Quiz 2 Deck 2 Flashcards

1
Q

what is MDC?

A

Minimal Detectable Change: amount of change that just exceeds the SEM

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

what is MCID?

A

Minimal Clinically Important Difference: smallest difference in measurement that signifies an important change in pt’s condition

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

for outcome measurement properties that are easy to interpret, what do MDC & MCID indicate?

A
  • below MDC = not stat significant & not important.
  • btwn MDC & MCID = stat significant, but not important.
  • above MCID = stat sig & important
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

for outcome measurement properties that are challenging to interpret, what do MDC & MCID indicate?

A
  • below MCID = not stat significant & not important.
  • btwn MCID & MDC = important change but cannot be distinguished from measurement error.
  • above MDC = stat sig & important
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

what is pooled standard deviation

A

calculation of effect size for 2 groups w/in a trial

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

3 ways to calculate effect size for single group studies

A
  • effect size, ES = (mean posttest - mean pretest) / SDpre
  • standardized response mean, SRM = (mean post - mean pre) / SDchange
  • Guyatt’s = MCID / sqrt(2 x MSE)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

what does Cohen’s d interpret?

A

effect size & SRM between control & experimental groups

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

Cohen’s d criteria

A
  1. 8 = large treatment effect.
  2. 5 = moderate
  3. 2 = small
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

if difference in means = variance, effect size is ___

A

large

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

if difference in means > variance, effect size is ___

A

large

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

if difference in means is 50% of variance, effect size is ___

A

mod

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

effect size correlation (r)

A

r = d / sqrt(d^2 - 4)

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

experimental success rate (what it means & how to calc)

A

how likely the Rx group will have a successful outcome.

exp success rate = 0.5 + r/2

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

control success rate (what it means & how to calc)

A

how likely the Ctrl group will have a successful outcome.

exp success rate = 0.5 - r/2

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

relative success rate

A

= exp success rate / ctrl success rate

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

what is benefit increase?

A

how much the intervention increases the chance of a desirable event

17
Q

how is relative benefit increase (RBI) calculated?

A

[(% therapy group w/ outcome - % control w/ outcome) / % control w/ outcome] x 100%

18
Q

how is absolute benefit increase (ABI) calculated?

A

% therapy group w/ outcome - % control w/ outcome

19
Q

how is number needed to treat (NNT) determined?

A

1/ABI or 1/ARR

20
Q

what is risk reduction?

A

how much the intervention reduces the risk of an unwanted event

21
Q

how is relative risk reduction (RRR) calculated?

A

[(% ctrl w/ problem - % therapy w/ problem) / % ctrl w/ problem] x 100%

22
Q

how is absolute risk reduction (ARR) calc?

A

% ctrl w/ problem - % therapy w/ problem

23
Q

what qualifies as “benefit” or “harm”?

A

benefit: # subjects who must receive rx to produce 1 (+) outcome OR avoid 1 adverse event.
risk: # subjects who must receive rx to produce 1 adverse event.

24
Q

what is EER

A

experimental group event rate (% subjects w/ outcome)

25
Q

what is CER

A

control group event rate (% subjects w/ outcome)

26
Q

what does it mean if NNT = 1

A

every pt will have the successful outcome

27
Q

what does it mean if NNT = 100

A

1 out of every 100 pts will have desired outcome

28
Q

what is an important detail to pay attention to when considering NNT in a study?

A

how they define “successful outcome”

29
Q

what is NNH?

A

number needed to harm: # harmed by the treatment

30
Q

what does it mean if NNH = 8?

A

1 out of 8 will be harmed by Rx