Answer the following to that best of your knowledge and we will review in class.

An epidemiologist wants to study the causes of a rare cancer, which of the following study designs would it be best for him to use?

A. Retrospective Cohort

B. Prospective Cohort

C. Cross Sectional

D. Case-Control

An investigator is interested in understanding how roadside air pollution may cause COPD in older adults. She randomly samples older adults living in Los Angeles with residences near and far from major highways. She then follows the study participants for 15 years and records who develops COPD. What measure of effect should she use in her data analysis?

A. Odds Ratio

B. Relative Risk

C. Prevalence Ratio

D. Prevalence Ratio

The main characteristic that distinguishes a clinical trial from a prospective cohort study design is which of the following?

A. The follow-up time is short (ex. 2 years or less)

B. Some study participants serve as the comparison group

C. Study participants are randomized to receive or not receive the exposure

D. Study participants are followed prospectively

Which of the following can introduce selection bias into a prospective cohort study?

A. Restricting the target population to control for confounding

B. Not randomly sampling the target population

C. Having lost to follow-up in the study

D. All answers are correct

Which of the following 95% confidence intervals associated with a relative risk of 1.6 is most precise?

A. 95% CI [ 1.2, 2.0]

B. 95% CI [ 1.0, 2.2]

C. 5% CI [ 1.0, 2.2]

D. 95% CI [0.9, 2.3]

Which 95% confidence interval listed in Question 5 suggests the RR =1.6 is not statistically significant?

A. 95% CI [1.1, 2.1] and 95% CI [0.09, 2.3]

B. 95% CI [1.0, 2.2]

C. 95% CI [1.0, 2.2] and 95% CI [0.09, 2.3]

D. 95% CI[0.09, 2.3]

A 90% confidence interval is associated with which of the following p-values?

A. 0.10

B. p-values are not related to confidence intervals

C. 0.05

D. 0.90

With which of the following measures of effect can an epidemiologist not infer causality?

A. Odds Ratio

B. Prevalence Ratio

C. Relative Risk

D. Hazard Ratio

In which section of a journal article should a reader look for potential selection bias?

A. Background/Introduction section

B. Methods section-measurement of the exposure

C. Results section-tables

D. Methods section-measurement of disease

E. Methods section-study population

In which section of a journal article should a reader look for potential misclassification of the exposure?

A. Methods section-study population

B. Background/Introduction section

C. Methods section-measurement of disease

D. Methods section-measurement of the exposure

E. Results section-tables

In which section of a journal article should a reader look for potential misclassification of the disease?

A. Methods section-measurement of the exposure

B. Methods section-measurement of the disease

C. Methods section-study population

D. Background/Introduction section

E. Results section-tables

In which section of a journal article should a reader look for variable thought to be confounders by the authors?

A. Background/Introduction section

B. Results section-tables

C. Methods section-measurement of the exposure

D. Methods section-study population

E. Methods section-measurement of the disease

In which section of a journal article should a reader look for the authors’ hypothesis?

A. Methods section-measurement of the disease

B. Methods section-measurement of the exposure

C. Background/Introduction section

D. Methods section-study population

E. Results section – tables

An epidemiologist is interested to know the 10-year survival probability of children diagnosed with acute lymphocytic leukemia (ALL).  The life table below represents 100,000 children who were diagnosed with ALL within the past year.  Each interval in the life table represents the number of years followed.  Use this life table below to answer the following questions.

 

Life table of 100,000 children diagnosed with acute lymphocytic leukemia
Years of follow-up	Number at Risk	# of Deaths	# of Lost to Follow-up or Competing Causes (Censored)	Effective Number at Risk	Risk of Death	Survival Probability	Cumulative Survival Probability
0-5	100,000	5,000	7,000
6-10		3,000	4,400

What is the effective number of children at risk for the first interval (0-5 years of follow-up)?

What is the risk of death for the first interval?

What is the survival probability for the first interval?

What is the number of children at risk for the second interval?

What is the cumulative survival probability at the end of 10 years of follow-up?

Directions for the rest of the questions. please see the following 2X2 table.  All of your calculations can be done on a separate piece of paper and do not need to be shown for the exam. 

*For these questions, please only enter your final calculation into the blank. 

*Remember to use the correct rounding as your answers will be automatically graded and therefore not sensitive to discrepancies in rounding.

What is the relative risk?

	Lung Disease	No Lung Disease	Total
Any Smoking	155	2,495	2,650
No Smoking	2,008	48,355	50,363

What is the ln(RR)?

What is the standard error of the ln(RR)?

What is the lower 95% confidence interval?

What is the upper 95% confidence interval?

Smoking increases the risk of lung disease. T/F

The relative risk is statistically significant.T/F

If you had calculated a 90% confidence interval, it would be more precise.T/F