Lecture 3 - Importance of evidence in practice of medicine

Question 1

What is EBM?

Answer 1

Clinicians should use critically appraised information in clinical practice for patient’s best care

Question 2

What is EBM defined as by Sackett?

Answer 2

The conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients

Question 3

What are the criticisms of EBM?

Answer 3

Impossible for any clinician to have enough time to critically appraise one article a week (need to appraise 3.5 every clinical session)
Gov healthcare commissioners use jargon of EBM to justify decisions that are seen as inappropriate by clinicians

Question 4

Why does EBM matter to clinicians?

Answer 4

Revalidation, Patient care, medical knowledge, Practice based learning+improvement, interpersonal and comm skills, professionalism

Question 5

What is EBM used for?

Answer 5

As a tool to help make decisions, NOT the final decision maker!

Question 6

What is the hierarchy of studies?

Answer 6

Systemic revies/meta-analyses > Randomised controlled trials > cohort studies > ecological studies > descriptive/cross-sectional studies > case report/series

Question 7

What is association?

Answer 7

Statistical dependence between 2 variables, that is degree to which rate of disease in persons with a specific exposure is higher/lower than rate of disease wihtout exposure

Question 8

What is chance?

Answer 8

An inference made from samples than from whole popn -> sample size, power calculations, P values and statistical significance

Question 9

What is bias?

Answer 9

A systematic error -> selection bias or measurement bias

Question 10

How do you evaluate a statistical association?

Answer 10

Considering chance, bias, confounding and cause

Question 11

What is confounding?

Answer 11

Mixing effects between exposure, disease and 3rd factor

Question 12

How do you account for confounding?

Answer 12

Use matching, randomisation, stratification and multivariate analysis

Question 13

What is causal effect?

Answer 13

Judgement of a cause-effect relationship - based on a chain of logic that addresses 2 areas: Observed association between exposure and disease is valid AND Totality of evidence taken from no of sources supports judgement of causality

Question 14

What factors should be considered when deciding between association and causation of a disease? Also known as the Bradford-Hill criteria

Answer 14

Strength, consistency, specificity, temporal relationship, dose-response relationship, plausibility, coherence, experimental evidence, analogy

Question 15

What is meant by strength of association? (association vs causation of disease)

Answer 15

Measured by relative risk magnitude - strong asso. more likely to be causal than a weak asso. which could be result of confounding/bias BUT weak asso. doesn’t rule out causal connection

Question 16

What is meant by consistency? (association vs causation of disease)

Answer 16

If similar results found in diff popn using diff study designs then asso. more likely to be causal BUT lack of consistency doesn’t rule out causal asso. as diff exposure levels and other conditions may reduce impact of causal factor in some studies

Question 17

What is meant by specificity? (association vs causation of disease)

Answer 17

If exposure ^ risk of certain disease but not others then its strong evidence in favour of cause-effect relationship BUT 1-to-1 relationships are rare and lack of specificity shouldn’t rule out a causal relationship

Question 18

What is meant by a temporal relationship? (association vs causation of disease) ESSENTIAL!!

Answer 18

Putative risk factor to be cause of disease it has to precede the disease - easier to establish from cohort studies and difficult from cross-sectional/case-control as both cause and effect are made at same time BUT reverse time order is not evidence against hypothesis

Question 19

What is meant by Dose-response relationship? (association vs causation of disease)

Answer 19

If ^ levels of exposure lead to ^ risk of disease then causal is more likely BUT some asso. can be caused by a single jump

Question 20

What is meant by plausibility? (association vs causation of disease)

Answer 20

If consistent with other knowledge, causal likely BUT lack of plausibility could reflect lack of scientific knowledge

Question 21

What does coherence mean? (association vs causation of disease)

Answer 21

Implies cause and effect interpretation doesn’t conflict with known natural history BUT lack of coherent info shouldn’t be taken as lack of causality

Question 22

What does experimental evidence mean? (association vs causation of disease)

Answer 22

On human/animals - human is seldom available and animals are different species with different exposures

Question 23

What is meant by analogy? (association vs causation of disease)

Answer 23

Provides a source of more elaborate hypothesis about asso in q BUT lack doesn’t mean falsity of hypothesis, just lack of imagination/experience