340.721 Epidemiologic Inference in Public Health I Exam with Complete Solutions

340.721 Epidemiologic Inference in
Public Health I Exam with Complete
Solutions


A. Validity and Predictive Value of a Screening Test for Heart Disease in School
Children

A school nurse examined a population of 2,000 school children in an attempt
to detect (preclinical) heart disease. Assume that the prevalence of heart
disease in this school- age population is known to be 10%. The sensitivity of
the nurse’s examination is 80% and specificity is 80%.


Fill in the 2x2 table based on the information provided above:

Heart Disease
Total
Yes No
P = 10%
Yes 160 360 520 10% of 2,000 = 200
Nurse's
Diagnosis
No
Sensitivity = 80% = 0.8
40 1440 148 Specificity = 80% = 0.8
0
Total
200
2000
1800


(HINT: Multiply the prevalence provided by the total population to determine
how many children have heart disease. Then use the sensitivity and
specificity to calculate how many children the nurse classifies as ‘positive’
and ‘negative’ for having heart disease.)




2

,Question 1
Based on the information provided in this assignment how many children
were labeled “positive” by the school nurse?

a. 360
b.
520
c. 1480
d. 1800



All school children who are labeled as “positive” (i.e., suspected of
having heart disease) by the school nurse are later sent for
examination by a physician. The sensitivity of the physician’s
examination is 90% and its specificity is also 90%.

Fill in the 2x2 table based on the information provided above.

Heart
Disease Yes Total
No


Physician's Yes 144 36 180
Diagnosis

No 16 324 340

Total 160 360 520


Question 2
Based on the information provided in this assignment, how many children
were examined by the physician (after they were examined by the school
nurse)? How many children were labeled “positive” by the physician?

a. 520 examined by the physician; 160 labeled positive by physician
b. 160 examined by the physician; 144 labeled positive by physician
c. 200 examined by the physician; 160 labeled positive by physician
d. 520 examined by the physician; 180 labeled positive by physician




3

, Divide the total number of children who had heart disease and were labeled as
“positive” by the physician by the total number of children with heart disease
who were screened and multiply your answer by 100%. This number is the
net sensitivity when two tests are administered one after the other (this
method of screening is referred to as ‘sequential’ screening in your
textbook.)
(HINT: It may help to copy the information from the 2x2 tables for Questions 1
and 2 to the 2x2 tables below.)


First, all children are screened by the nurse: Then, children who test positive on the
nurse’s
test are screened by the physician:

Heart Disease Heart Disease
Total Total
Yes No Yes No
Yes
Yes 160 360 520 144 36 180
Nurse's Physician's
Diagnosis Diagnosis 16 324
No 40 1440 148 No 340
0
160 360
Total 200
1800 2000 Total 520



Total number of children with heart Total number of children with heart
disease who disease who were screened are labeled “positive”
by the physician



Another way to calculate net sensitivity when tests are administered one
after the other is by using the following formula:

(𝑆𝑒𝑛𝑠𝑖𝑡𝑖𝑣𝑖𝑡𝑦 𝑜𝑓 𝑇𝑒𝑠𝑡 1) × (𝑆𝑒𝑛𝑠𝑖𝑡𝑖𝑣𝑖𝑡𝑦 𝑜𝑓 𝑇𝑒𝑠𝑡 2)
0.8 x 0.9 = 0.72


Question 3
What is the net sensitivity of the two tests when the all children are first
screened by the nurse and the children who test ‘positive’ are then screened
by the physician?
4