Principles of Epidemiology

SECTION A
Q1. Read the article: Systematic identification and treatment of smokers byhospital based cessation practitioners in a secondarycare setting: cluster randomised controlled trial;published in the BMJ in July 2013.

Complete a CONSORT checklist for this study. Briefly highlight relevant aspects of the study that you felt did not fulfill specific criteria as required by the CONSORT process. Mention if the authors have justified any of these aspects which did not meet the requirements, anywhere in the manuscript. (15 marks)
A sample CONSORT checklist for cluster randomised trials has been provided. Additional explanatory information is available in the accompanying guidelines document.

SECTION B
Q2. A colleague wishes to perform a case control study based at their hospital (a large private hospital in Brisbane). They are caring for 25 patients with a relatively rare cancer (cancer X) at their hospital and wish to test the hypothesis that the risk of cancer X is reduced by exercise. Your colleague wishes to use the 25 patients with cancer X at their hospital as cases, and asks for your help in selecting controls for the case control study.
A. What is the main function of the controls in a case control study? (1 mark)

a. To measure the exposure in the population the cases came from
b. To measure any confounders in the population the cases came from
c. To measure the incidence of the outcome in the population the cases came from
d. To check the accuracy of the recall of the exposure by the cases

B. Why are case control studies particularly useful for investigating rare outcomes? (1 mark)

C. Your colleague suggests two methods to select controls for the study. For each method comment on the potential for selection bias, and which direction (away from or towards null) any selection bias might bias the results:
1. Select other patients in their hospital as controls (4 marks)
2. Select controls randomly from the population of Brisbane (4 marks)