Jargon

Web site designed and maintained by Chris Cairns © SICS EBM Group 2004

EBM jargon

What it's all about ?:

'Evidence-based Medicine is the enhancement of a clinician's traditional skills in diagnosis, treatment, prevention and related areas through the systemic framing of relevant and answerable questions and the use of mathematical estimates of probability and risk' (A. Donald & T. Greenhalgh 2000)

It relies on the integration of three key elements:

Best research evidence
Clinical expertise
Patient values

How can we go about this ?:

Step one: Ask the question
- Which patients?, which treatment or test?, what outcome?
Step two: Find the best evidence
- search: electronic, paper, colleagues

Step three: Appraise the evidence you've come across
- Is the evidence any good ?
- Will it provide a sufficient benefit over your current practise ?
- Is it relevant to your patients?
Step four: Combine steps 1-3 with your
- clinical expertise and,
- considering the patients specific problems
Step five: Evaluate, change if indicated, re-evaluate

As you can see we can help with steps 1-3, the rest still relies on your clinical expertise. This should reassure the EBM sceptics. EBM is not designed to remove the need for clinical expertise but to enhance it.

Here is a quick guide to some of the terms used in this web site

Confidence interval (CI): Quantifies the uncertainty in measurement. It is usually reported as a 95% CI, which is the range of values within which we can be 95% sure that the true value for the whole population lies.

Event rate (ER): The proportion of patients in a group in whom the event is observed. The rates in the control and experimental groups are referred to as the control event rate (CER) and experimental event rate (EER) respectively. The patient expected event rate (PEER) refers to the rate of events we’d expect in a patient who received no treatment or conventional treatment.

Sensitivity / specificity / Likelihood ratio (LR)

	Disease positive	Disease negative
Test positive	A	B
Test negative	C	D

Sensitivity = a/(a+c) = the nearer 1.0 this figure the more sensitive the test is (low number of false negatives)

Specificity = d/(b+d) = the nearer 1.0 this figure the more specific the test is (low false positive rate)

LR+ = sensitivity / (1 – specificity) = the likelihood that positive test result is found in a patient with the disorder compared with a positive test result in a patient without the disorder.

LR - = (1 – sensitivity) / specificity = the likelihood that a negative test is found in a patient without the disorder compared with a negative test in a patient with it.

Negative predictive value (NPV): = d/(c+d) = proportion of patients with a negative test who do not have the disease. (ie a low number means a high false negative rate)

Positive predictive value (PPV): = a/(a+b) = proportion of patients with a positive test who do have the disease. (ie a low number means a high false positive rate)

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 process from 2% to 1% there is only an AAR of 1% but a RRR of 50%

Absolute risk reduction (ARR): = CER – EER. Normally expressed as a percentage it is the arithmetic difference is occurrence between the control and experimental groups.

Number needed to treat (NNT): = 1 / ARR. The number of patients needed to treat to achieve one additional good outcome. An excellent was of looking at interventions as this corrects for low occurrence rates. 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%. The RRR 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.

Number needed to harm (NNH): = 1 / ARR when outcome is worse in the experimental group.

Level of Evidence and Grade of Recommendation

Throughout this site validity is summarised using the Scottish Intercollegiate Guidelines Network grading recommendations. It is important to note at this stage that the grade of recommendation reflects the strength of the evidence (methodological quality) and not the clinical importance.

Level of Evidence

Level	Description
1⁺⁺	High quality meta analysis, systemic review of RCTs, or RCTs with a very low risk of bias
1⁺	Well conducted meta analysis, systemic review of RCTs, or RCTs with a low risk of bias
1^-	Meta analysis, systemic reviews of RCTs, or RCTs with a high risk of bias
2⁺⁺	High quality systemic reviews of case-control or cohort studies. High quality case-control or cohort studies with a very low risk of confounding, bias, or chance and a high probability that the relationship is casual.
2⁺	Well conducted case control or cohort study with a low risk of confounding, bias, or chance and a moderate probability that the relationship is casual.
2^-	Case control or cohort studies with a high risk of confounding, bias, or chance and a significant risk that the relationship is not casual.
3	Non-analytical studies, e.g. case report, case series
4	Expert opinion

Grade of recommendation

Grade	Description
A	At least one meta analysis, systemic review, or RCT rated as 1⁺⁺, and directly applicable to the target population: or A systemic review of RCTs or a body of evidence consisting principally of studies rated as 1⁺, directly applicable to the target population, and demonstrating overall consistency of results
B	A body of evidence including studies rated as 2⁺⁺, directly applicable to the target population and demonstrating overall consistency of results; or Extrapolated evidence from studies rated as 1⁺⁺or 1⁺
C	A body of evidence including studies rated as 2⁺, directly applicable to the target population and demonstrating overall consistency of results; or Extrapolated evidence from studies rated as 2⁺⁺
D	Evidence level 3 or 4; or Extrapolation of evidence from studies rated as 2⁺