Health democracy in Europe: Cancer patient organization participation in health policy

Abstract Background Patient organization participation in health policy decision making is an understudied area of inquiry. A handful of qualitative studies have suggested that the growing number of patient organizations in Europe and their increasing involvement in policy issues do not result in high political effectiveness. However, existing research is largely country‐specific. Objective To examine the degree and impact of cancer patient organization (CPO) participation in health policy decision making in EU‐28 and to identify their correlates. Methods A total of 1266 members of CPOs participated in this study, recruited from a diversity of sources. CPO participation in health policy was assessed with the Health Democracy Index, a previously developed instrument measuring the degree and impact of patient organization participation in various realms of health policy. Additional questions collected information about participants' and the CPO's characteristics. Data were gleaned in the form of an online self‐reported instrument. Results The highest degree of CPO participation was observed with respect to hospital boards, reforms in health policy and ethics committees for clinical trials. On the contrary, the lowest was discerned with regard to panels in other important health‐related organizations and in the Ministry of Health. The reverse pattern of results was observed concerning the Impact subscale. As regards the correlates of CPO participation, legislation bore the strongest association with the Degree subscale, while organizational factors emerged as the most important variables with regard to the Impact subscale. Conclusions Research findings indicate that a high degree of CPO participation does not necessarily ensure a high impact. Efforts to promote high and effective CPO participation should be geared towards the establishment of a health‐care law based on patient rights as well as to the formation of coalitions among CPOs and the provision of training to its members.


| INTRODUCTION
In contemporary health care, patients are envisioned to assume a more active role than in the past, a revolutionary paradigm known as "patient-centred care". [1][2][3] In particular, the term describes "a partnership among practitioners and patients to ensure that decisions respect patients' desires, needs, and preferences and that patients have the education and support they need to make decisions and participate in their own care" (p. 7). 4 Nonetheless, patient participation is not constrained on the individual level, as patients may collectively participate in decision making in various subjects, including guideline development, government policy and research agenda setting among others. [5][6][7][8] A number of theoretical arguments have been put forward to bolster patient participation at the collective level. 9 Above all, it advances democratic legitimacy. Patients are affected by the consequences of certain decisions, and therefore, they should have a say in the process. In this rationale, patient participation has been shown to have an empowering effect on patients and to increase their self-efficacy. [10][11][12] Equally important is patients' experiential knowledge of a disease and its treatment, which may enhance the quality of health-care decisions. 13 Patients may offer solutions consonant with their preferences, contributing thereby to the prevention of mistakes and the containment of costs. Thus, the overall effectiveness and efficiency of the health-care system are upgraded. Indeed, there are some studies to corroborate this reasoning, with evidence indicating that patient participation may promote optimal quality of care and patient safety, 14,15 curb health-care costs 16 and enhance population health outcomes. 17 In the context of patient-centred care, patient participation is interwoven with patient empowerment. For example, the Patient Empowerment and Centredness Committee considers patient empowerment to be a prerequisite for patient involvement, which in turn fosters the establishment of a patient-centred health-care system. 18 Indeed, published models of patient empowerment have underscored the need for patient involvement and participation. [19][20][21] Empowerment also acquires a central position in the European Charter of Fundamental Rights, 22 where citizens' right to health care is exalted. In a similar vein, many countries have launched raising awareness campaigns and have even introduced pertinent legislation. An illustration of this point is the French Act of 2 March 2002, which called for a "health democracy," paving the way for patients to exert influence on their own health.
Concomitantly, the European Commission has consistently conjured patients' rights in collaboration with various stakeholders, while a number of health and patient associations have promoted patient empowerment via various Bills of Rights or Declarations. 23 In countries outside Europe, serious obstacles in issuing and implementing similar legislations have been documented 24,25 ; however, not many reports have elaborated on these issues in low-and middle-income countries. At the same time, in spite of the importance of the topic, patient empowerment and participation on the meso-level (ie regional) and macro-level (ie national or international level) are largely understudied worldwide. 26,27 While patients' collective action has been identified as an indispensable vehicle for influencing health policy and service provision, there is a dearth of research on their associations. 12,[27][28][29][30][31][32] Perhaps the most well-known study in the field is the De Montfort study in UK, 30 which set out to explore a cross-section of health consumer groups, their relationship with each other and their impact on national policy.
It is noteworthy that the study addressed 5 disease conditions: arthritis, cancer, health and circulatory disease, maternity and childbirth and mental health. Findings corroborated that contact between health consumer groups and policymakers has risen in frequency the past years, while professional bodies and pharmaceutical companies have also included patient groups and consumers in discussions on policy proposals. Nonetheless, as the authors note, "this says little about the powers of health consumer groups either individually or collectively.
It may simply be that inclusion in the policy process leads to incorporation" (p. 753). 33 In other words, the inclusion of health consumer groups in health policy decision making may solely serve the purpose of adding legitimacy to governments, while advancing their own interests. Similarly, research from the Netherlands suggests that while the Dutch model aims to render patient organizations an equal party in health policy processes, this is not met in practice. 12 Similarly, evidence from the Mixed Advisory Committees in Italy underscores the limited impact of users' voice on decision making by health authorities. 34 Moreover, in a 2-year comparative study aiming to assess the political economy and effectiveness of 500 patient associations in UK and America, Wood 27 has argued that the proliferation of patient associations in both countries has not been tantamount to high political effectiveness.
In the same study, the authors delineated the different forms of political activity and advocacy work patient organizations engage into.
Newsletters and leaflets are often used to raise members' (patients' and relatives') awareness about a medical condition, new treatment techniques, available medication, existing health facilities and exemplars of good practice. Concomitantly, raising awareness activities may also target health professionals. To this end, some associations organize highly prestigious national and international conferences, contribute to the design and delivery of educational programmes and publish bulletins aimed at physicians. Lobbying of local health authorities and lobbying at the national and international level for various purposes, such as drug approval, research funding, are also common routes of political activity among patient organizations. Finally, campaigning initiatives may range from working with the government in campaigns targeting the public to exerting overt pressures in order to change government policy. In the realm of cancer, it is the advocacy work on the part of the National Coalition for Cancer Survivorship that succeeded in introducing the concept of "survivorship" as a distinct stage on the cancer control continuum in the jargon of oncology. 35 37 The workshop recorded high engagement of health consumer groups with policymakers and political institutions, however with marked diversity among countries. It was concluded that further comparative research is needed in Europe and patient organization activities should be analysed at the pan-European level.
In this context, this study builds upon previous work on the development and validation of a Health Democracy Index, [38][39][40]  The study explored the aforementioned aims with regard to cancer patient organizations. The reason for addressing only one disease group rather than adopting a more disease-general approach is justified by evidence indicating that health conditions display marked diversity in terms of their impact on the population, their priority status within a government policy, their media and public profile as well as the degree to which they are ingrained in the medical model. 30 Furthermore, a recent study has bolstered the moderator effect of type of disease in the association between patient empowerment and therapy compliance. 43 Cancer was chosen as the disease of interest, primarily because it constitutes a major public health concern incurring significant burden to European societies. 44 It is among the main causes of morbidity and mortality worldwide, while its treatment cost is high. 45 Furthermore, inequalities in accessing cancer treatment have been reported, while the need to tackle treatment barriers has been repeatedly stressed. 46 Concomitantly, cancer has a high political profile, as it attracts great media attention and influences more people. 37

| Sample
The present sample consisted of members of cancer patient organizations (CPOs). A CPO was defined as any patient group with a legal entity (formal organizations) that addressed cancer solely. To be in-

| Instrument
The self-reported questionnaire consisted of the following sections:

| Health Democracy Index 38-40
This is an original scale measuring patient participation in health policy decision making. Earliest versions of the instrument 38,39 encompassed 8 items; however, its extended form 40

| Individual characteristics
One section of the instrument entailed questions about respondents' individual characteristics. This included their socio-demographic profile (gender, age, family status and educational attainment) and their involvement in the CPO. With regard to the latter, individuals had to rate their familiarity with the disease and their knowledge about treatment/the country's health-care system/country's reimbursement processes (very low-low-moderate-high-very high) as well as their involvement in the organization (absent-very lowlow-moderate-high-very high). Moreover, they were enquired about their position in the organization (president or other board member-employed by the organization-voting member-nonvoting but active member and non-active member) and their membership duration.

| Organizational characteristics
One section of the questionnaire recorded information about the CPO characteristics. In particular, respondents had to indicate whether their CPO provides information material to its members (yes-no) and training (yes-no). Additionally, they were asked whether their CPO was a member of a national cancer federation (yes-no), of a national federation of chronic diseases (yes-no) and of a national federation for people with disabilities (yes-no).

| Country grouping
To explore potential association between contextual factors and CPO participation, countries were aggregated on the grounds of their existing legislation. Specifically, based on information emanating from the European Health Consumer Index, 47 countries were grouped with respect to the degree to which their health-care law is based on patient rights (low-medium-high).
It merits noting that the research instrument was in the native language of each country.

| Procedure
As already described in the Sample section, a sampling frame/list consisting of existing CPOs in Europe was composed. Contacts with these organizations were made through email or telephone. Specifically, an invitation was sent out to the president of the organization or a board member. After initial acceptance of participation, the Institutional Review Board of each organization approved the study protocol. The board members of each CPO were asked to distribute the questionnaire to their members. If a member agreed to participate, he/she signed the written informed consent and was referred to the online survey link. Data were gleaned online.
The present work was conducted with the valuable input of European CPOs delegates such as the European Patient Advocates Leadership Council in Oncology.

| Sample characteristics
The majority of respondents were women (57.8%) and their mean  Table 1.

| Degree and Impact of CPO participation
As illustrated in Table 2 Interestingly, the reverse pattern of results was observed with regard to the impact dimension of the HDI (Table 3). In particular, the highest impact was observed with respect to participation in panels/ It is noteworthy that mean values in the Degree subscale were conspicuously higher than the corresponding in the Impact subscale.

| Correlates of CPO participation
When multiple regression analysis was conducted with the HDI Degree subscale as the dependent variable (Table 4), the position in the organization and membership duration were found to be independently associated with the HDI Degree composite score. Concerning organization characteristics, the CPO being a member of a national cancer federation or a member of a national federation for chronic diseases was predictive for a higher score on the subscale. Additionally, both the "medium" and "high" country categories were found to have higher HDI Degree scores as compared to the "low" country category.
Country aggregation, personal position in the organization and the CPO being a member of a national cancer federation or a member of a national federation for chronic diseases were found to bear the strongest associations with the HDI subscale scores, as evidenced by the standardized regression coefficients. Both organizational and country characteristics increased the R 2 of the model, thus indicating that an additional proportion of variance was explained by the aforementioned factors.
Multiple regression analysis for the HDI Impact subscale (Table 5) revealed that among personal characteristics position in the organization and personal involvement in the organization had a significant association. The addition of organization and country characteristics on the block of variables concerning personal characteristics increased the R 2 of the model, thus indicating that an additional proportion of variance was explained by the model. Specifically, among organization characteristics, receiving information materials or training as well as the CPO being a member of a national cancer federation or a member of a national federation for chronic diseases was predictive for a higher score on the HDI Impact subscale. Also, a greater HDI Impact score was observed in the "high" country category as compared to the low category. Personal involvement in the organization, the position in the organization, receiving information materials and the CPO being a member of a national cancer federation or of a national federation for chronic diseases had the strongest association with the HDI Impact subscale score, as evidenced by the standardized regression coefficients.

| DISCUSSION
The present study is the first endeavour to systematically investigate the degree and impact of CPO participation in health policy across Europe. To this end, a previously developed instrument assessing the degree and impact of PO participation in various realms of decision making at the meso-and macro-level (regional, national and international) was employed.
Study findings indicate that a higher degree of patient participation is not necessarily translated into a higher impact. Arguably, the two are interlinked; however, a different pattern of results-in fact the reverse-emerged between the items of the Degree and the Impact subscale. Concomitantly, the multivariate linear models revealed both similarities and differences in the correlates of the two subscales, suggesting that the two dimensions are interwoven but should not be considered identical. In other words, a high degree of patient organization participation does not guarantee the effectiveness of this participation. A similar observation has been mooted by the handful of qualitative studies in the field. 12,[23][24][25][26][27][28][29][30]33,34 Evidence from UK, USA, Germany, Italy and the Netherlands substantiates the growing number of patient organizations in these countries as well as their heightened involvement in policy processes; however, the impact of their participation is disproportionately low. 12,[23][24][25][26][27][28][29][30]33,34 This echoes the concern raised by some authors that governments may capitalize on patient participation for adding legitimacy to their own interests, 30 thus resulting to what appears to be ostensible participation. A similar concern about the distinction between representation and representativeness has long been raised by Angela Coulter, 48 who has highlighted the dangers of tokenism entailed in patient representation committees, when they are not sufficiently independent from trust management in the UK.
This distinction between degree and impact of participation, as tapped by the HDI scale, can also be viewed in the light of the difference between the so-called input democracy and output democracy. 49 Sarah Hobolt draws upon the seminal quote of A. Lincoln on democracy: that is "government by the people, of the people and for the people". 50 In her stream of argument, input democracy refers to the procedures that allow citizen participation and input into the democratic process-the "by the people" clause, whereas output democracy stresses government effectiveness and performance-the "for the people" clause. 50 In this regard, findings from the present study raise concerns about the possibility of an output (health) democracy deficiency in EU-28.
The correlates of CPO participation (degree and impact) shed light on potential ways for tackling hindrances and thus facilitating patient participation in health policy. As regards the degree of CPO participation, the aggregation of countries on the grounds of a health-care law Bold values is smaller than P < .001. β = regression coefficient; SE = standard error; β* = standardized regression coefficient; R 2 = coefficients of determination.
T A B L E 5 Multivariate regression analysis for the HDI impact subscale promoting patient rights was found to be the most important correlate, as indicated by the marked rise in R 2 value, when this block of variables was included into the model, as well as by the highest standardized coefficient of the variable "high" (ie the health-care law is based on patient rights to a high degree). Even the "medium" category appears to confer a benefit, as it increased the degree of patient participation substantially as compared to the low category. In this reasoning, legislation is an important element for securing patient participation. Of secondary importance are the organizational characteristics. Among them, coalition among CPOs or among patient organizations for chronic diseases was found to contribute markedly to enhancing the degree of CPO

| Limitations
This is the first large-scale quantitative study to systematically investigate this issue; however, it has certain limitations that warrant consideration. In the absence of a sampling frame for CPOs in Europe, the representativeness of the present sample and thus the generalizability of the study findings cannot be ensured. In this rationale, the external validity of the study may be a shortfall. At the same time, in spite of systematic efforts to recruit respondents from different sources, it is assumed that the most active and motivated CPO members participated in the study. As a result of this, responders may systematically differ from non-responders and sampling bias may have arisen in the study. A future study should concentrate on constructing an adequate sampling frame to facilitate rigour in the field. Additionally, countries were grouped in terms of legislation. An alternative way would have been to take into account characteristics of the health-care system. A future analysis from our research group will try to address this alternative strategy of analysis.
It is noteworthy that findings of the present study are constrained to cancer patient organizations and cannot be extrapolated to other health conditions. While the literature suggests substantial variation among disease groups with respect to patient participation, 16,43 no study to date has aimed at directly addressing this objective. Some evidence on patient empowerment indicates that empowerment exerts no impact on therapy compliance of patients with most severe diseases, such as cancer, whereas it enhances therapy compliance among those with less severe conditions, such as diabetes. 43 At the collective level, the De Montfort study in UK suggested that cancer and mental health groups were well represented in influencing policy, as compared to other disease groups; however, it did not specifically assess the impact on health policy as a function of disease group. 16 Therefore, research on patient participation should take into consideration differences among disease groups. By utilizing the HDI, a further study may directly set out to examine this issue.
It merits noting that among the strengths of the present study is that data were based on responses made primarily by patients and their relatives (75.5%). However, the study could not have gone into depth with regard to the internal dynamics of CPOs and more specifically the degree of internal democracy in these organizations. It has long been posited that a dearth of internal participatory practices may result in elite decision making and a misrepresentation of interests within the organization. 62 Evidence from the De Montfort study indicates that the same people-disparagingly called as "usual suspects"-tend to appear as representatives of health consumer groups, while existing formal and informal internal participatory practices (such as elections to office and annual general meetings) within patient organizations have not been systematically explored. 30 Concomitantly, the government appears oblivious about how group representativeness has been achieved and it is likely in some cases that leaders may misrepresent their membership or that a group may be controlled by professional or commercial interests. In this regard, a further study-preferably a qualitative one to allow for more indepth responses-should aim to systematically explore members' views about their leaders and internal participatory practices within organizations.
In conclusion, quantitative research on CPO participation in EU-28 indicates that a high degree of CPO participation does not necessarily lead to high impact of CPO participation, raising concerns about an output health democracy deficiency in Europe. Initiatives to enhance the former should be geared towards promoting the establishment and enforcement of a health-care law based on patient rights. In addition, the formation of umbrella organizations and the provision of information and training on CPO members may foster the political effectiveness of the organizations.