Evaluating meta-ethnography: systematic analysis and synthesis of qualitative research

Health technology assessment

By the late 1990s, with the systematic review firmly established as a scientific method for bringing together findings from quantitative studies of effectiveness, interest was beginning to shift towards how qualitative data could be brought into the evidence base and a ‘science of synthesis’ developed for qualitative research. Whereas bringing together research information on effectiveness, costs and acceptability has always been fundamental to the health technology assessment (HTA) enterprise, the use of qualitative research methods within HTA has been a more recent development. In the mid-1990s, the HTA Methodology programme commissioned a review of the literature on the use of qualitative research methods in HTA, which noted that application of conventional systematic review methodology to qualitative research presented both philosophical and practical challenges. 9 Following publication of this report, the HTA recognised that further work was required to enable ‘users of qualitative research to be able to both appraise and to synthesise qualitative studies in a rigorous, replicable and formalised way’. 10

Aim

To appraise and synthesise qualitative health research for HTA using a meta-ethnographic approach.

Objectives

Content of this report

This report continues in Chapter 2 with a review of the methods of qualitative synthesis. We did not complete a review of methods of appraisal as two other projects commissioned by other bodies at a similar time had such a review within their remit. 19,20 To avoid unnecessary duplication, we focused our efforts on reviewing methods for the synthesis of qualitative research. Chapter 3 describes the literature searching strategies employed in the two meta-ethnographic syntheses undertaken. An evaluation of the methods of appraisal used in the syntheses is reported in Chapter 4. The precise methods used to conduct the syntheses are described in Chapter 5 together with an assessment of the reproducibility of the meta-ethnographic method. The two syntheses are presented in Chapters 6 and 7. Chapter 8 contains a discussion of all the issues raised by this evaluation of meta-ethnography, and a number of conclusions are drawn about its value as a method of qualitative synthesis and its potential role in HTA.

Aggregative synthesis

It seems paradoxical to begin this review of methods for synthesising qualitative research with numeric approaches. Nevertheless, a variety of methods have been proposed for converting data from qualitative research studies into a quantitative form for the purposes of analysis and synthesis. An early example is Yin’s25 case survey method, which was developed for the synthesis of case studies based on a qualitative or mixed-method approach. This method involves the initial application of a closed-ended coding instrument to each qualitative case study to abstract and record relevant data. The collective data from these codes are then tallied and analysed in much the same way as traditional survey data. Examples are analyses by Yin25 and colleagues of citizen participation in urban services26 and of innovations in urban services. 27,28 This approach to the synthesis of qualitative data mimics quantitative meta-analysis in which findings (usually from individual RCTs) are aggregated in order to have sufficient statistical power to detect a cause and effect relationship between a particular treatment and specific health outcomes. 28

Current interest in combining qualitative and quantitative syntheses has led to other types of numeric methods, with a key example being the Bayesian synthesis of qualitative and quantitative evidence. This involves pooling of the data and is illustrated by Roberts et al. ’s29 investigation of childhood immunisation and the probable effects of a variety of factors on uptake. Their approach involved the initial extraction of factors influencing the uptake of childhood immunisation identified from qualitative studies. These factors were then classified in terms of broader descriptive categories (e.g. ‘lay beliefs’ was used to capture such concepts as ‘parents’ beliefs’ and ‘opinions of parents and other family members about the value of immunisation’). Rankings of the relative importance of each factor by each reviewer yielded a probability of that factor being important in determining uptake. These factors were then combined with evidence from the quantitative data to form a posterior probability that each factor identified (e.g. lay beliefs) was important in determining uptake of immunisation. This synthesis, therefore, provided a more complete enumeration of factors associated with the uptake of immunisation through combining quantitative and qualitative studies. Bayes factor methods were then employed to compare two possible meta-regression models in explaining the log-odds of immunisation uptake for each possible factor. This approach forms an important attempt to combine qualitative and quantitative evidence using numeric methods and pooling data, but like other numeric methods is mainly concerned with what Hammersley14 identified as aggregative purposes in terms of the accumulation and generalisation of evidence.

Other approaches to the systematic analysis/synthesis of studies derive from Miles and Huberman’s30 cross-case techniques, which involve developing summary tables based on content analysis and noting commonalities and differences between studies.

Narrative synthesis

The aims of narrative synthesis are generally to achieve an aggregation of findings or check the comparability and replication of findings, based on narrative rather than numeric methods. A narrative synthesis may also lead to new insights and form the basis for further interpretive analysis and conceptual development, although this is not the primary focus.

There are two broad approaches to qualitative narrative synthesis. One is based on the traditional literature review and involves the use of informal but critical and reflective methods. This approach, lately described as ‘narrative review’,20 unusually aims to provide a commentary and summary of the main features and findings of a body of literature using methods that are not explicitly pre-defined and transparent. This method has been distinguished from ‘narrative synthesis’, which employs more transparent and systematic methods to identify studies, assess study quality and synthesise the findings. It usually involves a form of thematic analysis to identify the main, recurrent or most important themes in the literature, thus allowing the findings of studies to be summarised and grouped. Examples of narrative syntheses include McNaughton’s31 review of research reports describing the home-visiting practice of public-health nurses to identify common elements and differences between studies, and Finfgeld’s32 review of studies to better clarify how individuals overcome drug and alcohol problems without participating in 12-step-type self-help groups or formal treatment.

Narrative synthesis is an approach that has evolved largely from within the quantitative systematic reviewing tradition. Indeed, the Centre for Reviews and Dissemination at York University33 now encourages reviewers to first undertake a ‘descriptive or non-quantitative synthesis’ of studies to be included in a review, in other words a narrative synthesis, to help them think about what the ultimate method of synthesis should be. In order to further assist this development, a report has recently been compiled to provide detailed guidance on how to undertake narrative synthesis in systematic reviews. 34

Interpretive synthesis

This is generally regarded as ‘the’ approach to the synthesis of qualitative research, and has as its specific aim the achievement of the developmental goal of qualitative synthesis in terms of producing interpretations that go beyond individual studies and thus contribute to conceptual and theoretical development in the field14 and accords with Strike and Posner’s11 notion of qualitative synthesis as achieving ‘conceptual innovation’ or the ‘invention or employment of concepts not found in the characterization of the parts as a means of creating the whole’ (p. 346). The output is, therefore, a new interpretation or theory that goes beyond the findings of any individual study.

A number of approaches to the conduct of interpretive synthesis have been developed. These include meta-ethnography, meta-study, critical interpretive synthesis, realist synthesis and grounded formal theory. Each of these approaches is examined in detail in Approaches to interpretive synthesis, but a general feature is that the synthesis builds interpretation from original studies by firstly identifying interpretations offered by the original researchers (second-order constructs) and, secondly, enabling the development of new interpretations (third-order constructs) that go beyond those offered in individual primary studies.

Secondary data analysis

This approach is concerned with the analysis and interpretation of primary (i.e. ‘raw’) data rather than the findings of published studies. Secondary data analysis may be regarded as a form of interpretive synthesis if it fulfils two requirements: it involves multiple qualitative data sets rather than a single data set; and the aim is to provide a new perspective or conceptual focus rather than to undertake an additional in-depth or subset analysis of existing data.

There are currently only a few examples of secondary analysis of multiple qualitative data sets. These have mainly involved the re-analysis by researchers of their own data. An example is Bloor and MacIntosh’s35 study of techniques of client resistance to surveillance based on two earlier studies they conducted. Interest in the secondary analysis of qualitative data is growing, and the capacity to undertake such work is considerably enhanced by the establishment by the Economic and Social Research Council’s (ESRC’s) UK Data Archive of an archive of data from qualitative studies.

Secondary analysis of qualitative data raises a number of ethical and methodological issues, including issues of consent in relation to the reuse of another researcher’s data for a different purpose,36,37 and methodological questions such as the effects of distance from the data collection and of possible variations in study design on interpretation. 38 It is likely that the greater availability of qualitative data sets and increasing interest in secondary analysis will lead to new protocols and strategies to address these issues, thus, enhancing this approach to the synthesis of qualitative research.

Meta-ethnography

Noblit and Hare’s15 method of meta-ethnography was published in 1988 and is described as ‘an attempt to develop an inductive and interpretive form of knowledge synthesis’. Noblit and Hare15 developed meta-ethnography in response to the perceived failure of a synthesis of five ethnographic studies of educational desegregation that were undertaken to convey information to policy-makers. This educational synthesis took an aggregative, thematic approach that involved abstracting data and isolating factors in each study that appeared to be responsible for the failure of schools to desegregate. This process of abstraction de-emphasised the uniqueness of each site. The context therefore merely became a confounding variable in the search for common findings rather than contributing to an explanation of these findings. As a result, the synthesis did not provide researchers or policy-makers with an understanding of what went wrong and what could be done about it. Noblit and Hare15 aimed to overcome these limitations through developing a distinct method for the synthesis of qualitative studies that was informed by Turner’s39 theory of social explanation and is interpretive rather than aggregative. As they stated:

The nature of interpretive explanation is such that we need to construct an alternative to the aggregative theory of synthesis entailed in integrative research reviews and meta-analysis and be explicit about it.

Noblit and Hare15 (p. 18)

This aim of constructing adequate interpretive explanations required developing a way of ‘reducing’ and deriving understanding from multiple cases, accounts, narratives or studies while retaining the sense of the account. Noblit and Hare15 were themselves ethnographers who were concerned with long-term intensive studies that employed observation, interviews and documents, and termed the approach that they developed ‘meta-ethnography’. However, they described meta-ethnography as being applicable to qualitative research generally and as forming ‘a rigorous procedure for deriving substantive interpretations about any set of ethnographic or interpretive studies’ (p. 9). Noblit and Hare15 also noted that their particular approach was ‘a’ meta-ethnography and that it formed ‘but one of many possible approaches’ (p. 25).

Noblit and Hare15 identified seven phases in undertaking meta-ethnography (Box 2), but observed that in practice these phases may occur in parallel and overlap. The phases broadly correspond with other methods of synthesis, but differ in the assumptions and procedures involved. One difference is that the sample for research is purposively selected in relation to the topic of interest (and may involve maximum variation sampling), rather than being exhaustive. This reflects the general approach of qualitative methods and the aim of achieving interpretive explanation. A second difference is that the interpretations and explanations contained in the original studies are treated as data through the selection and analysis of key ‘metaphors’ (i.e. the themes, perspectives or concepts revealed by qualitative studies), with the aim of reducing accounts while preserving the sense of the account. Preparation for comparison between studies requires listing and juxtaposing the key metaphors, phases, ideas and/or concepts used in each account but retaining, as far as possible, the terminology used by the authors to remain faithful to the original meanings (phase 4). A third difference is that comparison between studies involves processes of ‘translation’, with the metaphors/concepts and their interrelationships in one account being compared with those in another account. This process of translation is idiomatic and focuses on translating the meaning of the text rather than a literal translation, with the aim of preserving original meanings and contextualisation. Noblit and Hare15 identified three possible types of relationship that guide translation and subsequent synthesis:

1. Reciprocal: when studies are about similar things, they can be synthesised as direct translations (i.e. in an iterative fashion each study is translated into the metaphors of the others – see Chapter 5 for a detailed description and illustration of this process). ‘These reciprocal translations may reveal that the metaphors of one study are better than those of others in representing both studies, or that some other set of metaphors not drawn from these studies seems reasonable.’ Noblit and Hare15 pointed out that the ‘uniqueness of studies may not make it possible for a single set of metaphors to adequately express the studies’, in which case more is often learned from the process of translation than from the metaphors alone.

2. Refutational: this is undertaken when studies refute each other. It requires a more elaborate set of translations ‘involving’ translations of both the ethnographic account and the refutations to examine the implied relationship between competing explanations.

3. Lines-of-argument: many studies suggest a lines-of-argument or inference about some larger issue or phenomenon. This involves first translating studies into each other and then constructing an interpretation (‘lines-of-argument’) that may serve to reveal what was hidden in individual studies (discovering a ‘whole’ among a set of parts). This is achieved through the use of the comparison and theory generation aspects of grounded theory as described by Glaser and Straus,40 and involves the detailed study of differences and similarities among studies to be synthesised with the aim of producing an integrated scheme and new interpretive context.

How translations are synthesised (phase 6), and the product of this process, depends on how studies relate to each other. Both translation and synthesis involve a continuous comparative analysis of texts until a comprehensive understanding of the phenomena is realised and the synthesis is then complete. Noblit and Hare15 did not specify this process in detail, but regarded it as akin to the general processes of qualitative research. The main difference is that the ‘data’ of meta-ethnography are the substance of several qualitative studies.

The final stage of meta-ethnography is ‘expressing the synthesis’ or communicating with an audience. This was given considerable emphasis by Noblit and Hare,15 who stated that ‘the worth of any synthesis is in its comprehensibility to some audience’ (p. 82). They described the needs of the audience (e.g. researchers, policy-makers) as influencing both the form and substance of the synthesis. Some understanding of the audience’s culture is therefore required to ensure that the translation of studies for the synthesis uses intelligible concepts to inform the final presentation of synthesis. They observed that if the data are inadequate or if the audience cannot see the connection between data and the argument then the study becomes unbelievable. Comprehensibility and believability are thus central to determining worth.

Noblit and Hare’s15 approach to meta-ethnography was based on a literary tradition of interpretivism and was, therefore, driven by a desire to achieve adequate interpretive explanations rather than by technical interests. This means that the authors were critical of the emphasis given by some qualitative researchers to the explicitness of processes employed to analyse data and refer to what Marshall41 described as the ‘bureaucratization of data analysis’. 15 Instead they emphasised the fluidity and interpretive aspects, and described the conduct of meta-ethnography as an ongoing process in which substantive interests may change as the synthesis proceeds. Similarly, translations were regarded as emergent and interactive, with the stages of translation and synthesis often interweaving rather than forming distinct phases.

Noblit and Hare15 illustrated the methods of meta-ethnography using brief excerpts of material from educational studies. Following publication of their monograph, there has been considerable interest in the application of meta-ethnography to a variety of fields including education, public policy and health. The greatest number of recent meta-ethnographic syntheses have come from nursing, reflecting the large number of qualitative studies in nursing and the particular interests and contributions of schools of nursing in the USA and Canada.

The present review involved a search for examples of meta-ethnography and other methods of interpretive synthesis in the health field. This initially involved a computer search undertaken in spring 2003 and updated in the summer of 2006 using MEDLINE, Cumulative Index to Nursing and Allied Health Literature (CINAHL), EMBASE, Web of Science (Social Science Citation Index and Science Citation Index), PsychINFO and Zetoc (British Library’s Electronic Table of Contents) electronic databases. A simple search strategy was used that combined the term ‘qualitative research’ with the terms ‘meta analysis’, ‘meta synthesis’, ‘meta ethnography’, ‘systematic review’, ‘data synthesis’ and ‘meta study’. Hyphenated terms (e.g. meta-ethnography) were also used. Further syntheses were identified via a published textbook of methods of qualitative meta-synthesis42 and later compared with Finfgeld’s43 review. However, the aim was not so much to achieve an exhaustive list as to identify representative syntheses using meta-ethnography and to examine their practical applications.

A total of 41 reports were identified that appeared to be qualitative syntheses. 16,17,44–82 The authors used various terms to describe the methods employed and occasionally used more than one term. We therefore broadly grouped syntheses according to what appeared to be their main approach. Syntheses employing meta-ethnography prior to the beginning of the present project demonstrated the potential value of this approach in producing conceptual developments. This is illustrated by the development of a wellness–illness model,61 a model of caring,74 models of diabetes management17,70 and explanations of states and behaviours such as moral distress among nurses, non-adherence among patients, the experience of post partum depression and help-seeking behaviours (Table 1). The studies included in these syntheses mainly collected data through interviews with only a few including focus groups or observational methods. They also exhibited a limited variety of theoretical perspectives, reflecting the dominant approaches to qualitative research in the field of health and illness.

Altogether we identified six syntheses that were clearly labelled as employing the methods of meta-ethnography. These included two papers on preliminary work published by the present team;16,17 a synthesis of 43 papers concerned with the experience of diabetes;70 a meta-ethnographic synthesis by Smith et al. 45 examining cancer patients’ help-seeking experiences; a synthesis by Feder et al. 44 of health-care professionals’ interactions with women in violent relationships; and another by Walter et al. 46 which was concerned with lay understandings of familial risk in common cancers.

The largest group of qualitative syntheses that we identified were labelled as meta-synthesis. Qualitative meta-synthesis was defined by Sandelowski et al. 13 as ‘the theories, grand narratives, generalizations, or interpretative translations produced from the integration or comparisons of qualitative synthesis’. Meta-synthesis, thus, seems to be, in effect, another term for qualitative synthesis and one that is not identified with any particular method. Many of the syntheses included within this category cited Noblit and Hare15 and claimed to be using meta-ethnographic methods, but what seemed to distinguish this group of studies from those labelling themselves explicitly as meta-ethnographies was that the latter group adhered more closely to the approach described by Noblit and Hare,15 included a smaller number of studies (maximum of 32) and, in general, achieved a higher degree of conceptual development. Meta-study and critical interpretative synthesis, which have their origins in meta-ethnography, are more recent developments and to date have been used only by their creators. The general tendency among all these syntheses was not to formally assess the quality of the research prior to synthesis. Where this did occur it was liable to result in just a couple of exclusions or in none at all. The reports of most syntheses aimed to draw out the implications for policy-makers and practitioners, but there was little evidence of wider policy/practitioner responses to these findings and assessment of worth in these terms.

The next section considers a development of Noblit and Hare’s15 meta-ethnography through its use as one aspect of a wider meta-study.

Meta-study

Paterson et al. 42 described an approach to synthesis that incorporates Noblit and Hare’s15 meta-ethnography as part of a meta-study. 70 Meta-study aims to go beyond meta-ethnography (or other methods of what they term meta-data analysis), as Paterson and colleagues do not regard meta-ethnography as giving sufficient attention to the way that theories and methods shape knowledge and confer meaning on findings. The enlarged framework of meta-study that Paterson and colleagues present derives from earlier sociological work that was concerned about understanding the ways in which theoretical, methodological or societal contexts shape reported results and bodies of knowledge. They draw particularly on the writings of Ritzer99 and of Zhao,100 who introduced the term meta-study and described its components: meta-data analysis (study of the processed data based on meta-ethnography or other methods of data analysis), meta-method (study of the appropriateness and rigour of particular methods used in the research studies and the implications of a range of epistemologically sound approaches for emerging data and interpretations) and meta-theory (analysis of the philosophical, cognitive and theoretical perspectives underlying the research that influences emerging data and interpretations). Paterson et al. 42 described each of these components as providing a unique angle of vision from which to deconstruct and interpret a body of qualitatively derived knowledge about a particular phenomenon, and suggests that meta-synthesis ‘… brings back together those ideas that have been taken apart or deconstructed in these three analytic meta-study processes’ (p. 13). Meta-synthesis, therefore, represents ‘the creation of a new interpretation of a phenomenon that accounts for the data, method and theory by which the phenomenon has been studied by others’ (p. 13) (Figure 1).

FIGURE 1.

Components of meta-study. Adapted from Paterson et al. 42

Paterson et al. 42 outlined the assumptions and methods for conducting a meta-study in their book, Meta-study of qualitative health research,42 and illustrated this in terms of the conduct of their own synthesis of the subjective experience of chronic illness. 70,101 Their initial phase of meta-data analysis was based on the methods of meta-ethnography, but whereas Noblit and Hare15 focused on the in-depth study of a small number of cases, Paterson et al. ’s42 synthesis began with the identification of close to 1000 reports, books, dissertations and research articles published during a 16-year period. These initial publications were reduced to 292 through assessment of their relevance to the review and appraisal of the sampling, data analysis and interpretation that was undertaken using Burns’86 standards for qualitative research. This appraisal involved at least three members of the team reviewing each report and reaching a consensus. Paterson et al. 42 stated that the number of studies was not reduced further as they regarded this number as necessary to achieve all components of the meta-study. The decision to include such a large number of studies required a systematised approach to meta-ethnography (i.e. meta-data analysis component) and was undertaken by the use of coded categories of themes, patterns, processes, etc. This was followed by a computer-based analysis in much the same way as the analysis of the original primary data. The use of computer software also facilitated comparisons between different groups (e.g. whether or not the disease was considered terminal, the method of sampling, gender, etc.). Paterson et al. 42 acknowledged that the handling of large numbers of reports in this way meant that there was an element of context stripping at this level, although their overall aim was to achieve greater contextualisation through finally integrating the different elements of a meta-synthesis (see Figure 1).

Although Paterson et al. 42 provided a clear and fairly detailed description of their approach to meta-data analysis, the stages of meta-method and meta-theory are less fully developed (particularly meta-theory). They also gave little guidance as to how the three components were to be integrated in the final stage of the meta-synthesis. Indeed, in terms of achieving the synthesis, they stated that they ‘… resist definitive procedural steps and encourage instead a dynamic and iterative process of thinking, interpreting, creating, theorising, and reflecting’. 42

Paterson et al. ’s42 own experience of conducting a meta-study has been their work on the subjective experience of chronic illness. 101 They acknowledged that the vast size of this undertaking presented considerable demands of organisation and data management, as well as requiring much time and a large team of researchers. What they termed the meta-data analysis component led to the depiction of the shifting perspective experienced by people with chronic illness from ‘wellness’ in the foreground to ‘illness’ in the foreground and the factors responsible for this shift and ‘bouncing back’. 80 Their attempt to achieve a meta-synthesis based on the three components of meta-study has not, however, led to a major new understanding of chronic illness, although they described the changes that have occurred in the conceptualisation of chronic illness over time. The approach they adopted to meta-synthesis also raises issues of practicability, with considerable time and resources required by the processes involved in first deconstructing and then reconstructing and synthesising the distinct components of meta-study; their own study involved seven researchers who contributed time over several years. Nevertheless, their work is important in its emphasis on the need for syntheses to take account of wider contextual factors, including the influence of theories and methods on the questions addressed and the interpretations developed. This is of particular importance in relation to the syntheses of studies conducted over a long time period and involving a broad area of research, as with Paterson et al. ’s42 own study of chronic illness. Their work also draws attention to the importance of meta-theory in terms of the study of the influence of cognitive, philosophical and theoretical perspectives on the questions addressed and interpretations as an aspect of study in its own right to complement more empirical data-based approaches.

Critical interpretive synthesis

More recently, Dixon-Woods et al. 102 used meta-ethnography to synthesise the diverse quantitative and qualitative literature on access to health care by vulnerable groups. In doing so, they suggest that it was necessary to introduce a number of innovations that were judged to be of sufficient substance to produce what they regard as a new methodological approach that they titled critical interpretive synthesis.

This synthesis was truly innovative in that it used meta-ethnography to synthesise a wide range of different paper types including primary quantitative, primary qualitative, mixed methods, editorial, review and theoretical papers. Books and reports were also included. Dixon-Woods et al. , citing Glaser and Strauss,40 present a compelling argument for there being no reason in principle why interpretive syntheses should not include different forms of evidence. The end result was that six syntheses were produced. The first was a synthesis that set out ‘a general taxonomic framework within which the literature on access to health care can usefully be organised’. 102 The remaining five syntheses comprised interpretive syntheses of access to health care for specific vulnerable groups (socioeconomically disadvantaged people, people of minority ethnicity, children and young people, and older people) and a synthesis examining the effects of gender. These syntheses led to a substantial number of findings and specific recommendations for policy and practice. In addition, concepts of candidacy and the porosity or permeability of services were proposed as useful aids to understanding access to health care. Candidacy was characterised as the processes of negotiation between people and health-care providers by which people’s eligibility for health care is established. Service porosity signalled ease of use and the number of qualifications for candidacy: the more porous a service the more comfortable it is for people to use and the fewer the qualifications required for candidacy. 102

Much of the claimed novelty of critical interpretive synthesis (CIS) rests on the way in which the three different types of syntheses identified by Noblit and Hare15 (reciprocal translation, refutational and lines-of-argument) were interpreted and understood within this particular synthesis. In terms of reciprocal translation, Dixon-Woods et al. 102 reported two problems. Firstly, they found it difficult to operationalise this process with such a large collection of papers. Although not specifically mentioned, given the wide variation in the types of papers included it seems possible that incommensurability was a problem. In other words, it is likely that the very different kinds of papers included had incompatible aims and therefore could not be integrated within the same frameworks. As we indicated after our first worked example of meta-ethnography, a key assumption of the process is that the studies are commensurable. 16 A second issue raised by Dixon-Woods102 et al. was that their experience suggested that reciprocal translation ‘provides primarily a demonstration of the extent to which concepts reported in papers are “translatable” into each other, but offers little in the way of insight … it provides only a summary in terms that have already been used in the literature, and the extent to which demonstrating “reciprocity” between the various terms and concepts is a useful exercise is questionable’. However, the authors also noted that in producing a lines-of-argument synthesis it may be necessary to employ ‘found’ or second-order constructs that are already reported in the literature ‘… to code evidence that was not previously categorised using such a construct’. 102 This use of second-order constructs from one study to code findings reported in another study102 which may not have been explicitly labelled in this way is, however, part of what we would recognise as reciprocal translation. In a more recent paper providing a reflexive account of the process of producing this synthesis, Dixon-Woods et al. 82 reflected further on the difficulties that they experienced with reciprocal translation and suggested that it may not have worked for them because of the ‘large and diverse body of literature’ that they were attempting to synthesise.

A further part of the uniqueness claimed for CIS is that it produces a ‘synthesising argument’ that ‘integrates evidence from across the studies in the review into a coherent theoretical framework comprising a network of constructs and the relationships between them’. 103 It is difficult, however, to see how this differs substantially from Noblit and Hare’s15 eloquent description of a lines-of-argument synthesis as having as its goal ‘to discover a “whole” in a set of parts’ and being akin to ethnographic accounts in that ‘it is emic … in its allegiance to the studies being synthesized; it is historical in that it uses time to give order and history-in-use to give context; it is comparative in that it constructs an analogy of relationships among studies and it is holistic in that it constructs an interpretation of all the studies; their interrelations, and contexts’. 15

The final original aspect of CIS, and the main rationale for use of the term critical, relates to refutation synthesis. In this new approach it is suggested that the synthesiser should ‘consider contradictions in the evidence in the course of producing the synthesising argument, in the same way as deviant or negative cases would be considered as part of the analysis that produces the theoretical output in primary research’. 103 Again this does not seem very different from Noblit and Hare’s15 suggestion that ‘the synthesis must “take into account” the implied relationship between the competing explanations. The implied refutation, then, is analyzed substantively and subsequently incorporated into the synthesis. Our approach treats the refutation itself as part of the interpretation to be synthesised’. 15

In their monograph, Noblit and Hare15 did appear to suggest, through the examples that they provided, that any individual synthesis would take just one of the three forms (i.e. reciprocal translation, refutational or lines-of-argument). In practice what we, and Dixon-Woods et al. ,103 have found is that one meta-ethnography may include elements of these different types of synthesis. The point at which variations on an existing method become sufficient to justify it being labelled as a new method is a matter or judgement.

The next sections describe two other approaches to meta-synthesis: grounded formal theory, that is closely linked with meta-ethnography; and realist synthesis, that describes an integrative theory-building approach developed in the health-policy field.

Grounded formal theory

Grounded formal theory is an approach to meta-synthesis that shares a common framework with Noblit and Hare’s15 meta-ethnography, but is specifically concerned with theory development and gives greater emphasis to the analysis and reporting of verbatim data. Examples of the synthesis of qualitative research using grounded formal theory are Kearney’s104 research aiming to develop a grounded theory of women’s addiction and recovery, and Finfgeld’s105 synthesis of courage among people with long-term threats to well-being that describes a grounded theory approach as forming the ‘epistemological and methodological basis’ for the study.

Kearney104 and Finfgeld105 both included dissertations in the reports that they synthesised because these include more first-hand data. Kearney104 described how she analysed the research reports as if each one was an interview transcript and coded descriptively (verbatim data) and theoretically. She developed a formal theory using a constant comparative approach that involved consulting the verbatim descriptions and the sociocultural and historical context. Following the requirements of grounded theory, once the formal theory was developed other kinds of qualitative reports on addiction and recovery were examined for relevance and to support or challenge the theory (e.g. personal accounts, autobiographies and ethnographies). This led to the development of a theory of truthful self-nurturing that regards the basic problem of addiction as self-destructive self-nurturing, and the basic process of recovery as truthful self-nurturing in which addiction gains meaning as a problem. Recovery work therefore involved three areas of psychological change: abstinence work, self-work and connection work. The analysis is illustrated by substantial reference to original data and the inclusion of verbatim material from the studies in terms of the respondent’s own words. Finfgeld’s105 approach gives a similar emphasis to coding verbatim data from the included reports. This material was then grouped into metaphorical codes and a process model was developed that consisted of the causal and contextual conditions of the phenomenon of becoming and being courageous.

Realist synthesis

This term was introduced by Pawson106 to describe a new methodology developed in the field of evidence-based policy to understand the changes introduced by programmes in terms of their generative mechanisms. Realist synthesis aims to achieve this through a process of theory building rather than assembling empirical generalisations, and draws on material regarding the workings of the same underlying mechanism (such as the public disclosure of information on performance) in different contexts. It therefore differs from other forms of synthesis in bringing together material from diverse sources that may include health, education, social policy or other sectors. In addition, the explicit aim of realist synthesis is theory development. Realist synthesis has so far not been widely employed, but forms one of the increasing armoury of approaches to developing new conceptualisations by drawing together the findings of individual studies and shares with other interpretive approaches the aim of going beyond individual studies to develop new theories and conceptualisations.

Defining the topic

We used the following definition when searching: ‘Papers whose primary focus is patients’ views of medicines prescribed and taken for the treatment of a long- or short-term condition (excluding medicines only taken for preventive purposes)’. The study had to use qualitative methods of data collection and analysis and be published in the English language.

Sampling strategy

We chose the 10-year period from 1 January 1992 to 31 December 2001 (the study began in spring 2002), believing that most of the relevant studies would be found within this period. We were aware that a small number of important early studies would be missed, e.g. Arluke,116 Trostle et al. 117 and Conrad,118 but needed to set limits on what was a potentially large number of available studies.

Electronic search strategies

We used the following electronic databases: MEDLINE, EMBASE, CINAHL, Web of Science, PsycINFO and Zetoc. The initial MEDLINE search was very detailed (Table 2) and produced a large pool of papers. Hoping for a more sensitive and focused search, we tried simply using the terms ‘patient compliance’, ‘adherence’, ‘medicine’, ‘medication’ and ‘qualitative’ (Table 3). The second, simpler, MEDLINE search produced a much smaller pool of potential papers, but all the relevant papers that had been identified in the first search (this simpler search produced no additional relevant papers). We repeated these two searches in EMBASE. Again, the second, simpler, search produced all the relevant papers that had been identified in the first, more detailed EMBASE search. The same strategy in CINAHL resulted in the second, simpler, search producing an additional three relevant papers to the first. Thus, the simple search was just as sensitive as the more comprehensive search in MEDLINE and EMBASE, and more sensitive in CINAHL.

For Zetoc, the search terms used were ‘prescri*’ and ‘qualitative’. For Web of Science, the following search was conducted: (medicine* or medication* or prescription*) and (compliance or adherence or concordance) and (qualitative or grounded same theory or patient* same interview* or patient* same perception* or patient* same experience* or content same analysis or ethnography). The databases searched were SCI-EXPANDED, Social Sciences Citation Index, Arts and Humanities Citation Index. The following search was conducted for PsycINFO: (medicine* or medication* or prescription*) and (compliance or adherence or concordance) and (qualitative or (grounded adj theory) or (patient* adj interview*) or (patient* adj perception*) or (patient* adj experience*) or (content adj analysis) or ethnography) and (PO= HUMAN).

Hand-searching strategies

For the hand-searches we asked colleagues and team members to suggest any papers they might think relevant, we searched the departmental Reference Manager (version 8; Thomson Research Soft, Carlsbad, CA, USA) database (which includes a record of all references held by academic and research staff in the School of Social and Community Medicine at University of Bristol), we checked reference lists of obtained papers and we searched the Medicines Partnership website. We also hand-searched the Arts and Humanities Library and the Medical Library of University of Bristol. All books in the RA–RC (medical) section and in the HM–HV (general sociology) section were searched, as were short-loan collections for the same class marks.

The key journals were hand-searched in case the electronic searches were not sufficiently sensitive or in case the individual articles had not been given adequate keywords. Hand-searching of journals was conducted for either the online or paper versions of journals and was shared between members of the team. The journals we hand-searched were Qualitative Health Research; Sociological Review; Psychology and Health; Journal of Health Psychology; Journal of Advanced Nursing; Medical Anthropology Quarterly; Culture, Medicine and Psychiatry; Medical Anthropology; Patient Education and Counselling; Family Practice; Sociology of Health and Illness; Social Science and Medicine; and Health.

The process of selecting/rejecting

For the electronic searching, PP was responsible for deciding whether or not to include studies. She examined titles and abstracts and obtained full copies of papers that looked promising. For hand-searching, the decision-making process was shared among team members and occurred as we went along. The process of selecting/rejecting was lengthy but was conducted thoroughly at this stage (rather than being initially overinclusive as with the RA synthesis), because of the much larger number of studies involved.

Results of searches

The electronic searches produced 21 studies for appraisal and the hand-searches 21, giving a total of 42 studies to be appraised for inclusion in the synthesis. (Further exclusions could be made following the appraisal.) Figure 2 gives a breakdown of the yield from each of the different databases.

FIGURE 2.

Search yields (medicines synthesis) for the 43 papers. ^aPapers could appear on more than one electronic database. Hnd, hand-searching; Med, MEDLINE; Emb, EMBASE; Cin, CINAHL; WoS, Web of Science; Psy, PsycINFO; Zet, Zetoc.

Reproducibility of electronic search strategy

A second researcher, GDW, conducted the electronic searches independently in order to assess the reproducibility of the selection/rejection process. GDW used the same search strategy with the same databases and examined titles and abstracts before drawing up a list of papers for possible inclusion (Box 3). Of the 21 papers that PP found electronically, GDW found 18. He also identified one additional useful paper that PP had missed, bringing the final number of studies to be appraised to 43 (22 produced electronically and 21 from hand-searching) (Box 3). (GDW also identified 22 other papers which PP had rejected. This may be accounted for by the fact that PP had time to obtain papers that she was not sure about and look at them in more detail, but for the purposes of this exercise, GDW was working with titles and abstracts only.)

Success of search strategy

As indicated above, the double screening of titles and abstracts ensured that one relevant study that might otherwise have been missed was included. Time limitations meant that only one person hand-searched any one journal. Late in the course of conducting the synthesis, we came across a study that could have been included and which should have been identified during the hand-searching. 119 Given that this discovery was made at a point at which the synthesis was well advanced, we did not feel it appropriate to try and incorporate it post hoc.

Defining the topic

Papers were sought that reported the results of qualitative studies concerned with lay experiences of RA, or lay perceptions of aetiology, treatment or management. The study had to use both qualitative methods of data collection and analysis. Papers were not restricted to English language only, because initial scoping exercises suggested that the number of papers produced for the RA synthesis would be manageable.

Sampling strategy

We searched intensively for papers published between January 1992 and December 2001, using the strategies below. However, because we knew that important work had been published before 1992, and because initial searches suggested that the number of potential papers would be manageable, we decided to further include so-called ‘classic’ studies in the field that were published prior to 1992. To decide which ‘classic’ papers to include, we consulted relevant experts in medical sociology, nursing, occupational therapy and rheumatology, including authors of some of the papers identified as a result of our searches below.

Electronic search strategies

We used the following electronic databases: MEDLINE, EMBASE, CINAHL, Web of Science, PsycINFO and Zetoc. The MEDLINE searches are detailed in Tables 4 (the detailed search) and 5 (the simpler one). The two different searches were conducted in order to maximise yield. Although they produced some of the same papers, they also each produced a distinct list of papers that met the inclusion criteria. The same searches were run in EMBASE and CINAHL. For PsycINFO and the Web of Science (Science Citation Index and Social Science Citation Index), the search terms were ‘arthritis and (qualitative or focus group or grounded theory or lay or perception or experience)’ for the period 1992–2001. For Zetoc, six separate simple searches were conducted:

1. arthritis qualitative

2. arthritis grounded theory

3. arthritis focus group

4. arthritis lay

5. arthritis interview

6. arthritis perception.

Hand-searching strategies

For the RA synthesis, the hand-searches were conducted in the same way as for the medicines synthesis, the only differences being that the journal Family Practice was not hand-searched for the RA synthesis, nor was the Medicines Partnership website.

The process of selecting/rejecting

For the electronic searching, GDW was responsible for deciding whether or not to include studies. He examined titles and abstracts and obtained full copies of papers that looked promising. For the hand-searching, decisions were shared among team members and were made as we went along. Because of the manageable number of papers, we erred on the side of inclusivity at this stage.

Results of searches

The electronic searches produced 38 papers for appraisal and the hand-searches four. A further nine ‘classics’ were obtained, giving a total of 51 studies to be appraised. (Further exclusions could be made following the appraisal.) Figure 3 provides a breakdown of the yield from each of the different databases.

FIGURE 3.

Search yields (RA synthesis) for 51 papers. ^aPapers could appear on more than one electronic database. Zet, Zetoc; Cin, CINAHL; Med, MEDLINE; Emb, EMBASE; WoS, Web of Science; Hnd, hand-searching; Psy, PsycINFO.

Reproducibility of electronic search strategy

The electronic searches were conducted independently by a second researcher, RC, in order to assess the reproducibility of the selection/rejection process. RC used the same search strategy with the same databases and examined titles and abstracts before drawing up a list of papers for possible inclusion (Box 4). RC’s searches did not produce any additional relevant papers.

Assessing reproducibility of appraisals

A second reviewer from the research team also appraised each of the papers, blind to the first appraisal, in order to assess the reproducibility of the process. (The results of this process are described in detail below.) The independent reviewers were allocated to each paper in a random manner using an allocation schedule drawn up in advance. This used random number tables and randomisation blocks of nine. If independent reviewers were allocated a paper that they had themselves authored, a different reviewer was chosen randomly from those remaining. The independent reviewers were MM, NB and LY for the medicine-taking synthesis and RC, RP and JD for the RA synthesis.

Inclusions/exclusions

Four of the papers in the medicine-taking synthesis were the subjects of disagreement (between the first and second reviewers) about whether or not to include them. This was because they were considered borderline, either in terms of methodology (were they sufficiently qualitative?) or because the findings were somewhat slight. Disagreements were resolved through a process of negotiation (by e-mail), with the result that one paper was excluded and three were included with the proviso that if they contributed nothing to the synthesis they would ‘fall out’ later on. In the RA synthesis, a team meeting was held to decide the fate of papers over which there was disagreement. If disagreement persisted or if studies were considered borderline according to the inclusion criteria, they were included.

For the medicine-taking synthesis, five papers were excluded following appraisal; for the RA synthesis, 22 papers were excluded following appraisal. The larger number of exclusions in the RA synthesis is largely because, at the literature searching stage, GDW decided to retain any papers that might prove relevant and enter them into the appraisal process. Papers in the medicine-taking synthesis were more heavily screened for relevance by PP at the literature searching stage. Thirty-eight papers were finally included in the medicine-taking synthesis and 29 papers in the RA synthesis, although further exclusions were made during the synthesis (Box 5).

Quality

We did not exclude papers on grounds of quality. The only reasons for excluding papers were either that they were not sufficiently focused on the topic or that they did not actually report qualitative research. Sometimes this was not immediately obvious. For example, some papers described work where qualitative data had been collected, but on closer examination it became clear that the data were not analysed qualitatively. In several papers that, strictly speaking, were relevant and the methodology was qualitative, the analysis was considered to be very superficial or few findings were reported. In these cases we did not exclude papers, but attempted to synthesise them.

Sometimes, it was an absence of adequate reporting that could confuse matters. For example, in the RA synthesis, some of the better theoretical papers (the ‘classics’) scarcely reported their research methods. These papers were conceptually very useful to the synthesis, but if strict quality screening had been employed these papers would have been excluded. Indeed, many of the RA studies would have been so excluded. Only four papers in the RA synthesis were characterised both by good reporting of the use of qualitative methods of data collection and by analysis. (This is discussed further in relation to specific studies in Chapter 7 in which the RA synthesis is reported.) On the other hand, for both topic areas there were papers that reported on methodology in great detail, but produced very superficial findings. We remained inclusive in our approach because, although some of the papers were slight, they were synthesisable in that at least they were descriptively confirmatory of previous findings.

In the medicine-taking synthesis we did not attach greater or lesser importance to any of the papers, we simply tried to put all the findings together. Some papers contributed more in terms of findings than others, but this is not necessarily a comment on their significance. Some papers were more interesting to read than others – usually implying that they interpreted their data rather than simply offering description and content analysis – but all papers were treated in the same way. Whatever they offered, whether description or interpretation, was synthesised. Some papers contributed more to particular areas than others. For example, the asthma papers contributed the least overall, but they made a strong contribution in one area, that of identity and the psychological aspects of medicine-taking.

Several of the individual questions posed in the appraisal were difficult to interpret (e.g. some of those on sampling) and others were felt to be redundant (e.g. one on the setting in which data collection occurred). The question on theoretical perspective was felt to be of limited value, particularly as many of the later studies that we found were primarily descriptive.

Appraising the papers in Access

After conducting the appraisals, both PP and GDW found it easier to import the findings and concepts from the Access database into Microsoft Word and work with the data in this format. Only one of the researchers (LY) found it necessary to access the database thereafter, in order to analyse the reproducibility of the appraisals. For quantitative analyses, therefore, the database was useful. However, the majority of the data were qualitative. Given the time that we invested in setting up the database and the small returns we obtained from doing so, we do not consider that it was a worthwhile endeavour to have conducted the appraisals in Access 2000. After conducting the appraisals, both PP and GDW found it much easier to revisit the papers themselves, rather than access the database, for any information that was subsequently needed. Those independent second reviewers who did not have access to the database conducted their appraisals in Word and reported that this was acceptable.

Inter-rater agreement

The data were firstly examined to determine inter-rater agreement (Cohen’s kappa) for each item. This calculation was based on six pairings of assessors, as there were two first assessors, each of whom was paired with three different second assessors. The number of papers rated by each of the six pairs for which complete data were available was too low (ranging from 6 to 14) to permit meaningful statistical comparison of the differences between kappa values for each pair of assessors, and so averaged kappa values for all pairs are presented in Tables 6–10.

It is evident that inter-rater agreement was typically rather low, albeit better for some classes of questions than others. There was moderate-to-good agreement on the majority of questions relating to design, sampling and recruitment (see Table 6), and modest agreement for all those relating to data analysis (see Table 8) and the value of the study (see Table 10). However, agreement was poor for questions concerning data collection (see Table 7), the interpretation and presentation of data (see Table 9) and the transferability of results (see Table 10).

Pooled ratings of papers

For all subsequent analyses, the data were pooled, such that ratings of the papers by all assessors were combined. The pooled data thus aggregated two sets of ratings for each paper: one set was derived from the two first assessors, and the second from the six second assessors. Consequently, it should be noted that the ratings of the two first assessors had a greater influence on the aggregated results than the ratings of the six second assessors, as each of the first assessors rated approximately three times as many papers as each of the second assessors.

Tables 6–10 show the percentage of pooled ratings that indicated that the assessors replied affirmatively to each of the CASP18 questions. The assessors answered positively to most questions about design and sampling (see Table 6); reporting of the setting of data collection and how data were recorded (see Table 7); clarity about how the analysis was conducted (see Table 8); satisfactory presentation of data and findings (see Table 9); and the value of the studies (see Table 10). However, few studies provided details such as justification of sample size, explanation of refusals to take part, how the research was explained to the participants or how the data presented were selected. Nonetheless, most papers were given positive ratings regarding the global evaluations (see emboldened question on the bottom row of each table) of their sampling strategy, data collection and data analysis. There was, however, little inter-rater agreement concerning which papers could not be rated positively on these questions. This pattern of findings may largely reflect the publication process, whereby papers would not normally be accepted unless the basic design, collection and presentation of data and value of the study were generally considered satisfactory, but space limitations may have prevented authors from providing more detail of the research process. It is perhaps of more concern that few studies appeared to have engaged in flexible or iterative data collection, reflexive consideration of the role of the researchers and their relationship to the participants, feeding back results to participants, or consideration of contradictory evidence or alternative interpretations. However, it must be noted that not all of these procedures for increasing the credibility and rigor of the research process would be appropriate for every study or would have been expected over the period of publication.

Correlations with ratings of the overall value of the papers

Ratings of the overall value of the papers had an approximately normal distribution, and so correlations were calculated between this item (see bottom line of Table 10) and all the other CASP18 items, in order to determine which aspects of the studies assessed by the CASP18 questions were associated with positive global evaluations (see final column of Tables 6–10). All those items that were significantly correlated with the global evaluation were then entered into a stepwise multiple regression (Table 11) in order to determine which items independently predicted the global evaluation of overall value.

These results show that the assessors’ global ratings of the value of the study were, unsurprisingly, associated with their judgements of its importance and contribution to knowledge. More significantly, global evaluations were related to positive ratings of a clear description of how conceptual interpretations were developed, adequate discussion of evidence for and against the interpretations presented, and a method of data collection appropriate to the aims of the research. Papers that had clearly identified quotations and whose findings were presented in an explicit and easy to follow manner were also more likely to be positively rated.

A later date of publication was correlated with more positive global evaluations (r = 0.24, p < 0.001). Additional features of the paper weakly associated (p < 0.05) with more negative global evaluations were type of publication other than peer-reviewed article and a later date of appraisal (possibly indicating papers identified by means other than electronic database searching).

Medicine-taking (PP lead)

PP’s synthesis

The four papers134–137 focused on antihypertensive medications, with the aim of describing the experiences of patients taking these medications and providing reasons for not taking medicines as prescribed. For PP, the first published, by Morgan,134 was the ‘index’ paper against which the others were compared. Morgan’s134 publication was a book chapter, and the others were articles published in journals: nursing (two135,136) and clinical (one137). PP then summarised and combined the findings from each of the papers around 13 issues that she felt emerged from the papers, some of which were inter-related. PP described the combined findings on issues such as the extent of taking and not taking medications as prescribed, lay testing of medication, and characteristics of patients taking and not taking medications as prescribed. In each case, she summarised the common findings from the four papers and indicated contrary or additional findings from particular papers. She began the process of exploring ‘higher order’ concepts, for example noting that several papers indicated that patients appeared to purposefully not take medications, although this was not overtly stated.

PP produced a ‘map’ (Figure 4) that explored links between the issues identified, using labels to indicate which papers contained the categorisations, and attempted to coalesce findings into common labels. For example, in one instance, a finding that patients did not take hypertensive medication as prescribed because they wished to avoid certain side effects when engaged in a particular activity was described by the authors as ‘incidental non-adherence’. 136 This was reclassified by PP as ‘purposeful non-adherence’,136 as specified in another paper.

FIGURE 4.

Map of antihypertensive medication papers (pilot synthesis – PP). AC, African Caribbean.

Rheumatoid arthritis (GDW lead)

PP’s synthesis

PP began with a short summary of the publications to be synthesised. 138–141 She felt that they fell into two groups: two were theoretical explorations of chronic illness,140,141 using RA as an example, whereas the other two focused on the day-to-day experiences of living with RA. 138,139 PP elucidated a number of key categories from the papers and then described each one, indicating how it was presented in each of the papers. PP indicated whether or not the concept was explicitly mentioned in each paper and then described how it was presented. For example, the concept ‘normalising’ was identified by two of the original authors explicitly138,140 and in terms of various activities in another. PP then drew several diagrams that placed the experience of RA on scales, with benefits and costs of normalising and renormalising presented as being in balance (Figure 5 and 6). There are then two tables (Tables 12 and 13), one presenting aspects of the key concept of uncertainty and the other focusing on issues related to medication use, its limitations and other resources which patients could draw on. PP focused the synthesis around the concepts of Wiener138 (the first and index paper) of normalising and renormalising, and synthesised the other papers around this. This produced the key concept of ‘achieving a balance’. She attempted to combine the papers around this concept, using a grounded theory approach, which appeared to work well.

FIGURE 5.

Normalising and costs of normalising (pilot RA synthesis – PP).

FIGURE 6.

Renormalising and costs of renormalising (pilot RA synthesis – PP). ^aRelated to resources.

TABLE 12 - Uncertainty (pilot RA synthesis – PP)

Acute flare-up	Author	Uncertainty	Author	Remission/relief	Author
When will it strike? How long will it last? How will it manifest?	Wiener138	Uncertainty about flare-ups and remission makes illness difficult to manage	Locker139	How long will it last? Timing?	Wiener138
What (if anything) will control symptoms?		Constant self-monitoring necessary		Remission mitigates against acceptance of disabled status
How much pain will there be?	Locker139	Continuous testing of tolerance to activity is necessary		Hope for remission balanced against dread of disease progression
How long will pain last? Which joints will be affected? When will it happen?
Which treatments (if any) will work?	Bury140

TABLE 13 - Medicine and inner resources (pilot RA synthesis – PP)

Medicine	Author	Limitations of medicine	Author	Inner resources	Author
Is a key cultural resource Referral to RA clinic and diagnosis gives relief and imposes order on symptoms	Bury140	People realise limits of medical knowledge regarding the cause and nature of RA	Locker139 and Bury140	Patients left to draw own conclusions about whether or not RA is inherited	Bury140
Diagnosis justifies people’s symptoms and behaviour	Locker139 and Bury140	Medicine cannot answer questions regarding aetiology		Patients have to construct own theories. Might search past for reasons or search family/personal biographies, or reinterpret doctor’s explanation	Locker139 and Bury140
				Narrative reconstruction	Williams141
People hope that medicine will provide information on how to manage RA, aetiology, fluctuating pain levels, prognosis	Locker139	Often medicine cannot provide satisfactory answers Medicine cannot allay fears about future with disability	Locker139 and Bury140	People have to impose own meaning and certainty in whatever ways possible	Locker139 and Bury140
Medicine provides treatment		People realise that regimens are difficult to follow, not very effective	Bury140
People hope it will help manage pain (because pain depletes energy, disrupts personal and social life, is difficult for others to understand, and is unpredictable and intangible, so it is difficult for sufferers to secure legitimacy)		Realisation that treatments are often unpleasant, painful, risky and of dubious efficacy	Locker139
		Realisation that pain can defy medical management		Development of own strategies to deal with pain (social withdrawal, hiding extent of pain to deal with problems of legitimation)	Locker139
		Worry regarding dependency on and side effects of painkillers		Adaptation of regimen to allow maximum relief with minimum side effects/dependency

Initial extraction of data from papers

The findings and concepts that we outlined for each of the studies constituted the raw data for the initial stages of the syntheses. (The modified CASP18 appraisal tool that we used had separate questions about findings and concepts.) For the medicines synthesis, the extracted findings and concepts were compared between the two independent reviewers for each of the papers. In the rare cases in which the second reviewer had identified a finding or concept that was missed by the first, this was included to ensure that the final record of findings and concepts was comprehensive. The process of extracting the findings and concepts inevitably involves a degree of organising and summarising; thus, to some extent, an initial process of interpretation was already under way at this stage, especially when organising descriptive findings that had not been interpreted. At this stage, it also became apparent that the distinction between findings and concepts was neither simple to make nor useful, and because it was felt to be unnecessarily complicated it was abandoned.

Organising studies into medicine groups

As noted above, Noblit and Hare15 suggested that the fourth stage of the synthesis might be determining how the studies are related, followed by (5) translating studies into one another, (6) synthesising the translations and (7) expressing the synthesis. However, Noblit and Hare15 used much smaller samples in their syntheses (two or three studies at the most, although the reports were probably quite lengthy) and, with 38 studies to synthesise, the need to organise the studies in some preliminary way became imperative. For this reason, we began by organising the studies in what seemed to us to be the most immediately obvious way, into groups according to the medicine studied, and then within medicine groups, by the date of publication (Table 15). This gave us seven groups: antihypertensive medicine, which was also used for the preliminary synthesis (four studies); human immunodeficiency virus (HIV) infection medicine (11 studies); psychotropic medicine (six studies); asthma medicine (five studies); studies on medicine in general (five studies); a group of medicines for miscellaneous illnesses (five studies); and proton pump inhibitors (PPIs) (two studies). We then proceeded to translate studies into each other within each of the medicine groups.

Translating studies into each other within medicine groups

Noblit and Hare15 suggested that the process of translating findings into each other goes something like, ‘one case is like another, except that …’ (p. 38). We followed this process in a systematic way. For example, paper 1 in the psychotropic medicine group might have findings X, Y and Z. Paper 2 in the same group might have finding w (something new that was not in paper 1), findings x and y (findings similar to findings X and Y in paper 1) and nothing like finding Z from paper 1. So this would produce a synthesis of papers 1 and 2:

• finding w (from paper 2)

• findings X and x (from papers 1 and 2)

• findings Y and y (from papers 1 and 2)

• finding Z (from paper 1).

This synthesis of papers 1 and 2 would then be compared with paper 3 in the same way. Then the synthesis of papers 1, 2 and 3 would be compared with paper 4, and so on until all the studies within, for example, the psychotropic medicine group had been translated into each other. Then the process would begin again for the next medicine group, and so on. This process of translating the findings into each other is one of ‘reciprocal translation’, as described by Noblit and Hare. 15 The process of ‘reciprocal translation’ was employed because the majority of the studies are essentially about similar issues. Another possible method of translation is ‘refutational’,15 but in this case none of the findings from the studies refuted each other.

Determining how findings relate to each other within medicine groups

The translation of findings from the studies into each other for each of the medicine groups resulted in a raw textual synthesis for each of the groups. We then needed to know how the findings related to each other within each of the groups. Noblit and Hare (p. 28)15 wrote: ‘We think it makes sense to create a list of the key metaphors, phrases, ideas and/or concepts (and their relations) used in each account and to juxtapose them’. We did this by summarising the key findings for each of the illness groups so that they fitted on one page and then drew the relationships between them. This resulted in a ‘map’ for each of the groups (Figures 7–12). These ‘medicine maps’ summarised the key findings, the way in which they translated into each other and the relations between them, for each of the groups.

FIGURE 7.

Proton pump inhibitors synthesis. OTC, over the counter.

FIGURE 8.

Highly active antiretroviral therapy (HAART) synthesis. QoL, quality of life.

FIGURE 9.

Miscellaneous medication (including for arthritis, cancer pain, mental-health problems, thalassaemia major and respiratory disease) synthesis. ntdap, not taking drugs as prescribed.

FIGURE 10.

General medication synthesis.

FIGURE 11.

Asthma medication synthesis.

FIGURE 12.

Psychotropic medication synthesis. CPN, community psychiatric nurse; ntdap, not taking drugs as prescribed; PRN, prescribed as needed; QoL, quality of life.

Determining how studies are related across the medicine groups

We now needed to know how the findings related to each other across the medicine groups. We determined this by comparing the medicine maps with each other, and developed a model that was able to encompass the findings from all of the studies in the synthesis. Some of North et al. ’s154 findings did ‘fit’ with this, but their typology of relationships to benzodiazepines did not. Our model changed slightly as we progressed with the synthesis, but the final version is shown in Figure 13. Dowell and Hudson,173 authors of one of the studies in the synthesis, had developed a model of medicine that was similar in many ways, but which was unable to ‘hold’, or account for all the findings from all the studies in our synthesis. Our model provides a useful overall map for the synthesis process. It was an invaluable organisational aid during the synthesis process, but it may also prove to be a useful model for understanding the route by which people decide whether or not and how to take their medicines. It also illustrates that seemingly disparate studies may be put together to create something new. The model, however, provides only an overview. As an abstract representation of the findings and the relationships between them, it lacks necessary detail.

FIGURE 13.

Model of medicine-taking.

Synthesising translations across the illness groups

The next stage was to return to the detail, to the textual translations that were conducted for each of the medicine groups, and synthesise these across the medicine groups. So whereas the medicine maps were ‘brought together’ to provide our overall model of medicine-taking, now the textual syntheses for each of the medicine groups were ‘brought together’, or synthesised, to provide an overall synthesis of medicine-taking. This initially involved reading and rereading the textual syntheses for each of the medicine groups (referring back to the original papers where clarification was necessary), then bringing together the data thematically. So, for example, if ‘worries about medicine-taking’ were identified in the asthma, HIV infection and hypertension textual syntheses, these data were brought together beneath this heading in the overall synthesis. Thus, the data for the overall synthesis were analysed thematically rather than according to medicine group. This overall synthesis is a ‘lines-of-argument’ synthesis,15 in that it says something about the whole based on studies of the parts.

Noblit and Hare (p. 64)15 explained the relationship between the translating of cases into each other and the subsequent expression of the synthesis as ‘lines-of-argument’: ‘In short, the translation of cases into one another sets the stage for a second-level inference about the relationship between the studies. It is the second-level inference that assigns interpretive significance to each study to be synthesized’. The ‘lines-of-argument’ synthesis, therefore, essentially involves a process of interpretation and conceptual advancement and is at the same level as Britten et al. ’s16 ‘third-order interpretations’. The entire process of the medicine-taking synthesis is summarised in Box 7.

Exclusions during synthesis

At the end of the synthesis, it became apparent that one of the papers in the medicines synthesis165 had contributed nothing at all to the final synthesis even though it had been included in the asthma medications map. The paper had been borderline because much of it was about children’s response to asthma as well as asthma medicine. On further reading, it became clear that its emphasis precluded it from contributing to the synthesis. The final number of papers that contributed to the medicines synthesis was therefore 37 (Box 8).

Exclusions during synthesis

Four papers were excluded at the synthesis stage,179–182 leaving a total of 25 papers in the RA synthesis, describing 22 studies (Box 9). One was excluded because the focus was doctor–patient communication, rather than the lived experience of RA. 179 Another was excluded because it focused on conditions other than RA,180 whereas the remaining two181,182 were marginal in terms of whether or not the research methods employed meant that the study fell within our definition of qualitative research.

Preliminary studies of reproducibility

Reproducibility of qualitative research is difficult to assess as it is not possible to collapse the detail and richness of the data into simple quantified components. Assessing the reproducibility of qualitative research is, then, a matter of judgement. There is considerable debate in the methodological literature about whether or not ‘reproducibility’ is a reasonable concept with which to assess qualitative research. It is always likely that different researchers with different backgrounds and different levels of experience will interpret qualitative data in different ways. It is likely, therefore, that the reproducibility of syntheses of qualitative research is also likely to be variable and related to the particular researchers involved. This can be seen in the comparison above, particularly in relation to researchers’ different levels of experience of synthesising: PP had undertaken this before and so did not feel the need to describe the process in as much detail as GDW who was doing this for the first time. Similarly, as the two researchers were going on to complete only one of the syntheses each, there were differences in their preparation of the material for continuing to synthesise beyond the first four papers.

Many of the concepts and findings in these two syntheses were nevertheless combined similarly by the researchers. There were differences of detail, but often these were related to the way in which the original authors presented the concepts, which presented the opportunity for terminology to be interpreted differently. The approaches to the syntheses were somewhat different. GDW, for example, preferred to place the findings in tables to facilitate comparison, whereas PP preferred to describe the process of combining findings more textually. Such differences are reflective of individuals’ particular modes of working, and it is unlikely that a single approach could be deduced and applied universally. Indeed, this is also true of primary qualitative research. The key to synthesising qualitative research lies in becoming immersed in the findings and concepts presented by the original authors, and then attempting to combine these to reach summaries of findings and higher level concepts.

Overall, this preliminary and exploratory attempt to investigate the reproducibility of the first stage of a qualitative synthesis has shown that the combination of concepts and findings is fairly similar, even when the approach to synthesising is quite different. It is possible to deduce reasons for differences related to approach and background of the researchers and a lack of explicitness on the part of the original authors. This investigation involved experienced qualitative researchers. The complexity of the task suggests to us that it should always be undertaken by experienced researchers. It is our view that less experienced researchers would find the tasks of translating concepts more difficult, with the likely result of less coherent syntheses – although, this remains to be tested. This investigation examined reproducibility only at an early stage with only four papers, and it is not clear whether the syntheses would have become more or less similar with the inclusion of further papers. Only two researchers independently synthesising the same full set of papers would show this. We would suggest that the evidence from this investigation of reproducibility should assure researchers and policy-makers that the process of qualitative synthesis is reasonably robust when undertaken by experienced researchers.

Full syntheses

As with the preliminary syntheses, the two full syntheses were conducted differently from each other, primarily reflecting the differing amounts of literature to be synthesised. Because there was more literature for the medicines topic, additional work had to be done to organise the literature and make the synthesis process manageable. Another research group might simply have synthesised the medicine papers chronologically, but it is likely that an analysis by medicine or disease would still have been conducted at some later stage. More detailed reflections on the full syntheses will be offered at the conclusion of the synthesis findings (see Chapter 6).

Sociological and anthropological studies challenging compliance

The earlier sociological and anthropological studies in this synthesis challenged decades of quantitative research that had worked within the paradigm of compliance. The first three studies153,166,171 noted that previous research rested on an ideology of compliance which implied medical dominance. Donovan and Blake,166 in the context of a study into lay experiences of treatment for RA, argued that the concept of ‘compliance’ lays the responsibility squarely on the patient, and questioned the assumption that patients should comply with their medical treatment, observing that ‘non-compliance’ is a problem for doctors, not patients. Roberson,171 whose study is about general experiences of taking medicine, also highlighted the assumptions upon which earlier research was based: that people’s health will improve if they follow their doctors’ orders, that biomedicine is the only option and that those who do not follow medical advice are irrational, deviant or uncooperative. Kaljee and Beardsley,153 writing about psychotropic medicines, noted that the term ‘compliance’ assumes the authority of doctors over patients, the latter being represented as ignorant and unable to understand the doctors’ orders. Like Roberson,171 Kaljee and Beardsley153 noted that the biomedical approach is only one among many other ways of achieving health (a point that is seldom made after these two early papers). Both Roberson171 and Kaljee and Beardsley153 concluded that if health professionals want to understand compliance it is important to understand the patients’ perspective, and Donovan and Blake166 recommended the ‘development of active, co-operative relationships between patients and doctors’ (p. 512).

Later studies by sociologists Adams et al. 161 of treatment for asthma, Rogers et al. 156 of neuroleptic medicine, Lumme-Sandt et al. 175 of older people’s experiences of medicine-taking and Atkin and Ahmad169 of the ways in which young people respond to chelation therapy for thalassaemia major also gave primacy to lay perspectives and tried to illustrate that patients’ behaviour could be understood as rational and strategic, rather than passive and powerless. Like the earlier studies, they suggested that qualitative investigations of lay explanations and experiences would provide clues as to how people can be better supported in taking their medicine. Thus, although they challenged the ideology of compliance for its paternalism, these studies nevertheless displayed varying levels of attachment to the biomedical paradigm.

Sociological studies not focusing on compliance

A small number of sociological studies investigated medicine-taking, but without a focus on compliance. North et al. 154 developed a typology of relationships to benzodiazepines in order to understand how people use them, whereas Britten172 took Cornwell’s193 work on ‘private’ and ‘public’ accounts as a starting point for her analysis of ‘orthodox’ and ‘unorthodox’ lay accounts of medicine-taking. Morgan134 was primarily concerned with how ethnicity influences the way people manage their antihypertensive medicines. She noted that reducing blood pressure through antihypertensive medicines was regarded as a major preventive strategy for decreasing stroke and heart disease, but she did not ally herself with attempting to resolve what the medical profession regarded as a problem of ‘non-compliance’. Rather, she suggested that reliance on medicine as a response to ill health can be regarded as culturally rooted. Prout et al. 162 investigated the role of medicine in the lives of families with asthma and concluded that people much prefer it to the ‘non-medical’ preventive strategies that are advocated alongside the use of inhalers.

Nursing study not focusing on compliance

Usher,158 writing in a nursing journal, explored the experience of taking neuroleptic medicine. She considered the issue of compliance, but her focus was on understanding what it is like to take the medicine.

Economically motivated studies

Four studies set out to answer specific and largely economically motivated questions about medicines. 155,159,160,167 All but the first of these were published in the journal Social Science and Medicine. Aiming to reduce benzodiazepine use, Barter and Cormack155 investigated why people continue to take this medicine despite a reduction in prescribing. Dowell et al. 167 explored patients’ experiences of having their prescriptions changed in order to determine whether or not expensive prescriptions can be switched for cheaper ones without upsetting too many people. Boath and Blenkinsopp159 and Pollock and Grime160 set out to discover, respectively, an explanation for the dramatic increase in prescribing for PPIs and ways of decreasing the costs of PPI prescribing.

Professional agendas

Watson and Mitchell,174 writing in a nursing journal, investigated how older women use medicines, claiming a dearth of information on women’s health. Their discussion, however, was more suggestive of an interest in expanding the role of community nurses in the care of older women. Similarly, Smith et al. ,170 writing in a pharmacy journal, investigated long-term users’ experiences of medicine use, but they seemed most interested in whether or not there was an opportunity for pharmacists to extend their information-giving role.

Increasing adherence

Just over half of the studies in the synthesis were primarily concerned with finding out how to increase people’s use of medicines. These papers were published mainly in medical or nursing journals and varied in the extent to which they questioned the concept of compliance.

Dowell and Hudson173 writing in Family Practice about medicine-taking in primary care, noted that compliance is based upon problematic assumptions and acknowledged that many people have good reasons for not taking their medicine as prescribed. Nevertheless, they aimed to increase compliance as long as this is effective, safe and acceptable to patients and they suggested that doctors could help by supervising the changes patients make to their regimens. Ersek and Miller,168 writing in Cancer Practice, aimed to find ways of overcoming cancer patients’ non-adherence to analgesic therapy. They concluded, somewhat unusually, that although patient education may help some, assisting people to use non-pharmacologic pain strategies may be more successful. Angermeyer et al. ’s157 study published in a psychology journal investigated people’s perceptions of neuroleptic medicine in order to find ‘hints’ for increasing compliance with treatment.

In addition to Morgan’s134 study, three other studies investigated the experience of taking antihypertensive medicine. 135–137 The first two were published in nursing journals and the third in a cardiology journal. All encouraged adherence and none questioned the assumption that medicines are the only way of managing high blood pressure and preventing strokes and heart disease. Van Wissen et al. 135 took the view that ‘Compliance with treatment is a fundamental prerequisite for therapeutic benefit’ (p. 568) whereas Johnson et al. 136 suggested, ‘Poor adherence to prescribed hypertensive pharmacotherapy is responsible for unnecessary complications, leading to increased health care spending, premature progression of complications, and early death’ (p. 319). Interestingly, Johnson et al. 136 seemed to be attributing the morbidity and mortality to non-adherence rather than to the hypertension itself, illustrating the extent to which health is seen to be allied to taking medicine. Svensson et al. ,137 on the other hand, although promoting adherence, suggested that modifying the regimen was ‘sound behaviour’ and should be encouraged if it represented an attempt to gain control of the condition.

In addition to the studies by Adams et al. 161 and Prout et al. 162 (above), two further studies dealt with taking medicine for asthma. Buston and Wood,163 writing in Family Practice, were concerned with surmounting the barriers to compliance in the case of adolescents with asthma, and Walsh et al. 164 were interested in how the doctor–patient relationship influenced compliance. Although the psychotherapeutic framework they adopted was unique in the synthesis, Walsh et al. 164 shared with Buston and Wood163 the assumption that compliance is desirable. However, Walsh et al. ’s164 emphasis that patients’ behaviour is rational and ‘sense making’ recalled the earlier sociological studies that stressed the importance of lay meanings.

Eleven studies142–152 (10 of which were American142–148,150–152) focused on adherence to antiretroviral therapy for HIV infection. Most of the authors of these studies took a fairly strong position, repeatedly citing the benefits of taking the medicine (increased survival, delayed disease progression, improved immune functioning, reduced incidence of opportunistic infections) and the problems associated with not taking it (vertical transmission to babies, viral resistance, limited future treatment options). Siegel and Gorey142 investigated women’s experiences of, and views about, taking azidothymidine (AZT; zidovudine) in the hope of increasing their uptake of the drug. Stone and Clarke143 recommended simplifying the antiretroviral regimen to improve adherence but suggested that only patients who committed themselves to the regimen should be treated. Erlen and Mellors144 suggested that health professionals need to know what patients’ experiences are so that they can individualise regimens to suit the patient. Both Halkitis and Kirton145 and Proctor et al. 146 focused on the obstacles that prevent adherence to antiretroviral therapies and the strategies people use to enhance adherence.

Siegel et al. 147 took the position that patients were misinterpreting the symptoms of HIV infection as the adverse effects of antiretroviral therapy, either through ignorance or because they could not accept that the HIV infection was progressing. They recommended that health professionals help people distinguish between symptoms and adverse effects, provide strategies for alleviating symptoms and educate patients about the harmful consequences of non-adherence. Siegel et al. 147 categorised those who did not adhere to antiretroviral therapy as offering ‘excuses’ or ‘justifications’ for their behaviour and recommended that misconceptions be debunked and the importance of adherence promoted. They noted that ‘… a set of beliefs may be growing among patients that may support non-adherence’ (p. 39). Similarly, Barton Laws and Wilson152 explored people’s experience of taking antiretroviral therapy and concluded that people come to believe that whatever standard of adherence they can achieve is good enough. Murphy et al. 150 investigated factors that facilitate and prevent adherence, and McDonald et al. 149 investigated the differences between men’s and women’s experiences of treatment. Noting that women with HIV are less likely than their male counterparts to use antiretroviral therapy, the authors concluded that this is because women are less likely to believe in the treatment’s efficacy. Johnston Roberts and Mann150 set out to explore the unique barriers to adherence to antiretroviral therapy faced by women. They wrote from the standpoint that adherence should be encouraged, but were sensitive to the social factors that impinge uniquely on women, such as motherhood.

In summary, the earlier studies in the synthesis can be understood as reacting to the concept of compliance that had held sway in the preceding decades. They advocated a qualitative methodology for the purpose of providing an alternative, lay perspective and emphasised the rationality of lay behaviour. A small number of sociological studies investigated medicine-taking in its own right, but the absence of compliance as a starting point was rare in the studies. A few studies answered specific economic questions and a couple seemed to be influenced by specific professional agendas. Among the later studies, few were sociological and more were conducted from the perspective of increasing adherence. Taken as a whole, the studies in this synthesis illustrated that the employment of a qualitative methodology, or a sociological perspective, does not necessarily imply abandonment of the medical agenda. The majority of the studies focused on the reasons why people do not take their medicine as prescribed. Few considered the experience of those who reject medicine outright or those who accept it uncritically. Additionally, the vast majority of studies were concerned with chronic illness. Because the protracted nature of chronic illness presents more challenges to Western medicine than acute illness, this is probably a reflection of medical concerns. The ideology that people should take their medicines as prescribed remained dominant in the studies.

Lay evaluation of medicines and its difficulties and limits

Stopping the medicine and observing the effect

Siegel et al. ,147 referring to patients as ‘naive scientists’, suggested that people formulate hypotheses about medicines causing effects and then test these hypotheses by altering the dose or stopping the medicine in order to observe the effects. This seems to be a popular method with several different types of medicine: antiretroviral therapy,148 PPIs,159 antihypertensive medicine,136 inhalers163 and non-steroidal anti-inflammatory drugs. 166 Although several of these studies implicitly referred to this type of test, none outlined it in much detail. Dowell and Hudson173 suggested that the lay testing of medicines may be either an explicit or a subconscious act and that people are more likely to conduct tests if the medicine is to be taken in the long term. They gave an example of a woman with systemic lupus erythematosus who was trying to manage without her steroid therapy:174

Prednisolone, I felt that I could be off them, I thought you know, I was feeling well at the time, I didn’t need to be taking this lot (…) and, err, I mean I did it, they knew I was doing it but it didnae work and I mean I had to go back on them.

This woman suffered rapid weight loss following the cessation of her medicine, which convinced her that she needed to commit herself to the regimen. Another woman, quoted by Buston and Wood,163 reached the same conclusion after experimenting with her asthma medicine:163

I stopped taking my inhalers for a while [laughs] cos I thought I didnae need them so I just stopped taking them. I never told anyone but I just stopped taking them. I’d been all right for a wee while but then I was in hospital. I was in hospital at the end of the year and I think it’s because I stopped taking them for a wee while and then it just built up.

Donovan and Blake166 found that many of the people with RA in their sample conducted a process of testing or evaluating their medicines before deciding whether or not to take the medicine as prescribed. The authors found that 4 of 41 people taking non-steroidal anti-inflammatory drugs eventually reduced their doses and six gave up altogether. One woman said:166

Feldene [piroxicam] worked OK but they make me feel sick and dizzy. I just stopped taking them because I couldn’t do anything when I was feeling giddy … I’d rather just cope with the pain than the pain and the side effects as well.

The authors gave another example of a person who gave up the drug having experimented with it:166

[The doctor] gave me anti-inflammatory tablets. I tried them for a week and they didn’t seem to help at all and I don’t particularly like taking tablets anyway, so I stopped taking them a week after I started.

Donovan and Blake166 noted that some people did not give the drugs sufficient time to work, but were unaware of this.

Worries about medicine that lay testing and evaluation cannot resolve

Although many of the concerns that people have can be resolved through lay evaluation of medicines, some are less easily resolved and may linger, affecting decisions about medicine-taking.

Medicine and identity

Being reminded of the illness

Medicine can also be a constant and unwelcome reminder of illness, whether or not that illness is accepted. Four of the HIV studies found this. 144–146,151 One of the men in Halkitis and Kirton’s145 study said:

… it’s hard every time you take that pill, to recognize what you’re taking it for, and to know that this is going to be probably the course of my life …

One woman wrote in her journal for Johnston Roberts and Mann’s151 study:

Each time I take a pill it is a reminder that I am sick. I do not want to focus on my sickness or let it control my life. I want so badly to be ‘normal’ and healthy again.

Only one other study reported that taking medicine reminded people of their illness, in this case cancer. 168

Disclosing the illness to others

A separate but related issue in the case of HIV infection (and a significant reason why some people did not always take their medicine as prescribed) was fear of disclosing their illness to others. 143,145,148,150,151 Consequently, rather than take the medicine in front of people and risk disclosing their HIV status, people might postpone or forgo their medicine. 151 A participant in Stone and Clarke’s143 study said:

If you don’t want everybody to know, you know, you don’t take ‘em in front of people. It raises questions.

This could include public social situations such as parties and the work place, but also private, family life. One woman in Johnston Roberts and Mann’s151 study who had not disclosed her HIV status to her children found it hard to take all her pills because she had to do so out of sight of her children.

The stigmatising effects of medicine

Some of the people in Rogers et al. ’s156 study reported that neuroleptic medicine was stigmatising because it could mark out a person as having schizophrenia. Similarly, Smith et al. 170 found that those with mental-health problems reported feeling stigmatised and labelled by their medicine, as did Usher,158 again in the case of neuroleptic medicine. Her respondents reported that the adverse effects of the medicine drew attention to their illness, as one person illustrated:158

When you go out it’s like advertising you have a mental illness, so the side effects draw attention to the fact that you have a mental illness. And even though you might be quite well mentally, the side effects stigmatize you … you cannot even go over to your sister’s place and go out into the yard without the neighbours thinking she’s got someone there who is mentally ill … you know your legs are going up and down all the time and they think you’re a lunatic. It’s like wearing a sign on your forehead.

Other studies in the psychotropic medicine group did not cover this issue, although North et al. 154 noted that some people reported feeling ashamed of using benzodiazepines. In the context of HIV infection, Barton Laws and Wilson152 found that some people did not initiate treatment because they worried that the regimen would identify them as having HIV. Finally, Atkin and Ahmad169 found that for children with thalassaemia major, one of the most hated aspects of their chelation therapy was the fact that it marked them out as different from their peers.

Ways of taking medicine

A small number of studies categorised or characterised the various ways in which people take their medicines. This section begins with these studies, before going on to examine the range of modifications people make to their regimens, the reasons behind them and how far these are communicated to doctors.

Those who take their medicine as prescribed (Figure 14)

Passive accepters

Dowell and Hudson173 described people who accepted the regimen that they had been prescribed, who assumed a more passive role and who relinquished control to their doctors as ‘passive accepters’. Similarly, in the context of hypertension, Svensson et al. 137 reported that those who accepted the prescribed regimen were passive, indifferent and less involved in their care. The authors noted that many had ‘rather obscure motives for continuing with treatment, and often claimed not to have considered the possibility of changing or omitting medication’ (p. 162). 137

FIGURE 14.

Categories of those who do take their medicine as prescribed.

Active accepters

In the context of hypertension, Johnson et al. 136 developed a category of medicine-taking which they termed ‘purposeful adherence’. ‘Purposeful adherence’ followed a conscious decision to pursue the prescribed regimen and was dependent upon having first tested the medicine to ensure its effectiveness. Those Morgan134 described as ‘stable adherents’ and ‘problematic adherents’ – also in the context of hypertension – could be seen as practising ‘purposeful adherence’. These people took their medicine as prescribed but the ‘problematic adherents’ expressed concern about doing so and had worries about adverse effects and dependence, whereas the ‘stable adherents’ seemed unconcerned, even by the prospect of remaining on the tablets for all of their lives. Although ‘stable adherents’ seem similar to Dowell and Hudson’s173 ‘passive accepters’, they were not necessarily passive and may have reached their position after a period of testing their medicines.

Those who do not take their medicine as prescribed (Figure 15)

Rejecters/sceptics

Dowell and Hudson173 reported that some people rejected their prescribed medicine outright, without testing it first. These people, whom the authors describe as ‘rejecters’ or ‘sceptics’, preferred to maintain control by using ‘alternative’ therapies or by tolerating their symptoms. They completely rejected Western medicine and its emphasis on drug taking. In the context of hypertension, Johnson et al. 136 noted that some people, whom they termed ‘purposeful non-adherers’, made a conscious decision not to take the medicine, but in this case possibly following a period of testing.

FIGURE 15.

Categories of those who do take their medicine as prescribed.

Lay evaluation

At least two studies preceding our synthesis considered the topic of lay testing,118,194 but we found only one study that specifically set out to investigate this issue. In common with our findings, the participants in the study by Arluke,116 which also predated our synthesis, tested their medicines in a number of ways including observing the effect of stopping the drugs or reducing the dose, and observing the reactions of others on the same medicine. However, although some lay people conduct their own tests of medicines this does not imply that they are ‘experts’ in the field. Prior195 criticised the sociological inclination to treat lay knowledge as ‘every bit as valuable’ as professional knowledge, arguing instead that lay knowledge is often partial and restricted. An example from the synthesis that may support this is the gender difference evident in lay evaluations of treatment for HIV infection. McDonald et al. 149 argued that gay men collectively engaged with the AIDS (acquired immunodeficiency syndrome) epidemic and achieved an unusual level of biomedical expertise on the subjects of HIV infection and trial methodologies, as well as credibility within medical circles (see also Epstein196). On the other hand, women’s decisions tended to be based on ‘personal and idiosyncratic health histories or current or very recent health status. Women were simply unused to making judgements about the efficacy of HIV treatments and their frame of reference was not historically and communally situated, as it was for the men’. 149

However, we believe that the significance of the lay evaluation of medicines lies not simply in the fact that it occurs, but in the reasons why it does. The urge to evaluate suggests widespread caution with regard to taking medicine as well as the possibility that people do not completely trust what they have been told about their medicines. It may also reflect an (instinctive) understanding of the principle of individual treatment response and a desire to ensure that the medicine is suitable in their particular case. As Johannessen et al. 197 explained: ‘The controlled single subject trial is concerned with determining the efficacy of a drug compared with placebo or another drug in one particular patient. Such trials can also be employed to assess the optimal dose of a drug, to identify adverse effects, or to stop established, unnecessary treatment by convincing the patient that there is no benefit beyond the placebo effect. In N of 1 trials each subject serves as his or her own control’. 197

The principle of the ‘N of 1’ trial methodology is often compared with the strategy of ‘trial and error’ in clinical practice, where perhaps the effect of a therapy or the optimal dose is uncertain. Lay people, in the way in which they test their medicines, are attempting to determine whether or not there are any adverse effects and whether or not the medicine works for them and, if so, to establish the most suitable dose for themselves. This suggests that lay people may have an intuitive understanding that population set doses may not be appropriate for individuals. Recently, the worldwide vice president of GlaxoSmithKline acknowledged that most prescription medicines do not work on most people who take them. Steve Connor, the Science Editor for the Independent, wrote ‘It is an open secret within the drugs industry that most of its products are ineffective in most patients …’198 (see also Smith199). Although this may be an open secret within the drugs industry, it is well concealed from those who take the medicines. Fortunately, as this synthesis suggests, many lay people work this out for themselves and modify their regimens accordingly. In support of a flexible approach to medicine-taking, a study on the use of non-steroidal anti-inflammatory drugs found that an adaptable approach that involved varying the dose according to the presence of symptoms was less likely to result in hospital admission for upper gastrointestinal bleeding than was rigid compliance to prescribed doses. 200,201

Adverse drug reactions

The literature on adverse drug reactions (ADRs) suggests that people are right to be cautious about taking medicines. A meta-analysis of studies in the USA found the incidence of fatal ADRs (excluding those because of prescribing errors) to be between the fourth and sixth leading cause of death,202 whereas in England and Wales the Audit Commission estimated that just under 11% of hospitalised patients suffer an ADR, with mortality following ADR showing a marked upward trend. 203 Yet the very real risks involved in taking medicine are either ignored or treated dismissively in the literature. For example, the WHO183 report stated, ‘Concerns about medication typically arise from beliefs about side effects and disruption of lifestyle and from more abstract worries about the long term effects and dependence. They are related to negative views about medications as a whole and suspicions that doctors over-prescribe medicines …’ (p. 44, italics added).

It is worrying that the WHO183 is so dismissive of people’s concerns about medicines. The qualitative studies suggest that these concerns are real and sensible; they are not ‘beliefs’, ‘abstract worries’ or ‘suspicions’ and the statistics on ADRs bear testimony to this. Nevertheless, lay reports of the adverse effects of medicines are consistently dismissed, despite the fact that the current Yellow Card Scheme, which relies on doctors voluntarily reporting ADRs, has been estimated to capture only 1% of actual ADRs. 204,205 Heath206 pointed out that more people are admitted to hospital for ADRs than for problems relating to ‘non-compliance’. ‘Medication errors’ caused by poor prescribing, poor dispensing and poor labelling are also a significant problem for patients and the NHS, with the potential to cause death, morbidity, increased or extended hospital admissions, litigation and financial burden to the NHS. 199

Concordance

Pollock and Grime,160 Dowell and Hudson173 and Svensson et al. 137 suggested that doctors need to accept that patients modify their regimens and that the way forward is to encourage ‘authorised’ self-regulation. This is similar to Donovan and Blake’s166 earlier call for ‘active and co-operative relationships between patients and doctors’ and to the newer concept of concordance, which ‘… describes the process whereby the patient and doctor reach an agreement on how a drug will be used, if at all’. 207 The aim of ‘concordance’ is to involve patients in making decisions about their medicines, to ensure that they have enough information for doing this and to support them with any problems that they might have. Thus, doctors might help patients with their testing and their modifications, providing feedback and guidance. ‘Crucially, concordance advocates a sharing of power in the professional–patient interaction.’208

Concordance and related approaches would appear to make sense in cases where doctors can help patients determine appropriate individualised doses, and provide information on adverse effects and how to deal with them as well as the information necessary to conduct their evaluations safely and effectively. However, concordance depends on doctors sharing information honestly with patients. Unfortunately, Cox et al. 209 have found that doctors emphasise the benefits of treatment rather than discussing its possible harms or risks, despite patients regarding these topics as essential. Equally, the concordance model depends on patients sharing information with doctors, but a few of the studies in the synthesis found that people were unlikely to raise issues with their doctors that they felt would be frowned upon. 156,171,172 Furthermore, Rogers et al. 156 noted that people with mental illness were very cautious about what they revealed to doctors because they were aware of the power that health professionals held over their lives. This was confirmed more recently by Britten et al. 16 Socially excluded patient groups, such as those with poor mental health or HIV infection, may be less likely to achieve ‘concordance’ with their doctors. This could be because they are less normative in their attitudes or because they feel disempowered owing to their membership of marginalised groups. On the other hand, they might feel unable to reveal how they actually take their medicine; for example, people infected with HIV might feel unable to ‘admit’ modifying their regimen because of public health fears about viral mutation. This is not to suggest that health professionals should give up trying to acknowledge and deal with the concerns of marginalised groups, but to make the point that the power imbalance inherent in the doctor–patient relationship will not easily be resolved by the model of concordance.

It is has also been suggested that concordance is simply another way of encouraging people to take their medicine, except that, unlike the ‘compliance’ model, this time the coercion is concealed. 206 This is an important consideration. Nevertheless, although doctors hold the balance of power during the consultation, the medical profession is oddly powerless once the patient has left the surgery. Whether or not there is coercion and whether this is concealed or explicit, lay people have always exercised their power to reject prescriptions or modify their regimens. Undoubtedly they will continue to exercise this power, as this is ‘normal’ behaviour for them. Unless health professionals learn to acknowledge their lack of power over how people take their medicines, ‘concordance’ is unlikely to be achieved.

Relationship between health and medicine-taking

Arguably, the ideologies of compliance, adherence and concordance, because they revolve around the axis of pharmaceutical medicine, distract attention away from non-pharmaceutical approaches to health. Heath206 made the point that the rhetoric of both concordance and compliance uncritically endorses medicine-taking. This relates to the most profound and hidden assumptions in the medicine-taking debate: firstly, that taking medicine is good for you and, secondly, that taking medicine is the only way of achieving desired health outcomes. In fairness, the original conceptualisation of concordance, as set out in the Royal Pharmaceutical Society of Great Britain report,210 acknowledged that taking medicine was not always the best thing to do and that ‘Almost all medicines have the capacity to harm as well as to do good’. 210 However, the WHO182 report made unsubstantiated claims about the health effects of not taking medicines as prescribed: ‘Poor adherence to long term therapies severely compromises the effectiveness of treatment making this a critical issue in population health …’. 183

Both the WHO183 report and the Medicines Partnership report184 assumed that taking medicines as prescribed would lead to good health outcomes. The introduction to the latter report, for example, stated, ‘In the context of a rising drugs bill and the key role of medicines in promoting health, the review underlines the vital importance of maintaining a clear focus on improving compliance and making better use of medicines’ (p. 2). 184 Yet in the body of the report, the authors noted that the Cochrane Review of interventions to improve compliance (also published as a journal article)189 found little evidence that such interventions would lead to large improvements in treatment outcomes. 189 Similarly, the review of the literature contained in the report indicated that the evidence relating ‘adherence’ to improved biomedical outcomes varies according to the illness. 184 At the very least this suggests caution against generalising about the benefits of taking medicines, as clearly the evidence for the benefits of pharmacological treatments is stronger in some areas than in others.

Although the bulk of the medicine-taking literature uncritically assumes that pharmaceutical drugs and Western medicine are the only response to ill health, those people who reject their prescriptions outright clearly think otherwise. Presumably so too did the one in three Americans in 1990 who reported using at least one ‘unconventional therapy’ in the previous year, and who spent almost US$13.7B on alternative health treatments in the same year. 211 It is estimated that approximately half the general population in developed countries uses complementary and alternative medicine. 168 Harrison185 argued that user support for complementary and alternative medicines ‘… indicates in the clearest possible way that a significant number of people do not like conventional medicine or that it has failed them in particular ways’ (p. 36).

Sociology, qualitative methodologies and the adoption of the medical agenda

The sociological studies in this synthesis barely touched the assumptions underlying the compliance debate: firstly, that taking medicine leads to good health outcomes and, secondly, that it is the only possible response to ill health. As such, sociologists have to a large extent adopted the medical agenda. Only a few studies considered those who rejected medicine,172,173,175 despite the fact that it is this group that probably has the most to tell us about the reasons for the widespread reluctance to take medicine. Nevertheless, the early sociological and anthropological studies in this synthesis134,153,156,161,166,171 demonstrated that people have understandable and logical reasons for not taking their medicine as prescribed and challenged the assumptions inherent in the ideology of compliance. As such they undoubtedly influenced medical culture to the extent that, 10 years later, the term ‘compliance’ is no longer considered appropriate and there is a greater readiness to accept that some people will disregard their doctor’s advice and that they may be perfectly sensible for doing so.

There was a discernible tendency in some of the studies to regard qualitative methods as a means of obtaining information on the patient’s perspective that might ultimately be used to encourage people to take their medicine as prescribed. Broadly speaking, the sociological studies advocated the use of qualitative methods as a means of understanding why patients do not take their medicine as prescribed, whereas the medical and nursing studies advocated the methodology as a means of finding out why people did not comply. Angermeyer et al. 157 put it bluntly, ‘Our findings point to the necessity for psychiatrists to explore patients’ subjective views and motivations with regard to their medicine in order to improve treatment adherence’. 157 Given that qualitative methods are the preferred means of exploring ‘patients’ subjective views and motivations’ it is possible to argue that in some cases qualitative methods may have been used to advance the medical agenda.

Bringing medicine into the medicine-taking debate

In our view, one of the key conclusions produced by this synthesis is that the main reason why people do not take their medicines as prescribed is because of concerns about the medicines themselves. On the whole, the findings suggest that there is considerable reluctance to take medicine and a preference to take as little medicine as possible. This is in contrast to earlier theories that attributed ‘non-compliance’ to failings in patients, or current theories that attribute it to failures of the ‘system’ or the doctor–patient consultation. Although medicines are obviously at the centre of the medicine-taking debate and lay people have repeatedly demonstrated their wariness of medicines, their worries have tended to be marginalised or, as noted earlier, treated as ‘beliefs about medicines’, despite the well-documented existence of ADR and instances where concerns about the safety of medicines, for example over addiction to benzodiazepines, have been highlighted by patient groups based on their lay expertise, well before they were acknowledged by professionals (p. 145). 212

It is difficult to imagine how or why this most obvious factor has been overlooked, but it may be because of the dominance of the cultural belief in the benefit of pharmaceutical medicines. Or perhaps it was dismissed because of the view that patients, doctors and systems are easier to modify than medicines themselves. It may be because most of the research has ignored the lay perspective and has failed to involve consumers, resulting in a mismatch between the priorities of researchers and those of patients. 213,214 However, sociologists are also responsible here because we have tended to focus on ‘perceptions’ of medicines or the ‘meanings’ people attach to medicines. This tendency has significant consequences. It makes the person, rather than the medicine, the focus of attention, in much the same way as the WHO183 report (as noted in the introduction to this synthesis) considered anxieties about medicines to be ‘patient-related’ rather than ‘therapy-related’ factors. The result of this is that attempts are made to modify patient behaviour or the doctor–patient consultation rather than question the appropriateness of the medicine. Thus, because people’s accounts are not taken at face value, the more mundane issues about the physical reality of medicines and the effects that they have on people’s bodies and minds have become obscured, with the result that patients’ priorities and concerns are neglected.

Resisting medicines

We feel that many lay people’s response to medicine is best captured by the concept of resistance. In part this is because the term encapsulates the ways in which people take medicines at the same time as attempting to minimise their intake. However, the majority of people represented in this synthesis were not just unwilling or reluctant to take their medicines, nor simply cautious. The strategies they adopted to manage their medicine-taking indicate varying degrees of resistance to the prescription which they were given and, if the term resistance sounds strong, it should be remembered that the huge literature on ‘non-compliance’ exists only because so many people have continued to resist taking medicines in the face of sustained advice, interventions and admonitions. The term resistance also captures lay people’s active engagement with their medicines, as well as the ingenuity and energy that they bring to dealing with their medicines. Additionally, it carries the suggestion of a clandestine operation, which it is, as most people do not tell their doctors about the modifications they make to their regimens.

The term resistance is usually employed in relation to the exercise of power or coercion. As such, it entails acknowledging that the traditional approach to medicine-taking has been, and in varying degrees continues to be, coercive. In this context it should be noted that the noun ‘medicine’ refers both to the pharmacological substance as well as to the profession that prescribes it. Therefore, we could also consider the possibility that any resistance is not only to the pharmacological substance, but also to Western biomedicine. From this perspective, the evaluation and modification of drug regimens that people engage in could also be understood as ways of reasserting control over their bodies and the decisions about it.

Methodological features of included papers

As noted in the methods chapter, of the studies initially included, only four were characterised by both good reporting of the use of qualitative methods of data collection and good analysis. 138,177,229,233 One further paper contained good reporting of qualitative methods of data collection and analysis,223 but in this case the reviewers were concerned that the methods described were not actually reflected in the way in which data were analysed or presented in the paper. Studies by Bury,140,216 Williams,141 Williams and Wood178 and Pinder221 were conceptually very useful to the synthesis, but were characterised by poor reporting of methods of data collection and/or analysis. These papers described various forms of case study approaches that sought to build lines of sociological argument. It might not even be obvious to the casual reader that these studies employed interpretive or qualitative research methods, as this was not always stated explicitly.

The synthesis was principally constructed on the seven following studies. The studies by Bury,140,216 Williams141 and Williams and Wood178 stood out for their theoretical contribution alone, whereas that by Stephens and Yoshida229 stood out for good reporting of methods. The studies by Wiener,138 Shaul177 and McPherson et al. 233 managed to combine conceptual coherence with good reporting and/or use of methods. The last two papers were both also grounded in the results of earlier ‘classic’ studies. The main value of Locker’s139 approach was that the book format allowed for a higher degree of ‘thick description’ and contextualisation of findings than that is found in the shorter journal articles.

The main contribution of three later studies to the synthesis was primarily as descriptive confirmation or elucidation of previous findings. 222,227,231 The value of the studies reported by Williams and Barlow227 and Brown and Williams222 became apparent only during the reciprocal translation stage. This is as an important point because, following initial ‘assessment by checklist’, the papers might have been rejected on grounds of quality or relevance. The studies by Rao et al. 226 and Archenholtz et al. 228 did not look as if they had employed an interpretive approach, although their findings did not conflict with other papers in the synthesis and they were not excluded, although they could have been on these grounds. In addition, neither paper was focused on RA, the latter also including patients with systemic lupus erythematosus, although both clearly incorporated the views of patients with inflammatory arthropathy. Three papers were characterised by poor reporting and/or poor methods, and also added little conceptually to the synthesis, although again they were not excluded on these grounds. 176,224,232 However, and in a similar manner to the Williams and Barlow227 paper mentioned above, the worth of the paper by Stenström et al. 176 became apparent only during the reciprocal translation stage. The importance of the paper by Bjørner and Hansen219 was that its topic – sexuality – was ignored in most of the other papers.

Lay perceptions of the causes of arthritis

Three of the five factors mentioned by Rice and Young’s220 respondents (damp and/or cold weather, ageing and injury) were also found in the two patient focus groups conducted by Lambert and Butin. 230 Rice and Young’s220 participants also mentioned heredity and diet. In the Lambert and Butin230 focus groups, diet was not mentioned as a cause of arthritis, but rather specific foods (red meat, coffee, sugar and junk food) were identified as exacerbating arthritis pain. Other causes mentioned only in the focus groups were viruses, bacteria, bursitis, close-fitting shoes and air pollution. However, in comparing the two studies, the main problem was that Rice and Young’s220 participants included people who did not have arthritis and these responses were not separated out from those of the other respondents.

Rice and Young220 noted that lay accounts of the aetiology of arthritis fell broadly into two types of explanation: firstly, just those that suggested internal, ‘mechanical breakdown’ as responsible, where ageing and injury were identified as factors; and secondly, those that saw arthritis as resulting from an outside agent that infected the body. It was reported both by Shaul,177 and Brown and Williams222 that people had attributed initial symptoms to ‘commonsense’ explanations like minor trauma or overexertion, which also pointed towards ‘mechanical’ explanations.

For people with RA, the search for an understanding of disease aetiology became a more personal matter, as part of the broader ‘attempt to establish points of reference between body, self and society, and to reconstruct a sense of order from the fragmentation produced by chronic illness’. 141 In the perceived absence of biomedical explanations for the disease,226,234 ‘the actual nature of the disease remains elusive’. 140 Without adequate scientific–medical explanations for such devastating ‘physical and social breakdown’,141 people turned to lay explanations reflecting that ‘the body is defined by its relationship to the world of social action not in isolation from it’ (p. 182). 141 In the account to follow, it will be shown that particular features external to the body become significant precisely because of characteristics within the individual, his or her place in society or his or her social milieu.

Living with the symptoms of rheumatoid arthritis

Pain dominated patients’ accounts of the illness experience in RA to such an extent that one could wonder whether pain was the issue at the heart of the matter, rather than RA itself. Many studies contained detailed accounts of the nature of pain in arthritis, although this concept was noticeably absent in Bury’s140 study. However, the abundance of work had fundamentally little to add to Wiener’s138 study of the sociological aspects of pain management in chronic illness. Wiener’s138 was also the earliest published study considered in the synthesis. Whether or not later studies referenced Wiener’s138 work, all those that contained findings about ‘pain’ demonstrated conceptual coherence with her framework and concepts.

Wiener138 noted four features of pain in RA. Firstly, there is an ‘absence of predictability’ to RA’s symptoms, including pain, which is characterised by ‘flare-ups’219 that come without signs or warning. Secondly, there is a ‘variability of progression, severity and areas of involvement’. 138 In other words, people with RA do not know from one day to the next whether they will be feeling better, worse or the same as they do today, or whether or not the disease will affect the same body part(s) as last time. However, the one ‘certainty’ is that sooner or later there will be pain. 139 Thirdly, Wiener138 also found that pain in RA had a ‘circuitous’ relationship with fatigue, such that ‘pain drains energy and fatigue produces more pain’ (p. 98). In later studies, ‘fatigue’ was sometimes mentioned as a separate symptom in its own right, rather than a concept in relationship with pain. Fourthly, Wiener138 noted that these features of pain in RA led to a ‘reduction in personal resources’ for the individual.

Wiener’s138 findings led her to two key interpretations. Firstly, the key concept in living with arthritis relates to ‘tolerating variable uncertainty’. Secondly, the disease conditions engender a ‘dread of [future] dependency [on others]’. Wiener’s138 twin theories were confirmed and rediscovered by most of the later published studies, but no single study has captured these concepts with such clarity and groundedness. Wiener’s138 findings were consistently validated by the findings of the other studies included in the synthesis. Furthermore, her reporting of research methods and methods of data analysis compared favourably with that found in other included studies.

Other work added further findings to Wiener’s138 basic schema. Locker139 completed detailed case histories and interviewed people twice, 1 year apart. He noted a further feature of pain in RA: that it ‘is often not accompanied by any signs which would help to verify the sufferer’s complaints’ (p. 132). 139 This is similar to Williams and Wood’s178 assertion that the symptoms of RA are ‘often vague’ although, for Brown and Williams,222 ‘vague’ was rather used to describe earlier symptoms, which were also characterised as ‘mild’ and ‘non-disabling’. Similarly, Pinder221 noted that both pain and fatigue provide considerable ambiguity, as they are not visible forms of disability.

Interestingly, apart from pain and fatigue, the only other typification of the condition of RA in the earlier literature reviewed was as a ‘crippling disease’ by Bury (p. 173),140 who also identified RA as ‘an outside force’ (see Lay perceptions of the causes of arthritis, above). The focus of earlier work was much more on coping strategies and the effects on social relations, which will be considered later. However, by 1989, Donovan et al. ,217 although still describing pain as the major symptom, also included ‘stiffness’ and ‘burning’ (p. 59). Bjørner and Hansen219 noted that flare-ups of arthritis were associated with reduced libido for some patients. Rice and Young’s220 folk model of arthritis identified ‘painful movement’ as a core concept, and also included ‘inflamed’ and ‘deformed’ joints as conditions, although their model was not necessarily applicable to specific types of arthritis.

From the mid-1990s, the practitioners of qualitative research into RA changed, and there was a discernible move towards descriptive studies highlighting the symptoms and conditions of the disease and their impact on physical and social functioning. This probably reflected the fact that these studies were being undertaken by people also involved in clinical practice, whereas earlier work had focused more on RA as an example of a painful, disrupting, disabling, chronic illness. Wiener138 was interested in the sociological aspects of pain management; Bury140 in the disruption to biography and relationships by chronic illness; Locker139 in RA as a form of social disadvantage; Williams141 in the application of the idea that people reconstruct their personal biographies in light of subsequent events; and Donovan217 in RA as a clinical encounter.

In later qualitative studies undertaken by nurses and others (although we should note that Wiener138 also came from a nursing background), there was a definite move towards a ‘disease-specific’ approach. At their best, these studies confirmed the findings of previous work while presenting results to clinical audiences. However, the reviewers characterised many of these studies as being of ‘poor’ methodological quality. One exception was the article by Shaul,176 which stood out as a far better study. Hers was the first paper to discuss areas of the body involved in RA symptomatology in any depth. For the women in Shaul’s study, ‘Most common symptoms involved the hands, wrists, shoulders and feet. In some cases pain and swelling occurred simultaneously, in others, pain occurred alone, followed by swelling. In either case, there was usually weakness in the affected limb and general fatigue’ (p. 292). 177 Shaul177 identified fatigue as the most ‘pervasive’ symptom, although both fatigue and pain were described as the most ‘distressing’ symptoms. These findings were further developed in a later study, where the root dissatisfaction with RA for female informants concerned the ‘visible swollen and disfigured appearance of these painful body parts’. 227 As one respondent to this study put it, ‘The swellings are like a physical effect of pain. You don’t want people to see the pain you are suffering’ (p. 131). 227 Although these findings seem partly to conflict with those of other studies which noted that the primary symptoms of RA are not obvious, they most probably reflect that in each case findings derived from people in different stages of the disease course (but see Discussion). The primary symptoms of arthritis lead to secondary symptoms, largely related to pain, and identified by: exhaustion, anger, depression, despair, self-pity and [perception of] loss of control. 223 Bjørner and Hansen219 similarly identified ‘low self confidence’, reduced libido and difficulties reaching orgasm.

Consequences for identity and problems maintaining taken-for-granted activities

This section is largely about the problems of maintaining ‘taken-for-granted’ actions and behaviours and the concomitant challenges to norms of behaviour (as perceived by self or others). The best term to capture these issues in the context of RA is probably Bury’s140 ‘biographical disruption’. In a later reflection on these data, he noted: ‘The onset and development of arthritis is simultaneously an assault on the body and a disruption of social life’. 216 This ‘biographical disruption’ centred on the loss of previous life and work identities, or in the metaphor employed by McPherson et al. ,233 loss of the individual’s ‘life definition’. Another metaphor related to these issues is Brown and Williams’222 ‘role incompetence’ (p. 699).

The experience of people with RA was mediated by certain sociological and physiological needs. The metaphors relating to these needs are detailed in Table 20 and are grouped according to the focus on symptoms, personal biography, presentation of self and social roles, and further categorised into intrinsic and interpersonal needs. The studies showed how RA brought previously taken-for-granted actions and behaviours into sharp focus. As Pinder221 put it, ‘The conditions under which people take themselves for granted may be compromised’ (p. 610). As important in this context were the conditions under which other people take the person with arthritis for granted. Some of these conditions were ‘expectations governing behaviour in the wider culture’. 216 As Williams and Wood178 noted, such social or cultural ‘norms’ may be unspoken, which seems to be a large part of the problem for the person with RA: ‘Informal structures like families contain unstated norms of reciprocity or give-and-take, and for many people the experience of disablement is one of having these norms upset’ (p. 130). It is evident that there are as many expectations governing physical functioning or the ability to do things as there are in relation to fulfilling social obligations. Table 21 shows that many of the factors causing disruption in the social lives of people are related to expectations governing social and cultural actions whether at home or in the workplace.

So far as ‘the material world’ and physical activities were concerned, the most detailed exposition of the effects of RA ‘as a disablement’ was that provided by Locker. 139 Usefully, he followed Wiener,138 and noted that the uncertain course of symptoms in RA called for ‘continuous monitoring and self-management since little can be taken for granted’. 139 ‘Activity, however minimal, requires physical and psychological effort’ (p. 19). 139 The ‘erstwhile, taken-for-granted world of everyday life’140 becomes ‘fraught with danger’ and ‘everyday objects and events take on an alarming character and may give rise to feelings of insecurity’. 139 One possible response to such dangers was to never leave the house, although such a response may leave people feeling that they were ‘locked-in’. 223

Responses and coping strategies

The studies showed that the main factors that affect people’s experiences of RA are the symptoms of the disease, the nature of the cultural backdrop, and expectations governing norms of social roles and behaviour. Certain consequences derive from these factors, which affect the ways in which people can or cannot cope with the disease. The psychologically oriented data also pointed towards a ‘career’ model of adaptation to RA, although an attempt to synthesise the studies along these lines was unsuccessful. However, the attempt yielded insights into both the nature of the studies themselves and the experience of adapting to life with RA. As with psychological coping strategies, it seemed as though the idea of an illness ‘career’ probably reflected more the professional and academic standpoint of the authors of the studies than it did the data gathered during the interviews. That is, concepts such as ‘mastery’177 or ‘taking charge’230 were more fundamentally associated with clinical imperatives surrounding treatment, cure and rehabilitation than they were patient perceptions. (Both reviewers felt that Shaul’s177 concept of ‘mastery’ of disease was not clearly evidenced by the findings presented in her paper.)

A further problem derived from the fact that many studies were cross-sectional in nature and were focused on people with more problematic or advanced disease, problems or disability. However, the failed attempt to synthesise the studies around the theme of illness career may say something about the nature of RA itself. Thus, it may be that the pervasive uncertainty and unpredictability of RA means that some people never manage to ‘master’ or ‘take charge’ in RA and, in effect, the disease ‘wins’. Alternatively, people may think that they will one day ‘master’ the disease, but repeated failed attempts lead to a sort of progressive disillusionment. So long as studies remain focused on captive clinical populations, we may never find answers to these questions, as the views of people who can cope, whose flare-ups are not so bad, or for whom the disease is ‘in remission’ were not included in the available research, which was more fundamentally concerned with describing the lives of chronically disabled people trying to cope with severe and debilitating disease.

Tables 22a and b outline the main coping strategies described in the studies. Those described in Table 22a seem to constitute ‘grand’ strategies related to activity in general. Those strategies presented in Table 22b are specific to different arenas: intrinsic strategies, interpersonal strategies, public strategies, and strategies relating to the environment.

Perceptions and experiences of treatment

Surprisingly (given that the interview samples were derived largely from people in clinical settings), perceptions and experiences of care were not a major feature of these studies. An exception was Bury,140 who noted that as medical intervention appeared ‘both important and limited’ (p. 173), medical knowledge in the context of RA was both a resource and a constraint. In a similar manner, Locker139 found that because all treatments involved risks in terms of side effects, ‘the disadvantages may outweigh the advantages to be gained’ (p. 62). Within a focus on individuals’ needs to ‘identify the causes of arthritis’ (p. 191), Williams141 noted, ‘the limitations of medical science in delivering a satisfactory explanation’ (p. 175) and also asserted within this context that ‘medicine can support political bureaucracy in preventing the establishment of social justice’ (p. 185).

Similarly, Brown and Williams222 reported that some of the seven women nursed by the interviewer expressed frustration at the perceived ‘medical impotence’ (p. 698). However, they were also relieved at being referred to a specialist and receiving ‘an acceptable name or diagnosis for their symptoms’ (p. 698). Similar themes were evident in Shaul (p. 292):177 ‘In many cases, the first experiences of medical management were not helpful, as many were misdiagnosed or found that their complaints were not taken seriously’. Ryan224 reported complaints about ‘too much emphasis placed on the physical manifestations of the condition’ (p. 48), and not always being reviewed by the doctor, but these issues did not seem specific to patients with RA. Similarly, Archenholtz et al. 228 mentioned ‘lack of continuity of care, not always the same doctor, [and] difficulties making appointments’.

The most detailed consideration of the issues was that provided by Rao et al. ’s226 focus group study, which usefully compared RA patients with people with osteoarthritis. As in Lambert and Butin’s230 patient focus groups, ‘[some] believed that doctors could do very little to relieve pain or cure the underlying disease’ because ‘arthritis has no known cause or cure’. 226 Interestingly, when compared with the osteoarthritis patients, RA patients more frequently mentioned ‘their trust in the physician’s ability to treat their arthritis’ (p. 259). 226 Other findings were that medications were only effective for short periods of time (p. 255) and that ‘Several patients wished that a single medicine instead of multiple medications could be used to treat their arthritis’ (pp. 255–6). 226

Gender

Given that RA disproportionately affects women, it was perhaps unsurprising that gender roles were particularly emphasised in these studies. However, although some of the papers seemed to have chosen RA as a focus for study precisely because of the likely gender issues, the conceptual development in this field is weak and generally not related either to the existing studies in RA or to wider feminist or gender theory.

A generic weakness across the studies seemed to be the attempt to explain the experiences of women by analysing their accounts in isolation from those of men. A notable exception here was the study by Bjørner and Hansen. 219 It was particularly disappointing that the conceptually rich account provided by Pinder221 seemingly excluded men at the analysis stage. We suspect that the reason for excluding male accounts was ostensibly to ‘strengthen’ the focus of the accounts on the experiences of women, but it could be argued that this approach actually had the opposite effect.

Once again, the main issues were covered in Wiener’s138 ground-breaking study and related to the fact that, within the traditional domestic division of labour, men and children have a stake in women remaining active. Where men take on what were once women’s roles, ‘role reversal may result in a permanent change in the household’s division of labour’ (p. 102). 138 Ultimately, the relevant issues did not seem to point towards gender per se, but rather towards the more general loss of the ability to engage in activities once seen as mundane. Whether for worker221 or homemaker,222 it was precisely these taken-for-granted activities which constituted a large measure of self-identity. 222 Some of the women interviewed in the studies complained that the support they were receiving from partners was ‘too much’, as such support was felt to be eroding their social roles. 232 Reading across the studies, however, it was evident that similar issues also applied to men:223

I’m no longer the man I used to be. I am no longer the husband, the lover, you know, all of the aspects of my life, they’re no longer there.

The one attempt to seemingly systematically compare accounts from women with those from men was found in the study by Locker139 who, in contrast to the other studies already cited, Locker noted that ‘The women experienced job loss as acutely as the men and few were content to settle into the role of housewife’. Thus, we are left wondering whether or not the sorts of gender issues highlighted in the studies were particular to women, or whether or not they were made particular in the way that data were collected and analysed.

Interpersonal gender dynamics received more attention in the two Scandinavian studies,176,219 and the main findings seemed to further underline Bury’s140 assertion that much of the success in coping with RA seems to hinge upon the degree of flexibility possible in both formal institutions and informal relationships:219

In the beginning Soren was going to help me with the cooking, but that was no good. I wanted everything done my way. Now we each do our own things, but duties are not fixed. Soren just does the things he knows I can’t do, and when I am feeling rough he just takes over. It isn’t something we talk about much.

Stages in a meta-ethnography

The studies in both syntheses were concerned with two specific topics but were conducted in several different countries, with disparate samples and methods and for varying readerships. In the case of the medicine-taking synthesis, they also involved a diverse group of medicines. Nevertheless, it was possible, using meta-ethnography, to systematically draw these studies together to present something new. Having completed three syntheses, it is our considered view that reciprocal translation is at the heart of why meta-ethnography is an effective method of qualitative synthesis. The requirement of systematically relating findings from different studies to each other makes it possible to establish new relationships between concepts, such that it is possible to conceptualise an issue in a coherent but fresh way. Following the reciprocal translation stage, the necessary reordering, relinking and reanalysis of material that a synthesis entails represents in itself new material or new findings and may in this case correspond to what Noblit and Hare15 called a ‘lines-of-argument’ synthesis. In the medicine-taking synthesis, the line of argument was both written and summarised and depicted diagrammatically in the model of medicine-taking (see Figure 16). However, further concepts may advance from this level, what Britten et al. 16 called ‘third-order’ concepts. In the medicine-taking synthesis, there were two such concepts: the notion that the medicines themselves needed to be brought back into the debate about medicine-taking and that the way in which many lay people respond to medicines is best characterised by the concept of resistance. In the RA synthesis, the lines-of-argument synthesis was presented in both a written and a tabular form.

Noblit and Hare15 suggested that there were three possible types of synthesis: reciprocal translation, refutational and lines-of-argument. As the preceding paragraph suggests, our experience is that there are three stages of qualitative synthesis in a meta-ethnography with reciprocal translation the first crucial stage. It requires the synthesiser to consider all the concepts and findings in each study in relation to all other studies. One of the concerns about qualitative synthesis is its potential to destroy the integrity of the initial studies. We consider that reciprocal translation protects against this because the method requires taking account of all the findings and concepts in each study and assiduously considering what each means. The next stage involves a process of reordering, relinking and reanalysis, finally leading to a representation of the synthesised material as line of argument. Beyond this there may be a further level of abstraction and conceptual development.

Identification of studies to be included

A key requirement of quantitative syntheses is to achieve an exhaustive search of the literature. Other research groups, using modified versions of meta-ethnography, have conducted extensive literature searches on broad topics. For example, Paterson80 followed the procedures of quantitative synthesis and retained all studies that satisfied the appraisal criteria, resulting in the meta-synthesis of 292 studies of chronic illness. Noblit and Hare15 regarded wide-ranging literature search as useful but not a requirement, depending on purpose of the meta-ethnography. They emphasised that the selection of studies should be made on the grounds of relevance to the topic and should include a range of studies likely to enable new insights. They warned that ‘… generalizing from all studies of a particular setting yields trite conclusions’ (p. 28). 15 We searched only for papers within a 10-year time frame for both the RA and medicine-taking syntheses, although in the case of the former we subsequently included a purposive sample of earlier ‘classic’ papers. In addition, the medicine-taking synthesis included only English-language papers, whereas in the RA synthesis non-English-language papers were considered, and one such included paper required translation, thus increasing the resources and time required.

The reciprocal translation stage is very time-consuming. It and the subsequent stages of synthesis also require the synthesiser to be immersed in all the studies. This places practical limits on the number of papers that can be synthesised. In the medicine-taking synthesis, it was necessary to begin by grouping the papers according to the type of medication as a way of making the task more manageable. Even so, it was difficult to conceive of synthesising many more than the 38 papers included in that synthesis without there being a trade-off between the breadth of papers included and the depth of the scrutiny and thought invested in the various stages of synthesis.

The practicalities of handling large numbers of studies may result in using processes of data selection, management and computing that lead to context stripping, and a less detailed immersion of researchers in the studies to be synthesised. Specialist computer packages are being developed to assist the synthesis of qualitative research, and these are likely to encourage the synthesis of greater numbers of studies, but this approach will potentially change the nature of the meta-ethnographic synthesis towards a more procedural and less interpretive approach.

Noblit and Hare15 were mainly concerned with identifying the relevance of studies and their potential contribution to the question of interest. As we noted in Chapter 4, most syntheses employing meta-ethnography, or methods of synthesis derived from meta-ethnography, have followed Noblit and Hare’s15 approach and not undertaken a formal appraisal of quality. From our experience, an important benefit of the use of formal appraisal criteria was to encourage the close reading of studies as an aid to interpretation. We also found it helpful to be able to discuss papers within the research team. We did exclude papers on the grounds of relevance, or because they did not conform to our definition of qualitative research as involving both qualitative methods of data collection and data analysis. This latter criterion ensured that we excluded qualitative work that was fatally flawed. Yet, in spite of our very detailed appraisal of studies, we did not exclude any studies on quality grounds. Some studies made an important contribution to the synthesis because of their conceptual strength, although their methods, as formally assessed, were weaker (or inadequately described), whereas other studies largely fulfilled the quality checklist in terms of methods but made little conceptual contribution to the synthesis.

We regard the quality of qualitative research as a very important issue. However, we look on the appraisal of qualitative research as more of a concern for the funders and publishers of qualitative research than for the qualitative synthesiser. This is because we have found that weak studies either will not synthesise or will contribute only minimally to a synthesis, adding weight to the better studies. Although there is an argument that including weak studies gives them an unwarranted credibility, such studies do not unduly distort a qualitative synthesis in the way that a poor-quality, highly biased quantitative study could influence a meta-analysis. In a qualitative synthesis, it is the power of ideas that matters.

Reproducibility and validity

The commissioning brief for this project was for methodological work ‘to enable users of qualitative work to both appraise and synthesise qualitative studies in a rigorous, replicable and formalised way’. 10 Meta-ethnography, as described by Noblit and Hare,15 is a rigorous approach to synthesis but it is also an interpretive approach. Consequently, a high level of reproducibility would not be expected. Indeed, the authors acknowledged this and did not view it as problematic, rather the contrary. They advised that ‘Like all interpretations, a meta-ethnography is but a “reading” of what is studied. Other readings are possible and are to be encouraged. However, all interpretations must be grounded in the texts to be synthesized’. 15 Evidence from the reproducibility exercise included within this evaluation is equivocal. Without repeating an entire synthesis, it is difficult to know whether the similarities and differences between the two sets of initial syntheses produced would have become more or less pronounced. (Since the work for this report was completed, we have undertaken an exercise to explore this issue. PP, the key author of the medicines synthesis, undertook a synthesis of all the papers in the RA synthesis. As expected her ‘reading’ of these papers was a little different, but the overall result was the same as the RA synthesis included in this report: it did not produce any substantial new insights.) Pilot syntheses and comparability/reproducibility exercises may have another valuable function, particularly for the first-time synthesiser, as a rehearsal for the full synthesis.

Noblit and Hare15 took a subtle relativist stance and accepted that an interpretation is but one possible interpretation, but also noted that the ethnographer reveals a limited relativism because the range of perceptions is influenced by context and socialisation. Noblit and Hare15 therefore adopted a position held by many qualitative researchers that worth is judged by the understandings and explanations derived rather than by assessments of the validity of processes. Noblit and Hare15 acknowledged that their approach may be regarded as overly relativist as they were essentially dealing with ‘interpretations of interpretation’. Nevertheless, they identified five criteria drawn from writings of qualitative researchers for what constitutes a good synthesis. These criteria related to the adequacy of metaphors (we have always understood metaphors to mean concepts in this context) in synthesising diverse studies which they suggested needed to be assessed according to the following criteria: economy – a metaphor is adequate when it is the simplest concept that accounts for the phenomena and has a superior ‘ease of representation’;234 cogency – a metaphor is adequate when it achieves the explanation without redundancy, ambiguity and contradiction;234 range – refers to the power of the metaphor in incorporating other symbolic domains;234 apparency – ability of language to (seemingly) ‘show’ us experience rather than merely refer to it;235 and credibility – adequate metaphors must be credible and understood by the audience. 236 Noblit and Hare,15 therefore, regarded the worth of a synthesis to be judged by the quality of its metaphors and whether or not the synthesis was regarded as useful and insightful by the intended audience.

Audiences

The issue of the audience for qualitative syntheses is an important one. There are a number of potential audiences for the two syntheses contained in this report. For the medicine-taking synthesis the relevant audiences are health policy-makers, clinicians, research funders, academic health researchers and those involved in the manufacture and control of medicines. In the case of the RA synthesis, the most relevant audiences are health practitioners and others caring for and working with people with RA (for example those providing social, educational, environmental, employment or housing services), people with RA and academic health researchers. These audiences are varied and the format for presenting the syntheses needs to vary accordingly. There is still work to be done on how to present and effectively disseminate the findings from qualitative syntheses. There are examples from within the quantitative review tradition that provide possible models, for example the ‘Guides for the public’ produced by the National Institute for Health and Clinical Excellence, which complement the formal guidance issued to health-care providers and any systematic reviews on which that guidance is based (www.nice.org.uk).

Place of meta-ethnography

Qualitative syntheses published prior to the present HTA project (and subsequently) have identified the value of this approach in informing policy and practice by contributing to conceptual and theoretical developments in the field. Qualitative syntheses may also advance the evidence base by achieving an accumulation of findings and enhancing individual studies by checking their comparability and replication. Moreover, qualitative synthesis has a potential empowering function, as it involves methods for combining multiple voices to seek new interpretations, rather than dismissing single case studies as locally bound. 237

Meta-ethnographies of qualitative health research could be stand-alone or could be a component part of a systematic review incorporating a variety of different types of evidence. This latter role may be particularly useful in HTA in which meta-ethnographies of qualitative research could complement quantitative systematic reviews of clinical effectiveness and cost-effectiveness, thus ensuring that patients’ views and experiences are fully represented in the evidence base. The benefits of this can clearly be seen in the case of a review of the neuroprotective drug riluzole in the treatment of motor neuron disease. 238 The review found that the cost of the drug to the NHS was greater than usually considered acceptable by the National Institute for Health and Clinical Excellence. Qualitative research found mixed views about the acceptability of the treatment for patients: some people with motor neuron disease were resistant to a drug that might marginally extend life without improving its quality; others felt that any extension of life was extremely important. On the basis of these findings, recommendations were made about the kind of information that should be given to patients about the possible benefits of the drug, so that they could make informed choices relating to their own requirements. It also highlighted to the policy-makers that there would be a limit to the number of people who would want the drug if it were made available. 238

Different approaches to meta-ethnography

The synthesis of qualitative research has opened up a very large agenda, reflecting the range of approaches and theoretical traditions that characterise qualitative research. This breadth and inclusiveness has led to the development of differing approaches to synthesis that reflect the varying assumptions and approaches to qualitative research. In addition, the fundamental distinctions between different schools of qualitative research have led to debates regarding appropriate methods for the conduct of each of the phases of study selection, appraisal and synthesis.

There is often a concern to identify a single ‘best’ approach to synthesis. As Barbour and Barbour239 observed, this may result in formalised, proceduralist approaches becoming dominant, and a corresponding neglect of the systematic and meticulous methods of qualitative research and its particular challenges for synthesis, thus leading to a kind of ‘mission drift’. We argue that it is important to acknowledge explicitly these differences and to distinguish between interpretive syntheses based on qualitative methods as developed by Noblit and Hare,15 and more structured proceduralist approaches that transpose checklists, standards and other terminology and procedures from quantitative research, with implications for appraisal, sample size, assessment of validity, etc. Whereas proceduralist approaches may be particularly suitable for conducting broadly focused syntheses involving large numbers of reports, or the synthesis of qualitative research for purposes of combining this with a quantitative synthesis, the strength of an interpretive approach based on the qualitative paradigm is in contributing to conceptual and theoretical development. As Hammersley14 stated, ‘I don’t think there is a single, all-purpose kind of research synthesis, whether we call it a systematic review or meta-ethnography. I think there are different kinds with different functions that make different demands on us’. We endorse this view, and believe that what is therefore of paramount importance is to be clear about the aims and purpose of the synthesis and to select methods that correspond to these goals.

The value and general feasibility of qualitative synthesis has been demonstrated, although the development of methods for the synthesis of qualitative research is at present at an early stage and there is a need for further methodological work. Specific issues requiring clarification and/or further research include:

1. A need for an agreed terminology, as the rapid development of new approaches in the field has meant that key terms such as meta-synthesis and meta-ethnography are employed to refer to differing approaches and there is no single classification of methods of qualitative synthesis.

2. Qualitative syntheses have tended to focus on research adopting thematic and phenomenological approaches, and there is a need for further work that synthesises research based on other theoretical traditions and methods.

3. Further experience of undertaking meta-ethnography is required to provide guidance regarding specific aspects of this process. In terms of a qualitative interpretive approach, this includes questions about the advantages of defining phenomena of interest a priori compared with a broader approach to the field; ways of assessing the worth of qualitative synthesis; and the value of approaching and feeding back interpretations to the original researchers. Other questions that are particularly applicable to a more proceduralist approach relate to forms of purposive sampling, the number of studies that are manageable, and ways of reducing context stripping and the decontextualisation of studies.

4. A wider question that requires a comparative methodology is the investigation of how different methods of synthesis influence the outcome of the synthesis in terms of its focus and emphasis, which in turn will help to clarify the applications and contributions of different methods of qualitative synthesis.

5. Further developmental work is required on how to take account of relevant quantitative syntheses while retaining an interpretative approach to the synthesis of qualitative research.

Objectives

1. To conduct, using the meta ethnographic method, systematic appraisal and synthesis of quantitative research studies in 2 areas:

   1. patient experiences of living with rheumatoid arthritis and

   2. lay beliefs about medicine taking in chronic disease.

2. To test our modified version of the CASP criteria for appraising qualitative research for their appropriateness, ease of use and inter reviewer agreement.

3. To evaluate the meta ethnographic method of synthesis including its reproducibility.

4. To complete our review of the methods available for appraising and synthesising qualitative research.

5. To document the effectiveness of the different elements of the search strategy in identifying relevant qualitative research studies.

Design

Systematic appraisal and synthesis of qualitative research.

Search Strategy

The strategy will include electronic searches of specialist databases and databases of grey literature, hand searching of key journals and collections of qualitative studies published in book form, contacting other qualitative health researchers in relevant areas, searching reference lists of studies retrieved.

Review strategy

The study will test our modified version of the CASP criteria for appraising qualitative research for their appropriateness, ease of use and inter reviewer agreement. The two syntheses will be undertaken using the meta ethnographic method which will be evaluated overall and for its reproducibility.

Expected output of research

1. Report containing:

   1. a review of methods available for synthesis,

   2. results of the summative evaluation of our modified CASP criteria for appraising qualitative research,

   3. two qualitative research syntheses,

   4. an evaluation of the syntheses and a discussion of the contribution qualitative research synthesis could make to HTA.

2. Two papers on each of the syntheses and a methodology paper concerned with the processes of appraisal and synthesis of qualitative research to be published in peer reviewed journals.

3. Presentations of the synthesis findings to different audiences.

Expertise in team

Every member of the research team is a senior social scientist and an experienced qualitative researcher. The research team will be supported by a steering group whose multi-disciplinary membership is drawn from cognate subject areas including public health, epidemiology, anthropology, health economics and rheumatology.

Prioritisation Group

1. Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

2. Professor Imti Choonara, Professor in Child Health, Academic Division of Child Health, University of Nottingham

   Chair – Pharmaceuticals Panel

3. Dr Bob Coates, Consultant Advisor – Disease Prevention Panel

4. Dr Andrew Cook, Consultant Advisor – Intervention Procedures Panel

5. Dr Peter Davidson, Director of NETSCC, Health Technology Assessment

6. Dr Nick Hicks, Consultant Adviser – Diagnostic Technologies and Screening Panel, Consultant Advisor–Psychological and Community Therapies Panel

7. Ms Susan Hird, Consultant Advisor, External Devices and Physical Therapies Panel

8. Professor Sallie Lamb, Director, Warwick Clinical Trials Unit, Warwick Medical School, University of Warwick

   Chair – HTA Clinical Evaluation and Trials Board

9. Professor Jonathan Michaels, Professor of Vascular Surgery, Sheffield Vascular Institute, University of Sheffield

   Chair – Interventional Procedures Panel

10. Professor Ruairidh Milne, Director – External Relations

11. Dr John Pounsford, Consultant Physician, Directorate of Medical Services, North Bristol NHS Trust

    Chair – External Devices and Physical Therapies Panel

12. Dr Vaughan Thomas, Consultant Advisor – Pharmaceuticals Panel, Clinical

    Lead – Clinical Evaluation Trials Prioritisation Group

13. Professor Margaret Thorogood, Professor of Epidemiology, Health Sciences Research Institute, University of Warwick

    Chair – Disease Prevention Panel

14. Professor Lindsay Turnbull, Professor of Radiology, Centre for the MR Investigations, University of Hull

    Chair – Diagnostic Technologies and Screening Panel

15. Professor Scott Weich, Professor of Psychiatry, Health Sciences Research Institute, University of Warwick

    Chair – Psychological and Community Therapies Panel

16. Professor Hywel Williams, Director of Nottingham Clinical Trials Unit, Centre of Evidence-Based Dermatology, University of Nottingham

    Chair – HTA Commissioning Board

    Deputy HTA Programme Director

HTA Commissioning Board

1. Professor of Dermato-Epidemiology, Centre of Evidence-Based Dermatology, University of Nottingham

2. Department of Public Health and Epidemiology, University of Birmingham

3. Professor of Clinical Pharmacology, Director, NIHR HTA programme, University of Liverpool

1. Professor Ann Ashburn, Professor of Rehabilitation and Head of Research, Southampton General Hospital

2. Professor Peter Brocklehurst, Professor of Women’s Health, Institute for Women’s Health, University College London

3. Professor Jenny Donovan, Professor of Social Medicine, University of Bristol

4. Professor Jonathan Green, Professor and Acting Head of Department, Child and Adolescent Psychiatry, University of Manchester Medical School

5. Professor John W Gregory, Professor in Paediatric Endocrinology, Department of Child Health, Wales School of Medicine, Cardiff University

6. Professor Steve Halligan, Professor of Gastrointestinal Radiology, University College Hospital, London

7. Professor Freddie Hamdy, Professor of Urology, Head of Nuffield Department of Surgery, University of Oxford

8. Professor Allan House, Professor of Liaison Psychiatry, University of Leeds

9. Dr Martin J Landray, Reader in Epidemiology, Honorary Consultant Physician, Clinical Trial Service Unit, University of Oxford

10. Professor Stephen Morris, Professor of Health Economics, University College London, Research Department of Epidemiology and Public Health, University College London

11. Professor Irwin Nazareth, Professor of Primary Care and Head of Department, Department of Primary Care and Population Sciences, University College London

12. Professor E Andrea Nelson, Professor of Wound Healing and Director of Research, School of Healthcare, University of Leeds

13. Professor John David Norrie, Chair in Clinical Trials and Biostatistics, Robertson Centre for Biostatistics, University of Glasgow

14. Dr Rafael Perera, Lecturer in Medical Statisitics, Department of Primary Health Care, University of Oxford

15. Professor Barney Reeves, Professorial Research Fellow in Health Services Research, Department of Clinical Science, University of Bristol

16. Professor Martin Underwood, Professor of Primary Care Research, Warwick Medical School, University of Warwick

17. Professor Marion Walker, Professor in Stroke Rehabilitation, Associate Director UK Stroke Research Network, University of Nottingham

18. Dr Duncan Young, Senior Clinical Lecturer and Consultant, Nuffield Department of Anaesthetics, University of Oxford

1. Dr Tom Foulks, Medical Research Council

2. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

HTA Clinical Evaluation and Trials Board

1. Director, Warwick Clinical Trials Unit, Warwick Medical School, University of Warwick and Professor of Rehabilitation, Nuffield Department of Orthopaedic, Rheumatology and Musculoskeletal Sciences, University of Oxford

2. Professor of the Psychology of Health Care, Leeds Institute of Health Sciences, University of Leeds

3. Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

1. Professor Keith Abrams, Professor of Medical Statistics, Department of Health Sciences, University of Leicester

2. Professor Martin Bland, Professor of Health Statistics, Department of Health Sciences, University of York

3. Professor Jane Blazeby, Professor of Surgery and Consultant Upper GI Surgeon, Department of Social Medicine, University of Bristol

4. Professor Julia M Brown, Director, Clinical Trials Research Unit, University of Leeds

5. Professor Alistair Burns, Professor of Old Age Psychiatry, Psychiatry Research Group, School of Community-Based Medicine, The University of Manchester & National Clinical Director for Dementia, Department of Health

6. Dr Jennifer Burr, Director, Centre for Healthcare Randomised trials (CHART), University of Aberdeen

7. Professor Linda Davies, Professor of Health Economics, Health Sciences Research Group, University of Manchester

8. Professor Simon Gilbody, Prof of Psych Medicine and Health Services Research, Department of Health Sciences, University of York

9. Professor Steven Goodacre, Professor and Consultant in Emergency Medicine, School of Health and Related Research, University of Sheffield

10. Professor Dyfrig Hughes, Professor of Pharmacoeconomics, Centre for Economics and Policy in Health, Institute of Medical and Social Care Research, Bangor University

11. Professor Paul Jones, Professor of Respiratory Medicine, Department of Cardiac and Vascular Science, St George‘s Hospital Medical School, University of London

12. Professor Khalid Khan, Professor of Women’s Health and Clinical Epidemiology, Barts and the London School of Medicine, Queen Mary, University of London

13. Professor Richard J McManus, Professor of Primary Care Cardiovascular Research, Primary Care Clinical Sciences Building, University of Birmingham

14. Professor Helen Rodgers, Professor of Stroke Care, Institute for Ageing and Health, Newcastle University

15. Professor Ken Stein, Professor of Public Health, Peninsula Technology Assessment Group, Peninsula College of Medicine and Dentistry, Universities of Exeter and Plymouth

16. Professor Jonathan Sterne, Professor of Medical Statistics and Epidemiology, Department of Social Medicine, University of Bristol

17. Mr Andy Vail, Senior Lecturer, Health Sciences Research Group, University of Manchester

18. Professor Clare Wilkinson, Professor of General Practice and Director of Research North Wales Clinical School, Department of Primary Care and Public Health, Cardiff University

19. Dr Ian B Wilkinson, Senior Lecturer and Honorary Consultant, Clinical Pharmacology Unit, Department of Medicine, University of Cambridge

1. Ms Kate Law, Director of Clinical Trials, Cancer Research UK

2. Dr Morven Roberts, Clinical Trials Manager, Health Services and Public Health Services Board, Medical Research Council

Diagnostic Technologies and Screening Panel

1. Scientific Director of the Centre for Magnetic Resonance Investigations and YCR Professor of Radiology, Hull Royal Infirmary

2. Professor Judith E Adams, Consultant Radiologist, Manchester Royal Infirmary, Central Manchester & Manchester Children’s University Hospitals NHS Trust, and Professor of Diagnostic Radiology, University of Manchester

3. Mr Angus S Arunkalaivanan, Honorary Senior Lecturer, University of Birmingham and Consultant Urogynaecologist and Obstetrician, City Hospital, Birmingham

4. Dr Diana Baralle, Consultant and Senior Lecturer in Clinical Genetics, University of Southampton

5. Dr Stephanie Dancer, Consultant Microbiologist, Hairmyres Hospital, East Kilbride

6. Dr Diane Eccles, Professor of Cancer Genetics, Wessex Clinical Genetics Service, Princess Anne Hospital

7. Dr Trevor Friedman, Consultant Liason Psychiatrist, Brandon Unit, Leicester General Hospital

8. Dr Ron Gray, Consultant, National Perinatal Epidemiology Unit, Institute of Health Sciences, University of Oxford

9. Professor Paul D Griffiths, Professor of Radiology, Academic Unit of Radiology, University of Sheffield

10. Mr Martin Hooper, Public contributor

11. Professor Anthony Robert Kendrick, Associate Dean for Clinical Research and Professor of Primary Medical Care, University of Southampton

12. Dr Nicola Lennard, Senior Medical Officer, MHRA

13. Dr Anne Mackie, Director of Programmes, UK National Screening Committee, London

14. Mr David Mathew, Public contributor

15. Dr Michael Millar, Consultant Senior Lecturer in Microbiology, Department of Pathology & Microbiology, Barts and The London NHS Trust, Royal London Hospital

16. Mrs Una Rennard, Public contributor

17. Dr Stuart Smellie, Consultant in Clinical Pathology, Bishop Auckland General Hospital

18. Ms Jane Smith, Consultant Ultrasound Practitioner, Leeds Teaching Hospital NHS Trust, Leeds

19. Dr Allison Streetly, Programme Director, NHS Sickle Cell and Thalassaemia Screening Programme, King’s College School of Medicine

20. Dr Matthew Thompson, Senior Clinical Scientist and GP, Department of Primary Health Care, University of Oxford

21. Dr Alan J Williams, Consultant Physician, General and Respiratory Medicine, The Royal Bournemouth Hospital

1. Dr Tim Elliott, Team Leader, Cancer Screening, Department of Health

2. Dr Joanna Jenkinson, Board Secretary, Neurosciences and Mental Health Board (NMHB), Medical Research Council

3. Professor Julietta Patrick, Director, NHS Cancer Screening Programme, Sheffield

4. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

5. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

6. Dr Ursula Wells, Principal Research Officer, Policy Research Programme, Department of Health

Disease Prevention Panel

1. Professor of Epidemiology, University of Warwick Medical School, Coventry

2. Dr Robert Cook, Clinical Programmes Director, Bazian Ltd, London

3. Dr Colin Greaves, Senior Research Fellow, Peninsula Medical School (Primary Care)

4. Mr Michael Head, Public contributor

5. Professor Cathy Jackson, Professor of Primary Care Medicine, Bute Medical School, University of St Andrews

6. Dr Russell Jago, Senior Lecturer in Exercise, Nutrition and Health, Centre for Sport, Exercise and Health, University of Bristol

7. Dr Julie Mytton, Consultant in Child Public Health, NHS Bristol

8. Professor Irwin Nazareth, Professor of Primary Care and Director, Department of Primary Care and Population Sciences, University College London

9. Dr Richard Richards, Assistant Director of Public Health, Derbyshire County Primary Care Trust

10. Professor Ian Roberts, Professor of Epidemiology and Public Health, London School of Hygiene & Tropical Medicine

11. Dr Kenneth Robertson, Consultant Paediatrician, Royal Hospital for Sick Children, Glasgow

12. Dr Catherine Swann, Associate Director, Centre for Public Health Excellence, NICE

13. Mrs Jean Thurston, Public contributor

14. Professor David Weller, Head, School of Clinical Science and Community Health, University of Edinburgh

1. Ms Christine McGuire, Research & Development, Department of Health

2. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

3. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

External Devices and Physical Therapies Panel

1. Consultant Physician North Bristol NHS Trust

2. Reader in Wound Healing and Director of Research, University of Leeds

3. Professor Bipin Bhakta, Charterhouse Professor in Rehabilitation Medicine, University of Leeds

4. Mrs Penny Calder, Public contributor

5. Dr Dawn Carnes, Senior Research Fellow, Barts and the London School of Medicine and Dentistry

6. Dr Emma Clark, Clinician Scientist Fellow & Cons. Rheumatologist, University of Bristol

7. Mrs Anthea De Barton-Watson, Public contributor

8. Professor Nadine Foster, Professor of Musculoskeletal Health in Primary Care Arthritis Research, Keele University

9. Dr Shaheen Hamdy, Clinical Senior Lecturer and Consultant Physician, University of Manchester

10. Professor Christine Norton, Professor of Clinical Nursing Innovation, Bucks New University and Imperial College Healthcare NHS Trust

11. Dr Lorraine Pinnigton, Associate Professor in Rehabilitation, University of Nottingham

12. Dr Kate Radford, Senior Lecturer (Research), University of Central Lancashire

13. Mr Jim Reece, Public contributor

14. Professor Maria Stokes, Professor of Neuromusculoskeletal Rehabilitation, University of Southampton

15. Dr Pippa Tyrrell, Senior Lecturer/Consultant, Salford Royal Foundation Hospitals’ Trust and University of Manchester

16. Dr Nefyn Williams, Clinical Senior Lecturer, Cardiff University

1. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

2. Dr Morven Roberts, Clinical Trials Manager, Health Services and Public Health Services Board, Medical Research Council

3. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

4. Dr Ursula Wells, Principal Research Officer, Policy Research Programme, Department of Health

Interventional Procedures Panel

1. Professor of Vascular Surgery, University of Sheffield

2. Consultant Colorectal Surgeon, Bristol Royal Infirmary

3. Mrs Isabel Boyer, Public contributor

4. Mr Sankaran Chandra Sekharan, Consultant Surgeon, Breast Surgery, Colchester Hospital University NHS Foundation Trust

5. Professor Nicholas Clarke, Consultant Orthopaedic Surgeon, Southampton University Hospitals NHS Trust

6. Ms Leonie Cooke, Public contributor

7. Mr Seumas Eckford, Consultant in Obstetrics & Gynaecology, North Devon District Hospital

8. Professor Sam Eljamel, Consultant Neurosurgeon, Ninewells Hospital and Medical School, Dundee

9. Dr Adele Fielding, Senior Lecturer and Honorary Consultant in Haematology, University College London Medical School

10. Dr Matthew Hatton, Consultant in Clinical Oncology, Sheffield Teaching Hospital Foundation Trust

11. Dr John Holden, General Practitioner, Garswood Surgery, Wigan

12. Dr Fiona Lecky, Senior Lecturer/Honorary Consultant in Emergency Medicine, University of Manchester/Salford Royal Hospitals NHS Foundation Trust

13. Dr Nadim Malik, Consultant Cardiologist/Honorary Lecturer, University of Manchester

14. Mr Hisham Mehanna, Consultant & Honorary Associate Professor, University Hospitals Coventry & Warwickshire NHS Trust

15. Dr Jane Montgomery, Consultant in Anaesthetics and Critical Care, South Devon Healthcare NHS Foundation Trust

16. Professor Jon Moss, Consultant Interventional Radiologist, North Glasgow Hospitals University NHS Trust

17. Dr Simon Padley, Consultant Radiologist, Chelsea & Westminster Hospital

18. Dr Ashish Paul, Medical Director, Bedfordshire PCT

19. Dr Sarah Purdy, Consultant Senior Lecturer, University of Bristol

20. Dr Matthew Wilson, Consultant Anaesthetist, Sheffield Teaching Hospitals NHS Foundation Trust

21. Professor Yit Chiun Yang, Consultant Ophthalmologist, Royal Wolverhampton Hospitals NHS Trust

1. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

2. Dr Morven Roberts, Clinical Trials Manager, Health Services and Public Health Services Board, Medical Research Council

3. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

4. Dr Ursula Wells, Principal Research Officer, Policy Research Programme, Department of Health

Pharmaceuticals Panel

1. Professor in Child Health, University of Nottingham

2. Senior Lecturer in Clinical Pharmacology, University of East Anglia

3. Dr Martin Ashton-Key, Medical Advisor, National Commissioning Group, NHS London

4. Dr Peter Elton, Director of Public Health, Bury Primary Care Trust

5. Dr Ben Goldacre, Research Fellow, Division of Psychological Medicine and Psychiatry, King’s College London

6. Dr James Gray, Consultant Microbiologist, Department of Microbiology, Birmingham Children’s Hospital NHS Foundation Trust

7. Dr Jurjees Hasan, Consultant in Medical Oncology, The Christie, Manchester

8. Dr Carl Heneghan, Deputy Director Centre for Evidence-Based Medicine and Clinical Lecturer, Department of Primary Health Care, University of Oxford

9. Dr Dyfrig Hughes, Reader in Pharmacoeconomics and Deputy Director, Centre for Economics and Policy in Health, IMSCaR, Bangor University

10. Dr Maria Kouimtzi, Pharmacy and Informatics Director, Global Clinical Solutions, Wiley-Blackwell

11. Professor Femi Oyebode, Consultant Psychiatrist and Head of Department, University of Birmingham

12. Dr Andrew Prentice, Senior Lecturer and Consultant Obstetrician and Gynaecologist, The Rosie Hospital, University of Cambridge

13. Ms Amanda Roberts, Public contributor

14. Dr Gillian Shepherd, Director, Health and Clinical Excellence, Merck Serono Ltd

15. Mrs Katrina Simister, Assistant Director New Medicines, National Prescribing Centre, Liverpool

16. Professor Donald Singer, Professor of Clinical Pharmacology and Therapeutics, Clinical Sciences Research Institute, CSB, University of Warwick Medical School

17. Mr David Symes, Public contributor

18. Dr Arnold Zermansky, General Practitioner, Senior Research Fellow, Pharmacy Practice and Medicines Management Group, Leeds University

1. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

2. Mr Simon Reeve, Head of Clinical and Cost-Effectiveness, Medicines, Pharmacy and Industry Group, Department of Health

3. Dr Heike Weber, Programme Manager, Medical Research Council

4. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

5. Dr Ursula Wells, Principal Research Officer, Policy Research Programme, Department of Health

Psychological and Community Therapies Panel

1. Professor of Psychiatry, University of Warwick, Coventry

2. Consultant & University Lecturer in Psychiatry, University of Cambridge

3. Professor Jane Barlow, Professor of Public Health in the Early Years, Health Sciences Research Institute, Warwick Medical School

4. Dr Sabyasachi Bhaumik, Consultant Psychiatrist, Leicestershire Partnership NHS Trust

5. Mrs Val Carlill, Public contributor

6. Dr Steve Cunningham, Consultant Respiratory Paediatrician, Lothian Health Board

7. Dr Anne Hesketh, Senior Clinical Lecturer in Speech and Language Therapy, University of Manchester

8. Dr Peter Langdon, Senior Clinical Lecturer, School of Medicine, Health Policy and Practice, University of East Anglia

9. Dr Yann Lefeuvre, GP Partner, Burrage Road Surgery, London

10. Dr Jeremy J Murphy, Consultant Physician and Cardiologist, County Durham and Darlington Foundation Trust

11. Dr Richard Neal, Clinical Senior Lecturer in General Practice, Cardiff University

12. Mr John Needham, Public contributor

13. Ms Mary Nettle, Mental Health User Consultant

14. Professor John Potter, Professor of Ageing and Stroke Medicine, University of East Anglia

15. Dr Greta Rait, Senior Clinical Lecturer and General Practitioner, University College London

16. Dr Paul Ramchandani, Senior Research Fellow/Cons. Child Psychiatrist, University of Oxford

17. Dr Karen Roberts, Nurse/Consultant, Dunston Hill Hospital, Tyne and Wear

18. Dr Karim Saad, Consultant in Old Age Psychiatry, Coventry and Warwickshire Partnership Trust

19. Dr Lesley Stockton, Lecturer, School of Health Sciences, University of Liverpool

20. Dr Simon Wright, GP Partner, Walkden Medical Centre, Manchester

1. Dr Kay Pattison, Senior NIHR Programme Manager, Department of Health

2. Dr Morven Roberts, Clinical Trials Manager, Health Services and Public Health Services Board, Medical Research Council

3. Professor Tom Walley, CBE, Director, NIHR HTA programme, Professor of Clinical Pharmacology, University of Liverpool

4. Dr Ursula Wells, Principal Research Officer, Policy Research Programme, Department of Health

Expert Advisory Network

1. Professor Douglas Altman, Professor of Statistics in Medicine, Centre for Statistics in Medicine, University of Oxford

2. Professor John Bond, Professor of Social Gerontology & Health Services Research, University of Newcastle upon Tyne

3. Professor Andrew Bradbury, Professor of Vascular Surgery, Solihull Hospital, Birmingham

4. Mr Shaun Brogan, Chief Executive, Ridgeway Primary Care Group, Aylesbury

5. Mrs Stella Burnside OBE, Chief Executive, Regulation and Improvement Authority, Belfast

6. Ms Tracy Bury, Project Manager, World Confederation of Physical Therapy, London

7. Professor Iain T Cameron, Professor of Obstetrics and Gynaecology and Head of the School of Medicine, University of Southampton

8. Professor Bruce Campbell, Consultant Vascular & General Surgeon, Royal Devon & Exeter Hospital, Wonford

9. Dr Christine Clark, Medical Writer and Consultant Pharmacist, Rossendale

10. Professor Collette Clifford, Professor of Nursing and Head of Research, The Medical School, University of Birmingham

11. Professor Barry Cookson, Director, Laboratory of Hospital Infection, Public Health Laboratory Service, London

12. Dr Carl Counsell, Clinical Senior Lecturer in Neurology, University of Aberdeen

13. Professor Howard Cuckle, Professor of Reproductive Epidemiology, Department of Paediatrics, Obstetrics & Gynaecology, University of Leeds

14. Professor Carol Dezateux, Professor of Paediatric Epidemiology, Institute of Child Health, London

15. Mr John Dunning, Consultant Cardiothoracic Surgeon, Papworth Hospital NHS Trust, Cambridge

16. Mr Jonothan Earnshaw, Consultant Vascular Surgeon, Gloucestershire Royal Hospital, Gloucester

17. Professor Martin Eccles, Professor of Clinical Effectiveness, Centre for Health Services Research, University of Newcastle upon Tyne

18. Professor Pam Enderby, Dean of Faculty of Medicine, Institute of General Practice and Primary Care, University of Sheffield

19. Professor Gene Feder, Professor of Primary Care Research & Development, Centre for Health Sciences, Barts and The London School of Medicine and Dentistry

20. Mr Leonard R Fenwick, Chief Executive, Freeman Hospital, Newcastle upon Tyne

21. Mrs Gillian Fletcher, Antenatal Teacher and Tutor and President, National Childbirth Trust, Henfield

22. Professor Jayne Franklyn, Professor of Medicine, University of Birmingham

23. Mr Tam Fry, Honorary Chairman, Child Growth Foundation, London

24. Professor Fiona Gilbert, Consultant Radiologist and NCRN Member, University of Aberdeen

25. Professor Paul Gregg, Professor of Orthopaedic Surgical Science, South Tees Hospital NHS Trust

26. Bec Hanley, Co-director, TwoCan Associates, West Sussex

27. Dr Maryann L Hardy, Senior Lecturer, University of Bradford

28. Mrs Sharon Hart, Healthcare Management Consultant, Reading

29. Professor Robert E Hawkins, CRC Professor and Director of Medical Oncology, Christie CRC Research Centre, Christie Hospital NHS Trust, Manchester

30. Professor Richard Hobbs, Head of Department of Primary Care & General Practice, University of Birmingham

31. Professor Alan Horwich, Dean and Section Chairman, The Institute of Cancer Research, London

32. Professor Allen Hutchinson, Director of Public Health and Deputy Dean of ScHARR, University of Sheffield

33. Professor Peter Jones, Professor of Psychiatry, University of Cambridge, Cambridge

34. Professor Stan Kaye, Cancer Research UK Professor of Medical Oncology, Royal Marsden Hospital and Institute of Cancer Research, Surrey

35. Dr Duncan Keeley, General Practitioner (Dr Burch & Ptnrs), The Health Centre, Thame

36. Dr Donna Lamping, Research Degrees Programme Director and Reader in Psychology, Health Services Research Unit, London School of Hygiene and Tropical Medicine, London

37. Professor James Lindesay, Professor of Psychiatry for the Elderly, University of Leicester

38. Professor Julian Little, Professor of Human Genome Epidemiology, University of Ottawa

39. Professor Alistaire McGuire, Professor of Health Economics, London School of Economics

40. Professor Neill McIntosh, Edward Clark Professor of Child Life and Health, University of Edinburgh

41. Professor Rajan Madhok, Consultant in Public Health, South Manchester Primary Care Trust

42. Professor Sir Alexander Markham, Director, Molecular Medicine Unit, St James’s University Hospital, Leeds

43. Dr Peter Moore, Freelance Science Writer, Ashtead

44. Dr Andrew Mortimore, Public Health Director, Southampton City Primary Care Trust

45. Dr Sue Moss, Associate Director, Cancer Screening Evaluation Unit, Institute of Cancer Research, Sutton

46. Professor Miranda Mugford, Professor of Health Economics and Group Co-ordinator, University of East Anglia

47. Professor Jim Neilson, Head of School of Reproductive & Developmental Medicine and Professor of Obstetrics and Gynaecology, University of Liverpool

48. Mrs Julietta Patnick, Director, NHS Cancer Screening Programmes, Sheffield

49. Professor Robert Peveler, Professor of Liaison Psychiatry, Royal South Hants Hospital, Southampton

50. Professor Chris Price, Director of Clinical Research, Bayer Diagnostics Europe, Stoke Poges

51. Professor William Rosenberg, Professor of Hepatology and Consultant Physician, University of Southampton

52. Professor Peter Sandercock, Professor of Medical Neurology, Department of Clinical Neurosciences, University of Edinburgh

53. Dr Philip Shackley, Senior Lecturer in Health Economics, Sheffield Vascular Institute, University of Sheffield

54. Dr Eamonn Sheridan, Consultant in Clinical Genetics, St James’s University Hospital, Leeds

55. Dr Margaret Somerville, Director of Public Health Learning, Peninsula Medical School, University of Plymouth

56. Professor Sarah Stewart-Brown, Professor of Public Health, Division of Health in the Community, University of Warwick, Coventry

57. Dr Nick Summerton, GP Appraiser and Codirector, Research Network, Yorkshire Clinical Consultant, Primary Care and Public Health, University of Oxford

58. Professor Ala Szczepura, Professor of Health Service Research, Centre for Health Services Studies, University of Warwick, Coventry

59. Dr Ross Taylor, Senior Lecturer, University of Aberdeen

60. Dr Richard Tiner, Medical Director, Medical Department, Association of the British Pharmaceutical Industry

61. Mrs Joan Webster, Consumer Member, Southern Derbyshire Community Health Council

62. Professor Martin Whittle, Clinical Co-director, National Co-ordinating Centre for Women’s and Children’s Health, Lymington