Clinical review

ABC of health informatics

Keeping up: learning in the workplace
Jeremy C Wyatt, Frank Sullivan

The amount of biomedical knowledge doubles every 20 years,

This is the ninth in a series of 12 articles
and new classes of drug (such as phosphodiesterase 4
A glossary of terms is available at http://bmj.com/cgi/
inhibitors) become available when lectures at medical school content/full/331/7516/566/DC1
are over. Therefore, a practice risks fossilising after doctors
finish professional training. Many continuing medical education
or continuing professional development activities help doctors
carry on learning and improving their skills. These activities
include courses, conferences, mailed educational materials, Patrick Murphy is a 8 year old boy who has recently
returned home from a hospital admission. The discharge
weekly grand rounds, journal clubs, and using internet sites. In
letter asks you to prescribe inhaled steroids and a
many countries, evidence of this process is needed for doctors
phosphodiesterase 4 inhibitor
to continue to practice. Although these activities may increase
knowledge, their impact on clinical practice is variable
The aim of traditional medical education is to commit
knowledge to memory and then use this knowledge in the
workplace. The way knowledge is learnt influences its recall and Workplace learning means finding solutions to clinical
application to work. One tactic to improve the process is to problems when they arise, or soon after, with minimum
effort. When unsure about what has happened, why, or
ensure that learning happens in the clinical workplace. Lessons
what to do, answers should be looked up
are learnt faster and recalled more reliably when they originate
in everyday experience.
Learning in the workplace means spending a minute here
or three minutes there to find answers prompted by the clinical
questions and learning opportunities that come up in every
working day, rather than doing continuing medical education
for an intensive two hours a week, or a few days a year.
Workplace learning is hard to achieve. It emphasises problem
solving and learning skills—such as how to find relevant answers
fast—not learning facts.

Barriers and solutions

Nobody can find a satisfactory answer to every clinical question
or information need, especially as there are about two needs for
every three clinical encounters. Many important clinical
questions have no satisfactory answer—for example, what is the
cause of motor neurone disease? Other questions are simply
interesting rather than information needs. A range of practical
difficulties face doctors who follow the approach of learning in
the workplace. Some suggestions about how to overcome the
difficulties follow.

In the United Kingdom, the National electronic Library for Health aims to
Too many questions, not enough time provide answers within 15 seconds that take only 15 seconds to read

Doctors generate approximately 45 questions about patient

care every week, and they probably allow two minutes to answer
each one. This adds up to an extra hour and a half per week,
and even though it represents only 3% of their working time,
where do doctors find this time? Time is always short. They
often have to adjust the threshold for seeking answers, Prioritisation of clinical questions
prioritising questions that have the highest clinical impact and 1 Answers needed now
are quickest to answer. 2 Answers needed before patient is seen next
Prioritising clinical questions by the likely impact of the 3 Answers needed to guide care of other patients
or to reorganise clinical practice
answer means distinguishing between the questions in the box
4 Answers that have interest to doctor and
opposite. When doctors have time, they can pursue all answers. patient, but carry no obvious clinical impact
When under pressure, they pursue answers that are needed
now (category 1). If they never pursue other answers, they will
miss many clinical advances. It is often hard to recognise when
knowledge is lacking, and so it is important to sometimes
pursue answers even when only slightly uncertain of the answer.

BMJ VOLUME 331 12 NOVEMBER 2005 bmj.com 1129

Clinical review

To ease time pressure, clinicians can spend less time

answering a question by using knowledge resources that are
comprehensive, and can be instantly accessed and easily Oxford Centre for EBM question log book

searched. They could also increase the time available for

workplace learning. Individually, doctors can work for longer Patient ID: Date:
hours, reserving time for “reflective practice” with a preceptor Issue:
or mentor, exploiting “teachable moments,” perhaps by
answering an educational prescription. Overall, the medical
profession needs to recognise the sanctity of workplace learning Question (PICO):
throughout doctors’ careers: life long, self directed learning.

Search:
Lack of clear questions
Main Reference:
Asking clear questions is not easy. Sometimes doctors feel
uncertain and fail to formalise a question, which makes it Adapted from a
page from the
harder to find the answer. Immediate identification of clinical Oxford Centre for
questions is important, and is easiest to do on ward rounds or R Y/N/? A % M
Evidence Based
when teaching students. When working alone, some clinicians Clinical bottom line: Medicine’s logbook
log their questions (for example, on BMJLearning), then look (R=randomised and
representative,
up the learning resources on the website (the “just in time A=ascertainment or
learning” package on childhood asthma) or other sources, or Notes: follow-up rate
they discuss the answer with peers later. Structuring clinical percentage,
questions using the problem, intervention, comparison and M=measures
unbiased, relevant
outcome (PICO) model makes them easier to focus, recall, and
answer.

Turning clinical problems into easily investigated formats

Intervention (or cause,
prognostic factor,
Patient or problem treatment) Comparison (if necessary) Outcomes
Tips for building Starting with your patient Ask “Which main Ask “What is the main Ask “What can I hope to
ask “How would I describe intervention am I alternative to compare with accomplish?” or “What
a group of patients similar considering?” Be specific the intervention?” Be specific could this exposure really
to mine?” Balance affect?” Be specific
precision with brevity
Example (see In children with poorly “. . . would adding “. . . when compared with “. . . reduce the likelihood
scenario on controlled asthma . . . phosphodiesterase 4 inhibitor adding a long acting ß of readmission?”
p 1129) to inhaled corticosteroid . . .” agonist . . .”

Lack of answers
A source of answers needs to be available in the workplace. This
source should provide answers that are clinically relevant,
scientifically sound, and in a form that can influence decisions.
One solution is a library in the workplace that contains current
text and reference books, relevant reprints, and electronic
resources. The library must be close and organised for rapid
access. The material should be filtered for clinical relevance and
be evidence based, such as Clinical Evidence in book or CD-Rom
format or an indexed collection of systematic reviews.
These sources will not answer all questions. In Patrick
Murphy’s case (see scenario on p 1129) the treatment is not
indexed, and so online access to Medline will be needed,
preferably via the PubMed clinical queries search page that
provides answers useful to practicing doctors. Ideally, doctors
will then retrieve the full text of relevant articles because relying
on the abstract alone can be misleading. When Pitkin compared
the statements made in 264 structured abstracts in six medical
journals with the corresponding article, a fifth contained
statements that were not substantiated in the article and 28%
contained statements that disagreed with those in the article.
Clinical evidence is a useful resource in workplace learning
Thus, tempting though it may be to rely on abstracts
alone—especially because they are now so accessible through
PubMed—it can be dangerous.

1130 BMJ VOLUME 331 12 NOVEMBER 2005 bmj.com

 Clinical review

An alternative to carrying out the search yourself is to call

or email a question answering service, such as ATTRACT, for
clinicians working in Wales. For years, NHS poisons and drug
information services have provided similar services that give
instant answers to specialist questions. Some libraries, primary
care trusts and academic departments have services that cover
many topics. The service usually returns a telephone call or
sends a summary within two to four hours. Despite their
obvious potential, these services seem underused at present.

Parochialism
If doctors only look up answers to questions arising in their
own practice, their knowledge will depend on the local case
mix. Most doctors broaden their knowledge by reading a
general medical journal or looking up points raised in replies to
referrals, inpatient summaries, clinic letters, or laboratory
reports. Some participate in multidisciplinary clinics or ward A PubMed search filters for clinical queries
rounds, or join colleagues in an email discussion group. To be
ready for rare, serious problems that need an instant response,
some clinicians use patient simulators to practice managing
cardiopulmonary arrest, anaesthetic accidents, or brittle
diabetes. Although time spent on simulators does not yet count
towards doctors’ continuing education, taking part in interactive
cases in some journals does.

Lack of incentives
To maintain the enthusiasm to keep looking up answers to
clinical questions, doctors can keep a log book of questions and
answers, or conduct clinical audits that compare practice and
outcomes with results a year ago. Such log books and audit
reports will become part of doctors’ folders for accreditation
and annual appraisal.
Sharing insights is an incentive to learn, and giving a
presentation often prompts discussion, especially if it is short,
and it defines and deals with a real clinical problem (along with
sources searched, the answers found, and actions taken). This
activity can be formalised as a single page, dated, critically
appraised topic (CAT), and stored in a loose leaf folder or a
practice intranet for others. The CATmaker tool is used to create critically appraised topics

Cultural changes associated with workplace learning

Old think New think
x Passive listening to lectures x Active participation in self directed learning
x Educator decides topic x You decide topic
x Attend continuing medical education course you know most about x Seek out areas of ignorance and answers to your clinical questions
x Focus is on laboratory research, pathophysiology, drug mechanisms x Focus is on what works in practice, what to do, problem solving
x Read a journal or textbook x Carry out problem solving on real or simulated cases
x Education to learn facts, pass exams x Learning to solve clinical problems, improve team work, clinical and
information seeking skills
x Formal, timed courses x Informal, self directed, learning in the workplace
x Get continuing medical education or postgraduate education x Get continuing medical education or postgraduate education
allowance points for turning up allowance points for participating in workplace learning, using
learning materials, improving standards
x Case presentation, journal club x Work on an educational prescription, write a critically appraised
topic, use a clinical simulator
x Competition: keep knowledge to yourself x Sharing: open learning, exchange of knowledge and understanding
to benefit patients and the health system
x Knowledge belongs to the individual. Continuing medical education x Communities of practice: learning is an attribute of the team and
points accumulate to the individual. Recertify the individual organisation and is part of its quality and risk management
strategies. Accredit the organisation
x Patients are passive recipients of care x Patients are sources of questions and insights, learning collaborators
x Errors should be forgotten and denied x Errors are a learning experience to be treasured, discussed, and
understood
x Errors happen to “bad apples” x Errors happen to everyone

BMJ VOLUME 331 12 NOVEMBER 2005 bmj.com 1131

Clinical review

Lowering barriers is also motivating: an old BNF in a desk

Further reading
drawer will be used more often than a current version in the
practice library 10 m away, or one in the health library 5 km x Wyatt J. Use and sources of medical knowledge. Lancet
1991;338:1368-73
away. Electronic libraries and the internet bring the world’s
x General Medical Council. A licence to practice and revalidation.
literature to your desktop, but can take longer and yield fewer London: General Medical Council, 2003
answers to clinical questions than paper sources. This is x Mazmanian PE, Davis DA. Continuing medical education and the
changing. A German study found that clinical use of online physician as a learner: guide to the evidence. JAMA
learning was about ten times that of print journals. 2002;288:1057-60
x Lave J, Wenger E. Situated learning. Cambridge: Cambridge
University Press, 1991
Summary x Ebell MH, Shaughnessy A. Information mastery: integrating
continuing medical education with the information needs of
Barriers to workplace learning can be overcome, but a minor clinicians. J Contin Educ Health Prof 2003;23:53-62
culture change in the medical profession is needed. This shift is x The resourceful patient website. The e-consultation: vignette.
www.resourcefulpatient.org/resources/econsult.htm (accessed 30
already taking place in undergraduate medical education and in
October 2005)
primary care. Clinical governance, risk management, patient x Smith R. What clinical information do doctors need? BMJ 1996;
empowerment, and the National Programme for IT will further 313:1062-8
advance the change. x Ely JW, Osheroff JA, Ebell MH, Chambliss ML, Vinson DC,
Using clinical questions to guide workplace learning relies Stevermer JJ, et al. Obstacles to answering doctors’ questions about
on the motivation of individuals, teams, and organisations. It patient care with evidence: qualitative study. BMJ 2002;324:710
x PubMed clinical queries: www.ncbi.nlm.nih.gov/entrez/query/
goes hand in hand with an open attitude to clinical errors and
static/clinical.html (accessed 30 October 2005)
near misses. Motivation is especially necessary to fund the x Pitkin RM, Branagan MA, Burmeister LF. Accuracy of data in
instant access resources needed to provide knowledge during abstracts of published research articles. JAMA 1999; 281:1110-11
clinical work. Fortunately, electronic media provide a simpler, x ATTRACT: www.attract.wales.nhs.uk/index.cfm (accessed 30
cheaper method for workplace learning than paper libraries, October 2005)
although there is evidence that health librarians on site are still x Harker N, Montgomery A, Fahey T. Treating nausea and vomiting
needed to support better clinical use of these resources. during pregnancy: case outcome. BMJ 2004;328:503

Jeremy C Wyatt is professor of health informatics, and Frank Sullivan

is NHS Tayside professor of research and development in general The series will be published as a book by Blackwell Publishing in
practice and primary care, University of Dundee. spring 2006.

BMJ 2005;331:1129–32 Competing interests: None declared.

Symbolism in imagery and linguistics—perception and misperception

Puzzling consultations are not unusual and help make medicine Image analysis and object recognition have advanced in recent
so fascinating. In this case, however, the clinical issues were years. Functional magnetic resonance imaging has recently been
resolved by simple investigation and reassurance. My puzzlement used to investigate regions of the brain involved in face
was from my certainty that I had met this patient previously, but recognition 2. A face-selective region of the inferior temporal lobe
my inquiries elicited a confident reply that we had not. I struggled seems to be involved in facial identity, while a separate
with this for, like most physicians, I have always enjoyed a facility face-selective region of the superior temporal lobe may be used
for recognising my patients. For several weeks, the question to identify facial expression. My failure to “recognise” my patient
recurred in my mind; how could I be so certain of recognising a may reflect an absence of any associated memories such as
patient whom I had never met before? The issue was resolved altered facial expression or personality attributes usually
when I did meet my patient again; this time while sitting on the associated with face identity. This suggests that both face-sensitive
train I took daily to and from work—not, however, as a fellow regions identified by Andrews and Ewbank are important for the
passenger but as a face in a drug advertisement in the medical wider process of person identity and recall. My experience shows
journal I was reading. the power of images in advertising, but also their limitations; I
Two recent reports relating to images and their interpretation remembered the face but had no recollection of its context or
brought this incident to mind after an interval of 10 years. Scott circumstances.
et al discuss the use of symbolism in images in drug advertising, Capturing the attention of readers is challenging for advertisers
suggesting that the use of mythology may be used to mislead
and writers alike. Scott et al argue that “in law and science, words
doctors about therapeutic efficacy.1 They suggest that doctors
are precise and accountable” and imply that writing occupies
need support to resist drug advertising, and they recommend
higher ethical ground; but it is evident that symbolism is
regulation of journals’ advertising to clinicians and lament the use
frequently used to reinforce messages, whether in linguistic or
of rhetorical rather than rational argument in scientific media.
image form, and not only in advertising. Let the reader beware.
However, it is evident that the authors use the very techniques
they warn about to persuade their readers. The article, entitled D J Sheridan professor of cardiology, Imperial College School of
“Killing me softly; myth in pharmaceutical advertising,” borrows Medicine, Academic Cardiology Unit, St Mary’s Hospital, London
from the title of Fox and Gimbel’s hit song recorded by Roberta ([email protected])
Flack in 1973 to capture the reader’s attention. Juxtaposing the
word “Killing” with “pharmaceutical advertising” conveys an
1 Scott T, Stanford N, Thompson DR. Killing me softly: myth in
unhealthy and negative sense and transforms the metaphor from pharmaceutical advertising. BMJ 2004;329:1484-7.
a romantic to an aggressive one. The word “myth” in this context 2 Andrews TJ, Ewbank MP. Distinct representations for facial identity and
is also ambiguous, carrying a sense of deception as well as the changeable aspects of faces in the human temporal lobe. Neuro Image
classical meaning. 2004;23:905-13.

1132 BMJ VOLUME 331 12 NOVEMBER 2005 bmj.com