Quality or quantity in GP research projects?
In the first of his two-part series on GP research projects, Dr Stefan Cembrowicz explains why a qualitative study is often the best option
When you come into general practice you will for the first time have unprecedented and privileged access to opportunities for research. Despite some consultants' beliefs that good science and respectable academic research is only possible in secondary care, you will soon realise you have access not only to a computerised database of patients far more detailed and sophisticated than any hospital department, but more importantly to a large body of people who have an ongoing relationship with their local practice. But quantitative research such as gold-standard randomised controlled trials (RCTs) may need unrealistic amounts of manpower and time to produce powerful statistical results. Good science is not, however, confined to RCTs – even the excellent, evidence-based Cochrane Collaboration approach using meta-analyses does not generate new ideas and leaves many questions about day-to-day practice unanswered. General practice may be the forum for smaller projects of different types that can be just as revealing.
Quality or quantity?
Decide early on whether you are interested in a quantitative (positivist, questionnaire-based, controlled trial, number crunching) or a qualitative (interpretive, narrative, interview, or focus group-based) approach.
Remember that quantitative and qualitative research complement each other. Often qualitative research is a good way of starting to research a topic because it generates ideas and brings up issues and questions that you might not have thought of before. And you can choose a combination of methods; for example a questionnaire (quantitative) may identify a group with problems, which can lead to (qualitative) semi-structured interviews, or conversely a (qualitative) focus group or a small number of individual interviews may identify themes. You thus create a hypothesis. And then, in order to test the hypothesis, you can do an RCT in a wider (quantitative) survey.
A good example of qualitative research is by Deirdry Murphy: BMJ May 1999;318;1414. Women's views on the impact of operative delivery in the second stage of labour; qualitative interview study.
Quantitative work generally involves large amounts of data and statistics – a very major undertaking unless you are able to give up a significant part of your working week. However, qualitative research based on social anthropology or ethnography is becoming respectable (and publishable) and most peer-reviewed journals will nowadays include a proportion of qualitative studies.
Attitudes, experiences and patients' narratives are best looked at by a qualitative approach, which some consider more attuned to the values of GP care itself – this method, developed from anthropology and ethnography, looks at patients' own voices and lives in their own multifaceted contexts, rather than disease entities according to the positivist, rational rules of physical science. As GPs, we spend our working lives immersed in local culture (anthropologists call this participant observation). This can give us privileged insights into patients' lives which we can then explore further.
Participant observation and interviews of patients will give you insights and understanding based on relatively small numbers of people – 20 or fewer may be plenty. Rather than looking for a statistically significant majority result, a qualitative approach may unearth much valuable information to reflect on from small numbers or even individual cases. For example, you could explore why your patients decline MMR vaccination, request referral for counselling, or don't attend outpatient appointments, with a qualitative method. The explanations offered can then generate theory to generalise from.
In this way you do not approach the subject with a pre-formed hypothesis to prove or disprove, but generate and develop your hypotheses reflexively from insights into your data gleaned during interviews. Your role as researcher is implicit and has to be considered too – how will it will influence the data you obtain? This sort of interactive data collection may be spread out via individual interviews, or achieved all at once with a focus group. Grounded theory is the term used to define this approach, as the theory is grounded within, emerges from and is tested by the data.
Pros and cons of qualitative study
Objections to the qualitative approach include that it is subject to observer bias, may lack reproducibility, and is at risk of presenting anecdote and personal impression rather than insights deduced from data. (It has been said that quantitative methods are reliable but not valid, and qualitative methods are valid but not reliable.) Validity, reliability and generalisability have to be addressed by meticulous explanation and demonstration of how ideas were formed from the data, so that an independent researcher could come to the same conclusions. Results and discussion are therefore implicitly combined, rather than presented as separate sections of the write-up.
Funding and advice
Advice is available from your local research and development support unit or university department where you will find experience and information on both research methods and good sources of funding. Register with your local R&D committee, who will be pleased to have you on board, and may be able to put you in touch with experienced researchers with similar interests. Consider asking someone to act as an educational supervisor to keep you on track.
Useful short courses and day-workshops at university level are also available on research methods that can give you a good idea of the best tactics – and save you time.
Grants are available from bodies such as the RCGP, BMA, and local and national charities. A formal RCGP grant application is a considerable undertaking involving much photocopying, but smaller sums (under £2,000) can be awarded by the RCGP Scientific Foundation on the Chair's discretion, with a quick decision in a few weeks. Costs of transcription, typing and stationery may be covered in this way. Remember the biggest expense is your own time and midnight oil.
Other similar small amounts are available locally – for example, in my area the Primary Care Research Collaborative provides bursaries of up to £500 to help development work towards preparing a research protocol or grant application and writing a paper. Small project grants of up to £2,500 for peer-reviewed primary care research and primary care research grants of up to £10,000 for larger projects are also available. Similar grants are likely to be available in your own locality (see websites at top of page).
If you want to take advantage of these research opportunities, first pause and think. Research is very time-consuming so keep your project as basic as your enthusiasm will permit.
Stefan Cembrowicz is a GP and trainer in Bristol
Next week: how to get started on your qualitative research project