1.

A newly introduced C-TB (Cellular Tuberculosis) screening test was administered to

300 individuals. Among them, 60 people were confirmed to have tuberculosis (TB)
using CBNAAT (Cartridge-Based Nucleic Acid Amplification Test), which is
considered the gold standard diagnostic method. The C-TB test yielded 48 positive
results among the confirmed TB cases and also showed 20 positive results among
individuals who were actually free of the disease.

Using this information, construict a 2 × 2 contingency table and calculate

Sensitivity,Specificity,Positive Predictive Value (PPV),Negative Predictive Value
(NPV),Diagnostic Accuracy (DA), andDisease Prevalence in the tested population.

Disease Present Disease Absent Total

Test Positive 48 (a) 20 (b) 68
Test 12 (c) 220 (d) 232
Negative
Total 60 240 300

Parameter Formula (with a, Substitution Resul Interpretation

b, c, d) t
Sensitivity Sensitivity = TP(a)/ 48 / (48 + 12) 80% The test correctly detects 80%
[TP(a) + FN(c)] of true TB cases (misses
20%).
Specificity Specificity = 220 / (220 + 91.7% The test correctly identifies
TN(d) / 20) 91.7% of non-TB individuals.
[TN(d) + FP(b)]
Positive PPV = TP(a) / 48 / (48 + 20) 70.6% 70.6% of individuals who test
Predictive [TP(a) + FP(b)] positive truly have TB.
Value (PPV)
Negative NPV = TN(d) / 220 / (220 + 94.8% 94.8% of individuals who test
Predictive [TN(d) + FN(c)] 12) negative truly do not have TB.
Value (NPV)
Diagnostic DA = [TP(a) + (48 + 220) / 89.3% The test gives the correct
Accuracy (DA) TN(d)] / 300 result in 89.3% of individuals
[a + b + c + d] tested.
Prevalence Prevalence = [TP(a) (48 + 12) / 20% 20% of the tested population
+ FN(c)] / 300 actually had TB.
[a + b + c + d]
 2. A new screening test for a certain disease was administered to 520 persons.
70 of them were known to have the disease. The test was positive in 55
among diseased and in 25 persons among non-diseased. Find out the
Sensitivity, Specificity, false positive, False negative, Positive Predictive
Value, Negative Predictive Value and the Prevalence of the disease.
3.
Disease Present Disease Absent Total
Test Positive 55 (TP = a) 25 (FP = b) 80
Test 15 (FN = c) 425 (TN = d) 440
Negative
Total 70 450 520 (Total tested, a + b + c + d)

Parameter Formula (with a, Substitution Result Interpretation

b, c, d)
Sensitivity Sensitivity = TP(a) 55 / (55 + 15) 55 / 70 = The test correctly detects
/ [TP(a) + FN(c)] 78.6% 78.6% of true cases (misses
21.4%).
Specificity Specificity = 425 / (425 + 425 / 450 The test correctly identifies
TN(d) / [TN(d) + 25) = 94.4% 94.4% of non-diseased
FP(b)] individuals.
False Positive FPR = FP(b) / 25 / (25 + 25 / 450 = 5.6% of healthy individuals
Rate (FPR) [FP(b) + TN(d)] 425) 5.6% are incorrectly labeled as
diseased.
False Negative FNR = FN(c) / 15 / (55 + 15) 15 / 70 = 21.4% of diseased
Rate (FNR) [TP(a) + FN(c)] 21.4% individuals are missed by
the test.
Positive PPV = TP(a) / 55 / (55 + 25) 55 / 80 = 68.8% of individuals who
Predictive Value [TP(a) + FP(b)] 68.75% test positive truly have the
(PPV) disease.
Negative NPV = TN(d) / 425 / (425 + 425 / 440 96.6% of individuals who
Predictive Value [TN(d) + FN(c)] 15) = 96.6% test negative truly do not
(NPV) have the disease.
Diagnostic DA = [TP(a) + (55 + 425) / 480 / 520 The test gives the correct
Accuracy (DA) TN(d)] / [a + b + c 520 = 92.3% result in 92.3% of
+ d] individuals tested.
Prevalence Prevalence = (55 + 15) / 70 / 520 = 13.5% of the tested
[TP(a) + FN(c)] / 520 13.5% population actually has the
[a + b + c + d] disease.
 3. A new glucometer-based Random Blood Sugar (RBS) test for Type 2 Diabetes was
evaluated in 480 individuals. Among them, 60 were known to have diabetes
(confirmed by the gold standard). The test was positive in 50 of the 60 individuals with
diabetes and also in 15 individuals without diabetes.
 Total tested = 480
 Diseased = 60 → Non-diseased = 480 − 60 = 420
 Test positive among diseased = 50 → TP(a) = 50 → FN(c) = 60 − 50 = 10
 Test positive among non-diseased = 15 → FP(b) = 15 → TN(d) = 420 − 15 = 405

Disease Present Disease Absent Total

Test Positive 50 (TP = a) 15 (FP = b) 65
Test 10 (FN = c) 405 (TN = d) 415
Negative
Total 60 420 480 (Total tested, a + b + c + d)

Parameter Formula (with a, Substitution Result Interpretation

b, c, d)
Sensitivity Sensitivity = TP(a) 50 / (50 + 10) 50 / 60 = The test correctly detects
/ [TP(a) + FN(c)] 83.3% 83.3% of true diabetes cases
(misses 16.7%).
Specificity Specificity = 405 / (405 + 405 / 420 The test correctly identifies
TN(d) / [TN(d) + 15) = 96.4% 96.4% of non-diabetic
FP(b)] individuals.
False Positive FPR = FP(b) / 15 / (15 + 15 / 420 3.6% of healthy individuals
Rate (FPR) [FP(b) + TN(d)] 405) = 3.6% are incorrectly labeled as
diabetic.
False Negative FNR = FN(c) / 10 / (50 + 10) 10 / 60 = 16.7% of diabetic
Rate (FNR) [TP(a) + FN(c)] 16.7% individuals are missed by
the test.
Positive PPV = TP(a) / 50 / (50 + 15) 50 / 65 = 76.9% of individuals who
Predictive Value [TP(a) + FP(b)] 76.9% test positive truly have
(PPV) diabetes.
Negative NPV = TN(d) / 405 / (405 + 405 / 415 97.6% of individuals who
Predictive Value [TN(d) + FN(c)] 10) = 97.6% test negative truly do not
(NPV) have diabetes.
Diagnostic DA = [TP(a) + (50 + 405) / 455 / 480 The test gives the correct
Accuracy (DA) TN(d)] / [a + b + c 480 = 94.8% result in 94.8% of
+ d] individuals tested.
Prevalence Prevalence = (50 + 10) / 60 / 480 12.5% of the tested
[TP(a) + FN(c)] / 480 = 12.5% population actually has
 [a + b + c + d] diabetes.

4 A screening test (PAP smear) for cervical cancer was conducted in 5000 women. Out
of these, 500 women had cervical cancer confirmed by the gold standard test. The PAP
smear was positive in 400 women with the disease and 150 women without the
disease. The test was negative in 100 women with the disease and 4350 women
without the disease.

Using this data, we can calculate sensitivity, specificity, false positive rate, positive predictive
value (PPV), negative predictive value (NPV), diagnostic accuracy, and prevalence.

Screening Test (PAP smear) Disease Present Disease Total

Absent
Positive 400 (TP = a) 150 (FP = b) 550
Negative 100 (FN = c) 4350 (TN = d) 4450
Total 500 (a + c) 4500 (b + d) 5000

Parameter Formula (with a, Substitution Result Interpretation

b, c, d)
Sensitivity Sensitivity = 400 / (400 + 400 / 500 The test correctly detects
TP(a) / [TP(a) + 100) = 80% 80% of women with
FN(c)] cervical cancer.
Specificity Specificity = 4350 / (4350 4350 / The test correctly identifies
TN(d) / [TN(d) + + 150) 4500 = 96.7% of women without
FP(b)] 96.7% cervical cancer.
False Positive FPR = FP(b) / 150 / (150 + 150 / 3.3% of healthy women are
Rate (FPR) [FP(b) + TN(d)] 4350) 4500 = incorrectly labeled as
3.3% positive.
Positive PPV = TP(a) / 400 / (400 + 400 / 550 72.7% of women with a
Predictive [TP(a) + FP(b)] 150) = 72.7% positive test actually have
Value (PPV) cervical cancer.
Negative NPV = TN(d) / 4350 / (4350 4350 / 97.8% of women with a
Predictive [TN(d) + FN(c)] + 100) 4450 = negative test truly do not
Value (NPV) 97.8% have cervical cancer.
Diagnostic DA = [TP(a) + (400 + 4350) / 4750 / The test gives the correct
Accuracy (DA) TN(d)] / [a + b + c 5000 5000 = result in 95% of women
+ d] 95% tested.
Prevalence Prevalence = (400 + 100) / 500 / 10% of the tested
[TP(a) + FN(c)] / 5000 5000 = population actually has
[a + b + c + d] 10% cervical cancer.
5. A new automated blood pressure monitor was evaluated in 600 adults. Among them, 120
adults had hypertension confirmed by a standard clinical measurement. The new test was
positive in 100 hypertensive adults and 50 adults without hypertension. The test was
negative in 20 hypertensive adults and 430 adults without hypertension.

Screening Test (BP Disease Present Disease Total

Monitor) Absent
Positive 100 (TP = a) 50 (FP = b) 150
Negative 20 (FN = c) 430 (TN = d) 450
Total 120 (a + c) 480 (b + d) 600

Parameter Formula (with Substitution Result Interpretation

a, b, c, d)
Sensitivity TP(a) / [TP(a) 100 / (100 + 100 / 120 Correctly detects 83.3% of
+ FN(c)] 20) = 83.3% hypertensive adults.
Specificity TN(d) / [TN(d) 430 / (430 + 430 / 480 Correctly identifies 89.6% of
+ FP(b)] 50) = 89.6% non-hypertensive adults.
False Positive FP(b) / [FP(b) 50 / (50 + 50 / 480 10.4% of healthy adults are
Rate (FPR) + TN(d)] 430) = 10.4% incorrectly labeled as
hypertensive.
False Negative FN(c) / [TP(a) 20 / (100 + 20 / 120 16.7% of hypertensive adults
Rate (FNR) + FN(c)] 20) = 16.7% are missed by the test.
Positive TP(a) / [TP(a) 100 / (100 + 100 / 150 66.7% of adults with a
Predictive Value + FP(b)] 50) = 66.7% positive test truly have
(PPV) hypertension.
Negative TN(d) / [TN(d) 430 / (430 + 430 / 450 95.6% of adults with a
Predictive Value + FN(c)] 20) = 95.6% negative test truly do not have
(NPV) hypertension.
Diagnostic [TP(a) + (100 + 430) / 530 / 600 The test gives the correct
Accuracy (DA) TN(d)] / [a + b 600 = 88.3% result in 88.3% of adults
+ c + d] tested.
Prevalence [TP(a) + (100 + 20) / 120 / 600 20% of the population tested
FN(c)] / [a + b 600 = 20% actually has hypertension.
+ c + d]
6.A rapid antigen test (RAT) was performed on 1000 individuals during a screening drive. 200
individuals were confirmed COVID-19 positive by RT-PCR (gold standard). The RAT was
positive in 160 of the 200 infected individuals and 50 of the non-infected individuals. The
test was negative in 40 infected individuals and 750 non-infected individuals.
Screening Test Disease Present Disease Total
(RAT) Absent
Positive 160 (TP = a) 50 (FP = b) 210
Negative 40 (FN = c) 750 (TN = d) 790
Total 200 (a + c) 800 (b + d) 1000

Parameter Formula (with Substitution Result Interpretation

a, b, c, d)
Sensitivity TP(a) / [TP(a) 160 / (160 + 160 / 200 Correctly detects 80% of
+ FN(c)] 40) = 80% COVID-19 cases.
Specificity TN(d) / [TN(d) 750 / (750 + 750 / 800 Correctly identifies 93.8% of
+ FP(b)] 50) = 93.8% non-infected individuals.
False Positive FP(b) / [FP(b) 50 / (50 + 50 / 800 6.25% of healthy individuals
Rate (FPR) + TN(d)] 750) = 6.25% are incorrectly labeled as
infected.
False Negative FN(c) / [TP(a) 40 / (160 + 40 / 200 20% of infected individuals
Rate (FNR) + FN(c)] 40) = 20% are missed by the test.
Positive TP(a) / [TP(a) 160 / (160 + 160 / 210 76.2% of individuals with a
Predictive Value + FP(b)] 50) = 76.2% positive test truly have
(PPV) COVID-19.
Negative TN(d) / [TN(d) 750 / (750 + 750 / 790 94.9% of individuals with a
Predictive Value + FN(c)] 40) = 94.9% negative test truly do not have
(NPV) COVID-19.
Diagnostic [TP(a) + (160 + 750) / 910 / The test gives the correct
Accuracy (DA) TN(d)] / [a + b 1000 1000 = result in 91% of individuals
+ c + d] 91% tested.
Prevalence [TP(a) + (160 + 40) / 200 / 20% of the tested population
FN(c)] / [a + b 1000 1000 = actually has COVID-19.
+ c + d] 20%