SCREENING IN PUBLIC
HEALTH
1
SCREENING IN DISEASE CONTROL
Definition:-
• Screening is a public health intervention intended to
improve the health of a precisely defined target
population.
• The presumptive identification of unrecognized disease
by application of rapid tests or examinations.
• The early detection of disease
– Precursors of disease
– Susceptibility to disease in individuals who do not
show any signs of disease
Definition…
• Detecting disease or risk factors in asymptomatic
individuals
• The examination of asymptomatic people in order to
classify them as likely or unlikely to have the diseases
of interest.
– Diagnosis is not equal to Screening
• Screening sort out apparently well persons who
probably have a disease from those probably do not.
Diagnostic test
o Diagnostic tests are used to confirm the presence or
absence of illness, provide prognostic information, and
predict a response to treatment
o Identify and confirm a disease condition in individuals
– Diagnostic test is performed in persons with symptoms
or a signs of an illness
– Tests performed in patients
– The objective is case finding within a population that
is probably “diseased”
4
Screening and diagnostic tests
o May be based on:
–Standardized interviews
–Physical examinations
–Laboratory tests
–More sophisticated measurements
radiography
electro-cardiograph etc…
5
Natural history of diseases
Clinical Aim of Screening
• To reverse, halt, or slow the progression of disease
more effectively than would probably normally
happen.
• To alter the natural course of disease for a better
outcome for individuals affected.
• Reduce morbidity and mortality through early
detection and treatment
Public health aim of screening
program
o To protect society from communicable disease
o To reduce mortality
o For rational allocation of resources…
Other Use:
o Research: study on natural history of disease
o Selection of healthy individuals usually for employment
E. g: Military and driving license …
8
Screening…..
• Screening is only useful for disease which can be
detected in their early, pre-clinical stage, and for which
early treatment significantly improves the outcome
• Screening can also be of useful for the communicable
disease control, such as trachoma, Schistosomiasis, or
TB, for which early treatment helps to prevent
transmission
Types of screening
programme
1. Mass/population screening
– It is offered to all, irrespective of the particular risk factors
2. Multiple/multi-phase screening
– The purpose of two or more screening tests in combination to a
large number of people at one time
3. Case finding/opportunistic screening
– It is restricted to patients who consult a health professional for
some other purposes
4. Targeted screening
– High risk or selective screening
– It will be most productive if applied selectively to high risk
groups
10
Types of screening
Selective Vs Mass.
– Selective – screening of people with selective exposure
– Mass – screening of people without reference to specific
exposure
• Single Vs Multiple
• Multiple-Parallel Vs Series
– Parallel testing – applying two screening tests and a positive
result on either test is sufficient to be labeled as positive E.g. –
Breast ca screening
– Series testing – applying two screening tests and both must
be positive in order to prompt action
E.g. – HIV testing, Syphilis
Criteria for establishing screening program
1. The condition should be an important health problem (Life-
threatening diseases)
2. There should be a treatment for the condition.
3. Facilities for diagnosis and treatment should be available.
4. There should be a latent stage of the disease.
5. A suitable screening test must be available
Modified Wilson criteria for introducing a screening program
1. The disease condition
– Important health condition (prevalent, serious)
• Has early stage
– High prevalence at asymptomatic stage
– Adequately understood natural history
2. The test
– Simple, precise, validated
– Acceptable to population
– Agreed diagnostic test(s)
– Available, adequate, effective and acceptable
– Sensitive, specific, reliable, simple, cheap, safe
Modified Wilson criteria, cont..
3. The treatment
– Effective treatment Available, adequate, and
acceptable
– Evidence based policies on whom to treat
4. The program
– Acceptable to public and professionals
– Benefits outweigh risks
– Available facilities and staff
– Overall costs assessed
WHO criteria for screening
programme
Is it a health problem?
Is there treatment?
Are there facilities in place?
Is it detectable pre-clinically?
Is there a suitable screening test?
Is the screening test acceptable to people?
Is the natural history of disease understood?
Are the costs acceptable?
15
Evaluation of a screening
program
Reliability
Feasibility
Validity
Performance
Yields
Effectiveness
Characteristics to be evaluated
• These characteristics are:
1. Validity
2. Predictive value
3. Reliability
4. Yield
1. Validity is analogous to accuracy
– The validity of a screening test is how well the given
screening test reflects another test of known greater accuracy
– Validity assumes that there is a gold standard to which a test
can be compared
• Validity- does it truly measure what it sets out to measure?
• Variability/reliability -does it gives the same measurement each time
Accuracy of screening tests…
o A test should be unbiased, precise, and be able to
determine the disease status of an individual without
error
o Accuracy is the ability of the test to correctly classify
individuals according to their disease status
Examples:
o Classify exposed versus non-exposed individuals
o Classify infected versus non-infected individuals
o Classify abnormal versus normal individuals etc…
18
Characteristics of a screening test
Validity of a categorical test
Screening
Test result
Definitive Diagnosis(Gold Standard)
Diseased Non diseased Total
Positive True Positive
(TP)
False Positive
(FP)
TP + FP
Negative False
negative
(FN)
True
Negative
(TN)
TN + FN
Total TP + FN FP + TN TP +FP +
TN + FN
Two-by-two contingency table
Screening
Test
Diagnostic tests
Total
Diseased Non-diseased
Positive
True positive [TP]
(a)
False positive[FP]
(b) a + b
Negative
False negative [FN]
(c)
True negative[TN]
(d) c + d
Total a + c b + d n
20
Two-by-two contingency table…
True positive
o The number of individuals for whom the screening test is
positive and the individual actually has the disease
False positive
o The number for whom the screening test is positive but
the individual does not have the disease
True negative
o The number for whom the screening test is negative and
the individual does not have the disease
False negative
o The number for whom the screening test is negative but
the individual does have the disease 21
Measures of accuracy of screening
tests
o Validity is the ability of a test to come up
with a result which is closer to the actual
value
o Reliability is the ability of a test to come up
with similar values upon repeated
measurements in similar occasions
22
Measures of accuracy…
Validity:
o The application of a screening test is to identify correctly
individuals who do and do not have preclinical disease
o Those who have preclinical disease should test positive,
and those who do not have it should test negative
o The ability of a screening test to successfully separate
these two groups is a measure of its validity which is
expressed by its sensitivity and specificity
23
Reliability versus validity of a test
24
A. Sensitivity
– The ability of the test to identify correctly those who have the
condition, Sensitivity + Fn = 100%
– A test with high sensitivity will have few false negatives
%
100



Fn
Tp
Tp
y
Sensitivit
Sensitivity (detection rate) =a/a+c
=(Tp/Tp+Fn)X100
Proportion of true positive with diseases
B. Specificity
The ability of the test to identify correctly those who
don’t have the disease
–Specificity + false positive = 100%
–test that has high specificity will have few
false positives
%
100
y
Specificit 


Fp
Tn
Tn
Specificity=d/d+b=(Tn/Tn+Fp)X100
Proportion of true negatives identified
Validity of Screening Tests
11/4/2023 27
132
63650
45
983
Breast Cancer
+ -
Physical Exam
and Mammo-
graphy
+
-
Sensitivity: a / (a + c)
Sensitivity = 132 / (132 + 45) = 74.6%
Specificity: d / (b + d)
Specificity = 63650 / (983 + 63650) = 98.5%
Predictive values of screening
test
o It is the probability that a person tested positive or
negative will or will not have a disease respectively
o Predictive values of screening test depends on:
– Prevalence of preclinical disease
– Sensitivity
– Specificity
28
Positive predictive value
o The probability of a person having the preclinical disease
when the test is positive
o It is defined as the proportion of people with the condition
among all those who received a positive test result
o It is the probability of the presence of the disease in a
person with a positive (abnormal) test
o It is calculated as a percentage; the number of individuals
with preclinical disease who test positive is in the
numerator, and the total number of individuals who test
positive is in the denominator
29
Negative predictive value
o The probability of a person not having the preclinical
disease when the test is negative
o It is the proportion of people without the condition
among all those who received a negative test result
o It is the probability of not having the disease when the
test result is negative
o It is calculated as a percentage; the number of
individuals without the disease who test negative is in
the numerator, and the total number of individuals who
test negative is in the denominator
30
Validity and predictive values of
screening test
31
Screening test trade-offs
o Tests with dichotomous results – tests that give either
positive or negative results
E.g: HIV positive or negative
o Tests with continuous variables – tests that do not
yield obvious “positive” or “negative” results, but
require a cutoff level to be established as criteria for
distinguishing between “positive” and “negative” groups
E.g: Blood pressure measurement
32
Screening test trade-offs…
 There is always a trade-off between sensitivity
and specificity; as one increases, the other
tends to decreases
o The higher the cutoff used to say a test is positive, the
more specific the test becomes, but this higher specificity
comes at a price
o As the cutoff is increased, the sensitivity decreases and
the test is more likely to miss affected individuals (life
lost) i.e. false negative increases
33
Screening test trade-offs…
o A lower cutoff might be used to create a very
sensitive test, so as not to miss anyone with the
disorder in question =false +ve
o In other situations, when using a test to confirm a
diagnosis, a higher cutoff, making the test highly
specific, is more desirable, so as not to incorrectly
label anyone with the disorder =false –ve
34
Choosing appropriate test
– Choose a test with a high sensitivity and high
negative predictive value to rule out a diagnosis
in the early stage of the investigation
– When a False Negative is DANGEROUS or
SERIOUS
35
Choosing appropriate test…
o Choose a test with a high specificity and
high positive predictive value to confirm a
diagnosis
o When a False Positive is DANGEROUS or
SERIOUS
36
Choosing appropriate test…
High sensitivity High
specificity
37
Sensitivity and specificity are affected by the ‘cut-off’ value of
measure at
which a test is defined as positive.
Yield of screening test
o The proportion of truly diseased individuals identified by the
screening test among screened population.
o The yield of a test result is a function of:
 Specificity of the test
 Prevalence of the preclinical disease
Strategies to increase yield of a test
o Select disease with high prevalence of preclinical
stage
o Target high risk group for screening program
o Select a test with high specificity test
38
Yield…
• The number of cases of the condition detected by screening
test in relation to the total number of persons screened
• Affected by sensitivity, prevalence of unrecognized cases,
number of tests employed, frequency of screening and
participation of people in screening and follow-up
%
100





Fp
Fn
Tp
Tn
Tp
Yeild
%
100
screened
Total
test
by the
detected
dx
with the
persons


Yeild
,
Yields of screening test
Reference test
Screening test
positive negative total
+ve 1555 988 2543
-ve 514 1394 1908
T 2069 2382 4451
Example…,
Sensitivity =1555/2069 =75.2%
Specificity=1394/2382 =58.5%
PVPT=1555/2543 = 61.1%
PVNT=1394/1908 =_____
Yield=1555/4451=______
RELIABILITY
• The ability of a test to give consistent results when it is
performed more than once on the same individual under the
similar conditions
• The reliability of a test is its capacity to give the same result -
positive or negative, whether correct or incorrect - on repeated
application in a person with a given level of disease
• Affected by variation in the method, observer and the
characteristic to be measured
Reliability…
o Intra-observer reliability – agreement among
measurements made by the same observer
o Inter-observer reliability – agreement among
measurements made by different observers
o Test-retest reliability – agreement among
measurements on the same subject at different times
o Biological (or subject) variation
o Technical (instrumental) problems
o Reagent problems
43
RELIABILITY…
• The reliability of a dichotomous test can be described
using a 2×2 table in which the first and second test
results are presented on the left-hand side and top of
the table, respectively.
RELIABILITY…
Second test
First test
positive negative total
+ve tests agree tests disagree
-ve tests disagree tests agree
T
RELIABILITY…
• A simple way to summarize these findings is to
calculate the proportion of tests that agree:
• Percent agreement = Number of tests that agree
Total number of tests
• Percent agreement = a+d *100%
n
• However, test agreement due to chance alone is
possible, even with an imprecise test.
• The Kappa statistic is used to describe test agreement
beyond that expected from chance alone.
• Calculation of the Kappa statistic is presented below.
RELIABILITY…
RELIABILITY…
• The Kappa statistic ranges from +1 (perfect agreement), to 0 (no
agreement beyond that expected from chance), to −1 (perfect
disagreement).
• In general, a Kappa statistic <0.2 is considered poor agreement,
0.2 – 0.6 is considered fair agreement, and > 0.6 is considered
good agreement.
• A common application of the Kappa statistic is to measure
agreement between two different testers who evaluate a
subjective test characteristic
Example
Second test
First test
positive negative total
+ve 5 3
8
-ve 2 30
32
T 7 33 40
Potential Source of Bias in
Screening
• Self-selection (volunteer) bias
• Lead time bias (early diagnosis)
• Length Bias (chronicity and progression)
Self-selection (volunteer) bias
• Implies that those accepting screening are different from those
declining it.
• Comparing populations accepting and declining for all
relevant characteristics.
Lead time bias (early diagnosis)
• Bias caused by picking screened cases at an early
stage of the disease, i.e., before they develop signs
and symptoms of the disease
Length Bias (chronicity and
progression):
• Related to the variation in the speed of progression of the disease.
• Cases picked up by a screening may be less severe, and slow
progressive compared with others.
• It is very important to be aware of such a possibility in interpreting
survival gains from a screening program.
Success Factors for Screening
Programs
• The health problem should be important enough to be worth
detecting.
• Availability of an acceptable intervention which is effective.
• Presence of a recognizable latent or early "asymptomatic" stage.
• Presence of a suitable and acceptable test.
Success Factors…
• The natural history of the condition should be adequately
understood.
• Presence of policy regarding when the intervention is appropriate.
• The cost of detecting the problem and its remedy should be
reasonable.
• The screening program should be ongoing, and not a "one-time"
effort.
Study Design for Evaluation of
Screening
• Ideally experimental
• Most commonly used is cohort
Summary
 A screening program is defined as a set of procedures for
early detection and treatment of a disease
─ It includes both diagnostic and therapeutic components, and
includes the screening test and the follow-up evaluation for people
with positive tests
─ The therapeutic component consists of treatment of confirmed cases
of the disease
─ Thus, a screening program is much more than just a screening test
57
END
58

7 Lecture 8 Outbrek invesitgation.pptx

  • 1.
  • 2.
    SCREENING IN DISEASECONTROL Definition:- • Screening is a public health intervention intended to improve the health of a precisely defined target population. • The presumptive identification of unrecognized disease by application of rapid tests or examinations. • The early detection of disease – Precursors of disease – Susceptibility to disease in individuals who do not show any signs of disease
  • 3.
    Definition… • Detecting diseaseor risk factors in asymptomatic individuals • The examination of asymptomatic people in order to classify them as likely or unlikely to have the diseases of interest. – Diagnosis is not equal to Screening • Screening sort out apparently well persons who probably have a disease from those probably do not.
  • 4.
    Diagnostic test o Diagnostictests are used to confirm the presence or absence of illness, provide prognostic information, and predict a response to treatment o Identify and confirm a disease condition in individuals – Diagnostic test is performed in persons with symptoms or a signs of an illness – Tests performed in patients – The objective is case finding within a population that is probably “diseased” 4
  • 5.
    Screening and diagnostictests o May be based on: –Standardized interviews –Physical examinations –Laboratory tests –More sophisticated measurements radiography electro-cardiograph etc… 5
  • 6.
  • 7.
    Clinical Aim ofScreening • To reverse, halt, or slow the progression of disease more effectively than would probably normally happen. • To alter the natural course of disease for a better outcome for individuals affected. • Reduce morbidity and mortality through early detection and treatment
  • 8.
    Public health aimof screening program o To protect society from communicable disease o To reduce mortality o For rational allocation of resources… Other Use: o Research: study on natural history of disease o Selection of healthy individuals usually for employment E. g: Military and driving license … 8
  • 9.
    Screening….. • Screening isonly useful for disease which can be detected in their early, pre-clinical stage, and for which early treatment significantly improves the outcome • Screening can also be of useful for the communicable disease control, such as trachoma, Schistosomiasis, or TB, for which early treatment helps to prevent transmission
  • 10.
    Types of screening programme 1.Mass/population screening – It is offered to all, irrespective of the particular risk factors 2. Multiple/multi-phase screening – The purpose of two or more screening tests in combination to a large number of people at one time 3. Case finding/opportunistic screening – It is restricted to patients who consult a health professional for some other purposes 4. Targeted screening – High risk or selective screening – It will be most productive if applied selectively to high risk groups 10
  • 11.
    Types of screening SelectiveVs Mass. – Selective – screening of people with selective exposure – Mass – screening of people without reference to specific exposure • Single Vs Multiple • Multiple-Parallel Vs Series – Parallel testing – applying two screening tests and a positive result on either test is sufficient to be labeled as positive E.g. – Breast ca screening – Series testing – applying two screening tests and both must be positive in order to prompt action E.g. – HIV testing, Syphilis
  • 12.
    Criteria for establishingscreening program 1. The condition should be an important health problem (Life- threatening diseases) 2. There should be a treatment for the condition. 3. Facilities for diagnosis and treatment should be available. 4. There should be a latent stage of the disease. 5. A suitable screening test must be available
  • 13.
    Modified Wilson criteriafor introducing a screening program 1. The disease condition – Important health condition (prevalent, serious) • Has early stage – High prevalence at asymptomatic stage – Adequately understood natural history 2. The test – Simple, precise, validated – Acceptable to population – Agreed diagnostic test(s) – Available, adequate, effective and acceptable – Sensitive, specific, reliable, simple, cheap, safe
  • 14.
    Modified Wilson criteria,cont.. 3. The treatment – Effective treatment Available, adequate, and acceptable – Evidence based policies on whom to treat 4. The program – Acceptable to public and professionals – Benefits outweigh risks – Available facilities and staff – Overall costs assessed
  • 15.
    WHO criteria forscreening programme Is it a health problem? Is there treatment? Are there facilities in place? Is it detectable pre-clinically? Is there a suitable screening test? Is the screening test acceptable to people? Is the natural history of disease understood? Are the costs acceptable? 15
  • 16.
    Evaluation of ascreening program Reliability Feasibility Validity Performance Yields Effectiveness
  • 17.
    Characteristics to beevaluated • These characteristics are: 1. Validity 2. Predictive value 3. Reliability 4. Yield 1. Validity is analogous to accuracy – The validity of a screening test is how well the given screening test reflects another test of known greater accuracy – Validity assumes that there is a gold standard to which a test can be compared • Validity- does it truly measure what it sets out to measure? • Variability/reliability -does it gives the same measurement each time
  • 18.
    Accuracy of screeningtests… o A test should be unbiased, precise, and be able to determine the disease status of an individual without error o Accuracy is the ability of the test to correctly classify individuals according to their disease status Examples: o Classify exposed versus non-exposed individuals o Classify infected versus non-infected individuals o Classify abnormal versus normal individuals etc… 18
  • 19.
    Characteristics of ascreening test Validity of a categorical test Screening Test result Definitive Diagnosis(Gold Standard) Diseased Non diseased Total Positive True Positive (TP) False Positive (FP) TP + FP Negative False negative (FN) True Negative (TN) TN + FN Total TP + FN FP + TN TP +FP + TN + FN
  • 20.
    Two-by-two contingency table Screening Test Diagnostictests Total Diseased Non-diseased Positive True positive [TP] (a) False positive[FP] (b) a + b Negative False negative [FN] (c) True negative[TN] (d) c + d Total a + c b + d n 20
  • 21.
    Two-by-two contingency table… Truepositive o The number of individuals for whom the screening test is positive and the individual actually has the disease False positive o The number for whom the screening test is positive but the individual does not have the disease True negative o The number for whom the screening test is negative and the individual does not have the disease False negative o The number for whom the screening test is negative but the individual does have the disease 21
  • 22.
    Measures of accuracyof screening tests o Validity is the ability of a test to come up with a result which is closer to the actual value o Reliability is the ability of a test to come up with similar values upon repeated measurements in similar occasions 22
  • 23.
    Measures of accuracy… Validity: oThe application of a screening test is to identify correctly individuals who do and do not have preclinical disease o Those who have preclinical disease should test positive, and those who do not have it should test negative o The ability of a screening test to successfully separate these two groups is a measure of its validity which is expressed by its sensitivity and specificity 23
  • 24.
  • 25.
    A. Sensitivity – Theability of the test to identify correctly those who have the condition, Sensitivity + Fn = 100% – A test with high sensitivity will have few false negatives % 100    Fn Tp Tp y Sensitivit Sensitivity (detection rate) =a/a+c =(Tp/Tp+Fn)X100 Proportion of true positive with diseases
  • 26.
    B. Specificity The abilityof the test to identify correctly those who don’t have the disease –Specificity + false positive = 100% –test that has high specificity will have few false positives % 100 y Specificit    Fp Tn Tn Specificity=d/d+b=(Tn/Tn+Fp)X100 Proportion of true negatives identified
  • 27.
    Validity of ScreeningTests 11/4/2023 27 132 63650 45 983 Breast Cancer + - Physical Exam and Mammo- graphy + - Sensitivity: a / (a + c) Sensitivity = 132 / (132 + 45) = 74.6% Specificity: d / (b + d) Specificity = 63650 / (983 + 63650) = 98.5%
  • 28.
    Predictive values ofscreening test o It is the probability that a person tested positive or negative will or will not have a disease respectively o Predictive values of screening test depends on: – Prevalence of preclinical disease – Sensitivity – Specificity 28
  • 29.
    Positive predictive value oThe probability of a person having the preclinical disease when the test is positive o It is defined as the proportion of people with the condition among all those who received a positive test result o It is the probability of the presence of the disease in a person with a positive (abnormal) test o It is calculated as a percentage; the number of individuals with preclinical disease who test positive is in the numerator, and the total number of individuals who test positive is in the denominator 29
  • 30.
    Negative predictive value oThe probability of a person not having the preclinical disease when the test is negative o It is the proportion of people without the condition among all those who received a negative test result o It is the probability of not having the disease when the test result is negative o It is calculated as a percentage; the number of individuals without the disease who test negative is in the numerator, and the total number of individuals who test negative is in the denominator 30
  • 31.
    Validity and predictivevalues of screening test 31
  • 32.
    Screening test trade-offs oTests with dichotomous results – tests that give either positive or negative results E.g: HIV positive or negative o Tests with continuous variables – tests that do not yield obvious “positive” or “negative” results, but require a cutoff level to be established as criteria for distinguishing between “positive” and “negative” groups E.g: Blood pressure measurement 32
  • 33.
    Screening test trade-offs… There is always a trade-off between sensitivity and specificity; as one increases, the other tends to decreases o The higher the cutoff used to say a test is positive, the more specific the test becomes, but this higher specificity comes at a price o As the cutoff is increased, the sensitivity decreases and the test is more likely to miss affected individuals (life lost) i.e. false negative increases 33
  • 34.
    Screening test trade-offs… oA lower cutoff might be used to create a very sensitive test, so as not to miss anyone with the disorder in question =false +ve o In other situations, when using a test to confirm a diagnosis, a higher cutoff, making the test highly specific, is more desirable, so as not to incorrectly label anyone with the disorder =false –ve 34
  • 35.
    Choosing appropriate test –Choose a test with a high sensitivity and high negative predictive value to rule out a diagnosis in the early stage of the investigation – When a False Negative is DANGEROUS or SERIOUS 35
  • 36.
    Choosing appropriate test… oChoose a test with a high specificity and high positive predictive value to confirm a diagnosis o When a False Positive is DANGEROUS or SERIOUS 36
  • 37.
    Choosing appropriate test… Highsensitivity High specificity 37 Sensitivity and specificity are affected by the ‘cut-off’ value of measure at which a test is defined as positive.
  • 38.
    Yield of screeningtest o The proportion of truly diseased individuals identified by the screening test among screened population. o The yield of a test result is a function of:  Specificity of the test  Prevalence of the preclinical disease Strategies to increase yield of a test o Select disease with high prevalence of preclinical stage o Target high risk group for screening program o Select a test with high specificity test 38
  • 39.
    Yield… • The numberof cases of the condition detected by screening test in relation to the total number of persons screened • Affected by sensitivity, prevalence of unrecognized cases, number of tests employed, frequency of screening and participation of people in screening and follow-up % 100      Fp Fn Tp Tn Tp Yeild % 100 screened Total test by the detected dx with the persons   Yeild
  • 40.
    , Yields of screeningtest Reference test Screening test positive negative total +ve 1555 988 2543 -ve 514 1394 1908 T 2069 2382 4451
  • 41.
    Example…, Sensitivity =1555/2069 =75.2% Specificity=1394/2382=58.5% PVPT=1555/2543 = 61.1% PVNT=1394/1908 =_____ Yield=1555/4451=______
  • 42.
    RELIABILITY • The abilityof a test to give consistent results when it is performed more than once on the same individual under the similar conditions • The reliability of a test is its capacity to give the same result - positive or negative, whether correct or incorrect - on repeated application in a person with a given level of disease • Affected by variation in the method, observer and the characteristic to be measured
  • 43.
    Reliability… o Intra-observer reliability– agreement among measurements made by the same observer o Inter-observer reliability – agreement among measurements made by different observers o Test-retest reliability – agreement among measurements on the same subject at different times o Biological (or subject) variation o Technical (instrumental) problems o Reagent problems 43
  • 44.
    RELIABILITY… • The reliabilityof a dichotomous test can be described using a 2×2 table in which the first and second test results are presented on the left-hand side and top of the table, respectively.
  • 45.
    RELIABILITY… Second test First test positivenegative total +ve tests agree tests disagree -ve tests disagree tests agree T
  • 46.
    RELIABILITY… • A simpleway to summarize these findings is to calculate the proportion of tests that agree: • Percent agreement = Number of tests that agree Total number of tests • Percent agreement = a+d *100% n • However, test agreement due to chance alone is possible, even with an imprecise test. • The Kappa statistic is used to describe test agreement beyond that expected from chance alone. • Calculation of the Kappa statistic is presented below.
  • 47.
  • 48.
    RELIABILITY… • The Kappastatistic ranges from +1 (perfect agreement), to 0 (no agreement beyond that expected from chance), to −1 (perfect disagreement). • In general, a Kappa statistic <0.2 is considered poor agreement, 0.2 – 0.6 is considered fair agreement, and > 0.6 is considered good agreement. • A common application of the Kappa statistic is to measure agreement between two different testers who evaluate a subjective test characteristic
  • 49.
    Example Second test First test positivenegative total +ve 5 3 8 -ve 2 30 32 T 7 33 40
  • 50.
    Potential Source ofBias in Screening • Self-selection (volunteer) bias • Lead time bias (early diagnosis) • Length Bias (chronicity and progression)
  • 51.
    Self-selection (volunteer) bias •Implies that those accepting screening are different from those declining it. • Comparing populations accepting and declining for all relevant characteristics.
  • 52.
    Lead time bias(early diagnosis) • Bias caused by picking screened cases at an early stage of the disease, i.e., before they develop signs and symptoms of the disease
  • 53.
    Length Bias (chronicityand progression): • Related to the variation in the speed of progression of the disease. • Cases picked up by a screening may be less severe, and slow progressive compared with others. • It is very important to be aware of such a possibility in interpreting survival gains from a screening program.
  • 54.
    Success Factors forScreening Programs • The health problem should be important enough to be worth detecting. • Availability of an acceptable intervention which is effective. • Presence of a recognizable latent or early "asymptomatic" stage. • Presence of a suitable and acceptable test.
  • 55.
    Success Factors… • Thenatural history of the condition should be adequately understood. • Presence of policy regarding when the intervention is appropriate. • The cost of detecting the problem and its remedy should be reasonable. • The screening program should be ongoing, and not a "one-time" effort.
  • 56.
    Study Design forEvaluation of Screening • Ideally experimental • Most commonly used is cohort
  • 57.
    Summary  A screeningprogram is defined as a set of procedures for early detection and treatment of a disease ─ It includes both diagnostic and therapeutic components, and includes the screening test and the follow-up evaluation for people with positive tests ─ The therapeutic component consists of treatment of confirmed cases of the disease ─ Thus, a screening program is much more than just a screening test 57
  • 58.