Family Practice Advance Access originally published online on June 24, 2007
Family Practice 2007 24(5):511-517; doi:10.1093/fampra/cmm029
Feasibility and acceptability of screening for eating disorders in primary care
a OJ is now at the Department of Psychology, Institute of Psychiatry, King's College London, PO78, Addiction Sciences Building, 4 Windsor Walk, London SE5 8AF, UK
b Community Clinical Sciences Division, School of Medicine, University of Southampton, Aldermoor Health Centre, Aldermoor Close, Southampton, Hampshire SO16 5ST, UK
Correspondence to: Tony Kendrick, Community Clinical Sciences Division, School of Medicine, University of Southampton, Aldermoor Health Centre, Aldermoor Close, Southampton, Hampshire SO16 5ST, UK; Email: a.r.kendrick{at}soton.ac.uk
Received 29 September 2006; Revised 9 April 2007; Accepted 30 April 2007.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionDeclarationAppendix Topic guide for...References |
|---|
Background. Earlier diagnosis of disordered eating is linked to improved prognosis, but detection in primary care is poor.
Objectives. To assess the feasibility of screening for disordered eating within primary care, in terms of the proportion of patients accepting screening, yield of cases, action taken by staff and staff views on screening.
Methods. Data were collected in open GP surgeries, midwife (MW) antenatal clinics and health visitor (HV) child health surveillance clinics in two GP practices, using face-to-face surveys and semi-structured interviews. Female patients aged 16–35 were asked to complete the SCOFF questionnaire, which was scored by researchers and taken by the patient into their consultation. If the result indicated possible disturbed eating, the health professional (HP) running the surgery/clinic was asked to complete a questionnaire and interview. One hundred and eleven women were screened and 11 HPs (GPs, MWs, HVs) were interviewed.
Results. Forty-six percent of patients agreed to be screened. Of these, 16% produced a positive result. The staff survey suggested that HPs found screening acceptable. However, concerns arose in the interviews, principally over what action to take in response to positive results. Positive results were rarely recorded in medical notes, and treatment was rarely offered.
Conclusion. In order for a screening programme for eating disorders to be implemented in primary care, HP concerns about options for dealing with positive results would need to be addressed. Feasibility of screening would be enhanced by production of a protocol to be followed in the case of positive results.
Keywords. Acceptability, attitudes, eating disorders, feasibility, screening.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationAppendix Topic guide for...References |
|---|
Disordered eating is one of the three most common mental health difficulties (along with depression and anxiety disorders).1 Eating disorders are particularly common in young women, and are a significant cause of morbidity and mortality.2,3 Luck et al.4 note that ‘health care workers in primary care are at the forefront of screening and managing these disorders’ (p. 755). Earlier diagnosis of eating disorders is linked to improved prognosis,5 but detection of eating disorders in primary care is poor.6 The National Institute for Clinical Excellence (NICE) guidelines on the management of eating disorders7 point to the need for improved identification and screening of eating disorders in primary care settings. A survey by the Eating Disorders Association8 indicated that 42% of GPs did not make an early diagnosis, suggesting that services are failing to meet the NICE7 recommendation that ‘People with eating disorders seeking help should be assessed and receive treatment at the earliest opportunity’.
A number of factors are likely to contribute to these low rates of detection. Patients may be slow to present, or may not present to services at all. Individuals with diagnoses of disordered eating are renowned for their high levels of secrecy and denial, and are often reluctant to disclose information.9 Patients with bulimic symptoms are often of normal weight or slightly overweight, causing some patients in this group to be missed.10 Screening for disordered eating within primary care could be one way of improving rates of detection. A short, reliable and valid screening questionnaire for eating disorders has been developed, and its acceptability to patients has been demonstrated.4,11 However, previous research has not investigated the feasibility of screening in terms of the proportion of patients accepting screening or health professional (HP) views on screening.
The finding that women who self-report eating disordered behaviour, and those who go on to receive diagnoses of eating disorders (including partial syndromes), have higher levels of health care utilization in primary care settings12,13 suggests that primary care workers may not be seizing the opportunity to ask screening questions. Previous research has found variation among primary care physicians in their rates of asking about disordered eating.14,15 This points to the need for exploration of primary care HP views on screening for eating disorders. As well as exploring the views of GPs, there is a need to explore the attitudes of other primary care HPs such as midwives (MWs) and health visitors (HVs). Such HPs also have opportunities to detect eating disorders [e.g. when running antenatal (AN) clinics or child health surveillance (CHS) clinics], but there is less research on detection of mental ill health by primary care practitioners other than GPs.6
The present study investigated the feasibility (in terms of numbers of patients accepting screening, yield and action taken by staff) of using the SCOFF questionnaire4 to screen for eating disorders within primary care. The acceptability of screening to primary care HPs was also explored.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationAppendix Topic guide for...References |
|---|
Design
Cross-sectional face-to-face surveys, semi-structured interviews.
Setting
Open GP surgeries, MW AN clinics and HV CHS clinics in two GP practices.
Participants
Only female patients aged 16–35 were approached. Patients with currently diagnosed mental health difficulties, or those who were taking psychotropic medication, as well as those with severe or terminal illness and those who did not speak English well enough to complete the questionnaire were excluded by the HP before the session commenced. Eleven HPs (seven GPs, two MWs and two HVs) were interviewed.
Data collection
The medical students (GF and SC) were present in the waiting rooms of general practice clinics/surgeries (AN clinics, CHS clinics or open GP surgeries) at a time previously agreed with the attending HP. The HP checked his or her appointment list to ensure that any patients he or she deemed unsuitable would not be approached. As each patient arrived, they were handed a slip by the receptionist, asking them to indicate their willingness to be approached by the students. If a patient was willing to be approached, the study was discussed with them in a quiet area off the waiting room. Patients who were prepared to take part signed a consent form and then completed the written questionnaire. Patients were asked to take the completed SCOFF questionnaire into their consultation.
If a positive screening result was produced, the attending MW/HV (where applicable) and the patient's GP completed a brief questionnaire on their reaction and planned response, and their rating of the acceptability of screening. In the case of AN/CHS clinics, a copy of each screening questionnaire was forwarded to the patient's GP (with a blank staff survey form attached for completion by the GP in the case of positive screening results). GPs were asked to note the patient's involvement and screening result in their medical notes. Medical notes were accessed by the students at a later date (with informed consent from the patient) to determine what (if any) action was taken following screening. Interviews were undertaken at the end of the study with those HPs who had received a positive screening result for one of their patients. While 16 HPs had offered to take part (10 GPs, 3 MWs and 3 HVs), only 10 received a positive screening result for one of their patients. These HPs (seven GPs, two MWs and one HV) were interviewed, along with an additional HV who had not received a positive screening result but wished to participate in an interview due to her interest in the topic. Interviews took approximately 30 minutes and explored in more detail the HPs' brief survey responses. Questions covered four main topics: responses to specific positive screening results in patients the HP had seen during the study, general responses to suspected disordered eating, views on screening and responses to results of the overall screening results from the study (e.g. proportion of patients accepting screening, proportion producing positive results).
Materials
The SCOFF questionnaire4 was used to screen for disordered eating. This questionnaire is a short and effective screening tool in general practice. It has shown excellent validity in a clinical population11 and reliability in a student population.16 The SCOFF tool also performed well against 10 questions set by Greenhalgh17 to assess screening tests. It has been found to be reliable and replicable when administered as a written questionnaire rather than when undertaken as an oral interview, and may even provide enhanced disclosure of symptoms due to the less intimidating approach.16 The use of this questionnaire, both the questions and the term ‘SCOFF’, has been found to be acceptable to patients, and its use by HPs in primary care is recommended.4,11 It is intended to raise suspicion of likely cases of disturbed eating, rather than acting as a diagnostic tool. A positive result is produced if a ‘yes’ response is provided to two or more of the five questions. Luck et al.4 report sensitivity of 84.6% and specificity of 89.6%. A brief staff survey form was produced to collect information on HP responses to positive screening results, and a semi-structured topic guide was used to collect data during staff interviews (see Appendix).
Qualitative analysis
A framework approach18 was used to analyse the interview findings. Framework analysis was developed for use in applied qualitative research with pre-specified objectives, and so employs deductively derived categories.19 However, this method also allows for inductive exploration of participant accounts and so enables exploration of emergent themes not anticipated at the outset of the study. Ritchie and Spencer18 note that the method is grounded or generative (‘heavily based in, and driven by, the original accounts and observations of the people it is about’, p. 176) and dynamic (‘open to change, addition and amendment throughout the analytic process’, p. 176), but also that the analysis will draw upon a priori issues. The five stages of analysis have much in common with other methods of qualitative analysis, but are more strongly informed by a priori reasoning. The stages of analysis consist of familiarization with the data (reading through the transcripts in detail and making notes on any emerging ideas or themes), identifying the thematic framework (drawing up a list of themes using a priori issues and issues emerging from the data), indexing/applying the thematic framework to the data (labelling transcript data with codes from the thematic framework), charting (abstraction and synthesis, e.g. drawing up tables summarizing what each participant said about each theme in order to identify possible patterns) and finally mapping and interpretation (summarizing and interpreting overall patterns).
Trustworthiness
Results from the staff interviews were triangulated with findings from the staff survey to improve the comprehensiveness of the study. An audit trail was maintained in the form of notes from the process of framework analysis to increase the transparency of this process. Also, the medical students kept notes reflecting on the possible impact of factors such as their student status upon data collection and analysis.
Ethical approval
Granted by Southampton and South West Hampshire LREC.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationAppendix Topic guide for...References |
|---|
Patient demographics
The majority were white (96.2%) and in paid employment (62.9%). Patients' ages ranged from 16 to 35 years, and the mean age was 26 years.
Proportion of patients accepting screening
A total of 510 patients were identified from appointment lists. After exclusion of patients for reasons outlined in Table 1, 300 women were eligible for participation in the study. Of these, 138 (46%) agreed to complete the SCOFF. Twenty-seven were interrupted by being called into their appointments, and so, in total, the SCOFF was completed by 111 women.
|
Yield of cases
Of the 111 women who completed the SCOFF, 18 (16%) produced positive results. This finding is in line with the findings of Luck et al.4 who found that 13% of participants produced a positive result.
Action taken by HPs
An intention to offer treatment or referral was recorded on a staff survey form for only 2 of the 18 positive screening results produced. Comments on positive screening results were only recorded in the patients' medical notes for four of the patients. This excludes one patient who declined permission for her notes to be accessed by the researchers. The comments made in patients' notes are provided in Box 1.
BOX 1: Comments on positive screening results made in patients' medical notes.
|
HP questionnaires
Of the 16 participating HPs, 10 received a positive patient screening questionnaire (seven GPs, two MWs and one HV), and one HV who had not received a positive result was also interviewed. They were asked how acceptable they found screening for eating disorders. For those who saw more than one patient with a positive screening result and so completed more than one staff survey form, their average acceptability score was calculated. A score of 7 denoted ‘completely acceptable’ and a score of 1 ‘completely unacceptable’. Overall, the mean acceptability score was 5.9, the median 6.5 and the mode 7, indicating that the HPs found screening for eating disorders to be highly acceptable. Nine of the 11 participants produced mean acceptability scores above the midpoint of 4 on the scale, suggesting that overall they found screening to be acceptable. One HP produced a mean acceptability score, which fell at the midpoint on the scale (participant F), suggesting neutrality. One HP (participant A) produced a mean acceptability score of 3, slightly below the midpoint, suggesting some reservations about the acceptability of screening.
HP interviews
The general consensus was that screening for eating disorders was acceptable and could be useful, and would fit with participants' roles as primary health care workers. The SCOFF questionnaire was found to be an acceptable instrument (Box 2).
BOX 2: Acceptability of the SCOFF questionnaire to HPs.
|
However, participants noted that they did not routinely screen for eating disorders as part of consultations, and there was wide variation in the extent to which HPs felt comfortable approaching the topic with patients (Box 3).
BOX 3: HP views on asking patients about disordered eating.
|
It was pointed out that the cultural emphasis on slimness and dieting could make it difficult to determine pathological cases from the norm. There were concerns about the accuracy and effectiveness of screening, time, administration and financial constraints, as well as uncertainty surrounding the benefit of early diagnosis and treatment (Box 4).
BOX 4: Uncertainty about the benefits of early detection.
|
The overwhelming issue arising among HPs was what action to take once a patient had screened positively. Possibilities such as referral to an in-practice counsellor, generic mental health service, specialist eating disorders service or a dietician were mentioned, but available options were often seen as being inappropriate or unavailable. Although a number of participants reported positive experiences with the local specialist eating disorders service, it was noted that many patients would not meet the referral criteria for the service (age 18–65 years and meeting diagnostic criteria for anorexia nervosa, bulimia nervosa or eating disorders not otherwise specified/atypical presentations).
Participants noted that they did not have sufficient time or expertise to manage possible disordered eating alone, and a general uncertainty about the available options was often expressed (Box 5).
BOX 5: Uncertainty about available options for addressing a positive screening result.
|
Reference was also made to the issue of patient's denial of disordered eating or lack of motivation to address their disturbed eating (Box 6).
BOX 6: Patient lack of motivation as a barrier to addressing positive screening results.
|
In terms of suggestions for implementation, HPs suggested that screening would be best carried out face to face during a general patient health check, first appointment or AN booking visit. It was suggested that male patients should also be included but that an age limit as was used in the study should be adhered to (in order to focus screening upon the age range perceived as having the highest prevalence of disordered eating).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationAppendix Topic guide for...References |
|---|
Summary of results
In total, just under half of the women approached agreed to be screened, and of these, one in six screened positively on the SCOFF questionnaire. It should be noted that a positive score on the SCOFF questionnaire is not diagnostic, and not all of those who score positively will have an eating disorder. Luck et al.4 found that 4% of those screened (24% of those producing a positive result) met the criteria for an eating disorder using a diagnostic interview.
The quantitative measure of HPs' opinions indicated that the majority found screening for eating disorders acceptable. However, analysis of the more detailed interview findings suggests a more complex picture. This pattern is seen in other studies measuring acceptability using both quantitative and qualitative methods.20 The interviews suggested that HPs viewed screening as acceptable in theory but indicated that there were concerns over what action to take with patients who screened positively, due to the dissatisfaction of some HPs with the current treatment options as well as concerns regarding time and cost. Positive results were rarely recorded in patients' medical notes, and treatment was rarely offered.
Strengths and limitations
Steps taken to ensure rigour included triangulation of data and the availability of an audit trail for the qualitative analysis (enhancing transparency), as well as consideration of reflexivity. However, data collection took place in only two surgeries, in a single city, which raises the question of how well the quantitative results of this study would generalize to other settings. Also, the screening process was somewhat artificial in that patient participants were required to agree with the receptionist that they would speak to researchers and were required to complete a consent form, prior to screening. These factors may have decreased the proportion of patients willing to be screened, compared to routine practice.
Comparison with previous literature
Previous literature has highlighted the need for improved detection of eating disorders in primary care,7 as well as outlining the yield of cases and acceptability to patients of the SCOFF screening questionnaire.4,11 This study provides a detailed analysis of the feasibility of screening for eating disorders in primary care, by exploring the views and actions of HPs. While the SCOFF questionnaire has been widely adopted within the UK as a standard screening measure, the absence of clear care pathways poses an obstacle to its use. This state of affairs is likely to impede access of patients to evidence-based treatments as outlined in the NICE guideline.7 Treatments recommended for use in primary care (such as guided self-help) may not be accessed by many patients due to barriers to the detection of disordered eating. The present findings highlight a number of potential problems that must be addressed before screening can be successfully implemented. HPs' concerns about asking about disordered eating, and uncertainty about how to proceed in the case of positive results, as well as the tendency not to record positive screening results in medical notes and not to offer treatment would at present limit the utility of any screening programme.
The present results echo findings on screening for the similarly ‘hidden’ problem of domestic violence in primary care. Research by Richardson et al.21 has highlighted the reluctance of primary care HPs to engage in screening for domestic violence. In a later study,22 although it was found that 17% of respondents had experienced physical violence from a partner within the last year, the case for screening was weakened by factors such as the gap between women's experiences and their medical records (which may have been produced by under-recording of disclosure) and limited acceptability of screening to patients (at least 20% objected to screening for domestic violence).
Implications for practice
Before screening is adopted, it is a prerequisite that acceptable and accessible treatment facilities should be available. This does not seem to be the case in eating disorders, given the uncertainty over referral. The findings suggest that dissemination of a protocol outlining action to be taken in the case of positive screening results would be necessary in order to enhance the feasibility of screening for eating disorders in primary care.
HP education about management options (e.g. guided self-help) would also be beneficial. Some participants explicitly referred to their lack of training in the area of disordered eating, and indeed there is currently no widespread training of HPs in the screening and assessment of people with eating disorders. Most MWs and HVs undergo very little training in mental health. While around 40% of GP trainees complete a 6-month post in Psychiatry, for most this will not include any specific training on eating disorder detection and evaluation.
The findings also raise the issue of potential screening burden. The challenge to primary care is handling and sustaining the task of screening without affecting other priorities, and there is a need for evaluation of the extent to which this is possible. As noted above, the SCOFF does produce false positives, with only 24% of those producing a positive screening result in research by Luck et al.4 meeting diagnostic criteria for an eating disorder. While 16% of participants in the present study produced a positive screening result, previous research4 suggests that only 4% would meet criteria for disordered eating. Currently, all individuals producing a positive screening result would require referral to more specialist services for exclusion of an eating disorder diagnosis, creating difficulties in terms of patient anxiety and cost to services. The alternative option would be to have a second stage of assessment within primary care, using a more structured interview-based or questionnaire-based diagnostic test, for which the HPs would need considerable training. The results also suggest that screening may sometimes be impractical within the constraints of primary care, as suggested by the fact that 27/138 individuals did not complete the SCOFF due to interruption.
Implications for further research
Future research should explore whether the extent to which positive screening results lead to action by staff is improved by use of a protocol outlining action to be taken in the case of positive screening results, and education about management options. In addition, the burden upon services produced by screening should be evaluated in greater depth.
Conclusion
Patient uptake of screening is relatively low, although this could have been influenced by the slightly artificial constraints of the study. HPs have concerns about the options open to them if they suspect that a patient has an eating disorder, and may not take action in relation to positive screening results. Guidance on action to be taken in response to positive results could improve the feasibility of screening for eating disorders in primary care.
|
Declaration |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationAppendix Topic guide for...References |
|---|
Ethical approval: Granted by Southampton and South West Hampshire LREC, reference number 04/Q1704/67.
Funding: No project funding (4th year study in-depth project for the University of Southampton BM degree; administrative costs covered by the School of Medicine). This work was conducted while OJ was in receipt of a postdoctoral fellowship from the Health Foundation and the DH/NCC RCD Programme.
Contributors: All authors contributed to the planning of the study, analysis and reporting. Data were collected by GF and SC.
Guarantor: OJ.
Conflicts of interest: None.
|
Appendix Topic guide for HP interviews |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationAppendix Topic guide for...References |
|---|
Responses to positive screening results in this study
During the course of this study, you have been asked to fill in a staff survey form for any patient producing a positive result on the SCOFF screening questionnaire. I would like to start off by asking you a bit more about your thoughts on these positive screening results. [Following questions repeated for each positive screening result received by the HP]:
- When you were told about this result, what did it mean to you?
- How did the screening result compare to your own perception of whether or not the patient was experiencing disordered eating?
- On the survey form you indicated [what you planned to do in response to this positive screening result]. Can you tell me a bit about what influenced that choice?
- Why did you decide not to ... [choose the other options listed on the survey form, such as ...]?
- Probe around what influenced any decisions or plans about action before receiving screening result.
General responses to suspecting disordered eating
- In general, what influences how you will respond if you suspect that a patient has disordered eating?
- What options do you feel are open to you if you suspect that a patient has disordered eating?
- How do you feel about these options?
Views on screening in general
- Do you ever ask patients screening questions if you suspect they have disturbed eating?
- If no: Can you tell me a bit about why you tend not to do this?
- If yes: Can you tell me a bit about the circumstances in which you would ask these questions? What sort of questions do you ask? How do you find asking about this?
- How do you feel about the use of screening for disturbed eating in primary care?
- How do you feel screening for disturbed eating fits with your role as a GP/MW/HV?
- What do you think are the advantages and disadvantages of screening? If you envisage any problems with screening, how do you think these could be overcome?
- What method of screening, if any, do you think would be best (e.g. targeting specific groups/blanket screening, screening in general waiting rooms or in specific clinics, face to face or postal, oral or written, different instrument/questions)? Can you tell me a bit more about this?
Feedback general screening results (proportion of patients accepting screening, producing positive results, etc)
- Now that you have heard about these results, what do you think about the use of screening for disturbed eating in primary care?
- Does this change any of your views? If so, how have your views changed?
|
Acknowledgments |
|---|
Thanks are due to the participants who gave up their time to take part in this study, and to the medical receptionists who helped with recruitment.
|
Notes |
|---|
Johnston O, Fornai G, Cabrini S and Kendrick T. Feasibility and acceptability of screening for eating disorders in primary care. Family Practice 2007; 24: 511–517.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationAppendix Topic guide for...References |
|---|
1 Department of Health. National Service Framework for Mental Health: Modern Standards and Service Models (1999) London: Department of Health.
2 Kreipe R, Birndorf S. Eating disorders in adolescents and young adults. Med Clin North Am (2000) 340:1027–1049.
3 Pritts S, Susman J. Diagnosis of eating disorders in primary care. Am Fam Physician (2003) 67:297–304.[Web of Science][Medline]
4 Luck AJ, Morgan JF, Reid F, et al. The SCOFF questionnaire and clinical interview for eating disorders in general practice: comparative study. BMJ (2002) 325:755–756.
5 Hall A, Slim E, Hawker F, Salmond C. Anorexia nervosa: long term outcome in 50 female patients. Br J Psychiatry (1984) 145:407–413.
6 Department of Health. Women's Mental Health: Into the Mainstream. Strategic Development of Mental Health Care for Women (2002) London: Department of Health.
7 National Institute for Clinical Excellence. Eating Disorders: Core Interventions in the Treatment and Management of Anorexia Nervosa, Bulimia Nervosa and Related Eating disorders (2004) London: National Institute for Clinical Excellence. Clinical Guideline 9.
8 Eating Disorders Association. ‘Getting Better?’ Is the Quality of Treatment for Eating Disorders in the UK Getting Better? (2005) Norwich: Eating Disorders Association.
9 Raphael F. The prevention of eating disorders. In: The Prevention of Mental Illness in Primary Care—Kendrick T, Tylee A, Freeling P, eds. (1996) Cambridge: Cambridge University Press.
10 Walsh J, Wheat M, Freund K. Detection, evaluation and treatment of eating disorders. The role of the primary care physician. J Gen Intern Med (2000) 15:577–590.[CrossRef][Web of Science][Medline]
11 Morgan J, Reid F, Lacey JH. The SCOFF questionnaire: assessment of a new screening tool for eating disorders. BMJ (1999) 319:1467–1468.
12 Sansone RA, Wiederman MW, Sansone LA. Healthcare utilization among women with eating disordered behavior. Am J Manag Care (1997) 3:1721–1723.[Web of Science][Medline]
13 Ogg EC, Millar HR, Pusztai EE, Thom AS. General practice consultation patterns preceding diagnosis of eating disorders. Int J Eat Disord (1997) 22:89–93.[CrossRef][Web of Science][Medline]
14 Boule CJ, McSherry JA. Patients with eating disorders: how well are family physicians managing them? Can Fam Physician (2002) 48:1807–1813.
15 Stevens S, Savageau J, Luckmann R, Sattler A. Screening for eating disorders: a survey of primary care physicians. J Adolesc Health (1999) 24:96.
16 Perry L, Morgan J, Reid F, et al. Screening for symptoms of eating disorders: reliability of the SCOFF screening tool with written compared to oral delivery. Int J Eat Disord (2002) 32:466–472.[CrossRef][Web of Science][Medline]
17 Greenhalgh T. Papers that report diagnostic or screening tests. BMJ (1997) 315:540–543.
18 Ritchie J, Spencer L. Qualitative data analysis for applied policy research. In: Analysing Qualitative Data—Bryman A, Burgess R, eds. (1993) London: Routledge.
19 Pope C, Ziebland S, Mays N. Analysing qualitative data. In: Qualitative Research in Health Care—Pope C, Mays N, eds. (2000) London: BMJ Books.
20 Rogers A, Karlsen S, Addington-Hall J. All the services were excellent. It is when the human element comes in that things go wrong: dissatisfaction with hospital care in the last year of life. J Adv Nurs (2000) 31:768–774.[CrossRef][Web of Science][Medline]
21 Richardson J, Feder G, Eldridge S, et al. Women who experience domestic violence and women survivors of childhood sexual abuse: a survey of health professionals' attitudes and clinical practice. Br J Gen Pract (2001) 51:468–470.[Web of Science][Medline]
22 Richardson J, Coid J, Petruckevitch A, et al. Identifying domestic violence: cross sectional study in primary care. BMJ (2002) 324:274–279.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||