Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Y3/ EBM > Randomised Controlled Trials > Flashcards

Randomised Controlled Trials Flashcards

1
Q

What are the 4 explanations of an observation?

A
  1. Bias (linked to study design)
  2. Confounding (linked to study design)
  3. Chance (statistics)
  4. Hypothesis is correct –> effective
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is the difference between internal and external validity?

A 
Internal = study design, statistics, patients in study 
External = is it generalisable to the general population
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What are the types of intervention study?

A
  • before and after intervention
  • intervention group compared to control
  • randomised controlled trial: cluster or individual
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What does cluster randomisation aim to reduce?

A
  • contamination or crossover
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Explaining our observations: Before and After study

  1. bias
  2. confounding
  3. chance
A
  1. difference in measurement before and after
  2. some other factor has changed before and after
  3. change - is difference a fluke?
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is regression to the mean?

A
  • most things vary to some extent by chance
  • if we observe an extreme value, this is partly due to chance
  • therefore the next observation is likely to be closer to the mean
  • can give the appearance of improvement
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is one way of getting round regression to the mean in before and after studies?

A
  • introduce a control arm
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Explaining our observations: Controlled study

  1. bias
  2. confounding
  3. chance
A
  1. patients or researchers beliefs affect outcomes measurement
    • differences between groups at the start
      - some other difference in the way in which groups are treated
  3. chance
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

How does an RCT reduce bias and confounding and chance?

A 
Bias = patients, researchers, analysts can be blinded
Confounding = differences at the beginning are only due to chance, randomly distributes factors 
Chance = bigger study pop, decreases chance
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Define RCT

A
  • intervention study
  • participants allocated to two or more groups
  • groups receive different interventions
  • allocation to groups is random
  • groups followed up after randomisation
  • outcome assessed
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is the effect of randomisation?

A
  • ensures known and unknown charactertistics that might affect the outcome (confounders) are distributed by chance
  • any differences between groups at the start are due to chance
  • minimises confounding
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Explaining our observations: RCT

  1. bias
  2. confounding
  3. chance
A
  1. Bias: in data recording - due to differences in data collection
  2. Confounding: at the start - are groups systematically different, at finish - due to differences in follow up, something other than the intervention is different
  3. Chance - are results due to chance? T test to compare means
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What are the 7 potential issues with RCTs?

A
  1. choice of outcome measure
  2. ethics
  3. choice of control
  4. contamination or crossover
  5. bias in assessment of outcome
  6. losses to follow up
  7. sample size
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q
  1. Explain how ‘choice of outcome’ can affect a study
A

Ideally measure outcome relevant to patient
- if proxy are we confident is this linked to outcome
Clinical effectiveness
- measure clinical outcome e.g. cure
Patient experience
- QoL or disease specific measure

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q
  1. Explain how ‘ethics’ can affect a study
A

Are individuals disadvantages by being randomised to inrevetion or control?

  • if we believe one is better, we cannot ethically randomised
  • need EQUIPOISE - need to not know which is better
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q
  1. Explain how ‘choice of control’ can affect a study
A
  • Usual care e.g. structured vs standard care
  • No treatment - if theres is no usual care and we don’t believe new treatment is effective
  • Placebo (something atet appears similar to intervention) - distinguishes intervention effects from placebo effects, helps blind the patients and researchers
17
Q
  1. Explain how ‘contamination or crossover’ can affect a study
A
  • participants randomised to the control group may unintentionally receive the intervention
  • participants not randomised to the intervention group may not revere the intervention
18
Q

What are the 2 options if there appears to be contamination in your RCT?

A
  1. analyse according to Treatment Receive
    - but groups are no longer randomised!
    - this is ALLOCATION BIAS
  2. analyse by Intention to Treat Analysis
    - reduces risk of allocation bias
    - may dilute/underestimate treatment effect
19
Q

Underwhat circumstance is contamination a problem?

A
  • if the intervention is education or information

- if we cannot treat patients differently in the same clinic

20
Q

How can crossover be reduced?

A
  • cluster randomisation
  • randomised entire hospitals/clinics/schools
  • reduces crossover but statistical analysis more complex
21
Q
  1. Explain how ‘bias is assessing outcome’ can affect a study
A

Blinding and researcher bias
- non-blinded trials exaggerate outcome
Blinding and potential for analysts bias
- eg. stopping bias, stop trial at the point where outcome is favourable

22
Q

What are the 4 degrees of blinding?

A
  1. open
  2. single - patient blinded
  3. double - patient and researcher
  4. triple - patient and researcher and analyst
23
Q

What is allocation concealment?

A
  • when randomisation is undertaken the researchers should not now (or be able to guess) whether a participant will be allocated to intervention or control
  • allocation concealment prevents researchers from influencing which participants are assigned to the intervention or control group
24
Q
  1. Explain how ‘losses to follow up’ can affect a study
A
  • Trial participants may be lost
  • little effect on realists if due to treatment allocation or outome
  • does affect results if related to treatment allocation AND outcome
  • helpful to describe characteristics of those lost - are they different?
25
Q
  1. What are the sample size considerations?
A

Expected effect
- smaller effects need bigger sample size
Variation in measured outcome
- outcomes that are measured imprecisely need a larger sample size
Significance
- at what significance level will we accept there is a difference
Power
- how certain do we need to be to find a difference if there is one

26
Q

How do eligibility criteria affect generalisablity?

A
  • What criteria did the researchers set?
  • If criteria mean that only a select group (e.g. people with severe disease) are included then the findings won’t be generalisable to all patients with that condition
27
Q

What do we mean by ‘reliable, valid and responsive’?

A

Reliable - ability to produce consistent, reproducible estimates of true effect
Valid - measures the construct that it claims to measure
Responsiveness of a patient-reported outcome deals with the extent to which the change in the scores are associated with changes in the underlying clinical status of the subject

28
Q

Give examples of well regard methods of randomisation

A
  • sealed envelopes, telephone randomisation
29
Q

What type of bias is a consequence of non-concealed allcoation?

A
  • selection bias

- one ends up with baseline imbalances between the two study arms

30
Q

What is selection bias?

A
  • can occur when people are asked to take part in a trill with knowledge of which arm they will be in
  • assessing baseline characteristics, looks for selection bias
31
Q

What is detection bias?

A
  • when there are systematic difference between groups in how outcomes are determined
  • this is prevented by blinding
32
Q

What is performance bias?

A
  • systematic differences between the groups in the care that is provided, or in exposure to factors other than the intervention of interest
  • can be reduced by blinding
33
Q

What is attrition bias?

A
  • systematic difference between groups in withdrawals from a study, look for incomplete outcome data §
34
Q

What is reporting bias?

A
  • systematic differences between reported and unreported bias
  • looks fro selective outcome report, some outcomes have been measured by results not given
35
Q

What are the 5 main biases in RCTs?

A 
selection bias 
performance bias 
detection bias 
attrition bias 
reporting bias
36
Q

What is I^2?

A

ranging from 0-100% measures the degree of inconsistency across studies in a meta analysis that is not accountable by chance

37
Q

-
-

A
  • clinical differences between studies
  • methodological differences between studies
  • unknown study characteristics

Y3/ EBM (7 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions