical and psychological exhaustion More ‘day clinics’ should be developed; the existing beds are not enough Cancer centres should be built in the district, so that patients with special needs will not have to travel to Athens or Thessaloniki Appropriate places for chemotherapy treatment The procedure of supplying opioid analgesics to patients must be improved Psychological support for the patient and close relatives Specialised medical and nursing staff should be increased Nurses should treat cancer patients like all the other patients, making them feel comfortable and forget their problem

6.75

New treatment (1) Organisation (8)

Resources (11)

Finance (2) Technology (5) Extra beds (2) Day units (2) Special cancer units (7)

Pain management (6) Psychological support (11) Staff shortages (5) Staff skills (1)

identify possible cultural issues that may affect reports of pain among Greek cancer patients. The need for psychological support for the patients and close relatives was also a consensual priority, although it was not amongst the highest priorities. It is possible that due to the strong links among families in Greece, psychological problems resulting from cancer diagnosis and treatment are usually resolved within the family (Iconomou et al., 2001). In addition, a Greek study among 146 relatives of cancer patients revealed that 22% of the participants would not like to liaise with psychological support specialists (Mystakidou et al., 2002), providing a cultural attitude that needs to be further explored. Literature from other countries presents psychological support as a very high priority among healthcare users. In a Finnish study among 168 carers of cancer patients, psychological support was regarded as very important by half of the participants (Eriksson and Lauri, 2000). It was surprising that statements within thematic groups such as ‘communication and information’ or ‘prevention’ did not attract high ratings or achieve consensus. Communication and information are essential in healthcare, especially if patients have to remain autonomous in making

6.58 6.63 6.63 6.42 6.38 6.58

6.46 6.54 6.54 6.38 6.38 6.33 6.04 6.25 6.25 6.29 6.21

decisions about their health or participate in decision making for the development and improvement of cancer services. It is acknowledged that cancer patients participating in the current study were aware of their diagnosis and had probably received adequate information about their situation. In a study by Iconomou et al. (2002), 77% of 100 chemotherapy outpatients in a Greek university hospital also expressed satisfaction with the information they had received. With regard to prevention, it was not highly rated by the panel, although it may offer the greatest public health potential and the most cost-effective long-term cancer control (WHO, 2002). It is possible that the panel consisting of cancer patients or carers involved in the care of a cancer patient did not consider prevention important at that stage. However, it was expected that carers would prioritise the area of prevention as healthy individuals, but it has to be mentioned that the entire health system in Greece is treatment orientated and health promotion has not been emphasised (Petridou et al., 1999). With this study, through a process of consensus building, the panel of healthcare users have agreed on certain priorities for cancer care and services that need to be improved or developed in Greece. Although the findings

ARTICLE IN PRESS A Delphi study to identify healthcare users’ priorities represent ‘expert’ opinion, it is suggested that the findings of any Delphi study do not always offer indisputable fact (Powell, 2003). It could be further argued that the panel was not representative of the whole population in Greece since the sample was a convenience one and the results might be of limited value. However, every effort was made to provide a representative sample of patients and carers of patients with different types of cancer, residing in different parts of Greece. The decision to stop the study at the second round, based on the statistical analysis that indicated consensus, provided a long list of priorities. A third round might have provided a smaller and more comprehensive list of priorities. However, it was considered that another lengthy questionnaire could be tiring for cancer patients, especially during or after intensive treatments, with a possible effect on attrition and consequently the validity of the results.

Conclusion The acknowledged difficulties in defining and implementing user involvement in healthcare have not hindered research, which has shown its benefits. Although user involvement is widely promoted in Europe, with some countries already having in place policies to address and encourage user involvement, in Greece, the concept of user involvement remains unexplored and healthcare users are seldom involved in decisions for healthcare and health services. This study gave the opportunity to a panel of cancer patients and carers to convey and agree on certain priorities for cancer care and cancer services in Greece. The Delphi technique, which was used to collect data in this study, was designed and administered to ensure that the concerns mentioned in the literature on the validity and reliability of the Delphi technique were as possibly addressed. The results showed that early diagnosis, organised treatment in specialised hospitals or wards by properly trained personnel were the main priorities for the panel of healthcare users. The study also identified some areas pertinent to Greece that require further research, such as the general knowledge of the population about the benefits of cancer prevention and screening, the willingness of cancer patients to travel for treatment at specialised cancer centres and cultural factors affecting the reports of pain. Based on the consensual priorities, the available literature, the recommendations by WHO (2002) and the availability of resources for healthcare in Greece, it would be recommended that the organisation and fair redistribution of the existing resources may reduce cancer patients’ unnecessary movement for tests or treatments and may speed up the process of diagnosis. The development of public health education on areas such as cancer prevention and screening could reduce the number of people developing cancer and promote the diagnosis of cancer at an early stage with better patient outcomes. Although the recommendations for employment of more nurses and improvements in healthcare professionals’ education may require extra financial resources by the Greek state, it has to be considered that appropriately educated healthcare personnel may develop the required services and provide effective care both in the community and in the hospital settings. Finally, the Greek government should consider the formation

369 of legislation to encourage user involvement in healthcare, as it is not likely to occur without this support.

Conflict of interest statement There is no known conflict of interest for any of the authors associated with this paper.

Acknowledgements The study was funded by the European Oncology Nursing Society (EONS/Roche grant). We would like to thank Professor Donna Mead and Professor Laurie Moseley for their advice and supervision during the initial stages of the study and their involvement in the study conception.

References Baker, J., Lovell, K., Harris, N., 2006. How expert are the experts? An exploration of the concept of ‘expert’ within Delphi panel techniques. Nurse Researcher 14 (1), 59–70. Barnes, M., Bennet, G., 1998. Frail bodies, courageous voices: older people influencing community care. Health and Social Care in the Community 6 (2), 102–111. Beretta, R., 1996. A critical review of the Delphi technique. Nurse Researcher 3 (4), 79–89. Bowles, N., 1999. The Delphi technique. Nursing Standard 13 (45), 32–36. Broomfield, D., Humphris, G., 2001. Using the Delphi technique to identify the cancer education requirements of general practitioners. Medical Education 35, 928–937. Campbell, N., Ritchie, L., Cassidy, J., et al., 1999. Systematic review of cancer treatment programmes in remote and rural areas. British Journal of Cancer 80 (8), 1275–1280. Cormie, J., 1999. The Fife user panels project: empowering older people. In: Barnes, M., Warren, L. (Eds.), Paths to Empowerment. The Policy Press, Bristol. Corner, J., Wright, D., Hopkinson, J., et al., 2007. The research priorities of patients attending UK cancer treatment centres: findings from a modified nominal group study. British Journal of Cancer 96, 875–881. Crisp, J., Pelletier, D., Duffield, C., et al., 1997. The Delphi method. Nursing Research 46, 116–118. Crowford, M.J., Rutter, D., Thelwall, S., 2003. User involvement in change management: a review of the literature. National Coordinating Centre for the NHS and R&D Service Delivery and Organisation Programme, London. Davaki, K., Mossialos, E., 2005. Plus c- a change: health sector reforms in Greece. Journal of Health Politics, Policy and Law 30 (1–2), 143–168. Department of Health, 1997. Involving patients: examples of good practice. Department of Health, London. Department of Health, 2001a. Involving patients and the public in healthcare. A discussion document. The Stationary Office, London. Department of Health, 2001b. Shifting the balance of power within the NHS. The Stationary Office, London. Duffield, C., 1993. The Delphi technique: a comparison of results obtained using two expert panels. International Journal of Nursing Studies 30, 227–237. Edwards, C., Staniszewska, S., 2000. Accessing the users’ perspective. Health and Social Care in the Community 8 (6), 417–424. Efstathiou, N., 2004. Cancer care services in Greece: a Delphi approach to investigate the views of healthcare providers and users. Ph.D. Thesis, University of Glamorgan.

ARTICLE IN PRESS 370 Efstathiou, N., Ameen, J., Coll, A.M., 2007. Healthcare providers’ priorities for cancer care: a Delphi study in Greece. European Journal of Oncology Nursing 11 (2), 141–150. Eriksson, E., Lauri, S., 2000. Informational and emotional support for cancer patients’ relatives. European Journal of Cancer Care 9, 8–15. Evans, S., Tritter, J., Barley, V., et al., 2003. User involvement in UK cancer services: bridging the policy gap. European Journal of Cancer Care 12, 331–338. General Secretariat of National Statistical Service of Greece, 2004. Cencus 2001. http://www.statistics.gr/gr_tables/S1101_SAP_1_ TB_DC_01_04_Y.pdf (accessed 07/04/2008). General Secretariat of National Statistical Service of Greece, 2008. Social Statistics. Health: hospitals, beds, doctors and nursing personnel by speciality. Year 2005. http://www.statistics.gr/ eng_tables/S801_SHE_7_TB_AN_05_1_Y.pdf (accessed 08/02/ 2008). Giokas, D., 2001. Greek hospitals: how well their resources are used. Omega 29, 73–83. Gnardellis, C., Niakas, D., 2005. Factors influencing inpatient satisfaction. An analysis based on the Greek National Health System. International Journal of Healthcare Technology and Management 6 (3), 307–320. Goodman, C.M., 1987. The Delphi technique: a critique. Journal of Advanced Nursing 12, 729–734. Green, B., Jones, M., Hughes, D., Williams, A., 1999. Applying the Delphi technique in a study of GPs information requirement. Health and Social Care in the Community 7 (3), 198–205. Hasson, F., Keeney, S., McKeena, H., 2000. Research guidelines for the Delphi survey technique. Journal of Advanced Nursing 32 (4), 1008–1015. Hauck, Y., Kelly, R.G., Fenwick, J., 2007. Research priorities for parenting and child health: a Delphi study. Journal of Advanced Nursing 59 (2), 129–139. Higginson, I., Hearn, J., 1997. A multicenter evaluation of cancer pain control by palliative care teams. Journal of Pain and Symptom Management 14 (1), 29–35. Iconomou, G., Viha, A., Kalofonos, H., et al., 2001. Impact of cancer on primary caregivers of patients receiving radiation therapy. Acta Ongologica 40 (6), 766–771. Iconomou, G., Viha, A., Koutras, A., et al., 2002. Information needs and awareness of diagnosis in patients with cancer receiving chemotherapy: a report from Greece. Palliative Medicine 16, 315–321. Irvine, F., 2005. Exploring district nursing competencies in health promotion: the use of the Delphi technique. Journal of Clinical Nursing 14 (8), 965–975. Karamouzis, M.V., Ifanti, A.A., Iconomou, G., et al., 2006. Medical students’ views of undergraduate oncology education: a comparative study. Education for Health: Change in Learning & Practice 19 (1), 61–70. Keeney, S., Hasson, F., McKenna, H.P., 2001. A critical review of the Delphi technique as a research methodology for nursing. International Journal of Nursing Studies 38, 195–200. Linstone, H., Turoff, M., 2002. The Delphi method: techniques and applications. http://www.is.njit.edu/pubs/delphibook/delphibook.pdf (accessed 14/04/08). Lopez, V., 2003. Critical care nursing research priorities in Hong Kong. Journal of Advanced Nursing 43 (6), 578–587. McKenna, H.P., 1994. The Delphi technique: a worthwhile research approach for nursing? Journal of Advanced Nursing 19, 1221–1225. McKenna, H., Keeney, S., 2008. Delphi Studies. In: Watson, R., McKenna, H., Cowman, S., Keady, J. (Eds.), Nursing Research: Designs and Methods. Churchill Livingstone, Edinburgh. Mead, D., 1993. The development of primary nursing in National Health Service care giving institutions in Wales. School of Nursing, University of Wales, Swansea.

N. Efstathiou et al. Merkouris, A., Yfantopoulos, J., Lanara, V., Lemonidou, C., 1999. Developing an instrument to measure patient satisfaction with nursing care in Greece. Journal of Nursing Management 7 (2), 91–100. MOHAW, 2005. National Health System and Social Solidarity Legislation. Greek Ministry of Health and Welfare. http://www. mohaw.gr/gr/theministry/nomothesia/nomothesia4/Attachment 00054076/e-ygia-pap.pdf (accessed 07/04/2008). Mossialos, E., Allin, S., Davaki, K., 2005. Analysing the Greek health system: a tale of fragmentation and inertia. Health Economics 14 (Suppl. 1), S151–S168. Murphy, M.K., Black, N., Lamping, D.L., et al., 1998. Consensus development methods and their use in clinical guideline development. Health Technology Assessment 2 (3), 1–88. Mystakidou, K., Parpa, E., Tsilika, E., et al., 2002. The families evaluation on management, care and disclosure for terminal stage cancer patients. BMC Palliative Care 1 (3) (article available from http://www.biomedcentral.com/1472-684X/1/3). Mystakidou, K., Tsilika, E., Parpa, E., et al., 2005. Patterns and barriers in information disclosure between health care professionals and relatives with cancer patients in Greek society. European Journal of Cancer Care 14, 175–181. Mystakidou, K., Tsilika, E., Parpa, E., et al., 2007. Exploring the relationships between depression, hopelessness, cognitive status, pain and spirituality in patients with advanced cancer. Archives of Psychiatric Nursing 21 (3), 150–161. Newell, S., Sanson-Fisher, R., Girgis, A., Bonaventura, A., 1998. How well do medical oncologists’ perceptions reflect their patients’ reported physical and psychosocial problems? Cancer 83 (8), 1640–1651. Niakas, D., Gnardellis, C., Theodorou, M., 2004. Is there a problem with quality in the Greek hospital sector? Preliminary results from a patient satisfaction survey. Health Services Management Research 17 (1), 62–69. National Institute for Clinical Excellence, 2004. Guidance on Cancer Services: Improving Supportive and Palliative Care for Adults with Cancer. National Institute for Clinical Excellence, London. Norwood, S., 2000. Research Strategies for Advanced Practice Nurses. Prentice-Hall, London. Orton, H.D., 1981. Ward learning climate and student nurse response. Nursing Times 77 (23), 65–68. Panteli, V., Patistea, E., 2007. Assessing patients’ satisfaction and intensity of pain as outcomes in the management of cancerrelated pain. European Journal of Oncology Nursing 11 (5), 424–433. Payne, S., Jarrett, N., 2000. The impact of travel on cancer patients’ experiences of treatment: a literature review. European Journal of Cancer Care 9, 197–203. Petridou, E., Mossialos, E., Papoutsakis, G., et al., 1999. Public health policies and priorities in Greece. In: Holland, W., Mossialos, E. (Eds.), Public Health Policies in the European Union. Ashgate, England. Plati, C., Katostaras, T., Mantas, J., Lanara, V., 1998. Nursing manpower development and strategic planning in Greece. Image: The Journal of Nursing Scholarship 30 (4), 329–333. Polyzos, N., Yfantopoulos, I., 2000. Health manpower development and national health system staffing. Archives of Hellenic Medicine 17 (6), 627–639. Powell, C., 2003. The Delphi technique: myths and realities. Journal of Advanced Nursing 41 (4), 376–382. Sitzia, J., Cotterell, P., Richardson, A., 2006. Interprofessional collaboration with service users in the development of cancer services: the Cancer Partnership Project. Journal of Interprofessional Care 20 (1), 60–74. Taylor, G., Brown, K., Caldwell, K., et al., 2004. User involvement in primary care; a case study examining the work of one patient participation group attached to a primary care practice in North London. Research Policy and Planning 22 (1), 21–30.

ARTICLE IN PRESS A Delphi study to identify healthcare users’ priorities Vivilaki, V., Romanidou, A., Theodorakis, P.N., et al., 2005. Are health education meetings effective in recruiting women in cervical screening programmes? An innovative and inexpensive intervention from the island of Crete. Rural and Remote Health 5 (2), 8 (Available from: http://www.rrh.org.au). White, E., 1991. The future of psychiatric nursing by the year 2000: a Delphi study. Report, Department of Nursing, University of Manchester, Manchester. WHO, 2002. National Cancer Control Programmes: Policies and Managerial Guidelines, second ed. World Health Organisation, Geneva.

371 WHO, 2004. Strategies to improve and strengthen cancer control programmes in Europe. World Health Organisation, Geneva. WHO, 2007. Palliative care (cancer control: knowledge into action: WHO guide for effective programmes; module 5). World Health Organisation, Geneva. Zinn, J., Zalokowski, A., Hunter, L., 2001. Identifying indicators of laboratory management performance: a multiple constituency approach. Health Care Management Review 26, 40–53.