Reference no: EM132636392
Question 1:
Two radiologists were assigned to independently review the results of mammograms on a group of women attending a public health clinic for indications of breast cancer. The results were as follows:
|
Radiologist B
|
Radiologist A
|
|
|
Positive
|
Negative
|
Total
|
|
Positive
|
16 (a)
|
14 (b)
|
30
|
|
Negative
|
21(c)
|
95 (d)
|
116
|
|
Total
|
37
|
109
|
146
|
a) What percentage of the time did the radiologists agree?
b) Calculate kappa.
Note: Gordis provides a detailed explanation of calculating the percent agreement expected due to
chance (pa). I recommend that you read this to understand the derivation of p„ however, a more functional formula is:
pe = [(a + b)(a + c.) + (c + d)(b + d)]/((a + b + c + d)2)
c) Explain why the result in part (b) differs from the result in part (a)?
Question 2:
You wish to compare two different screening strategies for fetal Down syndrome:
Strategy 1: The Quad test uses maternal blood tests to provide a composite risk score. At a standard cut-off value, this test was found to have a sensitivity of 0.810 and specificity of 0.932.
Strategy 2: Ultrasound measurement of fetal nuchal fold thickness and proximal long bone length results in a test with a sensitivity of 0.900 and specificity of 0.969,
Strategy 2 is clearly more sensitive and specific than Strategy 1, but is more burdensome to the mother and involves the additional expense of ultrasound testing.
a) We have a hypothetical population of 100,000 pregnant women in whom the actual prevalence of fetal Down syndrome is 1/200. The 2x2 table below is for Strategy 1, using the sensitivities and specificities above Create an equivalent 2x2 table for Strategy 2.
Hint: The description of setting up Table 5-8 in the "Relationship Between Positive Predictive Value and Disease Prevalence" section of your textbook may be helpful.
Strategy 1
|
Test result
|
Down syndrome
|
|
Yes
|
No
|
|
Yes
|
405
|
6,766
|
|
No
|
95
|
92,734
|
b) Use the tables in part (a) to answer the following:
i. How many additional cases of fetal Down syndrome are detected by strategy 2 than strategy 1?
ii. Calculate the positive predictive value (PPV) for each strategy. In 2-3 sentences, interpret and compare your results. What do these numbers mean in the context of our scenario?
iii. This scenario might be a good opportunity for sequential screening. Describe how you would set up a sequential screening program using these two strategies. Which screening test would you use first? Who would receive the second screening test? What do you see as the benefits or limitations of using this sequential screening?
4-6 sentences should be adequate to fully describe your screening program, but feel free to write more if you wish.
Question 3:
The following information for a recently-approved serology-based test for COVID-19 comes from the Johns Hopkins Center for Health Security
Answer the following questions using the data in the table above:
1. How many patients tested positive and tested negative using this test in the hospital trial?
2. Communicating to the public about the validity of screening tests is critically important. The website includes the following note on sensitivity and specificity. In 3-5 sentences, do you think that this is adequate to explain these measures to the public? If yes, why yes? If no, what more is needed?
Note on sensitivity and specificity data
Here we list the manufacturer reported sensitivity and specificity data, where available. A highly sensitive test should capture all true positive results. A highly specific test should rule out all true negative results. These measures are not independently validated by the Johns Hopkins Center for Health Security. If a sensitivity or specificity is not listed, it was not listed/available at the time of posting. When available, the number of samples used for sensitivity/specificity definitions are listed in the product description.
Attachment:- Epidemiology questions.rar