Download Measurements - Dr Ted Williams

Measurements
Measure of Variability, Scale Levels
of Measurements, Descriptive
Statistics, Measures of Central
Tendency
Measurements need to:

Produce valid and reliable results
 be sensitive and specific
 be able to identify clinically important
changes
 have outcome measures and endpoints
defined
 be easy to interpret
Reasons for errors in
measurement:

Improper function or calibration of
equipment
 patients providing misleading or dishonest
answers to verbal/written questions
 Improper recording/transcribing of data
 Investigators recording or making
inaccurate measurements
Types of Errors

Random error
– Random in occurrence, often balancing out
over course of study
– mean or average of measurements still close to
true value
– Large patient size reduces random error
Types of Errors

Systematic error
– represents bias in measurements and does not
tend to balance out over course of study.
– Bias can be knowingly or unknowingly
– Good study design minimizes systematic error.
Measurement Terms

Validity- degree to which an instrument is
measuring what it is intended to measure.
– Predictive, Criterion, Face

Reliability- reproducibility of a test
 Sensitivity- ability to measure a small
treatment effect
 Specificity- how well the test can
differentiate between the effect resulting
from treatment and random variation
Validity terms used in
association with
measurements:
 Predictive validity:
– the extent to which a measurement or test actually


reflects or predicts the true condition.
Criterion (construct) validity:
– the degree to which a measurement or test agrees with
or obtains the same results as other proven tests
designed to measure the same.
Face validity
– the extent to which a measure appears reasonable or
sensible for measuring a desired outcome
Reasons for False Positive
Results

Patient related
– patients weren’t as ill as originally believed,
and drug was more effective in mildly ill pts.
– Patients were much more ill than originally
believed, and drug was more effective in
severely ill patients.
– A few patients had a very large response, which
skewed the overall results.
– Patients gradually improved independent of
drug treatment.
False Positive Results

Patient related
– More medicine was absorbed than anticipated
– Patients took excess medication.
– Patients felt pressure to report a positive
medicine effect
– Concomitant non drug therapy or other drug
therapy improved results
False Positive Results

Study Design and Drug Related
– Blinding was broken or ineffective
– open label study can sometimes produce a
–
–
–
–
larger positive response
no placebo control to help interpret
error occurred in dosing patients- gave more
drug than intended
inadequate wash-out period, carry over effect
inappropriate clinical endpoints, tests or
parameters were used
False Positive Results

Investigator related
– influenced response by great enthusiasm
– chose inappropriate tests to measure

Results and Data Related
– systematic error- reporting large drug effect
– high percentage of non-responders dropped out
– not all data was analyzed
False Negative Results

Patient Related
– were much more ill than realized
– responded less to the drug than anticipated
– study group had large number of non
responders
– non-compliance-- took fewer doses
– concomitant medicines- interactions
– exposed to conditions that interfered with study
False Negative Results

Drug Related
– not adequately absorbed
– kinetics were different in study group than in
other patient groups

Study Design Related
– Too few of patients
– inappropriate study design
– insufficient drug dose was tested
False Negative Results

Study Design Related (cont.)
– Ineffective tests or parameters used
– Inadequate wash-out period in previous
treatment period
– Concomitant non-drug therapy interfered

Investigator Related
– influenced patients with skepticism displayed
– chose inappropriate tests to measure effects
False Negative Results

Results and Data Related
– Patients who improved dropped out leaving
higher number of non-responders
– systematic error resulted in reporting of an
inappropriately small drug effect.
Outcome Measures



Example: A study is performed to compare the
effects two antihypertensives, atenolol and
propranolol in 2 groups of patients with mild
high blood pressure. 2 types of outcomes
measurements are selected for this study:
measures of efficacy and measures of safety
Measures of efficacy: BP, HR, symptom relief
Measures of safety: adverse effects, blood
glucose, electrolytes, serum lipids
Criteria Used for Outcome
Measures

Presence or Absence criteria: Is sign, symptom
present or absent?
 Graded or Scaled Criteria: the use of grading
on a scale to measure clinical symptoms
 Relative change criteria- measured changes
 Global assessment criteria- Quality of Life
 Relative effect criteria- change in time to
effect.
Measurement Endpoints

Endpoints are measurable points used to
statistically interpret the validity of a study.
 Valid studies have appropriate endpoints.
 Endpoints should be specified prior to start
of study (should be included in study
design)
 Quality studies have simple, few and
objective endpoints.
Endpoints

Objective- based on actual or measurable
findings or events (heart rate, BP, Temp.)
 Subjective- based on thoughts, feelings,
emotions (pain scale, mobility)
 Morbidity- quality or condition at the
present-- quality of life
 Mortality- causing death or a death rate
Endpoints Example

In a study determining the effects of
clonidine on quality of life, the researchers
determine the number of days a patient
misses work. Each patient is also asked to
complete a rating scale to describe the
degree of fatigue they experience.
 What type of endpoints are used?
 What type of criteria are used?
Surrogate Endpoints

These reduce the quality and validity of the study.

Surrogate or Substitute endpoint examples:
– CD4/CD8 ratios instead of “survival” in studies for
treatment of AID’s.
– Measuring volume of acne instead of proportion of
patient’s cleared of acne.
– Determining cardiovascular disease or
atherosclerotic disease instead of measuring blood
pressure in a study of antihypertensive drug
treatment
Hawthorne Effect

Refers to the influence that a process of
conducting a study may have on a subject’s
behavior
– Subject
– Environment
– Research design

Reasons for Clinical
Improvement in a Patient’s
Condition
Natural regression to the mean (most acute
and some chronic conditions resolve on
their own
 Specific effects of treatment (drug or
intervention)
 Non-specific effects- attributable to factors
other than specific drug/intervention effect.
– Called a Placebo Effect
Placebo Effect





A placebo is an intervention designed to simulate
medical therapy, but not believed to be a specific
therapy for the target condition.
A placebo is used either for it’s psychological effect or
to eliminate observe bias.
Placebo “response”= due to change in pt. Behavior
following admin. of a placebo
Placebo “effect” = change in pt’s illness due to the
symbolic importance of a treatment.
A placebo effect doesn’t require a placebo.
Why do we see a Placebo
Effect?

Three different theories:
 1. The effect is produced by a decrease in
anxiety
 2. Expectations lead to a cognitive
readjustment of appropriate behavior.
 3. The effect is a classical conditioned
Pavlovian response.
Placebo Effect

Expectations lead to behavior change
– Patient’s and providers expectations
– Patient’s positive attitude toward provider and treatment
– Providers positive attitude toward therapy
– Provider interest in patient (sympathy, time, positive
attitude)
– Compliant patients have better outcomes than noncompliant
patients even with a placebo.
– The placebo response is stronger when stronger drugs are
used.
– Crossover studies show a stronger placebo response when
given in the 2nd period of study.
Appropriate Statistical Tests




To determine whether appropriate statistical tests
have been used, you must know 3 things:
1. The specific research question or hypothesis being
addressed.
The number of independent and dependent variables
The scales or levels of measurement used for the
dependent variables
Variables in a Study:

Dependent variables:
– those variables whose value depends upon or is
influenced by another variable.
– It is the variable that is measured, and the one
that changes as the result of a drug action.

Independent variables
– Those variables which modify a dependent
variable (drug treatment)
Example

Patients given Lovastatin to lower
cholesterol.
 Dependent variable- lowering of cholesterol
 Independent variable- Lovastatin

There can be more than one independent
and dependent variable in a study.

Dependent/Independent
Variables
Example: A single blind study of 30 patients with poison
ivy dermatitis were randomized to receive either topical
hydrocortisone 1% or 2% and apply QID. Severity of the
dermatitis was evaluated daily using a 5 point scale, where
5- severe and 0-none.
 What is the independent variable? Dependent variable?
 Example: A study was conducted to compare the efficacy
of procainamide and quinidine for reducing ventricular
arrhythmias. The number of ventricular ectopic
depolarizations was determined in patients both before
and during therapy with either drug.
 What is the independent variable? Dependent variable?
Scales of Levels of
Measurement

Nominal Level
– variables are grouped into mutually exclusive
categories.



Gender as female or male
cured and not cured
response and no response
– include histograms (bar graphs)
– weakest level of measurement
– referred to as dichotomous data
Scales of Levels of
Measurement

Ordinal level
– ranked or ordered categories
 1-2-3-4
 severe, moderate, mild, none
 always, sometimes, never
– stronger level than nominal
– not measured quantitatively, but qualitatively
– distance between groups need not be equal
Scale Levels of Measurement

Continuous Measurement
 Interval level: exact difference between two
measurements is known and constant
–
–
–
–
has arbitrary zero point
highest level of measurement
quantitative data
Examples: BP (mm Hg) serum Theo levels
(ug/ml), WBC (cells/cu mm)
Continuous Level of
Measurement

Ratio level:
– exact differences between measurements is
known and constant
– true zero point (Centigrade temp scale)
– can make ratio statements (2:1) that denote
relative size
– Can be converted to an ordinal scale (but
ordinal scale can’t convert to interval)
Scale levels of Measurements
Baseline Pain Assessment
0
1
2
3
(absent) (mild) (mod) (severe)
Placebo
0
2
18
14
PainawayR
0
4
12
16
(number of subjects in each group with varying
degrees of baseline pain intensity.
What scale level of measurement?
Scale level of Measurements
Infectious Outcome Among 46 Patients
Infection
No infection Total
Oxacillin
2
20
22
Placebo
0
24
24
Column total 2
44
46
What scale level of measurement?
Types of Interval/Ratio Data

Discrete scale of data (non-continuous):
when a measurement has the interval
characteristics but can only be assigned
integer values. (HR, number of patients
admitted to hospital/day)
 Non discrete (continuous) scale of data:
each data point falls on a continuum with an
infinite number of possible subdivisions
(temp, BP, BG, weight)
Data Distributions

Once data is collected, it can be organized
into a distribution, or graph of frequency of
occurrence, or chart of the number of times
that each measurement value occurs.
– Bar Graphs
– Bar Chart (Histogram)
– Line Graphs
Data Distributions

Nominal and Ordinal level data use
histograms (Bar charts) because data
classified into distinct categories
 Continuous level data are distributed in the
form of curves and line graphs (normal
distributions and non-symmetrical
distributions)
Bar Chart (Histogram)
90
80
70
60
50
Flu cases
40
30
20
10
0
1st Qtr
2nd Qtr
3rd Qtr
4th Qtr
Continuous Level Data
Normal Distribution
Gaussian Curve
Non-Normal Distributions
Bi-Modal Curve
Weights of American Adults (women and
men)
Non-symmetrical distributions
Non-normal distributions
Continuous Distribution
Examples:
The distribution of GPA’s of college students:
1.0
2.5
4.0
Continuous Distribution
Example
Distribution of the ages of patients taking Digoxin
20
40
60
80
Descriptive Statistics
Measures of Central Tendency
Measures of Central Tendency

Mean– mathematical average of a set of numbers.
– Affected by extreme data points (outliers)
– Useful for continuous level data (interval/ratio).
– Ex: uric acid concentrations: 8,6,5,4,3,2,2,2.
Total number of samples =8. Sum of
measurements = 32. 32/8= 4 (mean).
Measures of Central Tendency

Median
– “Middle” number of a group of numbers in
which an equal number of responses above and
below that point exist. (called 50th percentile)
– Not affected by outliers. Useful for ordinal,
interval and ratio data and non-symmetrical.
– Ex: Uric acid concentrations: 8,6,5,4,3,2,2,2.
Since even number, median lies between 4 and
3 or median= 3.5.
How to Recognize “skewed”
data

If the magnitude of the difference between
the mean and median is none or small, the
data is approaching normal (symmetrical)
distribution.
 If the difference between the mean and
median is large, the data usually prove to be
skewed.
Measures of Central Tendency

Mode
– The most commonly or frequently occurring
–
–
–
–
value(s) in a data distribution.
Useful for nominal, ordinal, interval/ratio data.
Only meaningful measure for nominal data.
Can have more than one mode in set of data
Ex: Uric acid concentration: 8,6,5,4,3,2,2,2.
The mode = 2.
Measures of Central Tendency
Scale Level of
Normal
Non-Normal
Measurement
Distribution
Distribution
Nominal
Mode
Mode
Ordinal
Median=mode
Median /mode
Interval/Ratio
Mean=med=mode
Mean/med/mode
Descriptive Statistics
Measures of Variability
Measure of Variability

Two distributions can have the same mean,
median and/or mode and yet be very
different.
 Variability refers to how spread out (or
close together) the data are.
Example

2 groups of men w/ mean SBP in each
group is 120 mmHg. Are they similar?
 First group: BP: 110,120,120,130
 Second group: BP: 80,90,150,160
 Both have mean= 120
 Spread of data or range of data is much
different
Range

Range: the interval between the lowest and
highest values within a data group.
 Can be significantly influenced by outlying
data (extreme values)
 Used for ordinal, interval or ratio data
Interquartile Range

Measure of variability directly related to the
median. The median represents the 50th
percentile.
 The interquartile range is that range
described by the interval between the 25th
and 75th percentile values.
 Used for ordinal, interval/ratio data that
don’t have normal distributions
Standard Deviation (SD)

Standardized measure of the spread of
scores around the mean.
 Useful for continuous (ratio/interval) data
 When reported in a study: +/- 1SD
 Needs normal distribution of data
 The mean +/- 1SD includes 68% of data
points (34% on each side of the mean)
Standard Deviation

Mean +/- 2 SD include about 95% of data
points (47.7% of the values on each side of
the mean)
 Mean +/- 3 SD include about 97.7% of the
data points (49.8% of values on each side of
the mean)
 DBP 100 mmHg +/- 5 mmHg includes data
points from 95-105 mmHg(assume 1 SD
unless tells you differently)
Standard Deviation (SD)

The larger the SD, the further the data
points deviate from the mean (more variable
data).
 The smaller the SD, the closer the data
points are to the mean (less variable data).
 Ex: 0.9 +/- 0.2mg% and 1.1 +/- 0.6mg%
 Which has more widely scattered data?
 Answer: 1.1 +/- 0.6 mg% (larger SD)
Variance

Variance is estimate of the study data.
Obtained by calculating the differences
between each individual value and the
overall mean
 Needed for calculating the SD
 SD= variance
 SD2= variance
Standard Error of the Mean
(SEM)

SEM is a way of estimating the variability
of an individual sample mean relative to the
population as a whole.
 SEM= SD/ variance or SD/ sample size
 SEM is used to calculate the Confidence
Intervals (CI)
 Improperly used in place of SD because it is
a smaller number and “looks better”
SD versus SEM

Mean Serum Theophylline concentrations of a group of
patients was 13.6 +/- 2.1 (1SD).
– Conclude about 68% of patients had conc.
Somewhere in the range of 11.5-15.7.
 Serum Theo conc. of a group of patients was 13.6 +/2.1 (SEM)
– could assume that if several add’l samples of pts
with same characteristics were studied, their mean
values would fall between 11.5 and 15.7 68% of
the time.
Review

Variance= differences between each
individual value and the overall mean.
 Variance used to calculate the SD
 SD= variance
or SD2= variance
 SEM or SE is derived from the SD
 SEM= SD/ sample size
Review…SEM




SD= measure of the variability of individual
values about the sample mean
SEM/SE= measure or indication of the
variability of individual sample means about the
true but unknown population mean.
SEM is used to estimate the reliability
(precision) of a study sample in terms of how
likely it is that the sample mean represents the
true population mean.
SEM is used to calculate the CI
Example

In a study of the effectiveness of Drug X on
the Blood Sugar concentrations in 15
patients, the authors report the mean BS
values in the patients as 150 +/- 2.3mg%.
 Would this represent the SEM or SD?
Confidence Intervals (CI)



Represents a range that has a high probability of
containing the true population value.
The likelihood that a study samples’ value
reflects the true value of the population.
Calculated for a desired level of probability
(95%). A 95% CI means there is a 95%
probability that the true population value falls
within the CI range
Confidence Interval Example

The mean difference in healing rates
between placebo and penicillin was reported
to be 59% (CI = 24-72%).
Confidence Intervals

Can be calculated for nominal level data
(proportions) and continuous level data
 90% CI assoc. with narrower range of
values (don’t need to be as confident)
 99% CI assoc. with wider range of values
(more confident the CI will contain true
population value.
CI is influenced by:

1. Level of confidence selected
 2. SEM (larger SEM, wider the CI)
 3. Standard Deviation (SD) of the study
sample. (larger SD, then larger SEM, then
wider the CI)
 4. Size of the study group. (The larger the
sample size, the smaller the SEM, and
narrower the CI)
Confidence Interval Example
Two similar studies are published about efficacy of
Pravastatin for reducing cholesterol. Both sets of patients
are comparable. Study 1 enrolled 200 patients. Study 2
enrolled 50 patients.
The mean +/- SD treatment reduction in cholesterol
concentrations in the Study 1 patients was 15.2 mg% +/2.0 mg%. The corresponding values in the Study 2
patients was 17.1 mg% +/- 2.0 mg%
Which mean values (15.2 mg% or 17.1 mg%) would most
likely have a wider 95% CI associated with it?
Confidence Intervals (CI)

CI applies to continuous data, proportions
(nominal data), medians, regression slopes,
relative risk data, response rates, survival
rates, median survival duration, hazard
ratios, non-random selection or assignment
between groups.
Measures of Variability
Level of Measurement SD
SEM CI
Nominal
No
No
Yes
Ordinal
No
No
No
Continuous
Yes
Yes
Yes
Statistical vs. Clinical
Significance

Example: A new antihypertensive drug is
studied to determine whether it decreases
the rate of myocardial infarction. The
results indicate that the drug decreases MI
by 11% with a 95% CI= -2-25%.
Statistical vs. Clinical
Significance

The 95% CI for the relative risk of
headache development with a new diabetes
drug is reported as 1.20 (CI 0.95-1.50) and
for a placebo drug as 1.25 (CI 0.88-1.76).
 Are these showing statistical significance?
Ratios, Proportions and Rates

Ratio expresses the relationship between
two numbers. Men: Women (45:90)
 Proportion: specific type of ratio expressed
as a percentage. 12% experienced cough
when using this drug) (12% of total study
population)
 Rate: form of proportion that includes a
specific time frame. (18% died from
influenza in the US last year)
Incidence and Prevalence

Incidence Rate =


Number of new cases of a disease
Total population at risk

Prevalence Rate=

Number of existing cases of a disease
Total population at risk

per time
per time

Descriptive Statistics
Measures of
Risk/Association
Relative Risk
 Odds Ratio
 Relative Risk Reduction
 Absolute Risk Reduction
 Number Needed to Treat
 Number Needed to Harm
Measures of Risk

Relative Risk (RR)
– the risk or incidence of an adverse event
–
–
–
–
occurring or of a disease developing during
treatment in a particular group.
RR= # pts in treatment group w/ ADR
Total # of pts in treatment group
# pts in placebo group w/ ADR
Total # pts in placebo group
Relative Risk Example:

A new drug is being compared to placebo to prevent
development of diabetic retinopathy (DR).
 Treatment
DR
No DR
Total
 New drug
50
75
125
 Placebo
65
55
120
 What is the risk of DR developing during treatment in
patients taking the new drug?

50__ = 0.4= 40%
Risk in placebo? 65 = 0.54=54%
 125
120

RR= 0.4/0.54 = 0.74 or 74%
Relative Risk (RR)

RR = 1 : When the risk in each group is the
same
 RR<1: When the risk in treatment group is
smaller than the risk in the placebo group
 RR>1: When the risk in the treatment group
is greater than the risk in the placebo group
Relative Risk (RR)

Example: The risk of an adverse event
developing during therapy with an eye
medication compared to the placebo group
was listed as 1.5. What does this mean?
 Answer: That the eye med is 1.5 times more
likely to cause an adverse event than the
placebo being used.
Relative Risk Example

92 men and women who were recovering from
heart attacks were followed and surveyed a year
later. 14 of the 92 patients had died. When death
rates were calculated according to pet ownership,
only 3 of the 53 pet owners (5.6%) were no longer
living, compared to 11 of 39 (28%) patients who
were without animals.
 Relative risk = 0.056/0.28 = 0.2
 What does this mean?
Relative Risk...

Relative Risk does NOT tell us the
magnitude of the absolute risk.
 Example: A RR of 33% could mean that the
treatment reduces the risk of an adverse
event from 3% down to 1% or from 60%
down to 20%. These may or may not be
significant depending on the population and
adverse event (minor or major adversity)
Odds Ratio (OR)

Commonly reported measure in case control
designs. Case control starts with outcomes.
(looks back for risk factors)
 OR = # pts taking drug w/ ADR

# pts taking drug w/o ADR__

#pts not taking drug w ADR

# pts not taking drug w/o ADR
Odds Ratio cont.

The Odds Ratio could also be expressed as:

Treatment A Deaths

Treatment A Survival_____

Treatment B Deaths

Treatment B Survival
Odd’s Ratio (OR)
Odds of developing a disease or ADR if exposed (to drug)
Odds of developing a disease or ADR if not exposed (to drug)
OR:
Disease
Present
Absent
Exposed factor
A
B
Not exposed to factor
C
D
OR= A/C = A X D
B/D
B XC
OR = A/B = A X D
C/D
BXC
Odds Ratio Example:
A case control study reported that 35 of 120 chronic
renal failure patients took NSAID’s compared to only
20 of 110 similar patients without renal failure. What
would be the odds ratio of developing renal failure if
taking NSAID’s?
35 (taking NSAID’s w/ RF)
A=35 B = 20
20 ( taking NSAID’s w/o RF)
C = 85 D= 90
85 (not taking NSAID’s w/ RF)
90 (not taking NSAID’s w/o RF)
Renal Failure/NSAID’s
A/C
A/B
B/D
C/D
35/ 85 = 0.41
or
35/20 =1.75
20/ 90 = 0.22
85/90 = 0.94
0.41 = 1.86
1.75 = 1.86
0.22
0.94
Odds Ratio

OR= l : The odds of developing an adverse
event or disease in the exposed (treatment)
group is the same as the odds in the nonexposed (non-treatment) group.
 OR<1: Odds of developing ADR in exposed
group is less than odds in non-exposed.
 OR>1: Odds of ADR in exposed group
greater than the odds in non-exposed.
Odds Ratio (OR)

Example: The odds that ASA was taken by
children who developed Reyes Syndrome
vs. the odds that ASA was taken by similar
children who did not develop Reyes
Syndrome was reported as OR=3:l.
– The odds that Reyes Syndrome children had
taken ASA was approximately 3 times greater
than for the children who did not develop
Reyes Syndrome.
Interpreting the OR and RR

1. Degree of validity of the study design.
 2. The confidence interval (CI)
 3. Relative Risk Reduction (RRR)
Relative Risk Reduction
(RRR)

Ex: If a new drug is shown to reduce the
risk of cancer, what is the exact percentage
of this reduction?
 RRR: measure of the reduction in the
relative risk in the exposed group.
 RRR= Rate in control group-rate in tx group

Rate in control group
 RRR= 1-RR
Relative Risk Reduction
(RRR)

Incidence of cancer was 7% in treatment
group and 12% in placebo( control) group.
 RRR= 12%-7% = 0.42 = 42%

12%
 Disadvantage of RRR- doesn’t discriminate
between very large and very small actual
incidence rates in the groups.
Relative Risk Reduction
Example

A study is performed to determine the
efficacy of a new LMWH, Drug “H” in
preventing PE from post surgical patients.
299 post surgical patients are randomized to
receive Drug H and 355 receive placebo. 43
patients developed PE in the placebo group,
and 21 developed PE in the treatment
group. What is the relative risk reduction
by Drug H (reducing the risk of PE)
Drug H Example cont...
Incidence of PE in placebo group = 43/355 = 0.12 = 12%
Incidence of PE in Drug H group = 21/299 = 0.07= 7%
RRR =Rate in control group - rate in treatment group
Rate in control group
RRR= 12%-7% / 12% = 0.12- 0.07/ 0.12= 0.42 = 42%
OR another way to calculate is RRR= 1-RR
1- 21/299 / 43/355 = 1- 7/12 = 12/12-7/12 = 5/12 = 0.42
= 42% or 1- 0.07/0.12 = 1-0.58= 0.42
Absolute Risk Reduction
(ARR)

ARR= Incidence rate in control group incidence rate in treatment group.
 Ex: cancer: treatment 7%, placebo 12%
 ARR = 12%-7% = 5%
 For serious conditions though, a small ARR
can still be very clinically relevant.
Number Needed to Treat
(NNT)





NNT: number of individuals that need to be treated
in order to prevent one adverse event or one
outcome. NNT = 1
ARR
Ex: study determine efficacy of drug preventing
cancer. Incidence of cancer in placebo 12%, in
treatment group 7%
12%-7% = 5% 1/5% = 20=NNT (20 pts needed to
treat to prevent 1 case of cancer
NNT= 1/ placebo - treatment group
Number Needed to Harm
(NNH)

NNH= 1/ treatment- placebo group
 Ex: Headache occurred in 25% of placebo
patients and 75% of patients taking drug X.
 The NNH = 75%-25% = 50% 1/0.5 = 2
 Only 2 patients would need to be treated
with drug X in order to cause a headache
occurrence.
Review

In a diabetes study, 4% of Glucotrol users and
18% of placebo pts. Developed CHF within 10
years.
 RRR= 18%-4% = 14 = 0.77 = 77% RR
18%
18
ARR = 18%-4% = 14%
NNT = 1/0.14 = 7 pts
In Glucotrol group 26% had HA vs. 3% in placebo.
NNH = 26%-3% = 23% 1/0.23 =4