Research
Why do research?
Most people perform research for a combination
of reasons including:
- Intellectual curiosity
- Improving diagnosis and management
- C.V. points
What type of research?
The foundation years do
not offer much time for research, even if you are on an academic
program. This doesn’t mean you can’t take part in any research,
just that you have to plan and schedule it carefully, and think
about the pattern and frequency of your availability. For example,
can you commit to any regular daytime availability (such as likely
to be needed for laboratory work)? Do you have a block of time
free? Or do you need a project that can be done in your evenings
and weekends only?
Tips for a simple and flexible project:
- Act as a research
assistant for a project set up by someone else
- Analyse an existing data
set in a new way
- Literature reviews can
be very flexible as to when time is spent on them, but generally
take a lot of time overall
- Think carefully before
starting a new project requiring collection of new data from human
subjects – the time commitment is likely to be unpredictable and
ethical permission will be needed, which can take a significant
period of time
- Whatever question you
think of, try to refine it until it is as simple as you can make
it. Most projects get more complex as they progress, so aim to at
least start off simple.
How to do research?
A full guide on how to
perform research would require several tomes and ideally a
lifetime’s experience. However, some top tips are below.
- Find a friendly
supervisor
The single most important
piece of advice is to find someone experienced to work with. This
reduces your risk of reinventing the wheel, going off at irrelevant
tangents or missing an obvious problem and should help the research
be focussed and useful. Science is a team effort.
- Find an approachable
statistician
Most research is
quantitative, producing numerical output of some sort. If you are
planning to do anything more exotic with your numbers than total
them up, you should speak to a statistician as soon as possible.
Statisticians like to be approached during the planning stages of a
project, when they can conceivably ensure the right data will be
collected. For example, if you are collecting new data you will
almost certainly have to perform a power calculation to estimate
sample size before you start the research project. Statistics are
confusing and often poorly done, and the more help you can get the
better.
- Factor in lots of
waiting time
Research is a lengthy
process sandwiched between even lengthier processes of ethical
approval and peer-reviewed publication. If you are starting a new
project requiring collection of data from human subjects, you will
need to apply for ethical permission and potentially attend a
meeting with an ethics committee. Depending on how often the
committee meets and what their backlog is, the wait from
application to approval can be weeks to months. If you are
submitting a paper to a journal, there is also likely to be a wait
of weeks to months for the review process to occur. If it is
accepted, there is then likely to be a further wait until it is
actually published.
Like most activities,
research can be mundane and frustrating at times. It can also get
overly interesting, in that it hardly ever goes exactly to plan. If
you hit a problem, don’t take it personally; talk about it with
someone else. Very few problems are fatal to a project. Remember
the overall purpose of the research and why you thought the topic
was worth investigating in the first place. With enough
perseverance and luck, you will come out at the end with a new
contribution to knowledge and potentially benefit
patients.
Back to top
Audit
Why do audit?
- To improve practice.
- Essential competency for foundation year doctors.
What is audit?
Audit is comparing practice with predefined
standards.
It can be difficult to
differentiate from research as although the purposes are different
the techniques overlap. However, it is an important distinction as
research requires ethical permission and consent, whereas audit
generally does not.
Audit often involves case
note reviews, but a case note review is not necessarily an audit.
For example, looking through notes to assess which SSRI patients
were started on in order to compare choice against the NICE
guidelines is audit (CG90
Depression in adults: quick reference guide, NICE, 28 October 2009) . Looking through notes to assess
which SSRI patients were started on in order to assess treatment
responses to different initial SSRIs is research.
Audit vs research is well
known to be a grey area. The National Research Ethics Service has a
number of tables to help you categorise a project. (eg see
‘Defining Research Leaflet’ available from http://www.nres.npsa.nhs.uk/).
How to audit
- Identify areas important to the service you
are working in where you suspect practice may not be meeting
standards
- Choose an area that interests you
- Identify relevant guidelines
- Choose criteria from the guidelines to assess
against
- Choose the population to assess
- Collect relevant data from the relevant
population
- Compare practice against criteria and
identify any areas of difficulty
Steps necessary if audit has demonstrated a
gap between practice and standard:
- Intervene to improve
- Reaudit to assess effectiveness of
intervention
As with research, this
process can be greatly eased by speaking to someone with experience
who can direct you towards useful areas to assess and existing
guidelines. Some services have audit managers or departments who
may be able to provide advice and ask you to register your audit
with them.
There are sometimes
rolling audits you can get collect data for. These are likely to
have been set up by senior members of the department and so be of
value and interest to the unit. It is also helpful for the team if
you collect data for these audits. However, employers generally
hope you will have been involved with audit design, analysis and
presentation, not just data collection.
Which gbuidelines?
Guidelines can be local or national, or both.
Try to choose some that are relevant to the setting you are
auditing. Possibilities include:
- Local hospital
guidelines eg pharmacy, radiology departments
- College or Society
Guidelines eg British Geriatrics Society guidelines on
dementia
- National Guidelines eg
NICE, SIGN
What population?
This is a deceptively
simple question, best answered before you waste your time
collecting data on patients not covered by the guideline remit.
Ask yourself:
- Any age? Or only over or
under a certain age?
- Anyone on the ward? Only
patients under the home team, or boarders too?
- Any admissions?
Admissions with a certain presenting complaint? Admissions without
a certain presenting complaint? Consecutive admissions?
- Is the audit ‘snap-shot’
(patients now), retrospective (past patients), or prospective
(patients to come)?
Closing the loop
This is audit jargon for performing an
‘intervention’ then reauditing afterwards to assess for change.
The overall aim of audit
is not just to document performance, but to improve it. So if the
first loop of your audit identifies some areas where guidance is
not followed, intervene in some way to rectify it.
Possible interventions include:
- Education, via a
presentation and/or posters.
- Make materials
accessible, for example if a standardised assessment is not being
performed, are there any copies of it on the ward? Are copies kept
in a clearly labelled drawer?
- Reminders, for example
if risk for DVT prophylaxis is not being assessed, you could put
stickers on the front of drug kardexes with a list of
thromboembolism risk factors.
Useful resources for research and
audit
- National website with
information about how to gain ethical approval and the difference
between research and audit: http://www.nres.npsa.nhs.uk/
Katie Marwick, foundation year 2 doctor
in South-East Scotland
Back to top
Back to foundation
trainees' home page
Page last updated on 22 May
by E Baker-Glenn