5 - External Validity Flashcards

1
Q

what is internal vs external validity?

A

internal: the ability of the study results to support a cause-effect relationship btw the treatment and observed outcome
external: the applicability of the study results to patients

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

what is efficacy vs effectiveness?

A
  • both wrt external validity!
  • efficacy (aka explanatory/management): can intervention work under “ideal” circumstances?
  • effectiveness (aka pragmatic): does the intervention work under “usual” circumstances?
  • typically most trials will have both, but more efficacy
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

what are the 8 hierarchies within valid studies?

A
  • participant eligibility criteria
  • intervention flexibility
  • practitioner expertise
  • follow-up intensity
  • nature of outcomes
  • participant compliance
  • practitioner adherence
  • primary analysis
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

participant eligibility criteria

A
  • p 66
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

intervention flexibility

A
  • p 66
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

practitioner expertise

A
  • p 66
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

follow-up intensity

A
  • p 66
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

nature of outcomes

A
  • p 67
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

participant compliance

A
  • p 67
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

practitioner adherence

A
  • p 67
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

primary analysis

A
  • p 67
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

a statistically non-singificant result in a pragmatic trial can be interpreted as….

A
  • either there is no diff btw the groups (signal is too small)
  • or the variability in the outcome is too great to detect diff (noise too big)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

if a well designed and executed efficacy study shows that the experimental treatment is no better than the control, we can conclude that..

A
  • there is no diff btw treatment groups
  • bc efficacy study has done all it can to reduce noise, so if we still cannot detect signal in this case, we know signal is not there (not big)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly