Ref. Simple Biostatistics by Indrayan and Indrayan, 1/e p58

Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

Review of Preventive and Social Medicine

300
– Sensitivity = × 100 = 92%
300 + 25
d TN
• Specificity = 100
×= × 10
b+d TN + FP
75
– Specificity: × 100 =
43%
75 + 100
a TP
• Positive predictive value (PPV) = × 100
= × 100
a+b TP + FP
300
– PPV = × 100 =
75%
300 + 100
d TN
• Negative predictive value (NPV) = 100
×= × 100
c+d FN + TN
75
– NPV = × 100 =
75%
25 + 75
Therefore,
– Sensitivity > PPV OR NPV > Specificity
– PPV = NPV
24. Ans. (a) Prevalence  [Ref. Simple Biostatistics by Indrayan and Indrayan, 1/e p58]
• Baye’s Theorm: Gives relationship between PPV of a screening test and Sensitivity, Specificity and Prevalence
of disease in a population
Screening of Disease

Sensitivity × Pr evalence  × 100


PPV =
Sensitivity × Pr evalence  + ( 1 − Specificity ) ( 1 − Pr evalence ) 
• Actual Baye’s Theorm: Gives relationship between Post-test probability of a disease in a population (PTP = PPV) and
Sensitivity, Specificity and Post-test probability of a disease in a population (pTP = Prevalence)

Sensitivity × pTP
PTP = × 100
Sensitivity × pTP  + ( 1 − Specificity ) ( 1 − pTP ) 

– Post-test probability of a disease in a population (PTP) IS SAME AS PPV


– Pre-test probability of a disease in a population (pTP) IS SAME AS Prevalence

## Also Remember

• NPV is inversely proportional to Prevalence of disease in a population


Specificity × (1 − Pr evalence)
NPV = × 100
 Specificity × ( 1 − Pr evalence )  + ( 1 − Sensitivity ) × Pr evalence 

25. Ans. (c) 50%  [Ref. Simple Biostatistics by Indrayan and Indrayan, 1/e p58]
In the given question,
Sensitivity = 0.90 = 90%
Specificity = 0.90 = 90%
Prevalence = 10%
Thus,
90 × 10
PPV = × 100 =
50%(0.50) = 50% (0.50)
[90 × 10] + [(100 − 90)(100 − 10)]
• Alternate way of solving such questions: Construct a hypothetical table of screening test (FOLLOW RULES: Disease
on top of table, screening test results on left side of table). Always take round values (for e.g. 100, 1000, etc as total
population)

236

You might also like