1. Measuring disease and death frequency
Integrated Disease Surveillance Programme (IDSP) district surveillance officers (DSO) course

2. Outline of the session Ratio, rate, proportion Prevalence Incidence
Relation between prevalence and incidence
Mortality 2

3. Count, divide and compare:
1.Question during an outbreak of hepatitis A in Sioux City, IA, USA: Are native Americans at higher risk?
Number of new hepatitis A cases
Native Americans: 19 (8% of cases)
Others: 228 (92% of cases)
Can you compare these two groups with this information?
How can this information be used?
Who can use this information?
First we will provide an illustration of the count, divide, compare method.
The first step is count. We count cases of hepatitis A in two groups.
But, can you compare directly the number of cases of hepatitis A in two groups?
Introduction 3

4. Count, divide and compare:
2. Divide the number of cases by the population
New hepatitis A cases #
Year
Population
Native Americans 19
1996
1,697
Others
228
96,576
The case count is useful if you do not leave your own group, but if you want to compare two groups, you need to compare.
The population of other groups is larger than the Native American population
So before comparing, we divide the number of cases in each group by the population denominator of the group. This gives us a rate.
Native Americans: 19/1,697
Others: 228/96,579
Introduction 4

5. Count, divide and compare:
3. Compare indicators
Native Americans: 1,112 per 100,000
Others: 236/100,000/ year Rates among Native Americans are higher
Now that we have divided by the denominator, we are allowed to compare. While others had more cases, the frequency of the disease is higher in Native Americans, once the division corrects for the difference in the population size.
Introduction 5

6. A ratio places in relation two quantities that may be unrelated
The quotient of two numbers
Numerator NOT necessarily INCLUDED in the denominator
Allows comparing quantities of different nature: Female to male ratio
Now that we have seen the CDC principle, we will see what a ratio is.
The ratio is a crude measure: You can divide anything by any other thing.
= 5 / 2 = 2.5/1
Introduction 6

7. Examples of ratio Number of doctor per beds
1 doctor for 85 beds
Number of participants per facilitator
Sex ratio: Females / Males
Ratio of white blood cells to red blood cells
1/600. What does it tell?
Number of children with scabies / number of children with malnutrition
Does it make sense?
Examples. As you can see, a ratio is permissive. You can divide anything by any other thing. Now of course, it needs to make some sense.
Introduction 7

8. A proportion measures a subset of a total quantity
The quotient of two numbers
Numerator NECESSARILY INCLUDED in the denominator
Quantities have to be of the same nature
Proportion always ranges between 0 and 1
Percentage = proportion x 100
A proportion is a more elaborate indicator. The quantities have to be of the same nature because the numerator is part of the denominator.
2 / 4 = 0.5=50%
Introduction 8

9. Example of proportions
The proportion of children with scabies in a village
Tuberculosis cases in a district:
400 males, 200 females
Question
What is the proportion of males among all cases?
What is the proportion of females cases among all cases?
Note:
All proportions are ratios
Is the converse true?
Examples and exercises. All proportions are ratios, but not all ratio are proportions. There is a condition to be a proportion that is not required for a ratio.
Introduction 9

10. A rate measures the speed of occurrence of health events
The quotient of two numbers
Defined duration of observation
Numerator
Number of EVENTS observed for a given time
Denominator (includes time)
Population at risk in which the events occur
A rate provides an idea of speed. It relates the frequency to a time unit.
Observed in 2004 2
----- = 0.02 / year
100
Introduction 10

11. Example of rate Mortality rate of tetanus in country X in 1995
Tetanus deaths: 17
Population in 1995: 58 million
Mortality rate = 0.03/100,000/year
Rate may be expressed in any power of 10
100; 1,000; 10,00; 100,000
Self explanatory examples.
Introduction 11

12. Prevalence – (P)
Number of existing cases (old and new) in a defined population at a specified point of time
Number pf people with disease at a specified time
P = x 10n
Population at risk at the specified time
In some studies the total population is used as an approximation if data on population at risk is not available
Now that we have seen ratios, rates and proportions, we will study prevalence.
Self explanatory definition.
Can you think of examples of prevalence? Like the prevalence of hypertension among participants of the course?
By “at risk” we mean that the person can develop the disease. For instance, the prevalence of cancer of the cervix can only be calculated among females: Males are not at risk as they do not have a uterus.
Prevalence 12

13. Source and type of prevalence data
Surveys generate prevalence data
Prevalence data are expressed as proportions
Number affected / Number surveyed
The numerator is included in the denominator
The affected are only identified among the surveyed
Prevalence data are cross sectional in nature. They usually come from cross sectional methods like survey.
A prevalence is expressed as a proportion.
Prevalence 13

14. Example of point prevalence
150 children in a school
Screening for visual acuity at a given time
15 children require glasses
Prevalence of refractory errors
15 / 150 = 10%
A simple example of point prevalence.
Prevalence 14

15. Factors influencing prevalence
Number of new cases
Duration of the illness
If the disease is short, the prevalence is reduced
The prevalence of sudden infant death = 0 (At a given instant, nobody has sudden infant death because the disease has no duration)
If the disease is long, the prevalence is increased
Rare lifelong disease can accumulate to build up a large prevalence
Prevalence can be affected by a number of factors: The number of new cases and the duration of illness.
Prevalence 15

16. Causes of increase of prevalence
Long duration
Low cure rate
Low case fatality
Increase in new cases
Immigration of patients
Improved detection
Emigration of healthy people
Once we have understood this principle, we can identify a number of factors that will increase the prevalence. Can you say how?
Prevalence 16

17. Causes of decrease of prevalence
Shorter duration
High cure rate
High case fatality
Decrease in new cases
Emigration of patients
Improved cure rate
Immigration of healthy people
Once we have understood this principle, we can identify a number of factors that will decrease the prevalence. Can you say how?
So you see, prevalence is an indirect reflection and it requires the understanding of numerous factors to be interpreted.
Conclusion: Changes in prevalence may have many causes and may be difficult to interpret
Prevalence 17

18. Uses of prevalence data
Assessing health care needs
Planning health services services
Measure occurrence of conditions with gradual onset
Study chronic diseases
Prevalence data are useful when we want to estimate burden and plan resources accordingly. They are adapted to diseases of long duration.
Prevalence 18

19. Incidence – (I)
Number of new cases in a given period in a specified population
Time, (i.e., day, month, year) must be specified
Measures the rapidity with which new cases are occurring in a population
Not influenced by the duration of the disease
Now that we have seen prevalence, we can study incidence.
Incidence is about new cases.
Incidence 19

20. Cumulated incidence - (CI)
Number of new cases
CI = x 10n
Population at risk at the beginning
Also known as:
Attack rate
Assumes that the entire population at risk at the beginning was followed-up for the time period of observation
This is the formula of prevalence.
Again, incidence, like prevalence, should be calculated among those “at risk”. Those who had measles (and who are immune) before should be theoretically excluded from the denominator when we calculate the incidence of measles.
Incidence 20

21. Source and type of incidence data
Surveillance generate incidence data
Incidence data are expressed as rates
Number affected / Population / time
Dynamic measure (speed)
We saw that prevalence data come from surveys.
Incidence data rarely comes from survey (unless you ask the survey participants about their recent past).
Incidence data usually comes from cohort study or from surveillance data.
Incidence is expressed as a rate to provide a dynamic figure of speed.
Prevalence 21

22. Uses of incidence data Describe trends in diseases
Evaluate impact of prevention programmes
These are the use of incidence data.
Incidence 22

23. The dynamic of incidence and prevalence
New cases
Incidence
Prevalence
Death
Cure
This graphs shows the relation between incidence and prevalence.
Think of a bucket that gets water from a tap (new cases) and leaks (death or cure) at the same time.
Incidence is the speed at which the tap is filling the bucket.
Prevalence is the level of water in the bucket.
Prevalence will go up if the tap flows faster of if the leaks is slower.
Incidence and prevalence 23

24. The relation between prevalence and incidence
Prevalence depends on
Incidence (I)
Duration of the disease (D)
P = I x D
Change in prevalence from one time period to another may be the result of changes in incidence rates, changes in the duration of disease, or both
This intuitive understanding can be made into a mathematical formula.
Incidence and prevalence 24

25. Patterns of incidence and prevalence
High prevalence and low incidence
e.g., Diabetes Mellitus
Low prevalence and high incidence
e.g., Common cold
Here are example of disease with low incidence and high prevalence and with high incidence but low prevalence.
Can you think of others?
Can you find examples of diseases of high prevalence and incidence? Of disease of low prevalence and incidence?
Incidence and prevalence 25

26. Incidence and prevalence
Evolution of HIV prevalence in a country scaling up public health efforts
Increase in HIV prevention
Reduction in incidence (Difficult to measure)
Increase in HIV AIDS care and support (treatment)
Increase in disease duration (reduced mortality)
Increase in prevalence (Easier to measure)
Incidence measures the impact of prevention efforts
Prevalence may be used to plan care and support
The immediate consequence of the plan may be an increased prevalence
This example illustrates how prevalence is complex. In a country scaling up HIV public health efforts, the first measurable outcome could be an increase of prevalence because of lower mortality.
Incidence and prevalence 26

27. Crude mortality rate - (CMR)
Number of deaths in a specified period
CMR = x 10n Average total population
Does not take into account factors such as age, sex, race, socio economic status, etc.
Provides information on trends in a population’s health status
Self explanatory. Mortality is in fact the incidence of death.
Deaths 27

28. Disease specific mortality rate - (SMR)
Number of deaths from a disease in a specified period
SMR = x 10n Average total population
Reflect the impact of a disease on a population in terms of death
Should not be confused with case fatality
Likewise specific mortality is the incidence of death because of specific causes.
Deaths 28

29. Case fatality ratio of a given disease
Divide
Number of deaths from the disease
Number of cases of the disease
Example: Measles outbreak
3 deaths
145 cases
Case fatality ratio: 2.1%
Don’t mix up with disease-specific mortality!
Case fatality divides the deaths FROM A DISEASE by the number of cases of that disease.
It reflects the severity of illness.
Don’t mix it up with disease-specific mortality! 29

30. Take home messages Tell apart ratio, proportion and rates
Prevalence is a static measure taken at a point in time
Incidence is a dynamic measure taken over a certain time
Mortality is calculated using population denominators to reflect burden while case fatality is calculated using cases as denominators to reflect severity
These are the take home messages of the session. 30