Assignment 1 Epib 611

ASSIGNMENT 1 – EPIB 611

This assignment is to be handed in on Sept 22, 11:00 a.m.

Question 1

Classify the following variables into one or more of the following categories: continuous, discrete, dichotomous, polychotomous, nominal, ordinal, ratio, interval and pseudo-continuous.

1A) What is your age? a) 0-9

b) 10-19

c) 20-29

d) 30-49

e) 50+

1B) What was your declared income in the last year? ______

1C) Patient temperature taken at 7 AM:______

1D) Smoking status a) current smoker

b) has never smoked

c) has smoked in the past, but is not smoking now

1E) Drug user a) yes

b) no

1F) Severity of functional decline 3 – severe

2 – moderate

1 – mild

0 – absent

1G) In the previous week how many times have you injected drugs with a needle that had been used by others? ______

1H) During the past month have you been waking up fresh and rested?

a) every day

b) almost every day

c) fairly often

d) less than half a time

e) rarely

f) none of the time

1I) How do you feel right now?

Relaxed Stressed

1J) During the past month how depressed or cheerful have you been?

0 1 2 3 4 5 6 7 8 9 10

______

Very depressed Very cheerful

Question 2

You are co-investigator in a study on the predictors of drug use among high-school students that are using intravenous drugs. One of your colleagues proposes the following three ways of measuring attitudes toward needle sharing, and he asks for your comments. Below are presented the three scales proposed by your colleague.

Scale 1

Last time when you were asked to share a needle, how anxious or relaxed did you feel?

Very anxious Completely relaxed

Scale 2

Sharing needles is cool (circle one answer).

1 2 3 4 5

strongly moderately no moderately strongly

agree agree opinion disagree disagree

Scale 3

Sharing needles is:

A sure way to get infected

Very risky

Risky

Depends on your luck

Quite safe

Very safe

Completely safe

2A) Describe the type of scale for each of the three scales presented by your colleague

2B) Describe two advantages of scale 2 when compared to scale 1

2C) What would be a potential source of bias in scale 3?

Question 3

Diagnostic test A (DT-A) is intended to be used in clinical practice as confirmatory test for disease D. In order to assess its performance DT-A is tested on 83 subjects out of which 46 are shown to be diseased and 37 are disease-free with the currently best available diagnostic test B (DT-B). DT-A values range from 1 to 25. Lower values of the test indicate greater likelihood of disease.

The distribution of the study population is presented below:

3A) Compute the sensitivity and specificity of DT-A for the following cut-points: 10, 12, 14, 16 (include the cut- point in the diseased category). Consider DT-B as gold standard.

3B) Draw the ROC curve

3C) Interpret the ROC curve (Tip: comment on the area under the ROC curve)

3D) Name and briefly describe 3 potential sources of bias in these results?

3E) Assuming that disease D is a cancer with a prevalence estimated at 1 per 1,000 population, an appropriate cut-point of the test should be selected for use in clinical practice. Two strategies are suggested below.

a)carry out each strategy and comment on the advantages and disadvantages of each.

• Strategy A: Minimize the number of false test results (Tip: use the table below)

• Strategy B: Maximize the sensitivity and specificity on the ROC curve by choosing the point on the ROC curve that is the farthest away from the diagonal

b)What other considerations should be taken into account in selecting a cut-point?

3G) DT-A test results can be also be expressed using likelihood ratios corresponding to different cut-off points. Compute the likelihood ratio for a cut-point of 12. Assume that the test is going to be used in a population where the prevalence of the disease is 0.005. Compute the new +PV by using the +LR.

3H) The investigators have concluded that DT-A is inferior in performance to DT-B since it has lower sensitivity and specificity. Do you agree with this conclusion? Justify your answer.

3I) If DT-B is the only test available at the moment of the study to diagnose disease D, propose different approaches to test its validity, other than comparing it to a gold standard as there is no one.

Question 4

The following ROC curves are the result of a study that aimed to compare different screening tests for anxiety in cancer patients. The test results were validated against a gold standard psychiatric diagnosis of anxiety disorder.

The ROC curves illustrate the performance of the following self-report questionnaires: State scale of the State-Trait Anxiety Instrument (STAI), Trait scale of STAI, QLQ-30 (Emotional funtional scale), Cancer concerns check-list, and Anxiety scale for hospital anxiety and depression scale (HAD-A).

a)Compare the performance of the Trait scale of STAI with HAD-A.

b)Which one of the five screening testspresented is the worst? Justify your answer.

1