Screening test &

Evaluation

Presenter
Dr. Meghana Narendran
Assistant Professor
Department of Community
Medicine
Symbiosis Medical College for
Women
Iceberg Phenomenon of Disease

Clinically
apparent
Disease

Hidden
Mass of
Disease

2
 Plan of Presentation
• Introduction
• Aim and Objectives
• Uses of Screening
• Types of Screening
• Criteria for Screening
• Evaluation of Screening Tests
• Evaluation of Screening Programmes
• Conclusion
• References

3
 “
Introduction

4
8-May-23
5
 Screening
Definition
“ The search for unrecognized disease or defect by
means of rapidly applied tests, examinations or other
procedures in apparently healthy individuals.”

“The presumptive identification of unrecognized defect

or disease by the application of tests, examinations or
procedures which can be applied rapidly, to sort out
apparently well persons who probably have a disease,
from those who probably do not.”

6
 Historically

Individual
Annual Diseases
Screening
Health
Programmes
Examinations Selected
Groups

Screening is a preventive care function

It is capable of wide application
Relatively inexpensive
Requires little physician-time
It is a form of secondary prevention.

7
 Screening is not a Diagnostic Test
Screening Test Diagnostic Test
Done on apparently healthy Done on those with indications or
sick
Applied to groups Applied to single patients, all
diseases are considered
Test results are arbitrary and not Diagnosis is final
final
Based on one criterion or cut-off Based on evaluation of a number of
point symptoms, signs and laboratory
findings
Less accurate More accurate

Less expensive More Expensive

Not a basis for treatment Used as a base for treatment

The initiative comes from the The initiative comes from a patient
investigator or agency providing with a complaint 8
care
 Concept of Lead Time
Lead time is the advantage gained by screening
i.e.
the period between diagnosis by early detection and
diagnosis by other means.

Disease First Final Usual

onset possibl Critical Time of
detection e point Diagnosi Diagnosi
s s A

Screening B
Time
Lead Time

Fig: Model for early detection programmes

9
 “
Aim and Objectives

10
 Aim and Objectives
• Aim: To sort out those having the disease and those not
having the disease from a group of apparently healthy
individuals.

• Objective: To provide treatment to those detected persons,

so that the disease is controlled in community

11
 “
Uses of Screening

12
 1. Case Detection: Prescriptive Screening

• The presumptive
identification of unrecognized
disease

• People are screened primarily

for their own benefit

• Example:
Neonatal screening
Bacteriuria in pregnancy
Breast Cancer
Diabetes
Pulmonary Tuberculosis

13
 2. Control of Disease: Prospective
Screening

• People are screened for the benefit

of others.

• The screening programme may, by

leading to early diagnosis permit
more effective treatment and
reduce the spread of infectious
disease and/or mortality from the
disease.

• Examples:
Screening of Immigrants for
infectious diseases like SARS, ZIKA virus,
COVID etc COVID Check at Airports
Screening for streptococcal
infection to prevent rheumatic fever
Screening of blood donors 14
 3. Research Purposes

• Screening may aid in obtaining

knowledge about the natural history
of diseases.

• The initial screening provides a

prevalence estimate.

• Subsequent screenings provide the

incidence figure.

• It also gives information about risk

factors and risk groups.

• Example:
Chronic diseases like Cancer,
Hypertension 15
 4. Educational Opportunities

Screening
Programmes help in :
• Acquisition of information
of public health relevance

• Providing opportunities
for creating public
awareness

• Educating health
professionals

16
 “
Types of Screening

17
 1. Mass Screening
• It is the screening of the entire
population of an area for a
disease.

• It is offered to all, irrespective

of the particular risk individual
may run of contracting the
disease in question.

• It is not useful unless it is

backed up by treatment and
follow-up.

• Example:
Night blood smear
18
examination for microfilariae of
 2. High Risk /Selective or Targeted
Screening

• It is the screening of only those

groups of population, who are at
high risk of the disease.

• Example:
Screening of
women above 35years for CaCx

all obese people for HTN and DM

sex workers for HIV

family contacts of infectious pulmonary
TB
19
 3. Multipurpose Screening

• It is the screening of a group

of population by application
of two or more tests, at one
time to detect more number
of diseases.

• Example
Screening of pregnant mothers
during antenatal checkup.

20
 4. Multiphasic Screening
It is the screening of the
population by applying different
tests in different phases, for the
diagnosis of one disease.

Example:

whole population of an area is

screened by testing urine for sugar

those who test positive for glycosuria

are subjected for FBS

those with FBS>120mg/dL are

subjected for OGTT
21
 5. Opportunistic Screening

• It is the screening of a patient,

who consults the doctor for some
other purpose.

• It is also known as “Case finding

Screening”.

22
 “
Criteria for Screening

23
 Criteria for Screening

• Before initiating a screening programme, a

decision must be made whether it is ethically,
scientifically and financially justified.

• It is based on 2 considerations

1. DISEASE to be screened
2. TEST to be applied

24
 Disease
1. It is of public health importance.
2. It has recognizable latent or early asymptomatic stage.
3. The natural history of the condition, including
development from latent to apparent disease, is
adequately understood.
4. There is a test that can detect the disease prior to the
onset of signs and symptoms.
5. There are facilities available for confirmation of
diagnosis.
6. There is an effective treatment.
7. There is an agreed policy regarding whom to treat as
patients.
8. There is good evidence that early detection and
treatment reduces morbidity and mortality.
9. The expected benefits of early detection exceed the
risks and costs.
25
 Screening Test
1. Acceptable

2. Repeatable

3. Valid

4. Yield

5. Simple, Safe, Cheap and Rapidly applied

❖ The test should be simple to perform,

easy to interpret and, where possible,
capable of use by paramedics and other
personnel.
Example: Capillary Blood glucose and urine 26
 Acceptability

• Since participation in screening is

voluntary, the test must be acceptable to
those undergoing it.

• In general, tests that are painful,

discomforting or embarrassing are not
likely to be acceptable.

Example: Screening for prostrate cancer

might not be acceptable to a large
proportion of the community.

27
 Reliability
≈ Repeatability, Reproducibility, Precision

“It is the ability of the test to give consistent results

when repeated applications are made.”

Depends on three major factors

1. Observer variation,
2. Biological variation, and
3. Mechanical variation

28
 Observer Variation:

❖ Intra-observer variation:
Variation between repeated observations
by the same observer on the same subject or
material at the same time.
≈ within-observer variation.
Example: Blood pressure measurement.

❖ Inter-observer variation:
Variation between different observers on
the same subject or material.
≈ between-observer variation
Example: X-Ray film report by different
radiologist

29
 Kappa Statistic
• Type of inter-observer agreement.
• Takes into account agreement between two observers purely
by chance.
• It is defined as the extent to which inter-observer agreement
exceeds the agreement that is purely by chance.

> 0.75= Excellent Agreement Beyond

Chance
0.40-0.75 = Intermediate to Good
Agreement
Kappa=<(Percent
0.40 = agreement observed) – ( Percent agreement
Poor Agreement expected by
chance alone)
100% - (Percent agreement expected by chance alone)

30
Biological Variation
≈Intra-subject variation
Variation associated with many physiological
variables like blood pressure, blood glucose, etc.

Biological Variation of Human Populations

Bimodal Curve Uni-modal

Curve

31
 Mechanical Variation

• These are errors relating to technical methods.

• Repeatability may be affected by variations

inherent in the method.

• Example:
Defective instruments
Erroneous calibration
Faulty reagents
Inappropriate/ Unreliable test

32
 Reliability can be ensured by:

✔ Standardization of definitions of the disease.

✔ Using standard techniques for measurement.

✔ Using standardized instruments.

✔ Training of workers.

✔ Continued supervision

✔ Establishing a quality control system.

✔ Using good quality instruments and reagents.

33
 Validity
“Validity is the ability of the test to correctly
diagnose what it intends to diagnose.”

Validity is interpreted in terms of

* Sensitivity
* Specificity
* Positive Predictive Value
* Negative Predictive Value

34
 S S
E P
a b
N E
S C
POSITIVE PREDICTIVE VALUE
I I
T F
I I
V C
c I d I
T T
NEGATIVE PREDICTIVE VALUEY Y

35
 Sensitivity and Specificity:
• Sensitivity:
It is the ability of a test to identify correctly all those
who have the disease, i.e. True Positive

• Specificity:
It is the ability of a test to identify correctly those who
do not have the disease, i.e. True Negative

True Characteristic in the population

Test Have the Disease Do not have the Disease
Result
True Positive False Positive
Positive Have the disease and test Do not have the disease but test
positive positive
False Negative True Negative
Negative Have the disease but test Do not have the disease and test
Sensitivitynegative
= TP Specificity
negative =
TN___
36
TP+FN TN+
5/8/2023 37
Tests of Continuous variables

Fig: A–G, The effects of choosing different cutoff levels

to define a positive test result when screening for
diabetes using a continuous marker, blood sugar, in a
hypothetical population.
38
 Concept of Cut-Off

❖ If the diagnostic test is expensive or invasive:

- minimize false positives
- use a cut-point with high specificity

❖ If the disease is fatal and treatment exists, or

disease easily spreads:
- maximize true positives, minimize false
negatives
- use a cut-point with high sensitivity

39
 Use of Multiple Tests
Sequential Testing
After the first (screening) test is conducted, those
who tested positive are brought back for the
second test to further reduce false positives.

Fig: A–B, Hypothetical example of a two-stage screening program. A,

Findings using Test 1 in a population of 10,000 people. B, Findings
using Test 2 in participants who tested positive using Test 1.
40
 Sequential Tests:

❖ Net Sensitivity:
The proportion of those with the disease who test
positive on BOTH Test1 and Test 2
= (Sensitivity of Test1) X (Sensitivity of Test 2)
There is loss in net sensitivity in sequential screening.

❖ Net Specificity:
The proportion of those without the disease who test
negative on EITHER Test 1 or Test 2
Specificity of Test 1 Specificity of test 1
= + - x
Specificity of Test 2 Specificity of Test 2
There is gain in net specificity in sequential testing.
41
 Simultaneous Testing
Two tests are applied at the same time.

❖ Net Sensitivity:
The proportion of those with the disease who test
positive on EITHER Test1 and Test 2
Sensitivity of Test 1 Sensitivity of test 1
= + - x
Sensitivity of Test 2 Sensitivity of Test
2
There is gain in net sensitivity in sequential screening.

❖ Net Specificity:
The proportion of those without the disease who test
negative on BOTH Test 1 or Test 2
= (Specificity of Test1) X (Specificity of Test 2)
There is loss in net specificity in sequential testing.
42
Net sensitivity and Net specificity using Simultaneous
Screening

43
 Predictive values

❖ Positive predictive value

The proportion of patients who test positive,
actually have the disease in question.

❖ Negative predictive Value

The proportion of patients who test negative,
actually do not have the disease in question.

44
 Limitations of Predictive Values

• The positive predictive value is directly related to

the prevalence of the disease.

• Similarly, the negative predictive value is

inversely related to the prevalence.

45
 Likelihood Ratio

“Ratio of probability of the specific test result in

people who do have the disease to the probability
who do not.”

❖ Likelihood Ratio Positive

Ratio of probability of a person having a positive result
in a person with disease as opposed to a person without
the disease.
= Sensitivity
(1-Specificity)

❖ Likelihood Ratio Negative

Ratio of probability of a person having a negative result
in a person without disease as opposed to a person with
the disease. 46
Example: if sensitivity of ELISA is 99% (i.e.
0.99) and specificity is 90% (i.e. 0.90)

Here, LR + = 0.99 / (1 – 0.90 )= 9.9

The interpretation is “A positive result on ELISA for HIV

is 9.9 times (say, roughly 10 times) more likely to occur
in a subject with HIV infection as compared to a subject
who does not have HIV infection”.

Here, LR - = (1-0.99)/0.90 = 0.01

The interpretation is that a negative result is only one

hundredth times likely to occur in a person who really
has HIV infection as compared to a person who does not
have HIV infection.
47
Reliability and Validity

48
 Yield

Yield is the amount of previously

unrecognized disease that is detected and
brought to treatment as a result of
Screening.

YIELD = (TP + FP) / (TP + FP + TN + FN)

❖ It depends on prevalence of the disease and

sensitivity of the screening test, participation in
the programme.

❖ Hence, yield of a screening test is high in high –

risk screening. 49
 “
Evaluation of
Screening Tests

50
 Receiver Operating Characteristic
Curve
ROC Curve

It is constructed by plotting the values of True Positive

Rate against the corresponding values of False Positive
Rate at various cut-off levels.

Optimum Cut-Off: Which gives the best combination of

specificity and sensitivity
• Cut off for ROC is 0.5

• The curve should not

touch the diagonal line

• The curve should not be

below the diagonal line

• Area under the curve >

50%
51
 Utility of ROC:

• To discard the test which

does not help in diagnosis
beyond a chance

• To decide better cut-off

values while using
quantitative assays/tests
in diagnosis

• To select best diagnostic

test for application among
different tests currently
available
52
 Other methods:

• Fagan Nomogram

• Measurement of Test on a Continuous Scale

There is no specific cut-off value which will give 100%
specificity and 100% sensitivity.
If cut-off is lowered, the test will have higher sensitivity
value and low specificity.
If the cut-off level is raised, it will result in high
specificity and low sensitivity.
53
 “
Evaluation of
Screening
Programmes

54
Evaluation of Screening Programmes
❖ Experimental:
Conduct an RCT of the screening test to compare the
disease specific cumulative mortality rate between the
intervention and control group.
Problems include- Long follow up, Costs and record
keeping. Allows study of distribution of lead time,
effects of early treatment and identification of prognostic
factors.
Uncontrolled trials
❖ Non – experimental:
Cohort study: comparison of advanced disease or death
rates in those who choose to screen and those who do not
Case -control study: comparison of screening history in
those who have advanced disease and those who are healthy
Ecological study: correlation of screening pattern and
disease experience of several populations
55
 “
Conclusion

56
 Apparently Healthy
(Screening Tests)

Apparently Apparently Abnormal

Normal

a. Normal Periodic Re-

Periodic Re- Screening
Screening b. Intermediate Surveillance
c. Abnormal Treatment

Fig: Possible Outcomes of Screening

57
 Conclusion
❖ Screening can promote or impair health, depending on its
application.

❖ Unlike a diagnostic test, a screening test is done in

apparently healthy people which raises unique ethical
concerns.

❖ Sensitivity and specificity tend to be inversely related, and

the choice of the cut-off point for abnormal should indicate
the implications of incorrect results.

❖ Even very good tests have poor predictive value positive

when applied to low-prevalence populations.

❖ Lead-time and length bias exaggerate the apparent benefit

of screening programmes, underscoring the need for 58
 “
References

59
 References
1. Park K. Park’s Textbook of Preventive and Social
Medicine.24th ed. Jabalpur: M/s Banasidas Bhanot;2017.
Chapter 4, Screening for Disease; p. 145-52.

2. Gordis L. Epidemiology. 5th ed. Philadelphia: Elsevier;2014.

Section 1, Assessing the Validity and Reliability of Diagnostic
and Screening Tests; p. 88-113.

3. Bhalwar R editor. Textbook of Public Health and Community

Medicine. 1st ed. Pune: Department of community medicine
AFMC: 2009.

4. Suryakantha AH. Community Medicine with Recent

Advances.4th ed. New Delhi: Jaypee Brothers Medical
Publishers; 2017. Chapter 19, Epidemiology of Infectious
Disease; p. 304-08.

5. Grimes DA, Schulz KF. Uses and abuses of screening tests. 60

Epidemiological
Exercises
 Question 1
A rapid test for streptococcal pharyngitis was tested on
400 individuals who have culture proven streptococcal
pharyngitis and 400 others who were suffering from other
febrile illnesses. There were 320 test positive, but 25 of
them were false positive.

a. Calculate sensitivity
b. Calculate specificity
c. Calculate PPV
d. Calculate NPV
 Streptococcal Pharyngitis
New Test
Total
S S
Yes No
E P
Yes N E
295 25 320
S
POSITIVE PREDICTIVE VALUE C
a I b I a+b
T F
No 105 I VALUE
NEGATIVE PREDICTIVE 375 I 480
c V d C c+d
I I
Total 400 T 400 T 800
Ans: a+c
Y
b+d
Y

◎ Sensitivity: a/a+c = 295/400 * 100 = 73.75%

◎ Specificity: d/b+d = 375/400 * 100 = 93.75%
◎ PPV: a/a+b = 295/320 * 100 = 92.18%
◎ NPV: d/c+d = 375/480 * 100 = 72.12%

63
 Question 2
A group of migrants with a history of unprotected sex were
subjected to HIV testing using Western Blot and a new
testing. Out of the 360 migrants, 110 were positive for
Western blot. Out of this 110, only 60 were positive for the
new test. Among the remaining who showed negative for
Western Blot, 104 were positive with the new test.

a. Calculate sensitivity
b. Calculate specificity
c. Calculate PPV
d. Calculate NPV
 HIV ( Western Blot )
New Test Total
Yes No
60 104 164
Yes
50 146 196
No

Total 110 250 360

65
 Question 3
A programme for screening of DM was done in Hadinaru
village. Totally 500 people were screened. Based on blood
sugar measurement by digital glucometer, 120 people were
found to be positive. On further investigations by glucose
oxidase peroxidase method, 100 were diabetic of which
glucometer had detected 90 cases correctly.

a. Calculate sensitivity
b. Calculate specificity
c. Calculate PPV
d. Calculate NPV of digital glucometer .

66
Test Positive Negative Total
Positive 90 30 120

Negative 10 370 380

Total 100 400 500

67
 Question 4
A study was conducted during 2015 at a private medical
college of Venjaramoodu to determine utility of FNAC for
IDD. 1500 individuals with enlarged thyroid were subjected
to FNAC and biopsy examination, of which 450 cases were
positive for both examination whereas 90 more cases were
detected by biopsy examination. The FNAC positive for IDD
was 680.
a.Calculate sensitivity
b.Calculate specificity
c.Calculate PPV
d.Calculate NPV
FNAC BIOPSY
Total
Yes No
Yes 450 230 680
No 90 730 820
Total 540 960 1500

69
 Question 5
A new Screening test for a disease was administered to
520 patients, 70 of whom are known to have the disease.
The test was positive in 55 persons with the disease and 25
persons without the disease. Construct a 2x 2 table.

a. Calculate sensitivity
b. Calculate specificity
c. Calculate PPV
d. Calculate NPV
e. prevalence of disease.

70
 Disease
Screening Present Absent Total
test
Positive 55 25 80
Negative 15 425 440
Total 70 450 520

71
 Question 6
Pap smear test was being evaluated for cervical cancer.
This test was done for 480 females, 60 of whom were
known to have the disease. This test was found to be
positive in 50 of the 60 females with disease as well as 150
female who didn’t have the disease. Calculate following
parameters:
a. Calculate Sensitivity
b. Specificity
c. Positive predictive value
d. Negative predictive value
e. Percentage of false positive cases
f. Percentage of false negative cases
g. Prevalence of the disease.
72
 Disease

Test Present Absent Total

Positive 50 (a) 150 (b) FP 200

Negative 10 (c) FN 270(d) 280
Total 60 420 480

Prevalence = Number of people having the disease x 1000

Total No of people

60 x 1000 = 125 / 1000 population

480

73
 To be written in the Record
books
• Introduction - Definition
• Uses of Screening
• Types of Screening
• Difference b/w Screening & diagnostic test
• Criteria for Screening
• Evaluation of Screening Tests
• Evaluation of Screening Programmes
• Epidemiological exercises – 6 Question & answers

74
THANK YOU

75