Measures of Epidemiolgy GMC CLASS.pptx

MBBS.USMLE, DPH, Dip-Card, M.Phil, FCPS
Professor Community
Medicine/Epidemiolgy
Ex- Professor Community Medicine
UmulQurrah University Makka Saudi Arabia

Measures in Epidemiology
 Numbers of cases
 Proportional mortality
 Proportional mortality
ratio
 Actual/Crude
prevalence and
incidence rates
 Specific prevalence
and incidence rates
 Standardised rates
 Standardised ratios
 Relative risk
 Odds ratio
 Attributable risks
 Numbers needed to
treat and prevent
 Life years lost
 Disability adjusted life
year (DALY)
 Quality adjusted life
year (QALY)
2/4/2018 3
DR M Tauseef Javed SIMS Lahore

Components of epidemiology:-
Disease frequency.
Distribution
determinants.
Disease frequency.
Distribution.
Determinants.
2/4/2018 4

Scope of measurement:-
 Mortality.
 Morbidity.
 Disability.
 Presence & absence or distribution of..
 Medical needs.
 Utilization of health services.
 Demographic variables.
2/4/2018 5

Tools of measurements-
Rate
Ratio
Proportion
2/4/2018 6

Rate
 Measures the occurrence of an event or
disease in a given population during a given
period (one Year).
(Birth rate, growth rate, accident rate)
 Usually expressed per 100 or per1000
population.
 It has a time dimension, whereas a
PROPORTION does not.
2/4/2018 7

 A fraction is made up of 2 numbers.
 The top number is called the NUMERATOR
 and the bottom number is called the
DENOMINATOR.
 In the fraction ¾ the 3 is the numerator
and the 4 is the denominator.
2/4/2018 8

Rate comprises-
No of death in one year
 Death rate= -------------------- X 1000
Total mid year population
Numerator
Denominator.
Time specification
Multiplier
(Numerator is part of denominator )
2/4/2018 9

10
 Rate: when we say that there were 500
deaths from motor vehicle accidents in city
A during 2014, its nothing more than
counting deaths in that city during that
particular year.
 It conveys no meaning to an epidemiologist
who is interested in comparing the
frequency of accidents in city A with that in
city B.
 To allow such comparisons, the frequency
must be expressed as a rate. It is the basic
measure of disease occurrence
2/4/2018 DR M Tauseef Javed SIMS Lahore

1. Rate: = No. of events in a specified period x K
Pop. at risk in a specified period
A rate comprises a numerator, denominator, time
specification & multiplier. The time dimension is usually a
calendar year. Rate is expressed per 1000, 10,000 or
100,000 selected according to convenience to avoid
fractions
Rate is used to estimate probability or risk of occurrence of a
disease or to assess the accessibility or coverage of
healthcare system.
Example
Crude death rate= Number of deaths in one year X1000
Mid – year Population
11

Ratio
 The value obtained by dividing one
quantity by another- X/Y.
 Male to female ratio.
 A ratio often compares two rates,
death rates for women and men at a
given age.
2/4/2018 12

Ratio-
 Ratio also expresses relation of size
between the two quantities.
 Numerator is not part of Denominator.
 Expressed as X / Y.
Doctor : Population ratio.
Male : Female ratio.
2/4/2018 13

14
2. Ratio:
 Relationship b/w 2 numbers expressed as
x:y or x/y e.g ratio of males to females 2:3.
 The numerator is not a component of the
denominator.
 E.g. Ratio of WBCs to RBCs is 1:600 or 1/600
 Other examples include: doctor-population
ratio, child-woman ratio, etc.

Proportion -
 A part/share or number considered in
comparative relation to a whole.
 "the proportion of greenhouse gases
in the atmosphere is rising”
 Usually expressed as a percentage
%
2/4/2018 15

Proportion-
 This is also relation /magnitude between two
quantities, And numerator is always part of
denominator.
 And expressed as percentage
-Proportion of female students .
-Proportion of anemic mothers
(60% mothers are anemic)
2/4/2018 16

1. Proportion:
Specific type of ratio in which numerator is
included in the denominator and the resultant
value is expressed as %age. E.g 1: If there are
1000 boys and 800 girls in a school, the
proportion of boys:
Boys / Boys+ Girls= 1000 x 100 = 55%
1000+800
E.g 2: From 7,999 females aged 16 – 45 y, 2,496
use modern contraceptive methods.
The proportion of those who use modern
contraceptive methods = 2,496 / 7,999 x 100 =
31.2%
17

Measurement of morbidity-
Incidence
Occurrence of new cases
•
Prevalence
Existence of all new & old cases.
2/4/2018 19

Prevalence vs. Incidence
 Prevalence:- how many people in a
population currently have the
disease (Photograph)
 Incidence:- how many people are
diagnosed each year (Film)
2/4/2018 20

Cure rate
2/4/2018 21

Figure 7.4
Incident cases
Deaths, emigrations and recovery
Prevalent cases
2/4/2018 22

New cases
2/4/2018 23

Figure 7.7
Incident cases
Prevalent cases
Recoveries
Emigrant cases,
unmeasured cases
occurring abroad,
and deaths
Recoveries
Population reservoir Immigration
Births
Emigrant and non-
measured cases,
deaths
2/4/2018 24

INCIDENCE RATE
Like prevalence, divided into
two types:
1. Cumulative incidence rate
2. Incidence density
2/4/2018 25

Number of new cases of disease
occurring over a specified period
of time in a population at risk at the
beginning of the interval.
1. Cumulative incidence rate:
2/4/2018 26

EXAMPLE OF CUMULATIVE
INCIDENCE RATE
If we count all new cases of influenza
occurring in MSU undergraduates
from September 1, 1997 - August 31,
1998, and we take as the
denominator all undergraduates
enrolled in September 1, 1997, we
would be describing the cumulative
incidence rate of influenza.
2/4/2018 27

Number of new cases of disease
occurring over a specified period
of time in a population at risk
throughout the interval.
2. Incidence density:
2/4/2018 28

The numerator does not differ between the two types of
incidence
However, the denominator can differ in incidence density
from cumulative incidence because it takes account of
(in the example):
•Students who left school during the year
•Students who died
•Students who had influenza once and will not
have it again the same season
•Students who entered school later in the year
2/4/2018 29

Incidence density requires us to add up
the period of time each individual was
present in the population, and was at risk
of becoming a new case of disease.
Incidence density characteristically uses
as the denominator person-years at risk.
(Time period can be person-months,
days, or even hours, depending on the
disease process being studied.)
2/4/2018 30

Types of prevalence-
• Prevalence at any given
point of time.
• 4% TB cases on 1st April
Point
prevalence
• Prevalence at a
given period of time.
• Period will be 1year.
Period
prevalence
2/4/2018 31

Proportion of individuals in a
specified population at risk who
have the disease of interest at a
given point in time.
POINT PREVALENCE RATE
2/4/2018 32

Proportion of individuals in a specified
population at risk who have the
disease of interest over a specified
period of time.
For example:
 annual prevalence rate
 lifetime prevalence rate.
(When the type of prevalence rate is not specified
it is usually point prevalence, or its closest
practical approximation)
PERIOD PREVALENCE RATE
2/4/2018 33

Prevalence increased by-
 Longer duration of the disease.
 Prolongation of life, with treatment.
 If incidence increases.
 Immigration of new cases.
 Better reporting of cases.
 Emigration of healthy people.
2/4/2018 34

Prevalence- increased by-
Longer
duration of
disease
Incidence
increases.
Prolongation
of life
without cure.
35

Prevalence decreased by-
 Shorter duration of diseases.
 Improved cure rate.
 Incidence decreases.
 Emigration of new cases.
 Under reporting of cases.
 Immigration of healthy people.
2/4/2018 36

Prevalence decreased by-
Improved cure rate.
Short duration of
disease.
Incidence decreases
2/4/2018 37

USES OF INCIDENCE AND PREVALENCE
1. Incidence is generally used for acutely
acquired diseases, prevalence is used for
more permanent states, conditions or
attributes of ill-health.
2. Incidence is more important when thinking
of etiology of the disorder, prevalence when
thinking of societal burden of the disorder
including the costs and resources
consumed as a result of the disorder.
3. Incidence always requires a duration,
prevalence may or may not.
2/4/2018 38

4. In incidence, the unit of analysis is the
event, in prevalence, it is the person.
Thus incidence may exceed 100% (e.g.
annual incidence of colds) unless a
convention is adopted to count only first
episodes of an illness that can occur
more than once.
5. Prevalence can never exceed 100%.
6. Incidence generally requires an initial
disease-free interval before counting
starts, because incidence is measured
only in those at-risk of disease.
2/4/2018 39

USES OF INCIDENCE DENSITY AND
CUMULATIVE INCIDENCE
 Incidence density gives the best
estimate of the true risk of acquiring
disease at any moment in time.
 Cumulative incidence gives the
best estimate of how many people
will eventually get the disease in an
enumerated population.
2/4/2018 40

RELATIONSHIP BETWEEN INCIDENCE
AND PREVALENCE
In a STEADY STATE (i.e. if
incidence is not changing, and the
population is stable)
Prevalence rate = incidence rate
times the duration of disease
(P = I x D)
2/4/2018 41

Mortality indicators
 Crude Death Rate.
 Specific death rate.
 Case fatality rate.
 Proportional mortality rate.
 Survival rate.
 Standardized death rate.
2/4/2018 43

Crude Death Rate-
 Number of deaths from all causes, per 1000
estimated mid year population in one year in a
given place.
No deaths during one year
CDR = _________________________ X 1000
Mid year population
2/4/2018 44

Specific death rate-
 Cause Specific death rate like
disease death rate, Road accident…
 Age specific-IMR, Child Mortality rate
 Sex specific death rate – MMR/female
 Period specific death rate–Death in May
2/4/2018 45

Case fatality rate-
 Percentage of particular cases dying during
particular disease epidemic.
 Killing power of disease particularly acute diseases
No of deaths due to cholera
 CFR= ----------------------- X 100
Total No of cholera cases
2/4/2018 46

Proportional mortality rate-
 Proportion or % of deaths due to
particular cause out of total deaths.
 It measures the disease burden.
 Under 5, No of deaths below 5
years
proportional = -------------------- X 100
mortality rate Total No all of deaths
2/4/2018 47

Survival Rate-
 Percentage of the treated patients remaining
alive at the end of 5 years treatment.
 Yard stick for assessing the standard of therapy
in cancer.
 Survival pts alive at the end of 5 yrs
Rate = ---------------------- X 100
Total No of pts treated
2/4/2018 48

Standardized Death Rate.
(Adjusted Death Rate)
 CDR can not be useful for comparison.
 Death rate need to be standardized for
comparisons.
 Standardization can be done by-
:adjusting death rate age wise,
:also can be done sex/race wise
2/4/2018 49

Important Rates
 Incidence Rates
 Prevalence Rates
Important Ratios
 Odds Ratio
 Relative Risk
2/4/2018 DR M Tauseef Javed SIMS Lahore 50

Incidence Rates
Number new cases of disease during a
specified period of time
Total population at risk at the beginning of
that specified period
Incidence = x 10n
Assumptions
1. The observation on study population started at the same time
2. Single common end point (time) of observations
3. The population at risk will remain unchanged during the study period

 What do you understand by
population at risk?
 What is population at risk for cervical
carcinoma?

Population at risk in study of
carcinoma of cervix
All men
All
Women
0-25
25-60
70 +
25-60 years
Total
Population
All Women age groups Population
at Risk

Concept of Numerator and
Denominator
Numerator
“Number of events or disease cases are the
numerators
Denominator
“The population under study or population at risk is
the denominator
Include all persons affected and unaffected
Person-time observation

Brain Storming
How will you estimate the incidence if?
 It is not possible to start of observation
on large population at the same time
 It may not be possible to end the
observation at the same time
 The population at risk may not remain the
uniform over period of observation

Concept of Person Time
 Person years
 Person days
Person time = People at risk x duration of observation
Examples
 12 patient observed for 1 year = How many person Years?
 6 patient observed for 2 years = How many person years?
 3 patient observed for 4 years = How many person Years?

Incidence rate (“incidence density”)
Number of new cases
–––––––––––––––––––––––––––––––
Avg population at risk × Time interval
Number of new cases
= ––––––––––––––––––––
Population-time

Incidence Density (I.D)
Number new cases of a disease during a
specified period of time
Total person time of the observation
Incidence
Density =
X 10n

Subj Jan
2001
July
2001
Jan
2002
July
2002
Jan
2003
July
2003
Jan
2004
July
2004
Time
at risk
Per-
time
A 2/20 1.5 30
B 0/10 2.0 20
C 5/40 3.0 120
D 3/10 4.0 40
E 3/10 2.0 20
F 2/35 2.0 70
Estimation of Incidence density
(Chronic hypertensive pt followed for nephropathy)
Number of events = 15
Person time = 300
Incidence density = X 1000 = 50 Person-years
Period of observation from Jan 2001 to 2004

Prevalence
(Point prevalence)
Number of people with disease or having
vital event at a given point in time
Total number of persons at risk in that
point in time
Point Prevalence
=
x 10n

Prevalence Rate
(Period Prevalence)
Number of people with disease or having
vital event at a specified time period
Total population at risk in the middle
of the time internal
Prevalence Rate
= x 10n

Concept of Incidence, point prevalence
and period prevalence
62
Jan
2009
31 Dec
2009
Case 1
Case 2
Case 3
Case 4
Case 5
Case 6
Case 7
Case 8
The cases for incidence rate during year 2009?
The case for point prevalence on 1st Jan?
The cases for point prevalence at 31st Dec?
The cases for period prevalence?

Incidence and Prevalence
P varies as Product of I & D
(I = incidence and D = duration of illness)
 Severity of the illness
 The duration of the illness
 The number of new cases
 Availability / non-availability of treatment
 Effectiveness of Treatment

Objectives of Analysis
• Distribution by place person
and time
• Observing the differences and
developing hypothesis
Descriptive
Analysis
• Comparison of exposures/risk
factors in groups
• Comparison of
outcome/disease in groups
Analytical
analysis

Analysis
• Comparison of exposure rates
• Odds Ratio (OR) for
association
Case control
study
• Comparison of Incidence rates
• Relative Risk (RR) for
association
Cohort
study

2 x 2 Table for Analytic studies
Disease Status
Exposure
Status
Cases Control Total
Exp. Yes a b a + b
Total Exp
Exp. No c d c + d
Total non-Esp
Total a + c
Total Dis.
b + d
Total non-Dis.
a + b+ c+ d
Grand total

Case Control Study estimates
• Exposure rate
among the cases
How much
exposure among
diseased?
• Exposure rate
among the controls
How much
exposure among
control?
• Odds Ratio
What is the Risk
Ratio?

Case Control Study design
Non-diseased
Exposure Present
Exposure absent
Diseased
Exposure Present
Exposure absent
Compare
Case control study proceed from effect to the cause
Select
Trace back

2 x 2 Table
Case Control Analysis
Disease Status
Exposure
Status
Cases Control Total
Exp. Yes a b a + b
Total Exp
Exp. No c d c + d
Total non-Esp
Total a + c
Total Dis.
b + d
Total non-Dis.
a + b+ c+ d
Grand total

Relative Risk in
Case-Control Studies
 Can’t derive incidence from case-control
studies
 Begin with diseased people (cases) and
non-diseased people (controls)
 Therefore, can’t calculate relative risk
directly
 But, we can use another method called
an odds ratio
2/4/2018 70

Odds Ratio in
Prospective (Cohort) Studies
2/4/2018 71

Odds Ratio in
2/4/2018 72

Derivation of Odds
Cases Controls
Exposed a b
Not exposed c d
Total a + c b + d
Odds of exposure among cases =
Probability to be exposed among cases
Probability to be unexposed among cases
a / (a+c)
Odds E cases = ------------ = a / c
c / (a+c)
Odds of exposure among controls =
Probability to be exposed among controls
Probability to be unexposed among controls
b/ (b+d)
Odds E controls = ------------ = b / d
d/ (b+d)
a/c
OR = ---- = ad / bc
b/d

Odds Ratio in
Case-Control Studies (cont.)
2/4/2018 74

When is the Odds Ratio a Good
Estimate of Relative Risk?
 When cases are representative of
diseased population
 When controls are representative of
population without disease
 When the disease being studied occurs
at low frequency
2/4/2018 75

REMEMBER !!!
 An odds ratio is a useful measure of
association
 In a cohort study, the relative risk can
be calculated directly
 In a case-control study the relative risk
cannot be calculated directly, so an
odds ratio is used instead
2/4/2018 76

Cohort Study estimates
• Incidence among the
exposed
How much disease
in exposed?
• Incidence among the
controls
How much disease
in non exposed?
• Relative Risk or Risk Ratio
What is the Risk
Ratio?
• Attributable Risk (Extra risk
due to exposure)
What is the risk
difference

2 x 2 Table
Cohort Study Analysis
Disease Status
Exposure
Status
Cases
(Diseased)
Control
(Non disease)
Total
Exp. Yes a b a + b
Total Exp
Exp. No c d c + d
Total non-Esp
Total a + c
Total Dis.
b + d
Total non-Dis.
a + b+ c+ d
Grand total

RR from 2x2 Table
a
a + b
Incidence among exposed =
c
c + d
Incidence among non-exposed =
a /a + b
c /c + d
Relative Risk (RR) =

Relative Risk Calculations
2/4/2018 80

Determining Relative Risk
2/4/2018 81

Interpreting Relative Risk
2/4/2018 82

Relative Risk in
 Can’t derive incidence from case-control
studies
 Begin with diseased people (cases) and
non-diseased people (controls)
 Therefore, can’t calculate relative risk
directly
 But, we can use another method called
an odds ratio
2/4/2018 83

Attributable Risk
Definition:
The amount of disease that can be
attributed to a certain exposure.
2/4/2018 84

Concept of Attributable Risk
2/4/2018 85

Attributable Risk for an
Exposed Group
2/4/2018 86

OR, expressed as a proportion:
Attributable Risk for an
Exposed Group (cont.)
From previous relative risk example:
2/4/2018 87

Calculation for Proportional
Incidence in Total Population
First calculate A-R for
group from
Formulas 11.1 & 11.2
(previous slide),
then use Formula 11.3
For proportion of the
incidence in the
total population, use
Formula 11.4
2/4/2018 88

Calculations for Attributable
Risks (cont.)
2/4/2018 89

Summary
 Relative risk and odds ratio are important as
measures of the strength of association
 Important for deriving causal inference
 Attributable risk is a measure of how much disease
risk is attributed to a certain exposure
 Useful in determining how much disease can be
prevented
 Therefore:
 Relative risk is valuable in etiologic studies of
disease
 Attributable risk is useful for Public Health
guidelines and planning
2/4/2018 90

Measures of Epidemiolgy GMC CLASS.pptx

Related slideshows

More Related Content

Measures of Epidemiolgy GMC CLASS.pptx