DIAGNOSTIC STUDY
Dr. A. PRECILLA CATHERINE, MD,DO,Ph.D
Professor of Physiology,
Govt. Thiruvannamalai Medical College
Specific Learning Objectives
At the end of the session, the participants will be able to
• Discuss about the various applications of DIAGNOSTIC Tests
• Able to outline the design of a diagnostic study
• Draw a diagnostic 2 X 2 contingency table for Summary statistic
• Discuss about different parameters used to assess Diagnostic
tests
Diagnostic tests establish PRESENCE or ABSENCE of Disease
• WHERE are they useful ??
Selection of Diagnostic
SCREENING test Depends on-
DIAGNOSIS • Disease
STAGING • Population
MONITORING & SURVEILLANCE
• Cost of treatment
• WHY should they be assessed ?? • Test Parameters
NO test is IDEAL, NO test is 100 % Reliable
IMPORTANT TERMINOLOGIES
• PREVALENCE
• REFERENCE TEST/ GOLD STANDARD
• INDEX TEST
• VALIDITY
• RELIABILITY
PICO Framework for Research question for
Diagnostic studies
P - POPULATION
I - INDEX Test
C - COMPARATOR- Reference Test -Gold Standard
O - OUTCOME
Select Study SAMPLE
Perform INDEX Test
Compare Outcome with
REFERENCE Test - GOLD Standard
Parameters to assess Diagnostic tests
• SENSITIVITY • LIKELIHOOD Ratio (LR+, LR-)
• SPECIFICITY • ROC Curve & AUC
• Positive Predictive Value
(PPV) • Youden’s Index
• Negative Predictive Value • Kappa statistic
(NPV)
Diagnostic 2 X 2 Contingency TABLE
ID Number- 1 2 3 4 5 6 7 8 9 10
Reference test- P P P N P N N P N N
Index test- P N P N P P N P N P
TP FN TP TN TP FP TN TP TN FP
Reference Test Reference Test
POSITIVE NEGATIVE
Index Test TRUE POSITIVE - TP FALSE POSITIVE - FP
POSITIVE a b
Index Test FALSE NEGATIVE - FN TRUE NEGATIVE - TN
NEGATIVE c d
Diagnostic 2 X 2 Contingency TABLE
ID Number- 1 2 3 4 5 6 7 8 9 10
Reference test- P P P N P N N P N N
Index test- P N P N P P N P N P
TP FN TP TN TP FP TN TP TN FP
Reference Test Reference Test
POSITIVE NEGATIVE
Index Test TP - 4 FP - 2
POSITIVE a b
Index Test FN - 1 TN - 3
NEGATIVE c d
Diagnostic 2 X 2 Contingency TABLE
Reference Test Reference Test Reference Test Reference Test
POSITIVE NEGATIVE POSITIVE NEGATIVE
Index Test TP - 4 FP - 2 Index TP - 4 FP - 2
POSITIVE a b Test a b
POSITIVE
Index Test FN - 1 TN - 3 Index FN - 1 TN - 3
NEGATIVE c d Test c d
NEGATIVE
Sensitivity Specificity
4/5 3/5
Sensitivity TRUE POSITIVE Specificity TRUE NEGATIVE
= =
TRUE POSITIVE + FALSE NEGATIVE TRUE NEGATIVE + FALSE POSITIVE
Diagnostic 2 X 2 Contingency TABLE
Reference Reference
Test Test
POSITIVE NEGATIVE
PPV TRUE POSITIVE
Index Test TP - 4 FP - 2 PPV =
POSITIVE a b 4/6 TRUE POSITIVE + FALSE POSITIVE
Index Test FN - 1 TN - 3
NEGATIVE c d
Reference Reference
Test Test
POSITIVE NEGATIVE
Index Test TP - 4 FP - 2
POSITIVE a b
NPV TRUE NEGATIVE
Index Test FN - 1 TN - 3 NPV =
NEGATIVE c d TRUE NEGATIVE + FALSE NEGATIVE
3/4
Sensitivity & Specificity
Vs
PPV & NPV
Effect of PREVALENCE
Diagnostic 2 X 2 Contingency TABLE
Reference Test Reference Test Reference Test Reference Test
POSITIVE NEGATIVE POSITIVE NEGATIVE
Index Test TP - 4 FP - 2 Index TP - 4 FP - 2
POSITIVE a b Test a b
POSITIVE
Index Test FN - 1 TN - 3 Index FN - 1 TN - 3
NEGATIVE c d Test c d
NEGATIVE
Sensitivity Specificity
4/5 3/5
Sensitivity 1 - Sensitivity
LR + LR -
= 1 - Specificity = Specificity
ROC Curve - Receiver Operating Characteristic CURVE
• ROC curves plot the Sensitivity of a Test
Vs False Positive Rate ( i.e, 1
– Specificity ) for various cut-off
points
• AUC - Area under the ROC Curve
• AUC of 1.0 indicates theoretically
perfect test
• Test with AUC of 0.5 does not have
discriminative value
ROC Curve - Receiver Operating Characteristic CURVE
• ROC curves plot the Sensitivity of
a Test Vs False Positive Rate
( i.e, 1 – Specificity ) for
various cut-off points
• AUC - Area under the ROC Curve
• AUC of 1.0 indicates theoretically
perfect test
• Test with AUC of 0.5 does not
ROC Curve - Receiver Operating Characteristic CURVE
• ROC curves plot the Sensitivity
of a Test Vs False Positive Rate
( i.e, 1 – Specificity ) for
various cut-off points
• AUC - Area under the ROC Curve
• AUC of 1.0 indicates
theoretically perfect test
• Test with AUC of 0.5 does not
have discriminative value
CHECKLISTS for Reporting Diagnostic Studies
Standards for the Reporting of
Diagnostic Accuracy studies (STARD)
Quality Assessment of Diagnostic
Accuracy Studies 2 ( QUADAS -2
Tool )
We have to Reduce BIAS in Diagnostic Studies to improve Validity
Specific Learning Objectives
At the end of the session, the participants will be able to
• Discuss about the various applications of DIAGNOSTIC Tests
• Able to outline the design of a diagnostic study
• Draw a diagnostic 2 X 2 contingency table for Summary statistic
• Discuss about different parameters used to assess Diagnostic
tests
