INEQUALITIES OF ACCESS TO CANCER SCREENING

INEQUALITIES OF ACCESS TO CANCER SCREENING: A LITERATURE REVIEW Lai Fong Chiu

Cancer Screening Series No 1 December 2003

Published by: NHS Cancer Screening Programmes The Manor House 260 Ecclesall Road South Sheffield S11 9PS Tel: 0114 271 1060 Fax: 0114 271 1089 Email: [email protected] Web site: www.cancerscreening.nhs.uk © NHS Cancer Screening Programmes 2003 The contents of this document may be copied for use by staff working in the public sector but may not be copied for any other purpose without prior permission from the NHS Cancer Screening Programmes. ISBN 1 84463 010 2 Further copies of this publication can be ordered from the NHS Responseline quoting Cancer Screening Series No 1: Tel: 08701 555 455 Fax: 01623 724 524 Email: [email protected] A copy is also available as a PDF file on the NHS Cancer Screening Programmes website.

Typeset by Prepress Projects Ltd, Perth (www.prepress-projects.co.uk) Printed by Cambrian Printers

Inequalities of Access to Cancer Screening: a Literature Review

CONTENTS Page No EXECUTIVE SUMMARY

1

1 2 3 4 5 6 7 8 9

Introduction Concept of health inequalities Ethnicity and inequalities Inequalities of access Systematic reviews Current research studies (1998–2003) The problem with the current knowledge base Ethical issues raised Conclusion and recommendations

1 1 1 1 2 2 5 6 6

1.

INTRODUCTION

7

2.

SCOPE AND METHOD OF THE REVIEW

8

3.

RESULTS PART I: CONCEPT OF INEQUALITY

10

3.1 3.2 3.3 3.4 3.5

Concepts of inequality in health Concepts of access to healthcare Concepts of access to cancer screening Inequalities of access to cancer screening Summary

10 13 14 15 16

4.

RESULTS PART II: SYSTEMATIC REVIEWS

18

4.1 4.2 4.3 4.4

The determinants of screening uptake and interventions for increasing uptake: a systematic review, 2000 Comments Systematic review of ethnicity and health service access for London Summary

18 19 20 20

5.

RESULTS PART III: CURRENT STUDIES

22

5.1 5.2 5.3 5.4 5.5 5.6 5.7

Survey studies on factors influencing screening uptakes Health and cultural beliefs as determinants of screening behaviours Intervention studies on increasing screening uptakes Intervention using lay health educators Prostate cancer screening Colorectal cancer screening Summary

22 23 24 28 30 31 33

6.

CONCLUSIONS

35

6.1 6.2 6.3

Limitations of the review Development of concepts Who suffers from inequality of access?

35 35 36

NHS Cancer Screening Programmes

iii

December 2003


6.4 6.5 6.6 6.7

Health interventions System effect Information and communication Theory and practice

36 37 37 38

7.

RECOMMENDATIONS

39

7.1 7.2 7.3

System Users Information and communication

39 39 40

REFERENCES

41

APPENDIX

48


iv

December 2003


EXECUTIVE SUMMARY 1

Introduction

The National Health Service faces many new challenges as it works towards modernisation. One of these challenges is to ensure that all population groups enjoy fair access to its services. Although the cancer screening programmes have achieved good coverage, there are indications that some population groups, eg minority ethnic groups, people living in deprived areas and those with learning difficulties, are prevented from accessing services for a variety of reasons. Another challenge facing the cancer screening programmes is the extension of the existing breast screening programme to the older age group (65–70) and the possibility of new screening programmes for prostate and colorectal cancers. This review was commissioned by the NHS Cancer Screening Programmes to provide a foundation for meeting these challenges. The scope of this review covers the conceptual development of the issue of access to cancer screening, the relevant contents of two recent systematic reviews72,73 and the findings of 56 empirical studies published between 1998 and 2003,79–102,106–120,128–144 which were selected from electronic databases. The results of the review are summarised below.

2

Concept of health inequalities

Since the publication of the Black Report,2 the issue of inequalities in health has been a permanent feature of health research and practice. The relationship between income inequality and health status is well established, and more recently the contribution of individuals’ lifestyles has also been recognised. Researchers have drawn our attention to the way in which individuals’ health behaviours are embedded in their social and material circumstances. In particular, it has been suggested that individuals’ psychosocial make-up, social interactions and place of habitation have a synergistic effect on health outcomes. Concepts such as ‘social exclusion’ and ‘social capital’, although theoretically still ill defined and open to contestation, have become everyday working concepts among practitioners. Supporting communities to engage and participate in civic and social activities is seen as a key enabling mechanism for addressing health inequalities.

3

Ethnicity and inequalities

Ethnicity is another focus of the issue of inequalities in health. The relationship between ethnicity and health status was first established by epidemiologists and health researchers. Although the construct of ethnicity as a social category is problematic and the use of ethnicity as an independent variable in health research is strenuously contested by critical researchers, it has been suggested that ethnicity might assert an independent effect on health inequalities. Hence, it should not be collapsed into socioeconomic categories.

4

Inequalities of access

Although there is a link between people’s access to the health system and their health outcomes, this relationship has not been well understood. The notion of access is complex and multilayered and can be related to the equity and quality of the service. However, it is often conceptualised, either from the user’s or from the service’s perspective, although rarely as a dynamic process involving the relationships between individuals and health systems in a changing social context.


1

December 2003


The inability to conceptualise access as a dynamic process involving both the individual and the health system has generated a plethora of studies centred on the mapping of personal and social determinants of uptakes. In determining what factors affect people’s attendance of cancer screening programmes, many researchers have relied on large scale surveys of self-reported screening status, which has been found to be associated with ethnicity, deprivation variables and/or geographical location. Many researchers have also attempted to establish a link between screening uptake rates and mortality outcomes. Although these kinds of studies are frequently criticised as crude and incapable of offering explanations of inequalities, the underlying patterns upon which they are based persist. Consequently, the relationships between socioeconomic status and health outcomes are well established. However, to date, detailed information on patterns of access among different socioeconomic/ethnic groups is still unavailable. 5

Systematic reviews

The systematic review commissioned by the Health Technology Assessment Programme in 200072 was firmly rooted in the positivistic paradigm and dealt only with randomised controlled trial studies. The empirical studies selected were not only of poor quality generally but also were not underpinned by a coherent theoretical framework. Since these trials were conducted inappropriately, numerical meta-analysis to determine the aggregated effect of these trials became an artificial exercise. Conclusions drawn from such analyses should be regarded with caution. In addition, this review did not address variation in uptakes of screening among different social groups. Thus, its contribution to knowledge in this area is limited. However, a recent systematic review of access to health services in London by minority ethnic groups73 has provided some valuable insights. This review identified the need for ethnic monitoring to provide better information and recommended that more qualitative research should be undertaken in order to improve understanding of the process of access and to determine the independent effect of ethnicity on access.

6

Current research studies (1998–2003)

Fifty-five studies on inequalities and cancer screening conducted between 1998 and 2003 were selected from electronic databases for this review. Four types of cancer screening are included in the review: cervical, breast, prostate and colorectal. Nearly half of the studies (26) were survey studies, which examined determinants of screening uptakes. The rest of the studies were intervention studies, aimed at increasing uptakes of cancer screening. All but one of these studies were quantitative. Correlation and pre- and post-quasiexperimental trials were common methods used in these studies. The target groups of all the survey studies included Latinos from different Hispanic groups, Asian-Pacific Americans, including Koreans, Chinese and Vietnamese, and African-Americans. It is important to note that most of these studies are from the USA; only two studies reviewed here were conducted in the UK.115,145 The following are salient themes emerging from the review of these studies.


2

December 2003


6.1

Overemphasis on behavioural determinants

Mapping of utilisation rates in the USA is often associated with the mapping of personal, psychological and social determinants of utilisation, ie age, income, knowledge, attitude, beliefs, literacy, language proficiency and the degree of enculturation as well as ethnicity. Specifically, some of these concepts are ill defined. Ethnicity and uptake rates are often conceived unproblematically as independent and dependent variables, respectively, in survey studies. Most of these studies also collapse socioeconomic status and ethnicity into one variable. Concepts such as language proficiency, enculturation and health and cultural beliefs are often ill defined and operationalised as variables in surveys without clear and explicit working definitions.

6.2

System effects

From the psychosocial perspective, problems of uptake are always seen as emanating from the target groups themselves. Rarely were system effects such as configuration of the screening services, administrative procedures (how systems call and recall screening participants) and professional capacity in health education and promotion (knowledge and skills not only in screening guidelines but also in communication and education or working with community based organisations) seen as determinants worth investigating. Twelve of the 19 survey studies reviewed allude to the system effect. However, this mainly centres on access defined narrowly by whether or not women had a ‘usual source of healthcare’. This is obviously a relevant concern in the USA, where lack of insurance cover could deny women a usual source of care and therefore access to preventative services. In the UK, addressing inequalities would not be relevant to a large proportion of the population because the NHS provides free access to all. However, this issue is likely to be relevant for the rising number of homeless, asylum seekers and refugees in the UK. Owing to their social positions and circumstances, these groups are likely to constitute a highly mobile population in many inner cities; thus, the ‘usual source of care’ might also be a factor influencing access to services for these marginal groups. In general, these survey studies seem to confirm that, at least in the USA, screening services are underused by minority ethnic groups. The psychosocial determinants of low uptakes explored by these studies were knowledge of screening, literacy, age, attitude, beliefs, and social and economic status. However, results were often contradictory. In addition, the results of these studies have limited generalisability, as a consequence of their theoretical and methodological weakness.

6.3

Health beliefs as determinants

The studies of health beliefs as determinants of uptakes of screening behaviours reviewed here are of varying quality. There exists an inconsistency in the application of the stated theoretical frameworks, ie the Health Belief Model (HBM) and the Theory of Planned Behaviour (TPB). There is also a lack of methodological rigour. Thus, results generated from these studies need to be interpreted with caution. While these studies might offer some insights into the possibilities of different responses from different groups to cancer screening, overgeneralisation of these results might, at best, perpetuate misconceptions about cancers that exist across communities and, at worst, reinforce cultural stereotyp-


3

December 2003


ing. In addition, uncovering cultural explanations of health behaviours cannot be seen only in an entirely rationalist view. The problem with these models is that they ignore the affective and social components that may influence health behaviours. Moreover, each minority ethnic culture is not a monolithic, unchanging whole that exists in a social vacuum. Many minority cultures are likely to be influenced by their structural positions vis-à-vis the dominant ideology and society. These will have influence upon their day to day lives, thus in turn possibly reinforcing or changing their beliefs. 6.4

Intervention approaches

Since many implicit or explicit theoretical frameworks underpinning screening are based on behavioural and social learning theories, the majority of the intervention studies followed in the same vein. Most studies targeted low income minority ethnic groups and aimed to improve users’ knowledge and awareness of the screening programmes. Methods used included providing information or motivation counselling. Although most of these interventions lack scientific merit, they nevertheless offered some interesting ideas for intervention designs. In addition, many studies seemed to be implicit in their acknowledgement of the value of social support and social networks. Thus, they commonly employed bilingual researchers from target communities and/or incorporated lay health educators, peer educators or counsellors as part of their intervention design.

6.5

Lay health educators as interventions

Six studies were found to have explicitly evaluated the effect of lay health educators on improving uptakes. The problem of evaluating lay health workers as an intervention using a field trial approach is that it is notoriously difficult to standardise and control the applications of human endeavours. Without a theoretical understanding of the concept of lay involvement and its current practice in various social and political contexts, it would be difficult for researchers to devise a ‘good enough’ evaluation strategy to examine its effects. Moreover, the concept of the lay health educator is inherently bound up with concepts of social networks and social capital. The development of this popular approach without bridging the theoretical and conceptual gaps between community health development and individual health behaviour would likely be detrimental to the development of a pragmatic strategy to tackle health inequalities.

6.6

Complex interventions

Two large scale intervention projects reviewed here, the Forsyth County Cancer Screening Project116 and the ENCORE plus project,117 show that, even in the biomedical model, there is an increased awareness of the need to acknowledge complexity and developmental issues in health interventions. These projects had relatively long lifespans (up to four years) and covered many different sites and involved many agencies. Both projects appeared to recognise the system effect on access and sought to intervene systemically by using both ‘outreach’ and ‘in-reach’ methods. However, large scale intervention projects are notoriously difficult to evaluate. Although both projects mentioned above had systematic monitoring and evaluation strategies, evaluation results have not been reported so far.


4

December 2003


6.7

Prostate and colorectal cancer screening

Since the provision of prostate and colorectal cancer screening is under consideration within the NHS, experience from the USA in these areas might be helpful. The USA has long recommended that annual prostate and colorectal cancer screening should be available to citizens who are at risk. The studies reviewed here found that members of minority ethnic groups tend to underuse these services. Intervention studies on prostate cancer screening tend to target African-American men because they are twice as likely to be at risk as white men. Intervention methods used so far remain informational. However, two particular points need to be noted. Although it might be important to improve knowledge and raise awareness about prostate cancer and screening, the issue of the effect of screening on sexuality, revealed by the focus group study on participants’ (both sexes) reactions to risk information, needs to be better understood. The fear that treatments could have adverse effects such as impotence and incontinence may reduce users’ participation in the programme. If colorectal screening is to be implemented in the UK, it should be borne in mind that a recent study has shown that a public awareness campaign to improve knowledge about and recognition of the symptoms of colorectal cancer, particularly among those who are less educated, might be needed. Specific efforts might also be needed to alter women’s perception of the disease as a male disease. Qualitative studies on the response to such a screening programme and the procedure involved across different social groups would be necessary to ensure effective communication with the public.

7

The problem with the current knowledge base

As current public health and health promotion projects tend to be increasingly complex, involving a mixture of activities that aim to improve both structural and material development as well as capacity building, health promotion and access to health and social services, devising a comprehensive evaluation strategy in which the linkage of these components can be discerned and then assessing their relative contributions and impacts to health outcomes is now almost impossible. What we are left with is a perpetuation of the unproductive debate about the primacy of agency and structure among intellectuals and the confusion generated in public health and health promotion practice as to whether changes should be aimed at individuals’ knowledge and behaviour, their lifestyles and/or their social conditions. Given the present knowledge system, in particular in biomedicine, the ultimate goal of generalisability of outcomes means that the process of abstraction in the creation of knowledge is overvalued and flourishes at the expense of practical knowledge. Hence, many of the studies reviewed here conform to the experimental method. Given the uncertainty of knowledge creation and the complexity of social phenomena, there needs to be a new approach to generating practical knowledge. Complex health interventions require a range of substantive theoretical frameworks and research methods. Results obtained from projects could then be seen on different levels of complexity. Under some circumstances, some results might contribute to a wider applicability while others would be more appropriately seen as providing learning insights for practice rather than


5

December 2003


discovering general truths that can be applied across populations of a diverse social and cultural background. 8

Ethical issues raised

This review identifies important ethical issues in relation to approaches that seek to increase screening uptakes by communicating personal risk information to users in written form. Communicating risk is a complex subject; the content, means and ends of such communication depend on a host of different variables and upon different circumstances. As can be seen, communicating risk information could lead to adverse effects, ie anxiety and avoidance of screening. Practitioners should be mindful of the ethical issues involved when considering this approach.

9

Conclusion and recommendations

There are diverse explanations of health inequalities. However, because of the lack of systematic monitoring information, no irrevocable relationship between access and health outcomes can be established at this time. Despite development in the social and structural factors that might affect inequalities of access, the search for determinants is still confined to the rationalist and behaviourist traditions, in which psychosocial factors such as knowledge, attitude and beliefs are emphasised. Many intervention studies are based on this approach. To advance current knowledge and practice so as to address the issue of inequalities of access, there is a need to better integrate the notion of access with social conditions and the perspectives of users and services. The dynamic process between the individuals and the health systems in the changing social context needs to be better understood. Cancer screening is only part of the cancer care system. Access to the screening services may be only a first point at which some individuals find themselves needing to negotiate a complex and, in many respects, labyrinthine service system. Given that professionals are often ill prepared to communicate the complexities of cancer screening to their diverse populations, there is a need to understand the part played by cancer information and health communication in inequality of access. We also need to understand better the way in which ‘ethnicity’ can affect access and the ways in which social capital and social networks mediate access. Recommendations to tackle inequalities of access to cancer screening were given from the perspective of both the system and the user. These recommendations indicate a broad area of work that the cancer service as a whole needs to address. However, some of these recommendations, such as exploring users’ perceptions of proposed new screening programmes, surveying and evaluating information and communication and the evaluation of the community health educator model, might fall within the remit of the cancer screening programmes.


6

December 2003


1.

INTRODUCTION

This review was commissioned by the NHS Cancer Screening Programmes to inform the development of cancer screening services in England and Wales. A literature review is, by definition, an exercise in which critical judgements are made about relevant published information in a particular field of study. It is therefore difficult to give a true picture of the world of practice in that area unless its activities are reported. However, by recognising that there exists a considerable gap between research and practice, and that there are different paradigmatic assumptions in knowledge creation, the author’s approach to this review is a pragmatic one. This means that the studies included in the review are judged not only on the validity or generalisability of their results and whether they provide a narrative on conceptual development but also on whether they can provide insights into either the understanding of the phenomenon or the implementation of practical solutions. The main aims of this review are to examine factors affecting inequality of access to cancer screening and to assess the efficacy of different approaches to addressing the problem. In particular, it seeks to address the following questions: 1. How is inequality as a concept defined in the current literature in relation to cancer screening? 2. What factors are perceived by researchers to affect access to cancer screening? 3. Who suffers from inequality of access? 4. What approaches and methods have been implemented to address issues of access and improve uptake and informed choice? The review was conducted with the intention of combining the knowledge gained from it with the feedback from the National Dissemination Forum of the Straight Talking project, which took place on 12 May 2003 at the Royal Armouries in Leeds, to form a basis for the consideration of future work. The report that follows first outlines the concept of inequality as defined in current literature and the way in which the issue of access to healthcare, and specifically to cancer screening, has been understood by researchers and policy makers. The aim of reviewing the development of these concepts is to help placing cancer screening studies in a broader context. By illuminating the theoretical and conceptual assumptions made in research studies in this area, we will be able to uncover what form the problem of inequalities of cancer screening takes and what methods might be provided for its resolution.


7

December 2003


2.

SCOPE AND METHOD OF THE REVIEW

This review takes a broad approach to examining the issue of inequalities of access to cancer screening. It is divided into three parts. The first part is concerned mainly with conceptual development in this area. It refers to literature on inequalities in health and inequalities in access to healthcare and cancer screening, including examination of research papers published in journals as well as grey literature, such as policy documents, pamphlets and other relevant articles in the press or electronic media such as the Internet. In the second part, the content of two systematic reviews that include substantive materials on cancer screening is examined. Relevant materials reviewed in these reports are summarised and commented upon. Owing to resource constraints, the main body of literature concerning the investigation of factors affecting access and interventions to address the problem of uptakes is reviewed in Chapter 5. The literature reviewed here consisted mainly of peer reviewed articles found in electronic databases such as Medline, Cochrane Library, Cinahl, Psychlit, Helmis, BIDS and the National Research Register. The terms used in the searches were a combination of the following: ‘inequalities’, ‘health’, ‘access’, ‘cancer screening’, ‘cervical screening’, ‘breast’, ‘breast screening’, ‘socioeconomic’, ‘low income’, ‘Pap smear’, ‘mammography’, ‘minority ethnic’, ‘black’, ‘ethnic communities’ and ‘deprivation’(including searches that mapped onto related subjects in particular databases). Some reports and pamphlets reviewed were obtained through personal contacts. For this part, 129 abstracts of articles about cancer screening among low income and minority ethnic groups between 1998 and 2003 were identified at the initial search. Because time available for the review was limited, and the full versions of papers were not always available electronically, only 55 articles were subsequently selected for review. The main points discussed in the selected papers are summarised in the appendix. The papers were then categorised into survey studies, which examined the factors (including health and cultural beliefs) influencing uptake or service utilisation, and intervention studies aimed at improving uptakes. Prostate and colorectal screening are currently not available as organised programmes in the UK. However, if these programmes were to be implemented, there would be a need to understand how target participants respond to them. Thus, studies that examine prostate and colorectal cancer screening were reviewed as distinct categories to create space for relevant discussions. The following table summarises the review strategy and numbers of papers selected in each category:


8

December 2003


Survey studies Cervical screening Breast screening Cervical and breast screening

4 9 6

Beliefs as determinants studies Cervical and breast screening

5

Intervention studies Cervical screening Breast screening Cervical and breast screening

3 7 5

Lay health workers’ interventions Targeted population groups Targeted a particular district

5 1

Prostate cancer screening Intervention studies Survey studies Focus group studies

2 1 1

Colorectal cancer screening Survey studies

6

Total articles selected


9

55

December 2003


3.

3.1

Concepts of inequality in health

RESULTS PART I: CONCEPT OF INEQUALITY

Studying health inequality by establishing relationships between large scale survey data and health status is not new. It can be dated back to Chadwick and the public health reform movement of the 1840s.1 The link between social and economic status and patterns of morbidity and mortality established in the Black Report2 marked the beginning of the debate on health inequalities in modern Britain. Although the Black Report drew attention to the social determinants of health, some social researchers have suggested that the dynamics of the social environment that produce health inequality are not well understood. The recent independent inquiry on inequalities of health further affirmed the effect of poverty and suggested that alleviating poverty and improving the living standards of the poor might help to prevent ill health. However, the report also recognised the contribution of individuals’ lifestyles to health.3 The development of research in this area over the past two decades has been diverse. While some researchers favour the study of structural factors that affect health, many have been intrigued by the behaviours of people who are located within these specific social structures. Aspects of lifestyle and the health related behaviour of individuals have been highlighted as explanations of health variations, and there is evidence to suggest that there is a link between socioeconomic deprivation, poor health experience and health damaging behaviour (for example, drinking and smoking).4 To inform health promotion strategies and practice in tackling health inequalities, Rogers et al5 reviewed qualitative evidence on the effects of lifestyle and behaviours on the key areas set out in the White Paper Health of the Nation,6 ie coronary heart disease (CHD)/stroke, cancer and smoking, mental health, sexual health and AIDS/HIV, and accidents. Their article provides a lay perspective on health and health behaviours and highlights the complex social processes that might have contributed to the perpetuation of health inequalities. The authors suggest that health behaviours are deeply embedded in people’s everyday lives and their social and material circumstances. They also suggest that the use of some key concepts in health promotion practice, such as ‘lifestyle’, ‘healthy behaviour’, ‘risk’ and ‘empowerment’, is problematic unless local conditions are taken into consideration. Sociologists’ interest in the meaning of inequalities in everyday life has led to a range of research projects focused on the microlevel.7,8 While current strategies in tackling inequalities make explicit reference to the need for economic regeneration, the question ‘Why do people behave in the way they do?’9 has continued to attract the interest of researchers on both sides of the Atlantic. Much explanation of the pathway between income inequalities and health status has come from research on social capital by Wilkinson’s group. Kawachi et al10 posited the relative income hypothesis, which holds that ‘the distribution of income among members of society matters as much for their health and well-being as does their absolute standard of living’ (p. xi). Wilkinson11 also postulated that, when


10

December 2003


health inequalities in societies reach a certain threshold, health status becomes determined more by social (dis)advantage than by material scarcity. Drawing on Putnam’s work,12 Kawachi et al13 proposed that social capital is a mediator between income inequalities and health. While there are many definitions of ‘social capital’, the term generally refers to social resources that individuals or groups accrue for a variety of productive activities.14,15 These activities can be economic as well as social. Social networks, social cohesion, civic trust and engagement in a geographical location (neighbourhood) are often used as indicators in the study of social capital (eg refs 16 and 17). Although there is ambiguity and confusion surrounding the concept of social capital, it appears to have offered researchers ‘an intriguing explanation for one of the pathways linking income inequality and health status’ (ref 18, p. 414). There is evidence to suggest that people’s ability to participate and engage in social and civic activities is associated with perceived self-efficacy and self-rated health status.19 Therefore, despite the controversy surrounding it, ‘participation’ and ‘engagement’ are increasingly seen as key enabling mechanisms for health development. Another focus in the study of inequalities is the health experiences of the minority ethnic groups. Epidemiologists have long highlighted the health differentials between ethnic groups.20–22 In an attempt to understand the relationship between social factors and health among minority ethnic groups, researchers have also explored the relationship between income or class inequalities and the health status of some minority ethnic populations.23,24 However, the study of inequalities in health among minority ethnic groups is complex; researchers have been confronted with considerable theoretical and methodological difficulties. The construct of ‘ethnicity’ is itself problematic,25,26 and in addition the lack of theoretical debates in health research has perpetuated the crudity of ethnic categories. Thus, the picture of ethnic minority health has been consistently distorted, resulting in the pathologising of particular groups, such as African-Caribbeans, Indians and South Asians, while rendering other minority ethnic groups, such as the Chinese, Somalis, Turks and Yemenis, completely invisible. Another problem in studying inequalities in relation to ethnicity is the limitation in applying conventional socioeconomic measurements such as housing tenure, occupation, class and education to minority ethnic populations. For example, the high owner–occupier rate among minority ethnic groups may reflect discrimination in accessing public housing rather than affluence; and often racism may prevent migrants or their subsequent generations from converting educational qualifications into employment opportunities.27 Failure to consider the influence of other dimensions of ethnicity, ie migration patterns, racism, language and cultural differences, and other material conditions, means that any explanations of the inequalities of health status among minority ethnic groups will always be partial. Helping to refine the population perspective of income inequalities is the concept of social exclusion. It has been postulated that in late modernity the effect of post-industrialisation and globalisation will generate a huge underclass owing to the loss of employment in traditional industries and


11

December 2003


economic and political migration.28 A classic example of social exclusion is the devastating health consequences Native Americans experienced after the utter demoralisation brought about by the destruction of their communities through the loss of employment in their areas.29 In the UK, the term ‘social exclusion’ has been officially defined by the Cabinet Office as ‘a shorthand term for what can happen when people or areas suffer from a combination of linked problems such as unemployment, poor skills, low incomes, poor housing, high crime environments, bad health and family breakdown’.30 This definition implies a complex and multidimensional view that social exclusion can be experienced by individuals or groups and in different geographic areas. What has been brought to our attention is not only economic deprivation but also the dynamic of the experience of people and their social environment.31 The Social Exclusion Unit (SEU) is a policy initiative that attempts to encourage interdepartmental cooperation in tackling inequalities. The work of the SEU includes reporting on truancy and school exclusion, rough sleeping, teenage pregnancy and neighbourhood renewal. Although the concept of social exclusion is complex, it has been regarded by researchers and practitioners alike as a composite term to include multiple deprivation, concentrated poverty and a social phenomenon produced by the dynamic of poor people within poor localities over time. One could argue that the concept potentially provides researchers with a new way of thinking about the relationship between individuals and their social and material circumstances in a more dynamic way. Controversies concerning the relative significance of individuals’ experiences, social relationships, material conditions and place of habitation for variation in health status still exist.32–34 However, by acknowledging that there is no simple explanation for health inequalities, studies of social networks and social capital as mediators between socioeconomic factors and the effect of geography have flourished.35 In considering the issues of inequalities in health, it is important to recognise that the classic sociological problem of agency and structure is at the heart of the debate.36 This means that many researchers will invariably take the position that people’s health is determined by the social positions in which they find themselves or that they have the capacity to adopt behaviours that are health enhancing. Rather than seeing that a complex social phenomenon requires a complex conceptual framework and complex methods for its comprehension, diverse intellectual efforts can be mutually enhancing. This dualistic thinking often means that researchers are split into different camps according to their traditions and disciplines, resulting in competing explanations and competing recommended strategies for resolution. As Giddens37 has suggested, social knowledge is part of the reflexivity of modernity. Knowledge constitutes social practice and in turn is constituted by it. We have seen in the past three decades, in parallel with the debate on inequalities in health, the development of a community health movement. It began with the community development projects in the late 1960s, followed by women’s and black and minority ethnic movements throughout the 1970s and 1980s, and found its legitimacy through the


12

December 2003


Alma Ata Declaration. A great deal of practical work covering a broad range of initiatives addressing direct or indirect health determinants can be discerned.38 The interest in community health development among health promotion practitioners has also been fuelled by research evidence such as the Black Report,2 The Health Divide39 and Deprivation and Ill Health.40 However, there seems to be an obvious gap between theory and practice in community health development. In research, the growing knowledge base around inequalities and health is ridden with theoretical and methodological difficulties; emerging concepts such as social capital, social exclusion and social networks, which claim to embrace both agency and structure, are ill defined. Moreover, much evidence generated in this area is open to contested interpretations. In practice, engaging communities to participate in health initiatives and building social capacity and capital are seen by practitioners as key mechanisms for addressing inequalities. However, how the building of social capital can be operationalised in practice and the extent to which it can be evaluated is less well understood.41 3.2

Concepts of access to healthcare

Although there is an obvious link between people’s access to the health system and their health outcomes, this relationship has not been well understood. The concept is mostly closest described by the ‘inverse care law’.42 The ‘inverse care law’ states that people in deprived areas tend to have greater health needs but are less likely to be able to access healthcare. There is evidence to support this generalisation, particularly in the context of screening and preventative services provided in general practice.43 However, the notion of access is complex and many sided. It has been defined as ‘those dimensions that describe the potential and actual entry of a given population group to the healthcare delivery system’ (ref 44, p. 13) and as ‘the timely use of affordable personal health services to achieve the best possible health outcomes’.45 These definitions seem to reflect the perspectives of both services and users. Seen from the service perspective, the widening of access requires the service to be organised to reflect the needs, interests and priorities of its users.46 Physical accessibility of services such as those in rural settings is arguably a significant issue.47 From the users’ perspective, access to services appears to be related to the economic, social and cultural circumstances of the individuals.48 This means that the issue of inequality of access is complex. It goes beyond either simple structural explanations of the availability, accessibility and utilisation of services in locations or the psychosocial make-up of the individuals. The issue of access needs to be understood in terms of the dynamic relationship between individuals and service systems in specific social contexts. The recent scoping exercise report on access to healthcare, commissioned by the NCCSDO (National Co-ordinating Centre for NHS Service Delivery and Organisational Research and Development), defines access to healthcare as that concerned with ‘helping people to command appropriate health care resources in order to preserve or improve their health’. It posits that the concept of access has at least four aspects. These are (1) service availability in terms of adequate supply, (2) service utilisation influenced by socioeconomic, cultural and organisational barriers,


13

December 2003


(3) the relevance and effectiveness of service and (4) the evaluation of availability and utilisation of service from multiple perspectives that take account of the needs and material and social conditions of diverse groups in society.49 This definition does seem to offer a more dynamic and complex view of the concept of access. Accessibility is also one of the dimensions of quality of care. Within the different evaluative dimensions of care, effectiveness, efficiency and equity are often linked with service accessibility.50 However, these evaluative dimensions are often in conflict. The NHS service, because of limited resources, often faces a dilemma of equity versus efficiency. If services are not targeted to those who have the potential to reap the most benefit from them, the efficiency of services might be compromised even though these services might be perceived to be more equitable. In a recent report, Sassi et al51 investigated the relationships between equity and efficiency within the context of economic evaluation of services and found these relationships to be far from simple. In the case of cervical cancer screening, the programme’s aim of achieving maximum coverage by giving powerful economic incentives to general practitioners has been achieved at the expense of differential access among social groups. The authors suggested that a more even coverage among different social groups, in particular increasing uptakes among high risk and socially disadvantaged groups, would probably result in an increase in the number of cases of invasive cancer prevented. Thus, defining access purely from the service perspective to mean access for all rather than access for those at risk or in need can be problematic. 3.3

Concepts of access to cancer screening

Throughout the Western industrialised countries, despite the increasing financial burden on health systems, ‘access to medical care is regarded as a social good and the inalienable right of every person’ (ref 52, p. 1). As the cancer screening programmes are free to users, it is even more difficult for policy makers to understand why some groups fail to take up these services. Hence, many attempts to understand the lack of uptake by these groups are often focused on the users and are conceptualised on a microlevel. From this perspective, there is a flourishing study of psychosocial factors and health beliefs among some social groups, eg minority ethnic groups and low income groups, as barriers to uptake of preventative services. For example, Andersen’s53 behavioural model of access focuses on the psychosocial characteristics of individuals and healthcare providers or primary care providers; the communication between them and outcomes of care that can be measured are ‘utilisation and satisfaction’. Recently, Mandeblatt et al54 expanded these measurements to include categories such as disease free survival, survival, stage at diagnosis, incidence rates and rate of rescreening and quality of life (functioning). From the service perspective, if access to cancer screening is related to outcomes, we have to consider access along a continuum of care. An effective screening programme would be expected to increase survival rate from the disease and decrease the mortality rate. To achieve this, we need to take a holistic view of cancer care. Mandelblatt et al54 suggested that:


14

December 2003


‘… the foremost important thing is to ensure individuals or the population that screening is targeted to have access to early detection services. Next, if a screening test is abnormal, the diagnostic services must be available. For those who are diagnosed with cancer, staging evaluation precedes and often determines treatment. Patients surviving their disease also need to have access to on-going surveillance, so that recurrences can be recognised promptly and treated. For those who will die of their diseases, having access to end of life care is critical. Thus, access has different dimensions and outcomes across the spectrum of cancer care on the patient’s cancer journey’ (p. 2379). From this perspective, cancer screening can be regarded as part of cancer care. Thus, when considering the issue of inequalities of access to cancer screening, cancer screening must be viewed as part of a dynamic process in which interactions take place between the individuals and the healthcare system in a changing social, economic and political environment. 3.4

Inequalities of access to cancer screening

On both sides of the Atlantic it has been found that uptake of breast and cancer screening among minority ethnic groups is persistently low. However, owing to theoretical and methodological difficulties, an accurate picture of uptake among different social and ethnic groups remains elusive. In the UK, evidence of low uptake of the cancer screening programmes is provided by small local surveys relying on self-reporting or statistical data association between ethnicity and deprivation variables43,55–58 and information generated by national health lifestyle surveys tends to use crude categories of ethnicity and self-reporting of service utilisation24,59 At present, direct evidence of low uptake by some groups is unavailable owing to the lack of ethnic monitoring at the primary care level. The situation is similar in the USA. Information about low uptakes of cancer screening tends to come from local studies relying upon selfreporting of adherence. Many researchers often attempt to establish links between low adherence rates and incidence and mortality outcomes. Cancer incidence and mortality statistics among ethnic groups are apparently available through the National Cancer Institute’s Surveillance, Epidemiology, and End Results Program (SEER). However, they represent only broad categories, eg African-American, Native American, Latinos, Asian-Americans. These statistics belie the complexity of ethnic diversity in the USA; for example, the category ‘Latinos’ has been found to encompass individuals from approximately 30 countries of origin. Critics claim that these statistics are unreliable as they are based on false assumptions about the ethnicities of the local populations; in addition, misunderstanding of these categories results in misclassification and misidentification.60 A recent historical study by Yabroff and Gordis61 investigated the association between social class, breast cancer incidence, mortality and increased breast cancer case fatality. The results confirmed previous reported findings. The authors found that, allowing for disease stage at diagnosis, the incidence of localised breast cancer declined as social and economic status (SES) declined whereas the incidence of distant breast cancer increased. This is consistent with reports of positive associations


15

December 2003


between low SES and low rates of uptake of mammography, low adherence to recommended follow-up after abnormal mammograms and more advanced disease at diagnosis.62–65 Similarly, in the UK, breast cancer survival rates among women from deprived areas have been found to be poor; this may be attributable to late presentation and increased comorbidity among women from deprived areas.66 A study of how previous cervical screening experience affects future uptake expectations found that women from lower social classes tend to have an aversive view of the procedure and feel less personally obliged to attend screening.67 A Norwegian study demonstrated that, despite an advanced welfare state in that country, cancer survival varies according to social group. The author suggested that early diagnosis among the socially advantaged might have contributed to such differences.68 Similarly, a recent report by the National Statistics Office on cancer trends in England and Wales (1950–1999) found that, although the incidence of and mortality from breast and prostate cancer are negatively associated with deprivation (ie the highest incidence is found in the most affluent group), in the case of most other cancers there is a positive correlation between incidence and mortality and deprivation (the highest age standardised incidence and/or mortality occurs in the most deprived group). In the case of breast cancer, it was found that, although incidence and mortality were lowest in the most deprived group, survival rate was higher in the most affluent group. In the UK there is also geographical variation in cancer survival, reflected by a north–south divide, with the incidence and mortality from breast cancer being higher in the Northern, Yorkshire and the North West regions than in the southern regions of Anglia, Oxford and the Thames.69 Patterns of screening uptakes reflect incidence and mortality statistics. The most recent results from the NHS Cancer Screening Programmes show that there is a large variation in uptake rates among regions. The percentage of eligible women who present themselves for breast screening ranges from 50% in parts of inner London to 81% in rural areas such as Norwich and Suffolk. The situation in the Northern and Yorkshire regions is similar, with up to 81% coverage in Cumbria and only 74% and 73%, respectively, in Leeds and the Pennines, where minority ethnic groups constitute a substantial part of the population.70 According to the Straight Talking project,71 uptake also varies among general practices at the ward level: in Leeds, the uptake rate in some inner city practices is as low as 45%, while others have achieved an uptake rate of around 73%. 3.5

Summary

The literature cited above suggests that the concepts of access are complex and multidimensional. In general, studies of access to healthcare tend to focus on either users or services. Although attempts have been made to conceptualise access as a dynamic relationship between users and providers,53,54 the relationships that these models described remain mainly at the microlevel. So far, no conceptual framework has adequately captured the complexity of access that takes account of both individuals and structural influences.


16

December 2003


Studies of inequalities of access to cancer screening have been plagued by both theoretical and methodological problems. The lack of systematic quality information makes it difficult to obtain a true picture of inequalities of access. The pictures that we are presented with are only snapshots, provided by either proxy measures in terms of geography or inferences from incidence and mortality outcomes. Although not entirely reliable at the individual level, these snapshots generally suggest that there is a relationship between low socioeconomic status and low uptakes of cancer screening in geographical locations associated with high levels of deprivation. The relationship between ethnicity and access is often established by the association of higher concentrations of minority ethnic populations with lower socioeconomic status and deprived localities.


17

December 2003


4.

RESULTS PART II: SYSTEMATIC REVIEWS

Contents related to cancer screening and access from two systematic reviews have been included in this report. 4.1

The determinants of screening uptake and interventions for increasing uptake: a systematic review, 2000

The first systematic review used here was commissioned by the Health Technology Assessment Programme (HTA).72 Its aim was to examine factors associated with screening uptakes and to assess the effectiveness of interventions in increasing uptakes. The scope of the review included not only cancer screening programmes but other types of screening tests, ranging from antenatal tests, such as screening for Down’s syndrome, to screening for hepatitis B or diabetic retinopathy. Screening literature collected for review was divided into studies of determinants of uptakes and intervention studies. Only the results of the reviews of cancer screening programmes will be commented on here. The second systematic review aimed to examine studies of access to healthcare among minority ethnic groups; it was commissioned by a local health authority in the Greater London area. Using a descriptive approach, reviewers summarised the studies of factors influencing uptakes. The review concluded that, in the case of breast screening, women were more likely to attend if they had attended previously, had the intention to attend, had health insurance or were recommended to attend by their physicians. In the case of cervical screening, age and health insurance were determinants, but it was unclear whether older or younger women were more likely to attend. Age greater than 65 years, previous experience of screening and functioning normally in daily lives were determinants found to be associated with attending for a faecal occult blood test (FOBT). Having a high level of education and being African-American were found to be predictors of prostate cancer screening attendance. For the review of intervention studies, the approach to the literature was quantitative pooling of data. Relative risks and 95% confidence intervals (CIs) were calculated. A test for heterogeneity was also performed. The reviewers found a great number of quasi-randomised control trial studies. They concluded that invitation appointments, letters (not so effective in the case of mammography), telephone calls, telephone counselling and removal of financial barriers were effective interventions for improving uptakes. Other interventions, such as educational home visits, opportunistic screening, multimethod community interventions and using reminders etc., were found to have some effect, while print, audiovisual materials, educational sessions, risk factors questionnaires and face to face counselling were the least effective interventions. They concluded that there was neither good quality evidence nor sufficient evidence to enable satisfactory evaluation of the effectiveness of intervention studies to be conducted.


18

December 2003


4.2

Comments

The poor quality and methodological and statistical heterogeneity of the majority of the studies prevented the reviewers from carrying out ‘proper’ meta-analyses on both determinants of and interventions for screening. Consequently, the reviewers chose to carry out a descriptive analysis of the studies, at least in relation to determinants. However, a descriptive review would require a considered discussion of different methodological frameworks, ie sociological, behavioural, economic and organisational, in the overall analysis and interpretation of the results. None of these was covered in this review. Overall, the positive gain from the review is a broad description of trials conducted in this field of study and the varied effects found under different designs and methods. However, there is particular concern about how reviewers approached the selection of intervention studies for their review. The quality of the trials included was not systematically determined, thus the number of trials that were not amenable to meta-analysis was not made explicit. Since meta-analyses on both determinants influencing and interventions aimed at improving uptakes of screening programmes were based on few good quality trials, the reviewers’ conclusions on their effects should be treated with caution. Unlike drug trials, the investigation of interventions of preventative behaviours (uptakes of screening programmes) rests firmly in the realm of human and social sciences. In addition, a large proportion of these trials took place in the USA, where the context of cancer screening is very different from that in the UK. For example, although cervical and breast screening programmes are organised programmes and free at the point of access in the UK, individuals in the USA have to pay for their screening through insurance cover; screening intervals and recommended ages to begin screening also differ between the two countries. Without addressing these differences, conclusions drawn might lack ecological validity. Given that the purpose of the HTA programme is to provide high quality information to practitioners who work in and manage the NHS, this review is disappointing for the following reasons: a. The reviewers failed to address the question of variation in uptake rates of screening programmes, which they proposed in their protocol. Addressing the issue of variation in uptake would have given the systematic review a broader context from which readers could interpret the results obtained b. Most of the trials were drawn from the USA, Canada and Australia. The reviewers did not make explicit the possible influence of context upon methodology of interventions, for example whether call/recall systems of breast and cervical screening programmes are the same in all these countries and how they compare with those in the UK. If the call/recall system sends out routine letters of invitation, would this constitute a ‘baseline’ or an ‘intervention’? Without due regard to such questioning, the results obtained have no ecological validity and are of very little value to practitioners in the UK


19

December 2003


c. Although the reviewers looked at sociodemographic determinants, their narrative review was descriptive and lacked analytical insights. The review missed the opportunity of giving readers a picture of the complexity involved in understanding differences in uptakes across social and ethnic groups d. Based on the review, interventions such as invitation by telephone calls, telephone counselling, tagged notes and reminders were recommended to practitioners as a means of increasing uptakes. Tagged notes and reminders are already widely used in primary care. The recommendations to invite participants by telephone calls and offer counselling will require resources and training, which cannot be implemented without due consideration of their effect on the system. The authors in their final comments cautioned that recommendations made were based on evidence from actual uptakes rather than informed uptakes and suggested that full information be provided on all aspects of screening services to enable participants to make an informed choice. However, without the understanding of the concept of informed choice, how it can be made possible and through what mechanisms, the suggestion of full information provision would appear to be made without firm empirical basis. 4.3

Systematic review of ethnicity and health service access for London

Another recent systematic review on access to health services in London by minority ethnic groups conducted by Atkinson et al73 provided some interesting information in the area of cancer screening. Using a pragmatic approach, Atkinson et al identified a total of 64 papers about access of health services and ethnicity, including five journal articles on cervical and breast screening between 1991 and 1999.55,74–77 Conclusions that are relevant to our present review are that: 1. better quality information from ethnic monitoring is needed, and greater use should be made of this information in order to justify its collection 2. qualitative research that can illuminate the process of inequality of access to services is required 3. although socioeconomic factors are important, ethnic group membership appears to exert an independent effect on inequality and this should be measured.

4.4

Summary

The first systematic review was conducted firmly according to the biomedical tradition based on randomised controlled trials carried out on the subject of cancer screening. Its main concern is the establishment of generalisability of results. It demands consistency in theoretical assumptions and rigorous adherence to methods and procedures in a large number of studies.78 The review of determinants of and effective interventions for uptakes of screening is a complex task because of differences found in the theoretical and methodological underpinning, and in statistical analysis among these studies. Poor quality and methodological and statistical heterogeneity prevented the reviewers from carrying out proper meta-analyses on both determinants of and interventions for screening. The fundamental problem of this review is the dearth of high quality trials


20

December 2003


in this field of study. Faced with the problem of assessing quasi-experiments (trials gathered for the review) for their ‘true experimental’ effects, reviewers were strapped with various technical problems in carrying out a ‘true’ meta-analysis, eg difficulty in estimating the number of negative trials, lack of information from authors, difficulty in determining quality. As no systematic pooling of data was possible and the overall effect was based on relatively few high quality trials, the results presented should be treated with caution. The reviewers recommended that ‘better’ trials be conducted. However, when the method is obviously inappropriate, it is doubtful whether it is ever possible to conduct a ‘true experiment’ of such a complex social phenomenon. Although not a direct challenge to the biomedical approach to review, the second systematic review conducted by Atkinson et al on ethnicity and access contested such practice. Although reviewing a large amount of qualitative evidence is problematic, reviewers were able to draw insightful conclusions. In general, this review suggested areas of inadequacy that needed to be addressed. While the first review ignored completely the issue of variation in uptake of cancer screening because of the lack of relevant trial studies, the second review highlighted the possible reasons why such a topic would not be suitable for research (eg the lack of ethnic monitoring). The reviewers were also able to draw our attention to the independent effect of ethnicity on inequality of access and warned against collapsing ethnicity into the socioeconomic category.


21

December 2003


5.

5.1

Survey studies on factors influencing screening uptakes

RESULTS PART III: CURRENT STUDIES

A total of 19 survey studies of cancer screening published between 1998 and 2002 were subjected to detailed review.79–97 Four of the 19 studies were specifically related to cervical screening;79–82 nine were of breast screening exclusively, although some studies included clinical breast examination as a variable related to the uptake of mammography screening.83–91 Six survey studies reported on both cervical and breast screening, one of which included colorectal screening.92–97 Minority ethnic groups were the main focus of almost all studies. These groups include the usual broad ethnic categories defined by the US National Health Interview Survey (NHIS), ie black (African-American), Hispanic, American Indian and Asian-American (Japanese and Chinese). Data for specific minority ethnic groups are not routinely collected in the NHIS. The survey studies exhibited varying degrees of awareness of the problem created by such broad ethnic categories. Although some studies took these categories for granted, many challenged the diversity masked by such broad definitions of particular ethnic groups by conducting local surveys, using more culturally and linguistically sensitive methods.93,94 These include gaining entry to the communities through churches and community centres, recruiting respondents through subscription lists of community language newspapers, using snowballing to widen the sample frame and using trained bilingual interviewers or local women from the targeted communities to interview respondents.79–81,85,86,88,90 The majority of the surveys aimed to examine various variables that were hypothesised to be associated with the utilisation of cervical and breast screening by minority ethnic groups. Survey designs varied according to researchers’ interests and the study hypotheses. Although many of the studies used cross-sectional or prospective surveys with interviews, a good proportion of the studies were based on analyses of secondary data collected from national surveys, such as the NHIS and BRFSS (Behavioural Risk Factor Surveillance System). Although socioeconomic status was not singled out as a specific variable, many of the studies did include it, as a high proportion of members of ethnic minorities groups also have low income. Apart from ‘ethnicity’, other variables explored in these studies were literacy, knowledge of screening, education, access to healthcare, age, beliefs regarding preventative medicine, acculturation, physician recommendation for screening, socioeconomic status and continuity of care. In general, all these studies appear to have confirmed that cancer screening services are underused by minority ethnic women. Although studies have explored a range of psychosocial and economic determinants, including knowledge of screening, attitudes and beliefs to health insurance cover and having a usual source of healthcare, many of the results are contradictory. However, some variables were shown to be particularly salient across studies. For example, several studies found


22

December 2003


that low income, less educated and older women are less likely to attend for screening.90,95,96 Another consistent theme that emerged from these studies is the issue of ‘access’, defined variously as access to free care, having a usual source of healthcare and having a female doctor.79 Some studies suggested that physicians play a vital role in affecting uptake of screening, depending on whether or not they would communicate cancer risks to women who had first degree relatives who had suffered from cancer91 and whether or not they would recommend screening equally to low income women.81,87,95 It was difficult to generalise results obtained from these studies owing to methodological limitations and the lack of any explicit theoretical framework within which certain variables are defined. 5.2

Health and cultural beliefs as determinants of screening behaviours

Five studies in the selected literature focused on the investigation of health and cultural beliefs as determinants of participants’ screening behaviours. All but one98 of these studies had the common underpinning of an explicit deployment of the Health Belief Model (HBM)103 and Theory of Planned Behaviour (TPB).104 Target groups included African-American women, Latina women,99 Vietnamese women100 and Samoan women101 in the USA and Thai immigrant women in Australia.102 Structured and semistructured questionnaires constructed based on the HBM and TPB models were used by two studies,99,100 while audiotape one to one and focus group interviews were used by Phillips et al98 and Ishida et al.101 The Australian study used both survey and ethnography methods to explore the beliefs and behaviours of Thai immigrant women. Despite their theoretical incoherence and methodological failings, these studies offer some insights that seem to be commonly applicable across many minority ethnic groups (with the exception of African-American). Members of most of these groups believed that they were not at risk of either cervical or breast cancer. Cultural values such as modesty and shyness were personal barriers to accepting screening. The lack of female care providers was highlighted as a structural barrier for accessing screening. However, different studies emphasised different aspects of traditional health beliefs among the minority ethnic groups they studied. For example, the belief among Samoan women that cancer means death and the belief among Thai immigrant women that poor personal hygiene and benign breast tumours or cysts can cause cervical and breast cancers were highlighted. Common suggestions for interventions were improving the cultural sensitivity of services and the framing of health communications or that messages should take account of the health beliefs from different cultures. Although the results of these studies seemed to offer practitioners practical suggestions as to how they should approach health interventions among minority ethnic groups, the ‘scientific’ validity of these results is questionable as data interpretations and conclusions drawn were based on the erroneous assumptions about the functions of the theoretical models. The HBM and the TPB were originally constructed as predictive models of health behaviour and intention of health behaviour respectively. These


23

December 2003


models describe a non-linear process in which behaviours are affected by the interactions between various variables. However, all of the studies not only used these models inconsistently as a guided framework for describing beliefs variables that might influence women’s screening behaviours but also overinterpreted these variables as having linear, predictive values on behaviours. 5.3

Intervention studies on increasing screening uptakes

Fifteen intervention studies between 1998 and 2003 were included in the review. Three of the studies aimed to improve utilisation of cervical screening,106–108 while seven attempted to improve breast screening uptake rates among low income minority ethnic groups.109–115 The remaining five studies reported on interventions that dealt simultaneously with both cervical and breast screening.116–120 Almost all of the studies reviewed targeted low income minority ethnic groups, which include the usual major categories defined by the NHIS in the USA, ie African-American, Hispanic, American Indian and AsianAmerican. A few of the large scale community based studies defined their target populations geographically as areas where low income minority groups made up a large proportion of the population. However, none of these studies made explicit the basis on which these localities were selected. The studies varied from those that used one simple intervention, such as telephone outreach,108,109 telephone counselling110 or issuing personalised letters,115,120 to complex multimethod, multisector and multisite approaches.116,119,120 With the exception of the study by Valdez et al,114 single method interventions did not have any explicit theoretical framework. Having chosen their target group, ie Latinas, because it was one of the minority ethnic groups that had a low utilisation of the breast screening service despite the availability of subsidised mammograms for low income women through the National Breast and Cervical Cancer Early Detection Program, Valdez et al114 attribute low uptake to lack of knowledge, a low level of education and literacy and a low level of English proficiency. Hence, instructional video was the main method used in the health education intervention. Social learning theory and social marketing methodology were explicitly used for the construction of the video.114 Unlike the standard breast screening videos in the UK, which describe mainly the purpose and procedure of breast screening and the programme, Valdez et al’s video covers a range of educational materials, including risk of breast cancer, the importance of early detection, questions for doctors, options for access (insurance covers and without insurance covers) and healthy lifestyle advice. Part of the video was also designed specifically to address fears and concerns regarding breast cancer, and the consequences of not having a mammogram.114 The delivery mechanism of the video in an interactive, touch screen information kiosk is innovative. However, the video was considered to be too long for field testing, and the results showed that the majority of participants did not view all of the content, but only the basic informa-


24

December 2003


tion. The evaluation was a randomised controlled trial with a sample of 1197 low income Latinas at six locations, which included community health clinics and community centres. Although the results of the study did show some effect of the intervention on women’s knowledge about breast cancer and their likelihood of seeking further information about mammography after exposure to information on the video, women appeared to have possessed a good level of basic knowledge about breast screening prior to exposure to the video. A more significant finding was that the video had a greater effect on women who had never had a mammogram and who had not been screened recently. These results were consistent with some of the findings from the field test in the Straight Talking project.71 Unlike the Straight Talking project, in which the evaluation of the effectiveness of the intervention was based explicitly on a behavioural theoretical framework, ie the TPB, Valdez et al’s evaluation was limited to crude measurements of knowledge, attitudes and beliefs and thus the authors were unable to comment on other psychosocial factors that might have contributed to behaviours. This highlights the fact that effective evaluation, ie illuminating both the theory and practice of health education/promotion, requires a well considered theoretical framework and methodology. Using a crude model, ie KAP (knowledge, attitude and practice), and using it simplistically and implicitly, Valdez et al missed the opportunity to provide better insights into the relative contribution of determinants of health education directed at low income, low literacy women. Thus, their study has a limited contribution to the improvement of both theory and practice of health education. Telephone helplines have been successfully used on both sides of the Atlantic as a form of intervention (for example, Childline and smoking helplines). However, these often fail to reach low income and/or minority ethnic groups. Both Crane et al108,109 and Duan et al110 used the medium of the telephone to reach low income and minority ethnic groups through ‘outcall’, whereby volunteers made ‘cold’ telephone calls to eligible women to promote breast screening. Crane et al’s was the larger intervention, making initial contact with 19 389 households based on the commercial database INFORUM, which provides neighbourhood profiles based on census and demographic information and consumer characteristics. Although the outcall system was efficient in contacting older women with low incomes, and who were relatively less educated, it was more successful in reaching white women than members of other ethnic groups. Crane et al also concluded that this method is labour intensive and potentially expensive to run. Working on a smaller scale, Duan et al’s intervention, using a telephone outreach method, was an interesting approach despite the many limitations of its evaluation. The intervention, although implicit, was based on the important role played by social support networks and social institutions such as churches in the life of ‘vulnerable’ groups. Female church members were recruited as peer counsellors to provide participants with one session of mammography counselling over the phone annually for


25

December 2003


two years. The intervention was a partnership between public health departments and 30 churches. However, the intervention had only a limited effect on the maintenance of uptakes and a small and insignificant effect on improving uptake among non-adherents (ie women who do not adhere to the recommended guidelines for breast screening, which is once a year in the USA). By failing to carry out a process evaluation, which would have enabled the effect of social support and networks provided by churches in reaching vulnerable women to be better understood, Duan et al missed the opportunity to enable others to better understand how such valuable partnerships between the health sector and the communities can be cultivated.110 In other words, practical knowledge that is crucial to this kind of intervention was lost because of the way in which the study was formulated. An inherent problem in measuring behavioural effects of this kind of ‘single method’108,109 or ‘multimethod’ intervention113 is that many confounding variables are present in the process. The claim to validity (Does it work? And can it be generalised?) is often hampered by methodological difficulties, such as true randomisation of experimental and control groups, sample size and the practical and ethical impossibility of ‘blinding’ participants. Intervention design based on ill conceived ideas may have effects that are more far reaching than negative results. The study by Jibaja-Weiss et al120 on tailoring messages for breast and cervical cancer screening using medical record data is an example. The intervention involved sending women in the intervention group personalised letters containing specific individual risk information. These individualised risk profiles were based on data extracted from the women’s medical records and took into account six factors: age, ethnicity, family history, parity, body mass index (BMI) and smoking status. Unexpectedly, the number of acceptances was higher in the control group than in the intervention group. The authors suggested that the women in the intervention group may have been alarmed by personalised risk information and the resulting anxiety induced avoidance rather than protective behaviour, thus decreasing the likelihood of women accepting screening. Although the researchers were aware that the intervention might have induced anxiety among women, they did not discuss the ethical implications of this intervention design. Although some interventions might have limited ‘generalisation’ value when viewed from the positivistic perspective, their innovative approaches could provide worthwhile learning experiences that could usefully be shared among practitioners. The innovative use of the peer counselling method by Bernstein et al111 in the Boston Medical Center project is a case in point. In an attempt to improve mammography utilisation among older minority ethnic women in inner cities, Bernstein et al joined forces with an inner city community project to recruit a team of multilingual peer educators to deliver breast screening counselling, using a simplified form of motivational interviewing, to older women who visited the emergency department of Boston Medical Center. After counselling, peer educators would arrange for these women to have access to free mammography promptly (the day after the interview) at a


26

December 2003


facility adjacent to the hospital, thus removing language and physical barriers to access, ie having to make an appointment and arrange transport. This intervention yielded a success rate of 69% among first time users. Although this study suffers from the same problem of generalisability as other intervention designs, it highlights the effect of culturally appropriate communication provided by peer educators and the removal of barriers to access as important strategies for improving utilisation. Two other complex interventions worth commenting on here are the Forsyth County Cancer Screening Project (FoCaS)116 and the ENCORE plus project.117 The FoCaS project is a project designed to test the effectiveness of interventions in both communities (outreach) and healthcare centres (in-reach). Outreach activities aimed to improve knowledge of and attitudes to breast and cervical screening and to increase participation among the target communities. In-reach activities aimed to improve professional knowledge and practice regarding follow-up recommendations for abnormal findings from mammography, clinical breast examinations and cervical smear tests. The project had a lifespan of four years. It moved systematically from assessing baseline conditions in the communities and in the health centres through to implementing interventions and evaluating effectiveness to, finally, transferring knowledge of such intervention to the comparison community. The project covered nine housing communities and four health centres. Most of the residents of these communities were older women from minority ethnic groups. The project involved local community agencies led by one of the participating health centres. Collection of data for final evaluation is not yet complete, therefore it is difficult to ascertain objectively the overall effectiveness of this intervention. However, the two pronged approach to intervention is of special merit, as the project recognises the effect that healthcare systems and professional practices might have upon service utilisation, in contrast to many other published projects, which often assume that problems stem from the communities, with the result that interventions tend to be directed only at them. This project takes a more balanced approach, in that improvement in professional practice was taken into account. Having established through surveys that some health professionals lacked knowledge and skills in some aspects of health promotion, training and practical support were provided and protocol for improving practice was established. The professionals were further supported by the implementation of computing technology (MedTrack) to facilitate the tracking of patients’ preventative activities, ie visits, test results, interventions and outcomes. Reports generated by the tracking system were then attached to the patient’s record and served as a cue for professionals to initiate offers of screening. Using an ecological approach, in which personal, social, economic, environmental and organisational determinants are integrated into the health promotion strategy, the ENCORE plus project aimed to improve the utilisation of screening among underserved women.117 The project was implemented at 78 YWCA sites across 30 states. This project illustrated some objective measures of success that could be designed to gauge effectiveness of the programme. By acknowledging that there


27

December 2003


were many difficulties associated with trial design, the evaluation of the project did not provide for controls or comparison groups. An estimate of improvement was made by comparing the project’s screening completion rates (SCRs) with data from other intervention projects obtained from a systematic literature review. At this point, we have to remind ourselves that a systematic review involves statistical pooling of data to generate significant results. Owing to the ecological conditions of communities, health intervention studies, unlike drug trials, seldom lend themselves to being repeated unproblematically. Thus, although it seems theoretically plausible to use data from other interventions, the likelihood that these data can be systematically pooled is suspect. However, the merits of this project lie not only in its outcomes but also in its ability to recognise that the cancer screening phenomenon is complex and requires a multidimensional approach to intervention. Within the framework of intervention, the project appears to have been sensitive to local conditions and to have honoured local staff’s knowledge and skills in tackling the issue as best they saw fit without insisting on rigid methods. At the same time, the project insisted on a standard protocol for data collection, thus ensuring the availability of some form of systematic measure of the programme’s effectiveness. Complex community based health interventions are always difficult to evaluate. Increasingly, the folly of a purely positivistic approach has been recognised. While attempts should be made to monitor and evaluate changes systematically, a pluralistic evaluation strategy, which takes account of different measures emanating from different perspectives, should be encouraged. In particular, the voices of those affected by the intervention should be heard. The results of the evaluation of these interventions are more usefully seen as learning tools rather than unchanging practice that can be applied indiscriminately across diverse populations with different social and cultural backgrounds. 5.4

Intervention using lay health educators

The concept of the lay health educator is not new. It can be broadly defined as a member of the community who is recruited by health agencies/ organisations to participate in health education and promotion initiatives. Reports of its use in developing countries first appeared in the 1980s.121 However, the practice of recruiting lay people from the communities to carry out health education was reported from the USA as early as the 1970s,122 and lay involvement in promoting cancer screening proliferated in the USA throughout the 1990s.123,124 At the same time, in the UK, the community health educator model in cancer screening developed out of three consecutive action research projects between 1990 and 2002,71,125,126 and the practice has been adopted by some health districts across the UK to address not only cancer screening but also other health promotion issues such as heart health and HIV/AIDS prevention.127 An electronic database search identified six journal publications between 1998 and 2002 that directly reported on the effectiveness of lay health workers in improving the uptakes of breast and cervical cancer screening in the USA. Two of the separate reports were of the same project and they are therefore treated as a single source for the purpose of this review.128,129


28

December 2003


Similar to the intervention studies reviewed above, the majority of these studies evaluated the effectiveness of lay involvement in cancer screening, using pre- and post-tests and cohort and comparison group or randomised or non-randomised controlled trial methods. However, a number of variations can be found among these studies. While five of the studies were community based and had explicitly constructed their interventions to target low income African-American, Latina, Asian-Indian-American and Vietnamese women, the sixth study based the intervention at a medical centre where a mixed population was recruited. Participants were subsequently stratified for evaluation purposes as white, African-American and Native American women.130 Across these studies, sample sizes varied from n = 146 to n = 1996, with intervention timescales varying between 18 months and four years. The many different terms, ie lay health advisers,128,129 senior aides,130 natural helpers (consejeras),132 community health educators133 and lay health workers,131 used in these projects reflect the many different interpretations of lay involvement by researchers. All these studies appear to have taken for granted the idea of using lay people in health interventions because no attempt was made to discuss even briefly any theoretical underpinning of this approach. However, the Pro La Vida model reported by Navarro et al132 did briefly refer to theoretical frameworks of ‘natural social networks in Hispanic communities’ espoused by Valle,134 who defines ‘natural helpers’ as ‘individuals who have a reputation in the community for good judgment, sound advice, a caring ear, and being discreet’ (p. 33). Earp and Flax128 and Earp et al129 have also referred to the concept of ‘natural helpers’ when describing the recruitment, training and activities of lay health advisers in their breast screening project in North Carolina. These studies claimed that lay involvement was effective in promoting utilisation of screening. As with the other interventions discussed above, they also suffered from many theoretical and methodological limitations in their claims for generalisability. For example, common problems faced were difficulties in randomisation, high attrition rates and no obvious method of controlling the quality of health education and promotion activities of the lay health advisers. However, while many researchers had acknowledged the problems in methods, there appeared to be no great interest in or understanding of the problems of conceptual difficulties in this area. Since the involvement of lay people relies on the implicit theory of social networks and social capital, the understanding of the wider social context of how individuals’ social relationships operate should be central to such research practice. The recent action research project (Straight Talking) using the community health educator model to inform women about breast screening in the UK has challenged this taken for granted idea.71 The results, from a scattered community, the Chinese, and an economically disadvantaged white community, have shown that the concepts of social networks, social capital and social exclusion are bound up with the concept of lay involvement in health promotion. Research and practice in this area are still at a nascent stage in the UK.


29

December 2003


5.5

Prostate cancer screening

Four studies on prostate cancer screening published between 1998 and 2002 were identified.135–138 Two were intervention studies aimed at improving knowledge of prostate cancer screening among AfricanAmerican men.135,136 One study investigated the relationship between the utilisation of prostate cancer screening services and actual and perceived knowledge of prostate cancer among adult men.137 A focus group study that reported on the responses of focus group participants to risk information is also included in the review.138 The two health intervention studies took place in Michigan and Chicago in the USA. The focus group study was conducted in Oklahoma City. All these cities have high concentrations of African-Americans and Hispanic-Americans. The study on actual and perceived knowledge of prostate cancer and utilisation is of poor quality. Not only did it fail to provide any contextual information, but its sampling strategy was unclear. The method of intervention was centred mainly on health education materials shown to respondents. The Michigan study recruited its participants through local primary care providers, community groups and minority churches. The Chicago study was essentially a hospital based study, with participants being recruited from the patients database of the University Health Services of the University of Chicago. Although men from many ethnic groups, including the white group, participated in these studies, African-American men made up by far the largest proportion of the participants. Pre- and post-questionnaire methods were the main methodological strategy in both evaluations. Baseline data from both studies confirmed that African-American and other minority ethnic participants, compared with white participants, had low levels of knowledge of prostate cancer and were most likely to be unable to identify its early symptoms. Both studies claimed that their interventions had been effective, in that knowledge of prostate cancer among participants had been significantly increased. However, both studies did acknowledge the limitations of their evaluation designs and were cautious in their claims to generalisability. While the health education materials used in the Michigan study were non-controversial, ie promotional flyers and a health education video, the personalised risk information used in the Chicago study does give rise to ethical concerns. In order to encourage African-American men to take up free prostate cancer screening on offer, Myers et al136 posted letters containing general information about prostate cancer and screening to both minimal and enhanced intervention groups and followed up with reminder calls from health educators. However, the enhanced intervention group also received a set of materials that was tailored to each participant. This set of materials included a personalised educational booklet on prostate cancer and early detection, together with a form containing information obtained from the baseline survey, highlighting tailored risk factors (eg the individual’s age, race, family history of prostate cancer and presence or absence of prostate cancer symptoms). This approach is similar to that of Jibaja-Weiss et al discussed earlier. While Jibaja-Weiss et al’s study120 reported that personalised risk information decreased the likelihood of participants attending cancer screening, Myers et al136 claimed that this


30

December 2003


approach was effective in influencing the acceptance of prostate cancer screening among African-American men. Setting aside the methodological weaknesses that perhaps render these contradictory results irrelevant, the issue at stake here is one of ethics. Should personalised risks be communicated to participants in such an impersonal way? What measures should the studies take in ensuring that undue anxieties are not raised by such information? In addition, JibajaWeiss et al acknowledged in their discussion that the risk information was based on medical records that might be neither accurate nor reliable. Should they have been more reflective on such an important ethical issue in their research practice? Similarly, should Myers et al have been concerned about the risk to confidentiality posed by posting personalised risk information to participants? In addition, did they consider the accuracy of this information, bearing in mind that it was based solely on the self-perception of participants taken from the baseline survey? Communicating personal cancer risks in such an impersonal way raises serious ethical questions in both research and practice, which researchers and practitioners should not ignore. 5.6

Colorectal cancer screening

Six research papers on colorectal cancer screening published from 1998 to the present day are included in this review.139–144 Five of the six studies were from the USA. These focused on exploring the relationships between psychosocial and cultural variables and uptakes of colorectal cancer screening. Minority ethnic groups targeted in these studies are Ashkenazi Jews, Korean-Americans, older American women and African-American women. While there is a high incidence of colorectal cancer among Chinese-Americans, 6% of the population of the Ashkenazi Jews were found to be genetically at risk of developing the disease. In general, a low level of knowledge and awareness of the screening guideline, ie the need for older men and women to be screened annually, was found among all the minority ethnic groups mentioned above. While Kim et al139 found that acculturation, ie duration of stay in the USA, was associated with awareness of colorectal screening among Korean men and women, Tang et al142 suggested in addition that acculturation might have influenced the initiation of screening behaviour among Chinese-American women. Both studies recommended outreach to the communities in the promotion of screening. As part of the Forsyth County Cancer Screening Project reported in the cervical and breast screening section, Brenes and Paskett141 investigated the relationship between knowledge, attitudes and beliefs and colorectal screening using the flexible sigmoidoscopy method among 202 low income African-American women in a randomised crosssectional survey. The uptake of colorectal screening among this group was low (28%) and physician recommendation was found to be significantly positively associated with uptake rates. Women who adhered to mammography screening guidelines were also likely to adhere to those of flexible sigmoidoscopy screening. Education was found not to be significantly related to screening behaviour. However, Friedman et al’s study140 of highly educated Ashkenazi Jews did find a relative effect of education. Physician recommendation was also a significant factor


31

December 2003


associated with obtaining faecal occult blood test (FOBT) among this population group. In the sixth study, McCaffery et al144 recently surveyed knowledge, attitudes and behavioural intentions in relation to colorectal screening in the UK. Using the TPB as a theoretical framework, a questionnaire was constructed to elicit the level of knowledge, attitude and behavioural intent of 1637 adults between the ages of 16 and 74. Sociodemographic characteristics included gender, age and marital status; educational level was measured by that age at which the respondent left full time education. The authors found that knowledge level was generally low, particularly among younger men and those who were less educated. They also found that women were more likely than men to have a negative attitude to screening. Low knowledge and negative attitudes were found to be associated with weaker intentions to be screened. The most noteworthy finding concerned participants’ recognition of early symptoms. It appeared that most respondents (90%) would seek medical advice if they noticed pain and bleeding, whereas only 77% would seek help if they noticed persistent changes in bowel habits. Although pain and bleeding might be better predictors of colorectal cancer, the disease can sometimes present without clinical symptoms. Thus, education on early symptoms such as change in bowel habits might be important. The study recommended that efforts should be made to increase the public’s knowledge and reduce negative perceptions about the disease, which in turn may influence the public’s decisions on whether or not to accept screening. Similar methodological problems as mentioned above have prevented the results of these studies from being generalised across population groups. Nevertheless, some insights can be gained from these studies. Low knowledge of colorectal cancer and lack of awareness of the need for screening are probably prevalent among minority ethnic groups. Uptake rates were low among these groups when compared with a recent study whose survey population mainly consisted of the white ethnic group (around 27–40% uptake found among minority groups versus around 61% in the white group). However, even among the white group, there was a clear gender difference in uptake, both in the type of test and in correlates of testing. Brawarsky et al143 found that men were more likely to have had screening overall by a colonoscopy or sigmoidoscopy and with a physician’s recommendation. The authors suggested that colorectal cancer was perceived as a male disease. Other insights provided by these studies were: 1. acculturation, ie the level of experience with the Western health system and the ability to speak fluent English, rather than cultural factors, ie modesty and traditional medicinal practice, were more likely to have an influence on acceptance of screening 2. women who have undergone mammography are more likely to accept colorectal screening. In considering the issue of colorectal screening in the UK, we have to be cautious in transferring knowledge obtained from these studies out of the context of the USA. At present many techniques are available


32

December 2003


for colorectal cancer screening, eg FOBT, FS (flexible sigmoidoscopy), colonscopy, DRE (digital rectal examination) and BE (barium enema) are offered in the USA. Most of the studies on minority ethnic groups’ responses to the colorectal screening programme were based on the American Cancer Society’s recommendation that individuals aged 50 years and older should undergo an FOBT annually and a sigmoidoscopy every five years. In the UK, FOBT is the only method available for the screening trials at present. It is also unlikely that other screening methods will be offered in primary care. The acceptance by both the general public and general practitioners of this screening procedure and programme in primary care needs to be investigated. The willingness and capacity of primary care professionals to promote the test to their at risk patients will also need to be considered. 5.7

Summary

In general, the survey studies reviewed seem to confirm that, at least in the USA, screening services are underused by low income minority ethnic groups. The psychosocial determinants of low uptake explored by these studies were knowledge of screening, literacy, age, attitude, beliefs, social and economic status, acculturation and ethnicity. Many of these variables are ill defined. Ethnicity and uptake rates are often conceived unproblematically as independent and dependent variables, respectively, in survey studies. Socioeconomic status and ethnicity are often collapsed into a single category. Although low income and minority ethnic status were consistently correlated with low uptake rates, other results, such as those for education and acculturation, were often contradictory. In addition, owing to the theoretical and methodological weaknesses of these studies, their results have limited generalisability. Many of the survey studies reviewed have suggested that a ‘usual source of healthcare’ and physician recommendation for screening are also important determinants for uptakes. Although, in the USA, having no insurance cover would deny women a usual source of care and therefore access to preventative services, the concept of usual source of care can also be seen to be applicable to an increasing number of homeless, asylum seekers and refugees in the UK. The studies of health beliefs as determinants of uptakes of screening behaviours were of various degrees of quality. There is inconsistency in application of the stated theoretical frameworks, ie HBM and TPB, and a lack of rigour in methods. Hence, the results of these studies need to be interpreted with caution. Overgeneralisation of the beliefs of a cultural group can run the risk of stereotyping, thus affecting quality of practice. Moreover, each minority ethnic culture is not a monolithic, unchanging whole that exists in a social vacuum. Many of the minority cultures are likely to be influenced by their structural positions vis-à-vis the dominant ideology and society. This will have influence upon their day to day lives, and might therefore, in turn, reinforce or change their beliefs. The majority of the intervention studies was also based on implicit social learning and behavioural theories, therefore most studies aimed to improve knowledge and awareness of the screening programmes among targeted populations. The methods generally used were provid-


33

December 2003


ing information or motivation counselling. Most of these interventions lack scientific rigour, and so their results cannot be generalised. However, these studies appeared to acknowledge implicitly the importance of social support and social networks in gaining access to target populations. Thus, they commonly employed bilingual researchers from target communities and/or incorporated lay health educators, peer educators or counsellors as part of their intervention design. Six studies evaluated the effectiveness of lay health educators in improving uptakes, and the results were positive. However, using a field trial approach to evaluate lay health workers as an intervention is methodologically inappropriate. Without a theoretical understanding of the concept of lay involvement and its current practice in various social and political contexts and awareness that the concept of the lay health educator is inherently bound up with concepts of social networks and social capital, it would be difficult for researchers to design an appropriate method for evaluation. Two large scale intervention projects, the Forsyth County Cancer Screening Project and the ENCORE plus project, took a more complex and developmental approach to health intervention design. These projects appeared to tackle both the health system and populations. However, evaluation results so far have not been reported. Studies on prostate and colorectal cancer screening in the USA have shown that members of minority ethnic groups tended to underuse these services. Intervention studies on prostate cancer screening tended to target mainly African-American men. Methods used so far remain informational. Participants’ anxiety about the adverse effect of treatments on sexuality and normal physical functioning need to be considered. A recent study on colorectal cancer in the UK has shown that a public awareness campaign to improve knowledge and symptom recognition of colorectal cancer, particularly among those who are less educated, may be needed. Specific efforts might also be needed to alter women’s perception of the disease as a male disease. Qualitative studies in response to such a screening programme and the procedure involved across different social groups will also be useful.


34

December 2003


6.

CONCLUSIONS

There are obvious moral and practical reasons as to why inequalities in health should be reduced.145 The relationship between socioeconomic status and health outcomes is well established,2,39 and recent studies show no sign that the gap between the advantaged and disadvantaged is closing.146 Recent national statistics on cancer highlight the unequal burdens shared by different socioeconomic groups and geographical regions.69 This literature review attempts to explore the concept of inequality, how it has been defined in the literature in relation to cancer and cancer screening and the perceived factors influencing access to cancer screening. It is hoped that, by understanding current approaches for improving access and informed choice, future strategies to tackle inequalities in the access to cancer screening can be made more effective and better targeted. 6.1

Limitations of the review

This review has the following limitations which readers need to take account of when considering its conclusions. 1. Because of the short time available for this project, only a limited number of research papers and other literature, in the form of books, book chapters, pamphlets and electronic material, could be included. Thus, in some ways the review is not comprehensive. It is, by and large, a snapshot of various related fields of knowledge available in a defined period of time 2. Because the review was undertaken from a pragmatic standpoint, with an emphasis on practical knowledge, it does not attempt to provide a thorough discussion of the theoretical issues from which many of the current concepts spring. Traditional researchers might find that discussion of some theoretical and methodological issues is inadequate 3. Some intervention methods and approaches were highlighted for their practical value and also in the light of the ethical issues that are identified in the review. Although it is my intention to offer insights to practitioners, they must take care in interpreting my judgement. More relevant insights might be gained by pursuing the research papers in question. The appendix contains detailed bibliographic information for those who wish to do so.

6.2

Development of concepts

A large scale survey has always been a tool to provide a big picture of social phenomena. Although the original concept of inequality of health mainly focused on the association between income inequalities and health outcomes, and therefore drew attention to social determinants of health, additional diverse explanations of health inequalities have since been developed. These include sociostructural factors, individual lifestyles and health behaviours. A recent development, which embraces both individual and social factors, is the recognition that disadvantaged groups may lack the resources to participate in the social and economic mainstream activity and, while there are income inequalities, there is also an uneven distribution of knowledge, power, prestige and social connections.147 Concepts such as social exclusion, social capital and


35

December 2003


social networks have recently emerged in health and social research as potential mechanisms linking social class and health status. Although there is still much confusion in the debate regarding the role of social capital and social networks for health among researchers, and many have doubts about how these concepts can be clearly defined and operationalised, these ideas appear to have gained popularity among public health and health promotion practitioners. Concepts such as engagement, empowerment and building of capacity have become key components of current practice in addressing inequalities in the community development and the new public health movements. However, how these practices are operationalised and how they can be evaluated is unclear. 6.3

Who suffers from inequality of access?

If the socially excluded have difficulties in engaging in the mainstream economic and civic activities of society, it is reasonable to assume that they would also have difficulties in engaging or negotiating with the healthcare system, particularly in its modern form. Official cancer and cancer screening statistics seem to support the possibilities of poor access by these groups. Ethnicity entered the health inequalities debate in the early 1990s. While there is some evidence to highlight the variations in health outcomes of different minority ethnic groups, generalisation is difficult because of the unavailability of systematic monitoring of ethnic status in primary care. Evidence of inequality of access among minority ethnic groups is often provided by small local studies and self-reported survey studies or is inferred from low uptake patterns in localities or regions where a higher concentration of low income and minority ethnic groups can be found. As ethnicity and socioeconomic status are often collapsed into one category among the studies reviewed here, it is difficult to establish the independent effect of ethnicity on patterns of access. However, the development of the concepts of social exclusion, social capital and social networks might offer a wider theoretical framework for capturing the synergistic effects of individuals’ behaviour, their structural position and geography. Innovative empirical studies based on this framework might be more useful in illuminating the process of how inequalities of access are produced or reproduced.

6.4

Health interventions

The majority of the interventions reviewed here, although having no obvious theoretical underpinning, are biased towards the behavioural model expounded by rationalists. Based upon these theoretical frameworks, changes are inevitably aimed at the individual. Rather than improving accessibility of services and supporting informed choice, women, specifically low income and minority ethnic women, are subjected to various forms of health education interventions. The application of purely behavioural interventions could be perceived as blaming the victims. Some of the larger scale interventions involving multiple agencies and multiple sites and those that took account of the system effects, albeit limited to the microlevel, do provide more useful ways to address the issue of inequality of access. However, evaluation of these complex interventions requires comprehensive evaluation strategies and adequate resources.


36

December 2003


6.5

System effect

The notion of access tends to be examined from either the user’s perspective or the service’s perspective. Seldom do investigations seek to understand the dynamic processes between the individual and the service system involved, let alone the wider social and economic context that might help to generate inequality of access. It has been highlighted, in recent action research projects on cancer and cancer screening information, that users can slip through the net (getting in)71 and/or find that their care is discontinued (getting through).148 The increasing differentiation in and specialisation of services in the modern health system tend to exacerbate inequality. The extent to which users’ access to services is affected by service organisation should be explored. To improve outcomes, interventions should not only be about improving access to cancer screening services among the vulnerable or at risk groups but also involve supporting individuals through different parts of the cancer care system once cancers have been diagnosed. Communication by health professionals is another aspect of the system effect revealed by the review of surveys and intervention studies in Chapter 5. The inability of doctors and clinicians to communicate with their patients about cancer risks and/or to recommend screening was consistently found to be associated with low utilisation. The Straight Talking project also highlighted missed health education opportunities in primary care.71 However, there are also lessons to be learned from some of these interventions, which assume that the effect of communicating personal cancer risks is always positive. The results of these studies are not only inconclusive but also raise serious ethical issues which researchers and practitioners should consider.

6.6

Information and communication

Communication is a two way process, and communicating cancer risks is no exception. If participation through informed choice is to be one of the fundamental principles of the cancer screening programme, an understanding of how risks can be communicated to the public and to individuals will need to be developed. More research will be needed, both on the presentation and communication of risk information and, crucially, on lay knowledge about cancer risks. The issue of communication is an important one. The development of genetic knowledge and technology in the information age will increase the potential for screening for many types of cancers. Correspondingly, there will be an increase in individuals’ demand for health information in support of their decisions to accessing or accepting screening services. Rapid social change will further exacerbate health inequalities, since those who are socially excluded may have difficulty in accessing information. However, there are signs that the innovative use of information technology can also help to support informed choice. For example, CHESS (Comprehensive Health Enhancement Support System) was developed for lay people. It is, therefore, important to take account of the development of health communication in the changing social and material conditions of the modern age, and to investigate the potential for new technology that might either reduce or perpetuate inequality of access.


37

December 2003


6.7

Theory and practice

There seems to be a gulf between conceptual development and empirical research studies in inequalities of health in general and in access to cancer screening in particular. While social networks and social capital are flourishing as components of an explanatory pathway linking socioeconomic status and health outcomes, little research has been carried out in exploring the value of these concepts in relation to access to health and social services. The complexity inherent in the task of intervention to improve access among the socially vulnerable demands a better conceptual integration between the notion of access and social conditions. Actions taken in large scale, community based projects are often embedded in a web of related social relations and social conditions. Traditional intervention research methods are not adequate for such a complex task. The development of an action research approach is growing nationally and globally; it has the potential to deal with complex processes and is sensitive to local knowledge and conditions. It should, therefore, be considered by grant-making bodies as an alternative method for improving both theoretical and practical knowledge in this area.


38

December 2003


7.

RECOMMENDATIONS

The advantage in taking a broad approach to the review of literature is that we can begin to understand how the problem of inequalities of access to cancer screening has been defined and learn of the range of approaches that have been taken to address the issue. Consequently, the results of the review indicate a broad area of work that needs to be undertaken if the issue of inequalities of access is to be adequately addressed. 7.1

System

On the national level, we need to have a broad view of how local health districts have planned to tackle inequalities of access to cancer screening. In particular, what initiatives have been implemented and what approaches have been taken and how are they evaluated? Both in research and in practice, there should be a better integration of the notion of access to social conditions. Therefore, a better understanding of how the concepts of social exclusion, social networks and ethnicity are related to inequalities of access is required. The relative contribution of system effects to inequality of access needs to be better understood. This will include the configuration, administration, implementation and evaluation of screening services and health educational and promotional capacity and capability of healthcare providers. In regard to building the public health and health promotion capacities to launch large scale community based interventions, we need to improve public health intelligence on the following fronts: • • • •

local information on patterns of access on both the primary care and secondary care levels availability of cancer and cancer screening information and patterns of access to healthcare services local health alliance and stakeholders’ information information on communities and voluntary organisations.

There is a dearth of information in this area in the UK. The bulk of the information from the USA, although insightful, is largely inapplicable in the context of the UK. Thus, there is an urgent need to generate knowledge and information in the UK to support and improve practices and research. The publication of results of local interventions should be actively encouraged. 7.2

Users

From the users’ perspective, we need to understand how vulnerable groups think and feel about cancer screening and how they access and respond to information that is currently available. There is also a need to understand the current lay knowledge and perceptions of risks of prostate and/or colorectal cancers, and the level of acceptance of screening among vulnerable groups, long before new cancer screening programmes are implemented.


39

December 2003


There is a need to understand how social networks and social capital can influence access to health services in general and cancer screening in particular. The community health educator model is inherently linked with concepts of social networks, social capital and capacity building. Therefore, it should be evaluated in respect of its potential and capacity to improve access to cancer screening in the wider context of health development. The current practice of this model in those districts where it has been implemented should be better understood. 7.3

Information and communication

There needs to be a better understanding of how published information on cancer screening is currently disseminated and utilised, and by whom and to what effect. There is also a need to understand how cancer and cancer screening is communicated to the public and individuals and the role played by risk information in bringing about informed choice and acceptance of screening among minority ethnic and low income groups.


40

December 2003


REFERENCES 1. Macintyre S. The Black Report and beyond: what are the issues? Social Science and Medicine, 1997, 44(6): 723–745. 2. Townsend P, Davidson N. Inequalities in Health: the Black Report. London: Penguin, 1982. 3. Independent Inquiry into Inequalities in Health Report (the Acheson Report). London: The Stationery Office, 1998. 4. Blaxter M. Health and Lifestyle. London: Tavistock/Routledge, 1990. 5. Rogers A, Popay J, Williams G, Latham M. Inequalities in Health and Health Promotion: Insights from the Qualitative Research Literature. London: Health Education Authority, 1997. 6. Department of Health. Health of the Nation. London: HMSO, 1995. 7. Popay J, Thomas C, Williams G et al. A proper place to live: health inequalities, agency and the normative dimensions of space. Social Science and Medicine, 2003, 57(1): 55–69. 8. Whitelaw S, Beattie A, Balogh R, Watson J. A Review of the Nature of Action Research. Sustainable Health Action Research Programme. Cardiff: Welsh Assembly Government Publication, 2003. 9. Wilkinson RG, Kawachi I, Kennedy BP. Mortality, the social environment, crime and violence. In: Bartley M, Blane D, Davey Smith, G (eds) The Sociology of Health Inequalities. London: Blackwell, 1998: 19–37. 10. Kawachi I, Kennedy B, Wilkinson R. Crime: social disorganisation and relative deprivation. Social Science and Medicine, 1999, 48: 719–731. 11. Wilkinson R. The epidemiological transition: from material society to social disadvantage? Daedalus, 1994, 123: 61–77. 12. Putnam R. The prosperous community: social capital and economic growth. American Prospect, 1993, Spring: 35–42. 13. Kawachi I, Kennedy B, Lochner K, Prothrow-Stith D. Social capital, income inequality, and mortality. American Journal of Public Health, 1997, 87: 1491–1498. 14. Coleman JS. Foundations of Social Theory. Cambridge, MA: Harvard University Press, 1990. 15. Bourdieu P. The forms of capital. In: Richardson JG (ed.) Handbook of Theory and Research for the Sociology of Education. New York: Greenwood, 1991: 241–258. 16. Kawachi I, Kennedy B, Glass R. Social capital and self-rated health: a contextual analysis. American Journal of Public Health, 1998, 89: 1187–1193. 17. Rose R. How much does social capital add to individual health? A survey study of Russians. Social Science and Medicine, 2000, 51: 1421–1435. 18. Macinko J, Starfield B. The utility of social capital in research on health determinants. The Milbank Quarterly, 2001, 79: 387–427. 19. Rico A, Fraile M, Gonzalez P. Regional decentralisation of health policy in Spain: social capital does not tell the whole story. West European Politics, 1999, 21: 180–199. 20. Balarajan R. Ethnic differences in mortality from ischaemic heart disease and cerebrovascular disease in England and Wales. British Medical Journal, 1991, 302: 560–564. 21. Balarajan R, Soni Raleigh V. The Health of the Nation: Ethnicity and Health, a Guide for the NHS. London: Department of Health, 1993. 22. Balarajan R, Soni Raleigh V. Ethnicity and Health in England. London: HMSO, 1995. 23. Navarro J. Race or class versus race and class? Lancet, 1990, 336: 1238–1240. 24. Nazroo J. The Health of Britain’s Ethnic Minorities. London: PSI, 1997.


41

December 2003


25. Ahmad WIU. Ethnic statistics: better than nothing or worse than nothing? In: Droling D, Simpson S (eds) Statistics in Society: the Arithmetic of Politics. London: Arnold, 1999. 26. Fenton S, Charsley K. Epidemiology and sociology as incommensurate games: accounts from the study of health and ethnicity. Health, 2000, 4: 403–425. 27. Davey Smith G, Charsley K, Lambert H et al. Ethnicity, health and the meaning of socio-economic position. In: Graham H (ed.) Understanding Health Inequalities. Buckingham: Open University Press, 2000: 25–37. 28. Castell M. The Information Age. Vol. 3: End of Millennium. Oxford: Blackwell, 1996. 29. Higgs P, Scambler G. Explaining health inequalities: how useful are concepts of social class? In: Higgs P, Scambler G (eds) Modernity, Medicine and Health. London: Routledge, 1998. 30. The Social Exclusion Unit Leaflet. London, Cabinet Office, 2000. Also available at www.cabinet-office-gov.uk/seu/index/march. 31. Social Inclusion and Community Development Practice. London: Community Development Foundation, 2001. Also available at www.cdf.org.uk/ html.socinc.html, 6.7.01. 32. Curtis S, Jones IR. Is there a place for geography in the analysis of health inequality? In: Bartley M, Blane D, Davey Smith G. (eds) The Sociology of Health Inequalities. London: Blackwell, 1998: 85–114. 33. Nazroo J. Genetic, cultural or socio-economic vulnerability? Explaining ethnic inequalities in health. In: Bartley M, Blane D, Davey Smith G (eds) The Sociology of Health Inequalities. London: Blackwell, 1998: 151–192. 34. Popay J, Williams G, Thomas C, Gatrell A. Theorising inequalities in health: the place of lay knowledge. In: Bartley M, Blane D, Davey Smith G (eds) The Sociology of Health Inequalities. London: Blackwell, 1998. 35. Cattell V. Poor people, poor places, poor health: the mediating role of social networks and social capital. Social Science and Medicine, 2001, 52: 1501– 1516. 36. Bartley M, Blane D, Davey Smith G. Introduction: beyond the Black Report. In: Bartley M, Blane D, Davey Smith G (eds) The Sociology of Health Inequalities. London: Blackwell, 1998: 1–16. 37. Giddens A. Modernity and Self-identity: Self and Society in the Late Modern Age. Cambridge: Polity Press, 1991. 38. Jones J. Community development and health education: concepts and philosophy. In: ‘Roots and Branches’. Papers from the OU/HEA 1990 Winter School on Community Development and Health. Buckingham: Open University Press, 1991: 27–36. 39. Whitehead M. The Health Divide. London: Health Education Council, 1987. 40. Deprivation and Ill Health. London: British Medical Association, 1987. 41. Petersen D. The potential of social capital measures in the evaluation of comprehensive community-based health initiatives. American Journal of Evaluation, 2002, 23(1): 55–64. 42. Tudor Hart J. The inverse care law. Lancet, 1971, i: 405–412. 43. Goddard M, Smith P. Equity of Access to Health Care. York: University of York, 1998. 44. Aday LA, Fleming GV, Anderson R. Access to Medical Care in the U.S.: Who Has it, Who Doesn’t. Chicago: Pluribus Press, 1984. 45. Millman ML. Access to Health Care in America. Washington, DC: Institute of Medicine, National Academy Press, 1993. 46. Hoof F, Weeghel J, Kroon H. Community care: exploring the priorities of clients, mental health professionals and community providers. International Journal of Social Psychiatry, 2000, l46: 208–220. 47. Watt IS, Franks AJ, Sheldon TA. Health and health care of rural populations in the U.K.: is it better or worse? Journal of Epidemiology and Community Health, 1994, 48: 16–21.


42

December 2003


48. Lee RG, Garvin T. Moving from information transfer to information exchange in health and health care. Social Science and Medicine, 2003, 56(3): 449–464. 49. Gulliford M, Hughes D, Figeroa-Munoz J et al. Access to Health Care. Report of a scoping exercise for the National Co-ordinating Centre for NHS Service Delivery and Organisation R & D (NCCSDO). London, 2001. 50. Maxwell RJ. Quality assessment in health. British Medical Journal, 1984, 288: 1470–1472. 51. Sassi F, Archard L, Le Grand J. Equity and the economic evaluation of healthcare. Health Technology Assessment, 2001, Vol 5, No 3. 52. Lupton D. Medicine as Culture: Illness, Disease and Body in Western Societies. London: Sage, 1994. 53. Andersen RM. Revisiting the behavioural model and access to medical care: does it matter? Journal of Health and Social Behaviour, 1995, 36: 1–10. 54. Mandelblatt JS, Yabroff KR, Kerner JF. Equitable access to cancer services: a review of barriers to quality care. Cancer, 1999, 86: 2378–2390. 55. Doyle Y. A survey of the cervical screening service in a London district, including reasons for non-attendance, ethnic responses and views on the quality of service. Social Science and Medicine, 1991, 32: 953–957. 56. Rudiman R, Gilbert FJ, Ritchie LD. Comparison of uptake of breast screening, cervical screening, and childhood immunisation. British Medical Journal, 1995, 310: 229. 57. Cancer Research Centre. Cancer Fact Sheet No. 13.5. London: Imperial Cancer Research Fund, 1994. 58. Cancer Research Centre. Cancer Fact Sheet No. 7.5. London: Imperial Cancer Research Fund, 1997. 59. Rudat K. Mori Health Research Unit. Report on the Health and Lifestyles Survey of Black and Minority Ethnic Groups in England. London: Health Education Authority, 1994. 60. Meyerowitz BE, Richardson J, Hudson S, Leedham B. Ethnicity and cancer outcomes: behavioural and psychosocial considerations. Psychological Bulletin, 1998, 123(1): 47–70. 61. Yabroff KR, Gordis L. Does stage at diagnosis influence the observed relationship between socioeconomic status and breast cancer incidence, casefatality, and mortality? Social Science and Medicine, 2003 (in press). 62. Breen N, Figureroa JB. Stage of breast and cervical cancer diagnosis in disadvantaged neighborhoods: a prevention policy perspective. American Journal of Preventive Medicine, 1996, 12: 319–326. 63. Chang SW, Kerlikowske K, Napoles-Springer A et al. Racial differences in timeliness of follow-up after abnormal screening mammography. Cancer, 1996, 78: 1395–1402. 64. Lannin DR, Mathews HF, Mitchell J et al. Influence of socioeconomic and cultural factors on racial differences in late-stage presentation of breast cancer. Journal of the American Medical Association, 1998, 279: 1801–1807. 65. McCarthy BD, Yood MU, Boohaker EA et al. Inadequate follow-up of abnormal mammograms. American Journal of Preventive Medicine, 1996, 12: 282–288. 66. Macleod U, Ross S, Twelves C et al. Primary and secondary care management of women with early breast cancer from affluent and deprived areas: retrospective review of hospital and general practice records. British Medical Journal, 2000, 320: 1442–1445. 67. Orbell S. Cognition and affect after cervical screening: the role of previous test outcome and personal obligation in future uptake expectations. Social Science and Medicine, 1996, 43: 1237–1243. 68. Kravdal Ø. Social Inequalities in Cancer Survival. Department of Economics, University of Oslo and Cancer Registry of Norway, 1998. http://swopec.hhs.se/ osloec/abs/osloec1998_003.htm.


43

December 2003


69. Quinn M, Babb P, Brock A et al. Cancer Trends in England and Wales (1950–1999): Study on Medical and Population Subjects No. 66. London: The Stationery Office, 2001. 70. Department of Health. Statistical Bulletin. Breast Screening Programme, England: 1999–2000. London: The Stationery Office, 2001. 71. Chiu LF. Straight Talking: Communicating Breast Screening Information in Primary Care. Leeds: Nuffield Institute for Health, University of Leeds, 2002. 72. Jepson R, Clegg A, Forbes C et al. The Determinants of Screening Uptake and Interventions for Increasing Uptake: a Systematic Review. Health Technology Assessment, 2000, Vol 4, No 14. 73. Atkinson M, Clark M, Clay D et al. Systematic Review of Ethnicity and Health Service Access for London. Warwick: University of Warwick, 2001. 74. Hoare T. Breast screening and ethnic minorities. British Journal of Cancer, 1996, 74(XXIX): S38–S41. 75. Box V. Cervical screening: the knowledge and opinions of black and minority ethnic women and health advocates in East London. Health Education Journal, 1998, 57(3): 3–15. 76. Atri J, Falshaw M, Gregg R et al. Improving uptake of breast screening in multi-ethnic populations: a randomised controlled trial using practice reception staff to contact non-attenders. British Medical Journal, 1997, 315: 1356–1349. 77. Chiu LF, Heywood P, Jordan J et al. Balancing the equation: the significance of professional and lay perceptions in the promotion of cervical screening among minority ethnic women. Critical Public Health, 1999, 9(1): 5–22. 78. Chalmers I, Altman DG (eds). Systematic Reviews. London: BMJ Books, 1995. 79. Kim K, Yu E, Chen E et al. Cervical cancer screening knowledge and practices among Korean-American women. Cancer Nursing, 1999, 22: 297–302. 80. Risendal B, DeZapien J, Fowler B et al. Pap smear screening among urban Southwestern American Indian women. Preventive Medicine, 1999, 29: 510–518. 81. Taylor VM, Schwartz SM, Jackson JC et al. Cervical cancer screening among Cambodian-American women. Cancer Epidemiology, Biomakers & Prevention, 1999, 8: 541–546. 82. Lindau ST, Tormori C, Lyons T et al. The association of health literacy with cervical cancer prevention knowledge and health behaviours in a multiethnic cohort of women. American Journal of Obstetrics and Gynecology, 2002, 186: 938–943. 83. Champion VL, Menon U, McQuillen DH, Scott C. Validity of self-reported mammography in low-income African-American women. American Journal of Preventive Medicine, 1998, 14: 111–117. 84. Giuliano A, Papenfuss M, DeZapien J et al. Breast cancer screening among Southwest American Indian women living on-reservation. Preventive Medicine, 1998, 27: 135–143. 85. Laws MB, Mayo SJ. The Latina breast cancer control study, year one: factors predicting screening mammography utilization by urban Latina women in Massachusetts. Journal of Community Health, 1998, 23: 251–267. 86. Wismer BA, Moskowitz JM, Chen AM et al. Mammography and clinical breast examination among Korean American women in two California counties. Preventive Medicine, 1998, 27: 144–151. 87. Friedman LC, Moore A, Webb JA, Puryear LJ. Breast cancer screening among ethnically diverse low-income women in a general hospital psychiatry clinic. General Hospital Psychiatry, 1999, 21: 374–381. 88. Risendal B, Roe D, DeZapien J et al. Influence of health care, cost, and culture on breast cancer screening: issues facing urban American Indian women. Preventive Medicine, 1999, 29: 501–509. 89. O’ Malley MS, Earp JA, Hawley ST et al. The association of race/ethnicity, socio-economic status, and physician recommendation for mammography: who


44

December 2003


90.

91.

92.

93.

94.

95.

96.

97.

98.

99. 100.

101.

102.

103. 104. 105.

106.

107.

108.

gets the message about breast cancer screening? American Journal of Public Health, 2001, 91(1): 49–54. Maxwell CJ, Bancej CM, Snider J. Predictors of mammography use among Canadian women aged 50–59: findings from three 1996/97 National Population Health Surveys. Canadian Medical Association Journal, 2001, 164: 329–334. Royak-Schaler R, Klabunde CN, Greene WF et al. Communicating breast cancer risk: patient perceptions of provider discussions. Women’s Health eJournal 2002, 7(2). http://www.mescape, com/viewarticle/429553_print. Coughlin SS, Uhler RJ, Blackman DK. Breast and cervical cancer screening practices among American Indian and Alaska Native women in the United States, 1992–1997. Preventive Medicine, 1999, 29: 287–295. Zambrana RE, Breen N, Fox SA, Gutierrez-Mohamed L. Use of cancer screening practices by Hispanic women: analyses by sub-group. Preventive Medicine, 1999, 29: 466–477. Maxwell AE, Bastani R, Warda US. Demographic predictors of cancer screening among Filipino and Korean immigrants in the United States. American Journal of Preventive Medicine, 2000, 18(1): 62–68. Foxall MJ, Barron CR, Houfek JF. Ethnic influences on body awareness, trait anxiety, perceived risk, a breast and gynaecological cancer screening practices. Oncology Nursing Forum, 2001, 28: 727–738. Coughlin SS, Uhler RJ. Breast and cervical cancer screening practices among Hispanic women in the United States and Puerto Rico, 1998–1999. Preventive Medicine, 2002, 34: 242–251. O’Malley AS, Forrest CB, Mandelblatt J. Adherence of low-income women to cancer screening recommendations: the role of primary care, health insurance, and HMOs. Journal of General Internal Medicine, 2002, 17: 144–154. Phillips JM, Cohen MZ, Moses G. Breast cancer screening and AfricanAmerican Women: fear, fatalism and silence. Oncology Nursing Forum, 1999, 26: 561–571. Jennings-Dozier K. Perceptual determinants of pap test up-to-date status among minority women. Oncology Nursing Forum, 1999, 26: 1327–1333. Schulmeister L, Lifsey DS. Cervical cancer screening knowledge, behaviours, and beliefs of Vietnamese women. Oncology Nursing Forum, 1999, 26: 879–887. Ishida DN, Toomata-Mayer TF, Braginsky NS. Beliefs and attitudes of Samoan women toward early detection of breast cancer and mammography utilisation. Cancer (Suppl) 2001, 91: 262–266. Jirojwong S, Manderson L. Beliefs and behaviors about pap and breast selfexamination among Thai immigrant women in Brisbane, Australia. Women & Health, 2001, 33(3/4): 47–66. Rosenstock I, Strecher V, Becker M. Social learning theory and the Health Belief Model. Health Education Quarterly, 1988, 15: 175–183. Ajzen I. The theory of planned behavior. Organizational Behavior and Human Decision Processes, 1991, 50: 179–211. Howe SL, Delfino RJ, Taylor TH, Anton-Culver H. The risk of invasive cervical cancer among Hispanics: evidence for targeted preventive interventions. Preventive Medicine, 1998, 27: 674–680. Jackson JC, Taylor VM, Chitnarong K et al. Development of a cervical cancer control intervention program for Cambodian American women. Journal of Community Health, 2000, 25: 359–375. Dignan MB, Michielutte R, Wells HB et al. Health education to increase screening for cervical cancer among Lumbee Indian women in North Carolina. Health Education Research, 1998, 13: 545–556. Crane LA, Leakey TA, Woodworth MA et al. Cancer information service – initiated outcalls to promote screening mammography among low-income


45

December 2003


109.

110.

111.

112.

113.

114.

115.

116. 117.

118.

119.

120.

121.

122. 123.

124.

125. 126.

127.


and minority women: design and feasibility testing. Preventive Medicine, 1998, 27: S29–S38. Crane, LA, Leakey TA, Rimer BK et al. Effectiveness of a telephone outcall intervention to promote screening mammography among low-income women. Preventive Medicine, 1998, 27: S39–S49. Duan N, Fox SA, Derose KP, Carson S. Maintaining mammography adherence through telephone counseling in a church-based trial. American Journal of Public Health, 2000, 90: 1468–1471. Bernstein J, Mutschler P, Bernstein E. Keeping mammography referral appointments: motivation, health beliefs, and access barriers experienced by older minority women. Journal of Midwifery & Women’s Health, 2000, 45: 308–313. Husaini BA, Sherkat DE, Bragg R et al. Predictors of breast cancer screening in a panel study of African-American Women. Women & Health, 2001, 34(3): 35–51. Nguyen T, Vo PH, McPhee SJ, Jenkins CNH. Promoting early detection of breast cancer among Vietnamese-American women. Cancer (Suppl), 2001, 91: 267–273. Valdez A, Banerjee K, Ackerson L, Fernandez M. A multimedia breast cancer education intervention for low-income Latinas. Journal of Community Health, 2002, 27(1): 33–51. Bell TS, Branston LK, Newcombe RG, Barton GR. Interventions to improve uptake of breast screening in inner city Cardiff general practices with ethnic minority lists. Ethnicity & Health, 1999, 4: 277–284. Paskett ED, McMahon K, Tatum C et al. Clinic-based interventions to promote breast and cervical cancer screening. Preventive Medicine, 1998, 27: 120–128. Fernandez ME, DeBor M, Candreia MJ et al. Evaluation of ENCORE plus: a community-based breast and cervical screening program. American Journal of Preventive Medicine, 1999, 16(3S): 35–49. Jenkins CNH, McPhee SJ, Bird JA et al. Effect of a media-led education campaign on breast and cervical cancer screening among Vietnamese-American Women. Preventive Medicine, 1999, 28: 395–406. Hiatt RA, Pasick RJ, Stewart S et al. Community-based cancer screening for underserved women: design and baseline findings from the breast and cervical cancer intervention study. Preventive Medicine, 2001, 33: 190–203. Jibaja-Weiss ML, Volk RJ, Kingery P et al. Tailored messages for breast and cervical cancer screening of low-income and minority women using medical record data. Patient Education and Counseling, 2003, 50: 123–132. Ofosu-Amaah V. National Experiences in the Use of Community Health Workers: a Review of Current Issues and Problems. Geneva: WHO Offset Publication No 71, 1983. Salber E. The lay advisor as a community health resource. Journal of Health Politics, Policy Law, 1979, 3: 469–79. Brownstein JN, Cheal N, Ackermann SP et al. Breast and cervical cancer screening in minority populations: a model for using lay health educators. Journal of Cancer Education, 1992, 7: 321–326. Eng E, Parker E, Harlan C. Lay health advisor intervention strategies: a continuum from natural helping to paraprofessional helping. Health Education and Behaviour, 1997, 24: 413–417. Chiu LF (ed.). Communicating Breast Screening Messages to Minority Women. A Conference Report. Leeds Health Promotion Service, 1993. Chiu LF. Woman-to-Woman: Promoting Cervical Screening Amongst Minority Ethnic Women in Primary Care. A Participatory Action Research Report. Rotherham Health Authority, 1998. WNCCC. Survey of Health Service Activity with Regard to Breast Screening and Women from Black and Minority Ethnic Groups. London: Women National Cancer Control Campaign, 2001.

46

December 2003


128. Earp JA, Flax VL. What lay health advisors do: an evaluation of advisors’ activities. Cancer Practice, 1999, 7(1): 16–21. 129. Earp JA, Eng E, O’Malley MS et al. Increasing use of mammography among older, rural African-American women: results from a community trial. American Journal of Public Health, 2002, 92: 646–654. 130. Margolis KL, Lurie N, McGovern PG et al. Increase breast and cervical cancer screening in low-income women. Journal of General Internal Medicine, 1998, 13: 515–521. 131. Bird JA, McPhee SJ, Ha N et al. Opening pathways to cancer screening for Vietnamese-American women: lay health workers hold a key. Preventive Medicine, 1998, 27: 821–829. 132. Navarro AM, Senn KL, McNicholas LJ et al. Por la Vida Model intervention enhances use of cancer screening tests among Latinas American. Journal of Preventive Medicine, 1998, 15(1): 32–41. 133. Sadler GR, Dhanjal SK, Shah NB et al. Asian Indian women: knowledge, attitudes and behaviours towards breast cancer early detection. Public Health Nursing, 2001, 18: 357–363. 134. Valle R. A natural resource system for health – mental health promotion in the Latino/Hispano communities. In: Valle R, Vega W (eds) Hispanic Natural Support Systems. Sacramento, CA: Department of Mental Health, 1980: 35–43. 135. Barker KR, Shaw R, Folts M et al. Differences between African-American and Caucasian men participating in community-based prostate cancer screening program. Journal of Community Health, 1998, 23: 441–451. 136. Myers RE, Chodak GW, Wolf TA et al. Adherence by African-American men to prostate cancer education and early detection. Cancer, 1999, 86: 88–104. 137. Agho AO, Lewis MA. Correlates of actual and perceived knowledge of prostate cancer among African-Americans. Cancer Nursing, 2001, 24: 165–171. 138. McFall SL, Hamm RM. Interpretation of prostate cancer screening events and outcomes: a focus group study. Patient Education and Counseling, 2002, 49: 207–218. 139. Kim K, Yu ESH, Chen EH et al. Colorectal cancer screening: knowledge and practice among Korean Americans. Cancer Practice, 1998, 6: 167–175. 140. Friedman LC, Webb JA, Richards CS, Plon SE. Psychological and behavioral factors associated with colorectal cancer screening among Ashkenazim. Preventive Medicine, 1999, 29: 119–125. 141. Brenes GA, Paskett ED. Predictors of stage of adoption for colorectal cancer screening. Preventive Medicine, 2000, 31: 410–416. 142. Tang TS, Solomon LJ, McCracken LM. Barriers to faecal occult blood testing and sigmoidoscopy among older Chinese-American women. Cancer Practice, 2001, 9: 277–282. 143. Brawarsky P, Brooks DR, Mucci LA. Correlates of colorectal cancer testing in Massachusetts’s men and women. Preventive Medicine, 2003, 36: 659–668. 144. McCaffery K, Wardle J, Waller J. Knowledge, attitudes, and behavioral intentions in relation to the early detection of colorectal cancer in the United Kingdom. Preventive Medicine, 2003, 36: 525–535. 145. Woodward A, Kawachi I. Why reduce health inequalities? Journal of Epidemiology and Community Health, 2000, 54: 923–929. 146. Shaw M, Dorling D, Gordon D, Smith GD. The Widening Gap: Health Inequalities and Policy in Britain. Bristol: Policy Press, 1999. 147. Link BG, Phelan JC. Understanding sociodemographic differences in health, the role of fundamental social causes. American Journal of Public Health, 1996, 86: 471–473. 148. Chiu LF, Wistow G. Improving Access to Cancer Information for Doncaster Residents: a Needs Assessment Project. Nuffield Institute for Health, University of Leeds, 2002.


47

December 2003


APPENDIX: SUMMARY OF SELECTED PAPERS


48

December 2003

Project Agho and Lewis (2001)

Barker et al (1998)

Bell et al (1999)

Bernstein et al (2000)

Ref no 137

135

115

111

To increase mammography rates among inner city women aged 50 years or older

To increase the uptake of breast screening in three inner city GP practices with a high proportion of ethnic minority patients

To evaluate the effectiveness of an educational prostate cancer screening programme among 944 men in an urban community in the Midwest, USA

Aim To investigate the effects of education, income, age and health insurance coverage on actual and perceived knowledge of prostate cancer. To explore the extent to which the use of prostate cancer screening services might be attributed to actual and perceived knowledge of cancer

A peer delivered intervention, which consisted of interview, mammography referral and the scheduling of a next day appointment, was conducted among a convenience sample of 151 culturally and racially diverse older women through a primary care referral project operating within an urban emergency department

Interventions: identification of ethnic language groups; GP endorsement letter; translated literature including multilingual leaflet, GP letter, screening invitation; transport to the screening centre; language support

An educational intervention in the form of a video that stressed the importance of early detection and treatment of prostate cancer was conducted. A brief questionnaire, administered before and after the video, assessed knowledge and attitudes concerning prostate cancer

The findings demonstrated that an interactive intervention among older women of colour has the potential to increase mammography uptake rates dramatically

The study found that translated literature, GP endorsement letters and language support by link workers were beneficial. The provision of free transport was ineffective and underutilised. Uptake was highest among Urdu and Gujarati speaking groups and lowest among Bengali and Somali speakers, who are hardest to reach. There is scope for improving the attendance rate among ethnic minority groups, but this can be costly

A relatively low level of knowledge regarding prostate aetiology and clinical factors was found among African-Americans, highlighting the need for educational programmes to target minority populations. There is a need to provide minority favoured access through private screening appointments

Method Key points A questionnaire survey study among a non-random The study revealed a moderately strong correlation sample of 108 African-American men between actual and perceived knowledge of prostate cancer. The use of prostate cancer screening services was positively associated with actual and perceived knowledge of the disease. Actual income and actual and perceived knowledge of prostate cancer were both correlated with having health insurance

Brenes and Paskett (2000)

Champion et al To compare self-reported (1998) mammography and medical records of mammography status in a low income AfricanAmerican sample

Coughlin et al (1999)

141

83

92

To explore barriers to cancer screening and underuse of cancer screening tests among American Indian and Alaska Native women

To investigate the predictors of adherence to colorectal screening among AfricanAmerican women

To evaluate correlates of current colorectal testing among individuals aged 50 and older, based on undergoing any of the tests recommended in the American Cancer Screening guidelines. To explore gender differences in colorectal cancer testing

Brawarsky et al (2003)

143

A controlled trial of a community outreach intervention to promote recognition, receipt and screening interval maintenance of clinical breast examinations, mammograms and Pap smears among VietnameseAmerican women

Bird et al (1998)

131

Secondary data analysis. The breast and cervical cancer screening practices of 4961 American Indian and Alaska Native women in 47 US states from 1992 to 1997 were analysed using data from the BRFSS

A total of 229 low income (at 150% of poverty or below) African-American women were interviewed regarding breast cancer screening. Self-reported mammography was compared with medical records at the facility named by respondents

A total of 202 predominantly low income and African-American women were interviewed to determine knowledge of, attitudes and beliefs about and practices related to flexible sigmoidoscopy

Analysis of secondary data. Questionnaire data collected from Massachusetts using the BRFSS (Behavioral Risk Factor Surveillance System) and a colorectal cancer screening survey

Over a three year period, indigenous lay health workers conducted small group sessions among Vietnamese women in a low income district of San Francisco, California. Women in Sacramento, California, served as controls

Older women and those less educated were less likely to be screened. Women who had seen these results emphasise the need for continued efforts to ensure that American Indian and Alaska Native women who are elderly or medically underserved have access to cancer screening services

Self-report alone may not provide accurate rates of mammography compliance. Further research among ethnic and low income women is necessary

Seventy-two per cent of the women in this study were non-adherent to flexible sigmoidoscopy screening guidelines. Psychosocial factors play an important role in screening for colorectal cancer

Men were currently more likely than women to be tested for colorectal cancer. Physician recommendation for testing was strongly associated with testing among men and women. A perceived high risk of colorectal cancer was more strongly associated with testing among men while other cancer screening was more strongly associated with testing among women

Trained Vietnamese lay health workers significantly increased Vietnamese women’s recognition, receipt and maintenance of breast and cervical cancer screening tests

Coughlin and Uhler (2002)

Crane et al (1998)

Crane et al (1998)

Dignan et al (1998)

96

109

108

107

To present a National Cancer Institute funded health education programme conducted among the Lumbee tribe in North Carolina that was designed to increase the proportion of women, age 18 and older, who have Pap smears performed to screen for cervical cancer

To test the use of telephone outcalls to extend the reach of the telephone information service to underserved women

To evaluate the impact of a telephone outcall intervention on screening mammography behaviour among lower income, older women

To examine the breast and cervical cancer screening practices of Hispanic women in 50 US states and Puerto Rico from 1998 until 1999 using data from the BRFSS Neither intervention had a significant effect on the main outcome. The intervention promoted minimal movement in the stages of change for mammography

These results highlight the continuing need for Hispanic women who are medically underserved to have access to cancer screening services

The health education programme was provided on a one on one basis in women’s homes by a trained lay health educator and included verbal, print and videotape information

The study concluded that the health education programme was associated with greater knowledge about cervical cancer prevention and higher proportions of Lumbee women having Pap smears performed

Neighbourhoods throughout Colorado were This approach successfully extended the Cancer identified using a geodemographic database Information Servce’s audience; however, its labour (INFORUM) that allowed selection of census intensity may limit its applicability block groups according to demographic characteristics. Households were assigned randomly to (1) a control group; (2) an outcallonly group, which received ‘cold’ telephone outcalls promoting screening mammography; or (3) an advance card plus outcall group, which received a card introducing the programme prior to the outcall

A geodemographic database, INFORUM, was used to identify low income and minority neighbourhoods. Households were assigned randomly to (1) a control group; (2) an outcallonly group, which received ‘cold’ telephone outcall promoting screening mammography; or (3) an advance card plus outcall group, which received a card introducing the programme prior to the outcall

Secondary data analysis from BRFSS database

Earp et al (2002)

Fernandez et al To assess the effectiveness of (1999) ENCORE plus in promoting mammography and Pap test screening among women who were non-adherent to screening guidelines

129

117

Thirty churches were randomised for telephone counselling and control conditions; telephone interview data were used in assessing intervention effects on mammography adherence. Separate analyses were conducted for baseline adherent participants (maintaining adherence) and baseline non-adherent participants (conversion to adherence)

To evaluate the effectiveness of the North Carolina Breast Cancer Screening programme among rural African-American women age 50 years and older using a community trial method

Information about lay health adviser activities serves several important purposes. Such information allows programmes to identify the types of messages that lay health advisers transmit and the number of contacts they make, while also identifying the groups that are more and less difficult to reach, and the topics and locations favoured by advisers and the women they contact. Activity data may indicate what resources or other support the advisers need, whether in-service training is necessary and how to enhance the recruitment and training of additional lay health advisers

Partnerships between the public health and religious communities are potentially effective conduits to promote maintenance of widely endorsed health behaviours such as regular cancer screening

Baseline data were collected from women participating in the programme. Follow-up occurred within six months of baseline to assess whether or not enrolees underwent recommended screenings. Screening completion rates were compared with rates from other published intervention studies

This study demonstrates that programmes implemented through community based organisations can be successful in increasing mammography and Pap test screening among low income and minority women

A stratified random sample of 801 AfricanA lay health adviser intervention appears to be American women completed baseline (1993–1994) effective in improving access to screening among and follow-up (1996–1997) surveys. The primary low income, rural populations outcome was self-reported mammography use in the previous two years

To ascertain the levels and types Health advisers were asked to complete, on a of activity that can reasonably be periodic basis, a standardised, self-administered expected from advisers activity report that asked about group presentations in the past three months and one on one contacts in the past week

Earp and Flax (1999)

128

To assess the effectiveness of telephone counselling in a church based mammography promotion, intervention trial

Duan et al (2000)

110

Foxall et al (2001)

Friedman et al (1999)

Friedman et al (1999)

Giuliano et al (1998)

Hiatt et al (2001)

95

140

87

84

119

A random sample of 314 women were interviewed

A total of 121 English speaking adults participated in a questionnaire survey

Eligible participants were 171 adult Ashkenazi Jews. A logistic regression showed that physician recommendation was related significantly to obtaining digital rectal examinations. Physician recommendation and education were related significantly to obtaining faecal occult blood tests

A total of 233 women were surveyed: 138 (59%) Caucasian, 37 (17%) African-American, 29 (12%) Hispanic and 29 (12%) American Indian (mean = 46.86 years)

To conduct breast and cervical 1. Community based lay health worker outreach cancer intervention studies using the controlled trial method in the San Francisco Bay area from 1993 to 1996

To determine rate of and predictors of participation in breast cancer screening services among American Indian women living on the Hopi reservation in Arizona, south-west USA

To investigate compliance with breast cancer screening guidelines among 121 ethnically diverse low income women recruited from a general hospital psychiatry clinic

To examine the psychological and behavioural factors related to annual colorectal cancer screening in a sample of Ashkenazi Jews

To examine ethnic influences on body awareness, trait anxiety, perceived risk and breast and gynaecological cancer screening practices

The strongest predictors of screening behaviour were having private health insurance and frequent use of medical services. Having a regular clinic and speaking English were also important. Race, ethnicity, education, household income and employment status were, overall, not significant predictors of screening behaviour. These baseline results support the importance of cancer screening interventions targeted at persons of foreign origin, particularly those who are less accultured

This study confirms the importance of breast screening knowledge and participation in routine medical services to rates of breast cancer screening

The results highlight the importance of physician recommendation of compliance with screening guidelines for early detection of breast cancer. Because psychiatric patients are typically treated by psychiatrists on a regular and frequent basis, psychiatrists are in a unique position not only to encourage cancer screening but also to monitor patient compliance with their recommendations

Physicians play a key role in motivating people to comply with colorectal cancer screening. Physicians need to encourage all asymptomatic patients aged 50 years and older to be screened for colorectal cancer

Hispanic and American Indian women reported a higher frequency of breast self-examination frequency than Caucasian and African-American women. Caucasian and African-American women reported more mammogram use than did Hispanic and American Indian women. The influence of psychosocial variables on these practices varied with different groups

Howe et al (1998)

Husaini et al (2001)

Ishida et al (2001)

105

112

101

To explore beliefs about and attitudes of Samoan women towards early detection of breast cancer and mammography utilisation

To examine the predictors of breast cancer screening participation in a panel study of African-American women over age 40

To examine the relationship of demographic, socioeconomic and acculturation factors to diagnostic stage prior to preventative interventions in a large Hispanic community (San Diego, California, 1988–1991)

An exploratory, qualitative design was used with a semistructured, open ended, talk story interview with Samoan women aged 40 years and older who spoke English and who had no personal history of breast cancer

A total of 364 participants were interviewed at recruitment and three months later

Age adjusted incidence rates were estimated for three subareas. A Hispanic community was targeted with interventions and a comparison made between the targeted Hispanic community and the remainder of county

3. Patient navigators for follow-up of abnormal screening results. Study design and a description of the interventions are reported along with baseline results of a household survey conducted in four languages among 1599 women aged 40–75

2. Clinic based provider training and reminder system

The study participants strongly believed that cancer meant death and no cure is possible. Samoan women place a firm emphasis on family health and education and their own role in family healthcare. Thus, healthcare providers should promote health and breast screening by emphasising that family healthcare also means caring for oneself. Although Samoan women reported that they used Western healthcare, most women were also willing to use traditional Samoan medicine if it was available

It was found that age, marriage and educational intervention, talking with friends and believing that early detection can lead to cure had a positive impact on attending for mammography between T1 and T2. In contrast, depression significantly reduces the odds of attending for mammography. A family history of breast cancer and church attendance have no effect on mammography uptake once other factors had been excluded

The results showed that efforts aimed at early detection of cervical cancer among Hispanic women are needed and well targeted. NonHispanic, white women in low education communities are also at risk. The population based methods used in this study could help focus efforts in communities needing preventative interventions

Jackson et al (2000)

Jenkins et al (1999)

JenningsDozier (1999)

106

118

99

To determine whether AfricanAmerican and Latina women were up to date with their annual Pap tests and whether behavioural, normative and control beliefs differed among women who were up to date compared with women in community based agencies located within the Philadelphia metropolitan area

To determine the effect of a media led education campaign on breast and cervical cancer screening among VietnameseAmerican women

To carry out an ethnographic study and survey on cervical cancer screening among Cambodian-American women

Questionnaire survey of a sample of 204 AfricanAmerican and Latina women

Over a 24 month period, a media led community education campaign was conducted to promote recognition, intention, uptake and currency of routine check-ups and clinical breast examinations was implemented. The uptake of mammograms and Pap tests was measured among VietnameseAmerican women in Alameda and Santa Clara Counties in northern California. Women in Los Angeles and Orange County in southern California served as controls

Unstructured interviews were conducted with 42 Cambodian-American women in target neighbourhoods. A multifaceted intervention programme was delivered by bicultural outreach workers including home visit and group meetings and logistic assistance

2. Differences within and between ethnic groups justify the need for culturally specific interventions and patient education geared towards promoting regular Pap test screening

1. African-American women whose smears were up to date were less concerned about receiving information or reminders regarding annual screening than those who were not up to date. Latina women also were more likely to report control beliefs concerning the need for family assistance, to have access to screening, to receive reminders and to have a female doctor than those who were not up to date.

A media-led education intervention succeeded in increasing recognition of and intention to undertake screening tests more than uptake of or currency with the tests. Copyright 1999 American Health Foundation and Academic Press

The results reinforce the importance of considering health problems within the context of a population’s traditional belief systems and daily routines

Jirojwong and Manderson (2001)

Kim et al (1998)

Kim et al (1999)

102

139

79

To examine cervical cancer screening knowledge and practices among KoreanAmerican women

To examine knowledge and practices of colorectal cancer screening among KoreanAmericans

To describe women’s beliefs and perceptions about the body, breast and cervical cancer and their perceptions of the causes and of the diseases

Jibaja-Weiss et To carry out a randomised al (2003) controlled trial to examine whether a personalised form letter containing generic cancer information and a personalised tailored letter containing minimally tailored individualised risks factor information based on medical records data affected breast and cervical cancer screening uptake

120

The 1987 Cancer Control Supplement questionnaire was translated into Korean and used to collect data

The National Health Interview Survey (NHIS) Cancer Control Supplement Questionnaire, prepared by the National Center for Health Statistics, was used to collect data regarding colorectal cancer screening knowledge and practices

In 1998, a cross-sectional study was conducted in Brisbane, Australia, among 145 Thai immigrant women, to explore cultural and social factors related to their use of Pap smear tests and breast self-examination

A total of 1574 African-American, MexicanAmerican and non-Hispanic white women, 18–64 years old, registered in two health centres were targeted

The findings from this study have important implications for health practitioners and policy makers who serve this ethnic population

2. The findings indicate that a majority of respondents were unaware of cancer screening examinations and did not attend these tests because they underestimated their importance

1. Gender, education, knowledge of the seven cancer warning signals and length of residence in the USA were significantly related to having heard of digital rectal examination

Some women believed that they would develop cervical cancer if their perineum or vulva was ‘dirty’. Some believed that they would have breast cancer because they had a history of benign tumour or cyst of a breast

Individualised tailored letters that contain cancer risk factor information may decrease the likelihood of attending cancer screening among medically underserved low income and minority women, but personalised form letters may improve uptakes among disadvantaged groups

Laws and Mayo (1998)

Lindau et al (2002)

Margolis et al (1998)

Maxwell et al (2000)

85

82

130

94

To assess mammography screening rates, related attitudes and intervention preferences in Filipino-American women, a group that has been neglected in cancer control research

To determine if women would have higher breast and cervical cancer screening rates if lay health advisers recommended screening and offered a convenient screening opportunity

To describe the relationship between health literacy, ethnicity and cervical cancer screening practices and to evaluate physician recognition of low literacy

To identify factors associated with screening mammography utilisation among Latinas aged 40 and older in two Massachusetts cities, who were predominantly of Caribbean origin

Face to face interviews were conducted in English and Tagalog with a convenience sample of 218 Filipino women aged 50 years and older residing in Los Angeles. Sixty-six per cent had ever had a screening mammogram, 42% had had one in the past 12 months and 54% in the past two years. These rates are about 20% lower than those found among African-American and white women in the 1994 California Behavioral Risk Factor Survey

A controlled trial among a sample from Urban County appointments in several non-primary care outpatient clinics. Lay health advisers assessed the participants’ breast and cervical cancer screening status and offered women in the intervention group who were due for screening an appointment with a female nurse practitioner

A prospective cohort study of English-speaking patients ≥ 18 years (n = 529) in ambulatory women’s clinics. Univariate and multivariate analyses were used to evaluate demographics, health practices and beliefs, and knowledge regarding cervical cancer screening and prevention. Physicians’ assessments of patients’ reading skills were obtained

A cross-sectional survey based on an area probability sample, using bilingual Latina interviewers. Sixty-one per cent of the sample were Latinas, 99% of whom were interviewed in Spanish

The following variables were negatively related to the outcome: concern over cost, the attitude that mammograms are needed only in the presence of symptoms, the perceived inconvenience of the time needed and difficulties in getting to the mammography facility and embarrassment. Implications for interventions to increase breast cancer screening are discussed

Breast and cervical cancer screening rates were improved in women attending non-primary care outpatient clinics by using lay health advisers and a nurse practitioner to perform screening. The effect was strongest among women in greatest need of screening

Poor health literacy was a better predictor of cervical cancer screening knowledge than ethnicity or education, yet physicians infrequently recognised low literacy. Improved physician awareness and development of low literacy interventions may improve cervical cancer screening, particularly for the most vulnerable women

Ethnicity, educational attainment and acculturation were unrelated to recent screening. While distinct healing traditions were found among some Latinas, their use was unrelated to screening. Higher fatalism scores had a bivariate association with lower probability of screening. The results suggest that universal medical insurance and efforts to influence physician practices are keys to promoting screening mammography among this population

Maxwell et al (2001)

McCaffery et al (2003)

McFall and Hamm (2003)

Myers et al (1999)

90

144

138

136

The study examined data from the 1996/97 National Population Health Survey to determine current patterns in mammography in Canada and factors associated with non-participation and time inappropriate use among women aged 50–69

To identify factors that predict adherence by African-American men to prostate cancer education and early detection

To carry out a focus group study to obtain reactions to numerical information about events and outcomes related to prostate cancer screening The authors identified 548 African-American men who were patients at the University Health Services of the University of Chicago, were aged 40–70 years and who did not have a personal history of prostate cancer. Baseline telephone survey data were collected from 413 men (75%). Participants were randomly assigned to either a minimal or an enhanced intervention group. Men in former group were sent a letter and reminder that invited them to a urology clinic for prostate cancer education and early detection. Men in the enhanced intervention group were sent the same correspondence and were also given printed material and telephone contacts, which were tailored to each recipient

The results of the focus group were used to develop health education material, ie a balance sheet to help people to make decision

To examine knowledge of An interview based survey in the UK of 1637 colorectal cancer and the adults aged between 16 and 74 years relationship between knowledge and attitudes to cancer, and intentions to engage in early detection behaviours for colorectal cancer in a national representative population sample

To examine the reasons for underuse of mammography among select groups of Canadian women

A tailored behavioural intervention can influence adherence to prostate cancer early detection among African-American men. Individual backgrounds and cognitive and psychosocial characteristics may also affect behaviour. Future studies should assess the impact of this type of intervention on cognitive and psychological correlates of decision making and behaviour along the continuum of prostate cancer care

The view of screening was less positive than initial opinions based on the media. Participants suggested shorter screening intervals, regular screening and different treatment methods. No single topic was sufficient to aid screening decisions

Increasing knowledge may reduce negative public perceptions of cancer, which may impact positively on intentions to participate in screening

Significant predictors of never having had a mammogram included higher age, residence in a rural area, Asia as place of birth, no involvement in volunteer groups, no regular physician or recent medical consultations, current smoking, infrequent physical activity and no hormone replacement therapy

Navarro et al (1998)

Nguyen et al (2001)

O’Malley et al (2001)

132

113

89

To investigate the association between physician recommendation for mammography and race/ ethnicity, socioeconomic status and other characteristics in a rural population

To carry out a promotional intervention in early detection of breast cancer among Vietnamese-African women

To describe the short-term impact of the intervention known as Por La Vida (PLV) on cancer screening for Latinas in San Diego, California

A total of 1933 black and white women aged 52 years and older in 10 rural counties were surveyed

Over a 2.5 year period, the following intervention activities were targeted at both Vietnamese women and physicians in Alameda County, California: neighbourhood based educational activities; dissemination of health education materials; a media campaign; and continuing medical education seminars for physicians. Women in Los Angeles and Orange County served as controls. Pre-intervention telephone interviews were conducted with 384 randomly selected Vietnamese women in the intervention community and 404 women in the control community in 1996, and post-test intervention interviews were conducted with 405 and 402 women, respectively, in 1998

Thirty-six lay community workers (consejeras) were recruited and trained to conduct educational group sessions. Each consereja recruited approximately 14 peers from the community to participate in the programme. The consejeras were randomly assigned to either a 12 week cancer screening intervention group or a control group, in which they participated in an equally engaging activity entitled ‘Community Living Skills’. Preand post-intervention self-report information was obtained from project participants regarding the use of cancer screening examinations. Outcome measures were changes in the percentages of women who had breast and cervical cancer screening test within the past year before and after the intervention was implemented

Socioeconomic status, age and other characteristics, but not race/ethnicity, were related to reports of physician recommendation, a precursor strongly associated with mammography uptake

Although the effect on the women in the intervention group was not significant, the intervention did have a modest positive impact on women who had more exposure to it

The key to the PLV intervention model is identification of natural helpers in the Latina community and their subsequent training in intervention based on social learning theory using culturally appropriate education materials. The model is an effective and viable approach for increasing the use of cancer screening tests in Latinas of low socioeconomic level and low level of acculturation

O’Malley et al (2002)

Paskett et al (1998)

Phillips et al (1999)

Risendal et al (1999)

Risendal et al (1999)

97

116

98

88

80

To identify predictors of Pap smear use

To identify predictors of recent clinical breast examinations and mammograms among those women aged 40 years and older

To explore the beliefs and practices related to breast cancer and breast cancer screening among low and middle income African-American women

To develop and implement strategies to improve the use of cervical and breast cancer screening among AfricanAmerican women aged 40 and older who resided in low income housing communities

To examine the effects of primary care, health insurance and HMO participation on adherence to regular breast, cervical and colorectal cancer screening These results suggest that clinic based interventions can improve the use of breast and cervical cancer screening and follow-up among low income women

Organising healthcare services to promote continuity with a specific primary care clinician, ensuring that a comprehensive array of services is available at the primary care delivery site, coordinating providers and improving patient– practitioner relationships are likely to improve participation

A random household cross-sectional survey was conducted in Phoenix, Arizona, to assess cervical cancer screening rates among 519 adult urban American Indian women

A random household cross-sectional survey of 519 adult American Indian women in Phoenix, Arizona, was conducted

The results of this study indicate that limited access to healthcare and lack of knowledge about the procedure were important barriers to Pap smear use. Improving cervical cancer screening participation rates is an important step in reducing the disease burden in this population

This study confirms the limited access of urban American Indians to preventative health services, and supports a role for cancer education in improving screening participation in this special population.

A qualitative study using focus group methodology Fear was found to be a primary reason for not engaging in breast cancer screening. Despite initiatives promoting breast cancer awareness, African-American women still hold misconceptions regarding the aetiology of breast cancer and fatalistic perspectives regarding breast cancer outcomes. Future research should address the influence of these factors so as to increase screening behaviours

Community health centre based strategies included educational interventions for providers and patients, follow-up inventions for abnormal screening tests and the implementation of a computer tracing system

Included in the survey were 1205 women over age 40, 82% of whom were African-American. Adherence was defined as reported receipt of the last two screening tests within the recommended intervals for age

Royak-Schaler et al (2002)

Sadler et al (2001)

Schulmeister and Lifsey (1999)

91

133

100

To describe the knowledge, beliefs and practices of cervical cancer screening of Vietnamese women who have migrated to the USA

To assess the degree to which ethnic subgroups are assimilating and acting on the proffered information

To investigate the relationship between breast cancer risk communication delivered by providers and patient knowledge, perceptions and screening practices

Data were collected by a bilingual nurse during face to face interviews conducted in the language preferred by the subjects (English or Vietnamese)

This paper describes the reported breast cancer knowledge, attitudes and screening behaviours of 194 Asian-Indian-American women

Telephone interviews were conducted with 141 African-American (n = 71) and white (n = 70) first degree relatives of breast cancer patients who received medical services at two university medical centres in North Carolina during 1994–95

The number of Vietnamese women who adhere to cervical cancer screening guidelines is low. Cultural beliefs and structural barriers influence the choices that Vietnamese women make

The majority of these reported that breast cancer knowledge is inadequate. They were willing to be called upon to share with others any knowledge they gained. There is a clear opportunity for public health nurses to provide Asian-Indian women with a more comprehensive understanding of breast health and disease. They can then share their knowledge with other women within their ethnic group

The study results indicate that provider discussions about family history and personal risk, accompanied by increases in risk perception, promote patient compliance with screening goals. The findings suggest that accurate knowledge about specific breast cancer risk factors may not be necessary to achieve screening compliance. However, additional studies are needed to investigate the relationship between knowledge of breast cancer risk factors and the adoption of behaviours associated with reducing breast cancer risk

Tang et al (2001)

Taylor et al (1999)

Valdez et al (2002)

Wismer et al (1998)

142

81

114

86

To collect baseline information on breast and cervical screening behaviour among KoreanAmerican women

To carry out a field test of a multimedia health education intervention designed to provide breast cancer education for low income Latinas

To assess factors predisposing to and reinforcing cervical screening participation using the PRECEDE model

To examine the factors associated with faecal occult blood test (FOBT) and sigmoidoscopy screening among Chinese-American women aged 60 years and older

A population based telephone survey of 1090 Korean-Americans in two California counties

Participants were randomly assigned to a control or intervention condition

Cambodian residents of Seattle were surveyed in person during late 1997 and early 1998

One hundred women were recruited from senior centres in two metropolitan areas on the East coast of the USA. Participants completed a questionnaire that included sections on demographics, health history, health insurance coverage, FOBT and sigmoidoscopy use, common and cultural barriers to colorectal cancer screening and acculturation

Planning and implementation of tailored programmes to improve screening are best undertaken using a community sensitive approach

Women in the intervention group exhibited a higher knowledge score than those in the pretest group (chi-squared test, P < 0.0001). Never screened women exhibited the largest differences in knowledge scores relative to the baseline sample. The intervention also increased the likelihood of women seeking information about mammography

Six variables independently predicted recent screening: age, beliefs about regular check-ups, cervical screening for sexually inactive women, the prolongation of life, having a female doctor and a previous physician recommendation for Pap testing. The study findings indicate that culturally specific approaches might be effective in modifying the cervical cancer screening behaviour of immigrant women

Logistic regression models found greater acculturation to be a significant predictor of having had a FOBT at least once, and found that both greater acculturation and physician recommendation to be significant predictors of having had a sigmoidoscopy at least once. This study found that older Chinese-American women underuse FOBT and sigmoidoscopy screening, as is recommended by the American Cancer Society colorectal cancer screening guidelines. These findings suggest that cultural factors may influence the initiation of colorectal cancer screening for Chinese-American women but are not predictive of adherence to screening over time

93

Zambrana et al To compare the use of three (1999) cancer screening practices (Pap smear, mammography and clinical breast examination) among five subgroups of Hispanic women and examine whether sociodemographic status, access, health behaviour, perception and knowledge and acculturation factors predict screening practices

Secondary data analysis from National Health Interview Surveys

Access and prior screening are more strongly associated with current screening than language and ethnic factors