Qualitative study
From Wikipedia
Contents |
Methods
Subjects
- Our goal was to interview doctors and patients from practices reflecting factors known to be associated with prescribing levels 6 29 30 rather than from statistically representative samples. We constructed a sampling frame by stratifying all practices in the Bro Taf Health Authority (n=133) into low, medium, and high deprivation groups (based on Townsend scores) and into small, medium, and large practices (based on numbers of partners). First, a practice was randomly selected from each of nine cells in the sampling frame, and then general practitioners were approached from the practice in random order. Each consenting general practitioner was asked to identify up to 10 consecutive patients (adults or children) consulting with sore throats within the two weeks before interview; we approached patients at random until one from each list had been interviewed. (Butler, 1998)
Interview procedure
- We conducted interviews with general practitioners (10-35 minutes) in surgeries after the patient list had been constructed. The practitioners were told that our purpose was not to audit practice but to understand their feelings about these consultations and antibiotic treatment. Interviews with general practitioners included questions about consultations for sore throats, their current practice, changes in their practice over time, the scientific evidence, and suggestions for action. We interviewed patients (10-15 minutes) within two weeks after interviewing their general practitioner, and these took place in their own homes. They were told that the purpose was to understand their feelings about going to the doctor for colds and sore throats. Interviews were relatively brief to minimise inconvenience, and confidentiality was assured. The patients were asked about their experiences and views about their recent consultation and previous ones; self care practices, consulting thresholds, expectations of the consultation, what happened, satisfaction, their attitude towards and knowledge about antibiotics, and suggestions for reducing prescribing levels. The local research ethics committee approved the study. We piloted the semi-structured interview schedules with four general practitioners and three patients. All questions were open. We added new questions as the interview process progressed. We stopped conducting interviews when no new themes were emerging, in line with the grounded theory approach.26 (Butler, 1998)
Analysis (quoted from Mays, 2000)
Interviews
- Audiotaped interviews were transcribed and read twice by CCB and SR, then discussed with an experienced qualitative researcher (RP). Coding schedules were agreed and piloted. All interviews were double coded, by a researcher trained in qualitative methods (FM-R) and by CCB or SR; ambiguities were resolved in discussion. Categories were reduced to major themes through ongoing discussion between researchers and the re-reading of transcripts. (Butler, 1998)
Triangulation
Triangulation compares the results from either two or more different methods of data collection (for example, interviews and observation) or, more simply, two or more data sources (for example, interviews with members of different interest groups). The researcher looks for patterns of convergence to develop or corroborate an overall interpretation. This is controversial as a genuine test of validity because it assumes that any weaknesses in one method will be compensated by strengths in another, and that it is always possible to adjudicate between different accounts (say, from interviews with clinicians and patients). Triangulation may therefore be better seen as a way of ensuring comprehensiveness and encouraging a more reflexive analysis of the data (see below) than as a pure test of validity.
Respondent validation
Respondent validation, or "member checking," includes techniques in which the investigator's account is compared with those of the research subjects to establish the level of correspondence between the two sets. Study participants' reactions to the analyses are then incorporated into the study findings. Although some researchers view this as the strongest available check on the credibility of a research project,8 it has its limitations. For example, the account produced by the researcher is designed for a wide audience and will, inevitably, be different from the account of an individual informant simply because of their different roles in the research process. As a result, it is better to think of respondent validation as part of a process of error reduction which also generates further original data, which in turn requires interpretation.11
Clear exposition of methods of data collection and analysis
Since the methods used in research unavoidably influence the objects of inquiry (and qualitative researchers are particularly aware of this), a clear account of the process of data collection and analysis is important. By the end of the study, it should be possible to provide a clear account of how early, simpler systems of classification evolved into more sophisticated coding structures and thence into clearly defined concepts and explanations for the data collected. Although it adds to the length of research reports, the written account should include sufficient data to allow the reader to judge whether the interpretation proffered is adequately supported by the data.
Reflexivity
Reflexivity means sensitivity to the ways in which the researcher and the research process have shaped the collected data, including the role of prior assumptions and experience, which can influence even the most avowedly inductive inquiries. Personal and intellectual biases need to be made plain at the outset of any research reports to enhance the credibility of the findings. The effects of personal characteristics such as age, sex, social class, and professional status (doctor, nurse, physiotherapist, sociologist, etc) on the data collected and on the "distance" between the researcher and those researched also needs to be discussed.
Attention to negative cases
As well as exploration of alternative explanations for the data collected, a long established tactic for improving the quality of explanations in qualitative research is to search for, and discuss, elements in the data that contradict, or seem to contradict, the emerging explanation of the phenomena under study. Such "deviant case analysis" helps refine the analysis until it can explain all or the vast majority of the cases under scrutiny.
Results
Subjects
- We approached 31 general practitioners by telephone; seven were too busy or declined without giving a reason, and four were on holiday or sick leave or were not contactable. Of the 21 general practitioners recruited, all but three had qualified in Britain, 13 held MRCGP, and all had been principals for between one and 28 years. While we successfully recruited practitioners from a wide range of practices (see table), we obtained fewer interviews from those in small practices in areas of high deprivation. Refusal by an individual practitioner in a small practice was more likely to result in the exclusion of the practice than was refusal of a practitioner working from a group practice. The general practitioners who participated provided an average of 7.5 patient names each; only one failed to provide a list. We telephoned randomly selected patients from these lists, but if contact could not be made within a few days we selected another patient on the list. We telephoned 35 patients: 11 had incorrect or unobtainable telephone numbers, and seven declined to participate, resulting in 17 successfully completed interviews. Each of the nine stratification groups contained at least one patient, in some cases two. Four of the 17 patients were male, and five of the women had consulted their general practitioner on behalf of a child. Since our goal was to enhance understanding and generate hypotheses rather than achieve significance in a statistical sense,26 our findings are not presented numerically. However, we give broad indications of the numbers of subjects who expressed each theme. (Butler, 1998)
Emerging Themes
- General perceptions of consultations for upper respiratory tract infections. Most felt that, while the diagnostic aspect of these consultations was easy and often boring, management decisions were potentially complex. Although they recognised that patients came for reassurance, doctors felt that patients definitely "wanted something done" by the time they consulted and usually expected antibiotics. A typical clinician's opinion was, "You can't just say, `It's viral, you don't need antibiotics, go away,' because they feel they're being fobbed off. They feel that their illness is not being taken seriously." Clinicians generally preferred to meet patients' expectations when reasonable to do so: "I really hate people leaving my room feeling really let down by not having their expectations met." All but one doctor felt that many patients had too low a threshold for consulting. If a prescription was issued then the consultation was short and the patient seemed satisfied. However, prescribing antibiotics for these usually viral infections made clinicians feel compromised: "It does make me feel uncomfortable. I do feel as though I've been slightly used. Sometimes slightly abused as well." Attempting to change patients' beliefs and expectations in the consultation was often perceived as time consuming and unrewarding. One practitioner said, "You spend 15 minutes trying to educate them, when they will go out disillusioned, come back the next day and see someone else, making you feel 5 minutes would be better spent just giving them a prescription and getting rid of them." However, some of the most satisfying consultations were when a patient expecting a prescription for antibiotics left the consultation accepting non-antibiotic management: "Obviously if I prescribe antibiotics then the patient is happy, but if they accept a rational explanation as to why they don't want antibiotics and seem happy enough with that, then that's equally, even more satisfying really."
