Knowledge Why bother?

1
KnowledgeWhy bother?

• Alan, Gavin and Neil

2
Todays Aims and Objectives

• To initially make you feel a bit uncomfortable
• To make you think and get your brains cogs
  whirring
• To (hopefully!) leave you feeling slightly
  better, and equipped with some practical tips
  that will help you in the exam, and your medical
  career

3
Medicine is an ocean of information

• How do we keep ourselves from drowning in the
  continuous flow of information poured upon us in
  the form of journals, papers, textbooks, courses
  etc. ?

4

• Discuss as a group possible strategies to keep
  updated

5
Some Practical Tips

• You dont have to know everything
• Limit yourself to a small number of relevant
  journals e.g. BMJ and BJGP
• Scan the contents, read the abstracts/conclusions
  of papers, then go back in more detail later if
  necessary/relevant

6
Some Practical Tips

• Do little and often
• Try to set realistic goals
• Try to learn around patients easier to
  retain/recall info
• Sign up to GPnotebook tracker (or similar)
  keeps a record of your learning good for PDP

7
The truth, the whole truth, and nothing but the
truth?

• Dismantling the misinformation in research papers

8
Why do we need to appraise research ourselves?

• Dont believe what you are told by anyone
  (including me)
• Spinning statistics is an industry standard
• Scientists do it to get papers published
• Governments do it to minimise expenditure
• Big Pharma does it to get you to prescribe their
  drugs
• Its important to know what you need to know
• Too much exists out there to read everything
• Discarding the rubbish is an important skill
• In an information world, you are going to be
  relied on more often to give patients advice (in
  effect appraise their evidence)

9
Lies, damn lies and Statistics

• You dont have to be able to do statistics to read
  a paper insightfully
• There are a few key questions to ask about any
  research paper
• Why was it done?
• How was it done?
• What havent they told you?
• Do the results support the claims of the authors?

10
Its all relative, innit?

• What does a Relative Risk mean? An example
• Relative risk of coffee drinkers versus tea
  drinkers acquiring vCJD 0.70 (95 CI 0. 68 -
  0.72) (ref 1)
• In plain English, I estimate coffee drinkers are
  30 less likely to acquire vCJD than tea
  drinkers.
• My margin of error in this estimate lies between
  24-30 less likely to acquire vCJD (Im 95
  confident of this)
• How would NESCAFE spin this information?

11
Its absolutely nothing.

• Its ABSOLUTE RISK what matters
• If risk of vCJD is 1 in 10 million for tea
  drinkers, and 1 in 7 million for coffee drinkers,
  would you really advocate every patient switching
  to coffee?
• What if it was 1 in 10 tea drinkers, and 1 in 7
  coffee drinkers?
• In both, the relative risk reduction is the same
  (30), but the absolute risk reduction is very
  different

12
Absolute risk

• Coffee 1 in 10 million 0.000000010
• Tea 1 in 7 million 0.000000014
• The absolute risk reduction by switching from tea
  to coffee is the difference in rates
  0.000000004
• How can we translate this into a meaningful
  figure for a patient to understand?

13
Numbers needed to treat (NNT)

• This is the most valuable tool you have at your
  disposal for making sense of statistics
• NNT 1/absolute risk reduction
• In this example, it is the number of patients you
  need to switch from drinking tea to coffee to
  prevent 1 additional case of vCJD
• NNT 250,000,000 (or the whole US population)

14
vCJD reference

• A multi-centre nested case-control study of the
  effect of beverage consumption on vCJD risk
  analysis from the NESCAFE cohort study.
• J. Swallow and P. Spitt (2006). Swedish Journal
  of Epidemiology, 43 69-99.
• And if you swallowed that, you will swallow
  anything!

15
Part 1 Take home message

• Beware studies that sell their message on
  relative risk reductions or improvements
• Always try to determine what the absolute risk
  difference is
• Convert this to NNT for a meaningful value you
  can explain to patients (or yourself)

16
Small group work

• Split into 4 groups
• Pharmaceutical reps
• PCT pharmacy commissioners
• GPs
• Patients (this is the easy job!)
• Look at the abstract of this paper on Clopidogrel
  versus Aspirin, published in The Lancet, and
  follow the briefing notes for each group

17
You have 15 mins

• Work in groups to produce an argument for your
  case, and nominate a spokesperson to give a brief
  talk to the patient interest group.
• The patient group must then vote to decide who
  gives the most convincing argument, and state
  what convinced them to vote that way.

18
How did it turn out?

• The Drug reps will have pushed the relative risk
  rather than absolute risk, and side effect
  profile
• The PCT will have highlighted the cost to prevent
  each additional MI (NNT) and suggested patients
  should not receive Plavix
• The GPs will have been caught in the middle!
• The patients will have found it difficult to see
  beyond their own interests

19
Notes from the full paper

• In most cases, there was no significant
  difference in all cause mortality between the two
  treatment arms (not mentioned in abstract)
• The side effect profile is similar with no
  differences
• The study was funded by the makers of Plavix, and
  no conflict of interest was stated

20
Critical appraisal summary

• Consider the motives behind presentation of
  results by the authors
• Ask yourself if the study fits your patients
• Ask yourself if there is a conflict of interest
• Dont be frightened by high power statistics you
  should be able to ask the same question of any
  trial how many patients do I have to treat, and
  what is the cost to prevent one event.