Critical Appraisal Flashcards

1
Q

Define Validity

A

Is the finding true

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

Define Generalisability

A

Is the finding applicable elsewhere?

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

3 reasons for not being valid

A
Chance (random error)
—- (finding imprecise)
Bias (systematic error)
—- (finding inaccurate) 
Confounding (error of interpretation)
—- (beware, unknown confounders cannot be statistically accounted for)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Define
Efficacy
Vs
Effectiveness

A

Efficacy: Shows if something works under ideal conditions
Vs
Effectiveness: shows if something works under normal conditions

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

Types of research Qs:
Pragmatic research
Vs
Explanatory research

A
Pragmatic research
- demonstrates effectiveness 
Vs 
Explanatory research 
- demonstrated efficacy
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Aspects of

Pragmatic research

A

Unselected population
Pt centred outcomes
Does not interfere with clinical practice

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

Aspects of

Explanatory research

A

Can it work under ideal conditions
Specific staff/setting/population
Often clinical outcome

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

Null hypothesis

Define

A

No difference between active treatment and placebo

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

P value

Define

A

Chance null hypothesis is true

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

Type 1 error =

A

False positive (p value)

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

Type 2 error =

A

False negative

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

Statistical significance =
alpha + beta

Define a+b

A

Alpha
- maximum possibility of a false positive (p value)

Beta
- max possibility of false negative
(Determined by sample size)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Power of study =

A

1- beta

Beta = max possibility of a false negative
Usually set to 80-90%

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Study power is set by

A
  • Level of alpha
  • Sample size
  • The minimum clinical difference you wish to detect
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Issue with multiple hypothesis testing:

A

Increases false positive

- only use if there is a clear rationale to the test

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Benefit of randomisation

A

Removes confounders

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Concealment prevents

A

Allocation bias

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Blinding prevents

A

Measurement bias

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Outcome bias

More likely if

A

Soft outcomes

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Intention to treat

A

All analysis

Protects against bias

21
Q

Loss to follow up

Assume

A

Patients list are similar to analysed group.
Unless >30% LTFU
Or difference between test and control group

22
Q

Absolute risk =

A

Intervention - control

23
Q

Relative risk =

A

Intervention-control / proportion

24
Q

NNT=

A

1 / absolute risk

25
Q

Cluster randomisation

Issues

A

Allocation of concealment not possible

Therefore bias introduced

26
Q

Hawthorne effect =

A

Change due to being monitored

27
Q

Sustainability

A

What additional resources are required

What negative effects could occur due to intervention being rolled out

28
Q

Generalisability

A

X

29
Q

Independence of reference standard

A

X

30
Q

Work up bias

A

When a different reference standard is used

31
Q

Incorporation bias =

A

Diagnostic test forms part of reference standard

32
Q

Reliability =

A

Intraobserver error

33
Q

Effect of prevalence on
Sensitivity and specificity
PPV
NPV

A
Sensitivity and specificity
— remains constant 
PPV 
-increases with increasing prevalence
NPV
- decreases with increasing prevalence
34
Q

Likelihood ratio =

A

Value of information added
I =no value
>10 is diagnostic
<0.1 is rule out

35
Q

Forest plot =

A

Graph for heterogeneity

In a meta analysis

36
Q

Pros and cons

Meta analysis

A

Increased precision but same accuracy

Does not overcome bias of original data

37
Q

Areas of evaluation

For critical appraisal

A

Therapy
Service organisation
Diagnostic test
Systematic reviews

38
Q

Write an abstract

OD PICO RACS

A

Objective
Design

Population
Intervention
Comparator
Outcome

Results
Adverse events
Conclusions
Studies (further)

39
Q

Two main questions that critical appraisal tries to answer

A

1) is the result valid

2) is it generalisable

40
Q

Probability estimates of random error include

A

P value

Confidence intervals

41
Q

Main factor to consider in non randomised data

A

Confounders

42
Q

Inaccurate
Vs
Imprecise
Estimates

A

Inaccurate - due to bias
Vs
Imprecise - random error

43
Q

Confidence intervals and
Precision

Wide
Vs
Narrow

A

Wide - imprecise
Vs
Narrow - precise

44
Q

Statistics used to estimate effect of random error on results
2 types

A

P value
- for hypothesis testing

Confidence intervals
- for estimation of differences between groups

45
Q

P values and alpha values

And associations with false/true positive/negative

A

P value = probability of false positive (type 1 error)

Alpha value = probability of a true positive
(Max probability of false positive)

Beta value = probability of false negative

46
Q

Allocation or selection bias
How it occurs
And prevention

A

If patients or researchers can choose which treatment they get

Prevent via randomisation and concealment of allocation

47
Q

How to randomise service organisation as a intervention

A

Cluster randomisation

Disadvantage - no concealment if allocation

48
Q

80/20 rule

As a chart

A

Pareto chart
Bar chart with
cumulative line as percentage

49
Q

Likelihood ratio

A

The Likelihood Ratio (LR) is the likelihood that a given test result would be expected in a patient with the target disorder compared to the likelihood that that same result would be expected in a patient without the target disorder.