EvidenceBased Medicine

1
Evidence-Based Medicine

• Chris Cairns
• February 2007

2
Evidence-Based Medicine
3
EBM

• EBM is
• Integration of
• Best research evidence
• Clinical skills and experience
• Patient values
• EBM is not
• Just reading papers
• The removal of common sense
• Meant to be inflexible

4
EBM

• Only the clinician can decide whether any
  evidence is relevant to their patient
• There are plenty of alternatives for the
  practising physician in the absence of evidence.
  This is what makes medicine an art as well as a
  science. (David Isaacs)

5
Reading

• To keep up to date
• I need to read 17 articles a day, 365 days a
  year.
• I need to read.
• I dont (well not that much!).
• Nor does anyone else.
• So.I need to read effectively

6
To start with..

• Whats the study trying to do..
• What patients?
• What intervention?
• What outcome?
• Primary
• Secondary
• Patient orientated vs non-patient orientated

7
so.

• Does that interest me.?
• No.Then move on.
• Yes..Then is it worth reading?

8
Key questions to ask

• Do the methods allow accurate testing of the
  hypothesis?
• Randomised
• concealment
• Controlled
• similar patients, other interventions identical
• Blind
• patients, clinicians, investigators

9
Key questions

• Do the statistical tests correctly test the
  results to allow differentiation of statistically
  significant results?
• Outcomes
• Meaningful
• Measured in standard, valid, reliable way
• Stats.

10
Absolute Risk Reduction

• ARR CER EER.
• Normally expressed as a percentage
• It is the arithmetic difference in occurrence
  between the control and experimental groups. 

11
Relative Risk Reduction

• RRR (CER EER) / CER.
• Normally expressed as a percentage. Similar to
  ARR but gives a proportional reduction.
• For example if an intervention cuts the mortality
  from a disease from 2 to 1 there is only an AAR
  of 1 but a RRR of 50  

12
Number Needed to Treat

• NNT 1 / ARR.
• The number of patients needed to treat to achieve
  one additional good outcome.
• An excellent way of looking at interventions as
  this corrects for low occurrence rates.

13
NNT

• For example an intervention may have a fantastic
  RRR of 50 but only reduces the rate of a rare
  complication from 0.2 to 0.1 (ARR 0.1)
• The RRR ARR may be misleading.
• The number of patients needed to treat to avoid
  one complication would be 1000. This makes
  judgments of risk / benefit / cost more
  straightforward.

14
Van den Berghe, NEJM, 2001

• Death (ICU)
• CER 63/783 8
• EER 35/765 4.6
• ARR CER EER 8 - 4.6 3.4
• RRR (CER-EER)/CER 3.4/8 43
• NNT 1/ARR 1/0.034 29

15
key questions

• Did results get omitted, and why?
• Intention-to-treat
• Per Protocol..

16
key questions

• Are conclusions valid in light of the results?
• Methodologically sound?
• Statistically correct?
• Conclusion sound?
• Scientifically
• Clinically

17
key questions

• Did they suggest areas of further research?
• Multi-centre?
• Wider population?
• Larger study?
• Difference regimen?

18
key questions

• Did they make any recommendations based on the
  results and were they appropriate?
• Credible link between results and clinical
  recommendations?

19
key questions

• Is the study relevant to my clinical practice?
• Control group
• Similar to your patients?
• Similar to what you do?

Applicability
20
Applicability

• Type of patients
• Severity of illness
• Management prior to study entry
• When randomised
• Inclusion criteria
• Exclusion criteria
• i.e. Are these similar to my patients?

21
key questions

• What level of evidence does this study represent?
• What grade of recommendation can I make on this
  result alone?
• What grade of recommendation can I make when this
  study is considered along with other available
  evidence?

22
Levels of Evidence
23
Levels of Evidence

• 1,2,3,4..purely descriptive
• - , , .. risk of bias
• Gives you little idea of clinical impact on the
  patient (remember back to NNT).

24
Grade of recommendation
25
Grade of recommendation

• Recommendation graded according to the risk of
  bias.
• Again the clinical impact needs to be weighed up
  when considering which recommendations are the
  most important.

26
key questions

• Should I change my practice because of these
  results?
• Should I audit my current practice because of
  these results?

27
Changing practice

• Were the results valid?
• Was it relevant to your practice?
• Control group comparison?
• Your interventions?
• Your outcomes?
• Your costs?

Yes?
28
Changing practice..

• Local factors..the Killer Bs
• Burden of illness
• Beliefs of patients or communities
• Bargain
• Barriers
• Geographic
• Organizational
• Traditional
• Authoritarian
• Legal
• Behavioural

29
A possible solution

• www.sicsebm.org.uk

30
Information Sources for Use at the Point of Care

• Everything is based on the usefulness equation
• Usefulness Relevance x Validity
• Work

31
What you really want to know

• In .. Patients managed with, rather than..,
  one additional life will be saved for
  every..patients treated.
• Some indication of risk of bias
• Further information if you want to read on.

32
The Solution

• One of these

33
(No Transcript)
34
Critically Appraised Topic (CAT)

• A one page summary
• Declarative title
• Question
• Name of paper
• Search terms
• Design
• Setting
• Patients

• Intervention
• Outcome Measures
• Results
• Table
• Conclusion
• Commentary

35
Bottom line read in seconds
36
Trial details read in minutes
37
Read trial details (minutes)
38
Particularised for your own practice, integrate
with your expertise
39

www.sicsebm.org.uk
40
2003

• Website launched
• 4 Reviews
• ARDS I
• Gastric Prophylaxis
• Insulin
• Sepsis
• JICS editorial

41
2004

• 5 Reviews
• Therapeutic Hypothermia I
• Making Changes to Practice
• Prevention of VAP
• Guillian-Barre Syndrome
• NO Proning in ARDS
• 11 CATs of published papers
• Equates to 47 new CATs
• Research/EBMG meeting

42
2005

• Reviews
• CVCrBSI
• New projects started
• METs
• Outreach
• 20 free CATS, 1 review
• 56 in 2005 in total
• 136 now on website

43
2006

• JICS partnership
• One review
• Early tracheostomy
• 20 CATS (11 in JICS)
• 163 CATS on the website

44
JICS / ICS

• JICS now using SICS format
• JICS CATS published on SICS web site
• SICS EBM editing input with JICS
• SICS EBM publications listed in JICS
• EBM updates in ICS e-letter

45
Website

• Evidence-Based Intensive Care
• Google 1st
• BT.yahoo 2nd
• Alexa.com 2nd
• Lycos 3rd
• Yahoo 10th
• Several other features
• Feedback
• Downloads
• Updates to all SICS members
• Discussion Forum
• Jargon

46
Website Visitors
20,152
47
Website Visitors
47,988
20,152
48
Website Visitors
2005 79,541
2004 47,988
2003 20,152
49
Monthly Website Visitors
2006 gt110,00
2005 61,326
2004 47,988
2003 20,152
50
Cumulative Sessions
51

www.sicsebm.org.uk