Lecture 27: Understanding Study Results Cont'd Flashcards

1
Q

what are indications of how large risks are? (4)

A
  • Absolute risk increase (ARI)
  • Relative Risk (RR)
  • Relative Risk Increase (RRI)
  • Number Needed to Harm (NNH)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Absolute risk

definition
how to calculate?

A

§ Report the risk of swelling & ischemia in the treatment group (i.e., EER)
§ Report the risk of swelling & ischemia in the control group (i.e., CER)

EER = a/a+b
CER = c/c+d
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Absolute Risk Increase (ARI)

definition
how to calculate?

A
  • Difference between the proportion who had an event in treatment group and the proportion who had an event in control group
  • ARI = c/c+d – a/a+b
  • ARI = EER – CER
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Absolute Risk Increase (ARI)

= 0.01

translation

A

Wearing compression stockings increased

the absolute risk of getting swelling/ischemia by 1%

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

Number Needed to Harm (NNH)

definition
how to calculate?

A

• The number of patients who must receive the therapy for one additional patient to be be harmed during a specific period of time

• Related to the ARI
§ NNH = 1/ARI
§ NNH = 1/0.01
§ NNH = 100

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

Number Needed to Harm (NNH)
= 100
translation

A

You need to treat 100 patients with
compression stockings for the duration of their trip to produce
1 extra episode of swelling/ulceration

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

Relative Risk (RR)

how to calculate?

A

Relative Risk = EER/CER
= 2/100
1/100

  • RR = 0.02 / 0.01
  • RR = 2
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q
Relative Risk (RR)
= 2

translation

A

§ The risk of developing swelling/ischemia was 2 times greater in those who wore compression stockings
§ The risk of developing swelling/ischemia was doubled by compression stockings

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

Relative Risk Increase (RRI)

definition
how to calculate?

A
An estimate of the proportion of baseline risk that is added
by the therapy
§ RRI = (EER – CER)/CER
§ RRI = (0.02 – 0.01)/0.01
§ RRI = 1

• Alternatively RRI = RR-1 = 2-1 = 1 (expressed as 100%)

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

Relative Risk Increase (RRI)
= 1

translation

A

Compression stockings increased the risk of

edema/ischemia by 100%.

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

odds ratio for risk

how to calculate

A

OR = ad/bc
= 299/981
= 2.02

eg. The odds of leg ischemia was doubled

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

Absolute measure used for

Reduces A BAD Event
Increases a BAD event
Increases a GOOD event

A

Reduces A BAD Event: Absolute Risk Reduction
Increases a BAD event: Absolute Risk Increase
Increases a GOOD event: Absolute Benefit Increase

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

relative measure used for

Reduces A BAD Event
Increases a BAD event
Increases a GOOD event

A

relative risk for all

odds ratio for all

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

relative CHANGE measure used for

Reduces A BAD Event
Increases a BAD event
Increases a GOOD event

A

Reduces A BAD Event: relative Risk Reduction
Increases a BAD event: relative Risk Increase
Increases a GOOD event: relative Benefit Increase

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

numbers needed to ______ measure used for

Reduces A BAD Event
Increases a BAD event
Increases a GOOD event

A

Reduces A BAD Event: Number Needed to Treat
Increases a BAD event: Number Needed to Harm
Increases a GOOD event: Number Needed to Treat

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

what 2 NNT limitations are there?

A
  1. Baseline Risk - NNT Depends Heavily On Baseline Risk

2. Time Frame - NNT is an Inherently Time Dependent Measure

17
Q
  1. Baseline Risk - NNT Depends Heavily On Baseline Risk
A

high risk patients and low risk pts - if they have the same RRR, the NNT for high risk patients is much lower

eg. NNT is 10 for high risk and 200 for low risk pts

Relative risks - don’t need to know characteristics of a pop that much

Absolute risk and NNT highly sensitive to baseline risk

18
Q
  1. Time Frame - NNT is an Inherently Time Dependent Measure
A

• The NNT depends on WHEN the outcomes are counted

• Are these NNTs all equal?
§ 30 over 5 years
§ 150 over 1 year
§ 15 over 10 years

• NNTs become more impressive (i.e., decrease) as the
duration of follow-up increases
Talk about 10 years for more impressive NNT

19
Q

when is RRR best used?

A

When people do not specify what they are talking
about…
§ e.g., Drug therapy was 30% effective in reducing the risk of death
§ …Likely talking about RRR as it gives the most “impressive” numbers
§ The relative efficacy of drug therapies is often similar across patient subgroups
at different risk & usually is assumed to be constant over time.
• Applying across diseases, various time points, personalizing risk
calculations;

20
Q

Communicating with health care providers; not ideal for

communicating with patients

A

NNT

21
Q

what’s best for communicating with patients?

A

EER; CER; ARR:

EER, CER, ARR, NNT: All highly dependent on baseline risk & time point!