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

BACKGROUND: Methods for reviewing and synthesising findings from quantitative research studies in health care are well established. Although there is recognition of the need for qualitative research to be brought into the evidence base, there is no consensus about how this should be done and the methods for synthesising qualitative research are at a relatively early stage of development. OBJECTIVE: To evaluate meta-ethnography as a method for synthesising qualitative research studies in health and health care. METHODS: Two full syntheses of qualitative research studies were conducted between April 2002 and September 2004 using meta-ethnography: (1) studies of medicine-taking and (2) studies exploring patients' experiences of living with rheumatoid arthritis. Potentially relevant studies identified in multiple literature searches conducted in July and August 2002 (electronically and by hand) were appraised using a modified version of the Critical Appraisal Skills Programme questions for understanding qualitative research. Candidate papers were excluded on grounds of lack of relevance to the aims of the synthesis or because the work failed to employ qualitative methods of data collection and analysis. RESULTS: Thirty-eight studies were entered into the medicine-taking synthesis, one of which did not contribute to the final synthesis. The synthesis revealed a general caution about taking medicine, and that the practice of lay testing of medicines was widespread. People were found to take their medicine passively or actively or to reject it outright. Some, in particular clinical areas, were coerced into taking it. Those who actively accepted their medicine often modified the regimen prescribed by a doctor, without the doctor's knowledge. The synthesis concluded that people often do not take their medicines as prescribed because of concern about the medicines themselves. 'Resistance' emerged from the synthesis as a concept that best encapsulated the lay response to prescribed medicines. It was suggested that a policy focus should be on the problems associated with the medicines themselves and on evaluating the effectiveness of alternative treatments that some people use in preference to prescribed medicines. The synthesis of studies of lay experiences of living with rheumatoid arthritis began with 29 papers. Four could not be synthesised, leaving 25 papers (describing 22 studies) contributing to the final synthesis. Most of the papers were concerned with the everyday experience of living with rheumatoid arthritis. This synthesis did not produce significant new insights, probably because the early papers in the area were substantial and theoretically rich, and later papers were mostly confirmatory. In both topic areas, only a minority of the studies included in the syntheses were found to have referenced each other, suggesting that unnecessary replication had occurred. LIMITATIONS: We only evaluated meta-ethnography as a method for synthesising qualitative research, but there are other methods being employed. Further research is required to investigate how different methods of qualitative synthesis influence the outcome of the synthesis. CONCLUSIONS: Meta-ethnography is an effective method for synthesising qualitative research. The process of reciprocally translating the findings from each individual study into those from all the other studies in the synthesis, if applied rigorously, ensures that qualitative data can be combined. Following this essential process, the synthesis can then be expressed as a 'line of argument' that can be presented as text and in summary tables and diagrams or models. Meta-ethnography can produce significant new insights, but not all meta-ethnographic syntheses do so. Instead, some will identify fields in which saturation has been reached and in which no theoretical development has taken place for some time. Both outcomes are helpful in either moving research forward or avoiding wasted resources. Meta-ethnography is a highly interpretative method requiring considerable immersion in the individual studies to achieve a synthesis. It places substantial demands upon the synthesiser and requires a high degree of qualitative research skill. Meta-ethnography has great potential as a method of synthesis in qualitative health technology assessment but it is still evolving and cannot, at present, be regarded as a standardised approach capable of application in a routinised way. FUNDING: Funding for this study was provided by the Health Technology Assessment programme of the National Institute for Health Research.

Time to talk? Patient experiences of waiting for clinical management of knee injuries.

BACKGROUND: People with knee problems face long waits for elective surgery in many parts of the world. However, there is little evidence about the impact of delays in such treatment, especially for patients with mechanical knee injuries. OBJECTIVE: To conduct a detailed exploration of patient experiences of waits for specialist diagnosis and surgery for knee injuries at one UK centre. RESEARCH DESIGN: In-depth qualitative paired interviews with a range of patients at baseline and 6 months later. SUBJECTS: Patients awaiting imminent therapeutic arthroscopy of the knee (n = 20) or recently referred from primary care for specialist opinion or imaging for a knee injury (n = 19). Sample stratified to maximise variation by gender and age. RESULTS: 36 patients completed both interviews. Four topic areas were identified. First, problems in the healthcare system were highlighted, including a lack of adequate information, which made it difficult for patients to make decisions about their lives. Second, patients experienced a social and psychological cost of waiting. Third, patients varied in their ability to cope and demonstrated both passive and proactive coping strategies. Fourth, patients described the management effectiveness of clinicians and their ability or otherwise to provide support. CONCLUSIONS: A detailed qualitative approach has identified broad physical and psycho-social consequences for patients with knee injuries experiencing delays in clinical management. An overarching theme was the important potential of both systemic and interpersonal communication to improve patient well-being. A managed care pathway which enhanced information provision may provide immediate opportunities for improving patient well-being.

Public attitudes towards the use of primary care patient record data in medical research without consent: a qualitative study.

OBJECTIVES: Recent legislative changes within the United Kingdom have stimulated professional debate about access to patient data within research. However, there is currently little awareness of public views about such research. The authors sought to explore attitudes of the public, and their lay representatives, towards the use of primary care medical record data for research when patient consent was not being sought. METHODS: 49 members of the public and four non-medical members of local community health councils in South Wales, UK gave their views on the value and acceptability of three current research scenarios, each describing access to data without patient consent. RESULTS: Among focus group participants, awareness of research in primary care was low, and the appropriateness of general practitioners as researchers was questioned. There was general support for research but also concerns expressed about data collection without consent. These included lack of respect and patient control over the process. Unauthorised access to data by external agencies was a common fear. Current data collection practices, including population based disease registers elicited much anxiety. The key informants were equally critical of the scenarios and generally less accepting. CONCLUSIONS: This exploratory study has highlighted a number of areas of public concern when medical records are accessed for research without patient consent. Public acceptability regarding the use of medical records in research cannot simply be assumed. Further work is required to determine how widespread such views are and to inform those advising on confidentiality issues.

Welsh women's comments about breast problems and the care given: a qualitative study in the community.

The aim of this study was to undertake a systematic analysis, using qualitative software, of the free text comments from a postal survey exploring women's experiences of breast symptoms, their expectations of treatment, knowledge of breast cancer risk factors and perceptions of risk, in 34 group general practices in South Wales. The 959 women who returned the questionnaire, out of 1126 (response rate 85%), comprised 497 women who consulted their general practitioner (GP) with a new breast symptom during the baseline data collection period and 462 controls who had not. When the survey was conducted the researchers did not know whether these women had cancer or had previously been treated for it. One-third (33.1%) of those returning the survey (n = 318) wrote comments. Compared to the rest, they were significantly more likely to have consulted their GP for a new breast symptom and to have stayed on at school and/or gone on to more education or training. The majority wrote about their own experience of breast symptoms and/or the care received in primary and secondary settings. The general tone was factual and when evaluation took place positive comments were more frequent than negative ones. Nothing suggested that the respondents had been upset or made more anxious by the preceding questions on such a potentially sensitive topic. Free text comments gathered in surveys can provide valuable data if systematically analysed. Doctors, particularly GPs, can be reassured that more women in this community sample who expressed an opinion on care were positive. The negative comments, however, highlight issues that still need to be addressed in therapeutic relationships.

The development of a new site-specific measure of quality of life for breast problems: the Cardiff breast scales.

The management of women presenting to primary care with symptoms of breast disease is of increasing interest given recent organisational changes aimed at improving accuracy and speed of referrals. As part of a randomised controlled trial, 1063 women were recruited following a primary care consultation for a variety of breast-related problems. In the absence of a suitable outcome measure for such women, a site-specific instrument was developed to complement a generic quality of life scale (SF-36). Items were generated using key informant interviews with health professionals. Draft scale items were piloted using a postal questionnaire and subsequent patient debrief interviews. A sample of respondents were also sent the same questionnaire I month later to assess test-retest reliability. Across the whole sample (n = 848), three factors were identified: 'general well-being', 'concerns' and 'relationships'. These factors accounted for 60% of total variance. Evidence of scale validity, reliability and responsiveness are reported for this new outcome measure for use in women presenting with breast problems.

Lay attitudes to professional consultations for common mental disorder: a sociological perspective.

How, why, and under what kinds of circumstances lay people consult for symptoms of emotional distress are topics that have commanded various degrees of attention from secondary and primary care professionals. We argue below that many of the responses made by such professionals to these issues carry within them a set of very important assumptions about how members of the lay public view psychiatric symptoms. Whether such assumptions are justified by the evidence is, however, a matter of some debate. In what follows we draw on some recent, sociologically informed research on lay attitudes to emotional distress so as to highlight the debates and to suggest some ways in which they might be resolved.

Presenting risk information--a review of the effects of "framing" and other manipulations on patient outcomes.

Discussing risks and benefits of treatments or care options is becoming an increasingly important part of modern health care. This paper reviews the literature about manipulations of risk and benefit information in the clinical setting. There is a paucity of evidence in this field, particularly when examining specific manipulations. Only three categories of manipulation had three or more studies. The available evidence shows that the way information is presented can have significant effects on decisions made. The largest effects are evident when relative risk information is presented, as compared with absolute risk data. In addition, "loss framing" is more effective in influencing screening uptake behaviors than "gain framing" (odds ratio 1.18 [95% confidence interval 1.01-1.38]). There is also a pattern of evidence from studies comparing simpler with more complex information, more data with less, and those comparing numerical with verbal descriptions of risks. These studies suggest that providing more information, and which is more understandable to the patient, is associated with improved patient knowledge and a greater wariness to take treatments or participate in trials. These findings can contribute to efforts to improve communication between professionals and patients.

Qualitative study of views of health professionals and patients on guided self management plans for asthma.

OBJECTIVES: To explore the views held by general practitioners, practice nurses, and patients about the role of guided self management plans in asthma care. DESIGN: Qualitative study using nine focus groups that each met on two occasions. SETTING: South Wales. SUBJECTS: 13 asthma nurses, 11 general practitioners (six with an interest in asthma), and 32 patients (13 adults compliant with treatment, 12 non-compliant adults, and seven teenagers). RESULTS: Neither health professionals nor patients were enthusiastic about guided self management plans, and, although for different reasons, almost all participants were ambivalent about their usefulness or relevance. Most professionals opposed their use. Few patients reported sustained use, and most felt that plans were largely irrelevant to them. The attitudes associated with these views reflect the gulf between the professionals' concept of the "responsible asthma patient" and the patients' view. CONCLUSIONS: Attempts to introduce self guided management plans in primary care are unlikely to be successful. A more patient centred, patient negotiated plan is needed for asthma care in the community.

The effectiveness of one-to-one risk communication interventions in health care: a systematic review.

OBJECTIVES: To assess whether risk-communication interventions are associated with changes in patient knowledge, attitudes, and behaviors, and to identify aspects of these interventions that modify these effects. DESIGN: Systematic review. DATA SOURCES: 96 studies from the period 1985-1996 retrieved by electronic searching of eight databases, hand searching of four journals, contacting key authors, and reference list searching. MAIN OUTCOME MEASURES: The effect size of the principal outcome was identified from each study. Outcomes measuring behavioral change were preferred; if these were not available, knowledge, anxiety, or risk perceptions were used, according to the focus of the study. Data were available to calculate the principal effect sizes for 82 of the studies. ANALYSIS: Meta-regression. RESULTS: The methodologic qualities of the studies varied. Nevertheless, risk-communication interventions generally had positive (beneficial) effects. Interventions addressing treatment choices were associated with larger effects than were those in other contexts, such as prevention or screening. Interventions using individual risk estimates were associated with larger effects than were those using more general risk information. Two design variables were identified as effect modifiers: randomized controlled trials were associated with smaller effects than other designs, and dichotomous outcomes were associated with larger effects than continuous outcomes. CONCLUSIONS: Risk communication interventions may be most productive if they include individual risk estimates in the discussion between professional and patient. Patient decisions about treatment appear more amenable to change by these interventions than attendance for screening or modification of risky behavior.

Does experience predict knowledge and behavior with respect to cutaneous melanoma, moles, and sun exposure? Possible outcome measures.

A prospective survey of 3105 adults in 16 randomly selected group practices was conducted to test whether individuals with a personal experience, family history, or other contact with malignant melanoma display more knowledge and appropriate behavior with regard to melanomas and sun exposure than those with no such experience. Although patients with previous contact with the condition were more likely to check for moles, were more aware of the significance of changes in the shape of moles, and were more aware of the necessity for prompt treatment, they were not more knowledgeable about other important signs or more likely to protect themselves from sun exposure. Direct experience or contact with melanoma appeared to have some positive effects on knowledge, but there is no room for complacency for those at increased risk or those treating them. Both primary and secondary prevention for people who are vulnerable could be undertaken in primary care.

Identifying which women will stop breast feeding before three months in primary care: a pragmatic study.

BACKGROUND: In Britain only 42% of women who initiate breast feeding are still breast feeding at four months, despite well documented health benefits. AIM: To explore whether sociodemographic and social support information collected routinely by health visitors at the new birth assessment can help predict which women will give up breast feeding before three months. METHOD: A survey of 279 consecutive births in three general practices in an inner-London borough. Health visitors collected sociodemographic, infant feeding, and social support data at the new birth assessment 10 to 14 days after birth and at an immunisation visit at three to four months after birth. A data collection form was piloted and used by health visitors as part of their routine clinical care. Stepwise logistic regression was performed on 160 women who initiated breast feeding to identify predictors for those who would still be breast feeding at three months. RESULTS: Three variables were found to be significantly associated with breast feeding at three months. Younger women and women with moderate to poor emotional support as assessed by their health visitor were less likely to still be breast feeding at three months. White women who left full-time education at age 16 years or below are least likely to be breast feeding at three months but educational level is not a significant predictor for women from other ethnic backgrounds. CONCLUSION: This pragmatic study illustrates how information collected during routine clinical care by health visitors can help predict which women will give up breast feeding before three months. This could be useful to identify women whose social support needs are not being met and who may benefit from local initiatives. Infant feeding researchers should consider the influence of ethnicity and levels of social support on breast feeding outcomes.

Efficient literature searching in diffuse topics: lessons from a systematic review of research on communicating risk to patients in primary care.

Using the example of communication about risk in a primary care setting, this paper puts forward a method of developing and evaluating a detailed search strategy for locating the literature for a systematic review of a 'diffuse' subject. The aim of this paper is to show how to develop a search strategy that maximizes both recall and precision while keeping search outputs manageable. Six different databases were used, namely Medline, Embase, PsychLIT, CancerLIT, Cinahl and Social Science Citation Index (SSCI). The searches were augmented by hand-searching, contacting authors, citation searching and reference lists from included papers. Other databases were searched but yielded no extra references for this subject matter. Of the 99 papers included, 80 were indexed on Medline. The Medline search strategy identified 54 of them and the remaining 26 were located on other databases. The 19 further unique references were found using the other databases and methods of retrieval. A combination of several databases must be used to maximize recall and to increase the precision of searches on individual databases, thus improving the overall efficiency of the search.

Can nurses learn to let go? issues arising from an intervention designed to improve patients' involvement in their own care.

The aim of the paper was to explore the reasons behind a perplexing observation that an intervention designed to alter professional behaviour in general practice consultations in South Wales was poorly sustained despite initial enthusiasm among the nurses. Nurses' baseline and continuing responses to the new approach with patients with non-insulin dependent diabetes are described, drawing on observations and interviews with the 18 practice nurses in the 15 experimental practices. The nurses fell into two groups: those showing greater understanding of the principles underlying the approach and a more reflective attitude to their work; and those who adapted the simple technology provided to their usual consultation style. This variation was related to their perception of their role and their attitudes to diabetic care, especially the challenge posed by the 'difficult patient'. They were noticeably less willing to allow the patient freedom to decide what to do once diabetic control was deteriorating according to biochemical measures. It is concluded that the failure of the nurses to sustain behaviour change over time stems from a basic dilemma, namely, what is the extent of my responsibility and how ought I to discharge it? There is scope for more research into professional perspectives on chronic care, the strategies they use to manage patients and to cope with their own feelings, and the implications of these strategies for patients.

Moles and melanomas--who's at risk, who knows, and who cares? A strategy to inform those at high risk.

BACKGROUND: Malignant melanoma is uncommon but potentially fatal. Knowledge and attitudes play an important part in the early detection of this skin cancer. AIM: To assess knowledge, risk perception, and intended behaviour related to melanoma compared with actual risk status. To suggest measures to improve the primary and secondary prevention of melanoma in general practice for a high-risk group. METHOD: A prospective questionnaire survey was carried out on consecutive adults attending in 16 randomly selected group practices. Applying MacKie's personal risk factor chart for melanoma, the study assessed self-reported risk, knowledge of skin cancer--especially malignant melanoma--and self-reported preventive activity. RESULTS: A total of 3105 (69%) attenders completed the questionnaire. The responders showed greater concern for minor rather than major clinical signs in pre-existing moles. Young people and the professional classes were the least knowledgeable about skin cancer and exhibited the most risky behaviour in terms of sun exposure, failure to check their skin, and to seek medical advice about new or changing moles. The majority of an 8.7% high-risk group showed lack of awareness of their increased risk, and women in this group reported the highest desire for a suntan and the use of sunbeds. In addition to showing less concern than their low risk counterparts about moles growing in size, they were also reluctant to seek medical advice about new moles. CONCLUSION: Consideration should be given to targeting primary prevention and selective screening in general practice towards a high-risk group for malignant melanoma. Young people and the professional social class should receive particular attention.

Qualitative study of decisions about infant feeding among women in east end of London.

OBJECTIVE: To improve understanding of how first time mothers who belong to a socioeconomic group with particularly low rates of breast feeding decide whether or not to initiate breast feeding. DESIGN: Qualitative semistructured interviews early in pregnancy and 6-10 weeks after birth. SETTING: Women's homes in east end of London. SUBJECTS: 21 white, low income women expecting their first baby were interviewed mostly at home, often with their partner or a relative. Two focus groups were conducted. RESULTS: Women who had regularly seen a relative or friend successfully breast feed and described this experience positively were more confident about and committed to breast feeding. They were also more likely to succeed. Exposure to breast feeding, however, could be either a positive or a negative influence on the decision to breast feed, depending on the context. Women who had seen breast feeding only by a stranger often described this as a negative influence, particularly if other people were present. All women knew that breast feeding has health benefits. Ownership of this knowledge, however, varied according to the woman's experience of seeing breast feeding. CONCLUSIONS: The decision to initiate breast feeding is influenced more by embodied knowledge gained from seeing breast feeding than by theoretical knowledge about its benefits. Breast feeding involves performing a practical skill, often with others present. The knowledge, confidence, and commitment necessary to breast feed may be more effectively gained through antenatal apprenticeship to a breastfeeding mother than from advice given in consultations or from books.

Communication about risk: diversity among primary care professionals.

OBJECTIVE: We aimed to identify the important clinical topics and to report current practice of communication about risk in primary care. METHODS: We carried out a qualitative study using six semi-structured focus group discussions with primary care professionals. The subjects were 36 primary care professionals from general practice, practice nurse, district nurse, community psychiatric nurse and health visitor disciplines. RESULTS: All clinical topics were felt to raise issues of effective risk communication. Participants expressed concern about the lack of accessible up-to-date information and the problems of conveying information to patients. They described the circumstances and contexts in which communication varies and is difficult, and identified media and medico-legal influences and professional uncertainty as also contributing to the problems of risk communication. Specific training in risk communication was identified as an important need. CONCLUSIONS: The complexity of risk communication and diversity of influences on it will require wide-ranging interventions in order to address them. Standardized communication may be a difficult goal to attain.

Communication about risk: the responses of primary care professionals to standardizing the 'language of risk' and communication tools.

OBJECTIVE: We aimed to gauge responses of primary care professionals to standardization of the 'language of risk' and risk communication tools. METHODS: We carried out a qualitative study using six semi-structured focus group discussions. The subjects were 36 primary care professionals from general practice, practice nurse, district nurse, community psychiatric nurse and health visitor disciplines. RESULTS: Between professionals, the standardization of the language of risk was felt to have potential benefit in making professionals consistent in their appreciation of risks and communication with each other. Between professionals and patients, standardized language was thought inappropriate or insufficient because of contextual variation in communication and interpretation of risk information by patients. The use of more-detailed comparisons of risks was felt to be a potentially effective development of risk communication in practice. CONCLUSIONS: A standard language of risk communication was perceived as being potentially helpful for communication between professionals, but many respondents were sceptical about its usefulness in communication with patients.

Understanding the culture of prescribing: qualitative study of general practitioners' and patients' perceptions of antibiotics for sore throats.

OBJECTIVES: To better understand reasons for antibiotics being prescribed for sore throats despite well known evidence that they are generally of little help. DESIGN: Qualitative study with semi-structured interviews. SETTING: General practices in South Wales. SUBJECTS: 21 general practitioners and 17 of their patients who had recently consulted for a sore throat or upper respiratory tract infection. MAIN OUTCOME MEASURES: Subjects' experience of management of the illness, patients' expectations, beliefs about antibiotic treatment for sore throats, and ideas for reducing prescribing. RESULTS: Doctors knew of the evidence for marginal effectiveness yet often prescribed for good relationships with patients. Possible patient benefit outweighed theoretical community risk from resistant bacteria. Most doctors found prescribing "against the evidence" uncomfortable and realised this probably increased workload. Explanations of the distinction between virus and bacterium often led to perceived confusion. Clinicians were divided on the value of leaflets and national campaigns, but several favoured patient empowerment for self care by other members of the primary care team. Patient expectations were seldom made explicit, and many were not met. A third of patients had a clear expectation for antibiotics, and mothers were more likely to accept non-antibiotic treatment for their children than for themselves. Satisfaction was not necessarily related to receiving antibiotics, with many seeking reassurance, further information, and pain relief. CONCLUSIONS: This prescribing decision is greatly influenced by considerations of the doctor-patient relationship. Consulting strategies that make patient expectations explicit without damaging relationships might reduce unwanted antibiotics. Repeating evidence for lack of effectiveness is unlikely to change doctors' prescribing, but information about risk to individual patients might. Emphasising positive aspects of non-antibiotic treatment and lack of efficacy in general might be helpful.

A randomized controlled trial of an intervention designed to improve the care given in general practice to Type II diabetic patients: patient outcomes and professional ability to change behaviour.

AIM: Our objective was to evaluate the effect of training in a patient-centred intervention for GPs and practice nurses on outcomes for patients with Type II diabetes. METHODS: We carried out a randomized controlled trial within general practices as the basis for randomization and a before-and-after design for measures of patient outcome. A parallel process study examined the use of the method by professionals. The study was carried out in 29 general practices in South Glamorgan who had participated for at least 2 years in a local scheme of audit and CME in relation to Type II diabetes care. The subjects were 252 Type II diabetic patients recruited by 15 experimental and 14 control practices. The main outcome measures were changes in glycosylated haemoglobin, patient satisfaction with care and treatment, functional health status and professional ability to apply the intervention. RESULTS: Professionals adopted the innovative method with enthusiasm, but after 2 years only 19% continued to apply the method systematically. The trial was, therefore, unable to demonstrate significant biochemical or functional improvements. This highlights the need to understand the factors associated with professional uptake and subsequent ability to sustain changes in behaviour. CONCLUSIONS: The efficacy of this behavioural intervention remains unproved, despite its acceptability to professional staff. Detailed and prolonged development and testing of behavioural interventions is an essential first step before embarking on randomized controlled trials which involve complex behavioural changes in professionals or patients.

Qualitative study of patients' perceptions of doctors' advice to quit smoking: implications for opportunistic health promotion.

OBJECTIVES: To determine the effectiveness and acceptability of general practitioners'opportunistic antismoking interventions by examining detailed accounts of smokers' experiences of these. DESIGN: Qualitative semistructured interview study. SETTING: South Wales. SUBJECTS: 42 participants in the Welsh smoking intervention study were asked about initial smoking, attempts to quit, thoughts about future smoking, past experiences with the health services, and the most appropriate way for health services to help them and other smokers. RESULTS: Main emerging themes were that subjects already made their own evaluations about smoking, did not believe doctors' words could influence their smoking, believed that quitting was down to the individual, and felt that doctors who took the opportunity to talk about smoking should focus on the individual patient. Smokers anticipated that they would be given antismoking advice by doctors when attending for health care; they reacted by shrugging this off, feeling guilty, or becoming annoyed. These reactions affected the help seeking behaviour of some respondents. Smokers were categorised as "contrary," "matter of fact," and "self blaming," depending on their reported reaction to antismoking advice. CONCLUSIONS: Doctor-patient relationships can be damaged if doctors routinely advise all smokers to quit. Where doctors intervene, a patient centred approach-one that considers how individual patients view themselves as smokers and how they are likely to react to different styles of intervention-is the most acceptable.