**1. Succinctly state the research question addressed in the study by Fontham et al.?

  2. The authors refer to the existing body of data as suggesting that there is a "small but significant elevation in risk of lung cancer" (35, col 2). What do you think they mean by "small" and "significant" in this context?

  **3. For each of the "unresolved issues" listed in the introduction, indicate why that issue is important to resolve in order to answer the research question. Could any of these account for the finding of an association between ETS and lung cancer?

  4. What are advantages of the study's being population-based?

  **5. What is the key information that the control group is intended to provide?

  6. What are relative advantages and disadvantages of population controls versus hospital controls?

  7. What population group will be unavailable for inclusion in the population-based control group in this study. Is this unavailability likely to be an important source of bias?

  **8. Interviews were conducted with 84% (431 of 514) of eligible cases and 72-73% of controls. What concern does this raise about the validity of the overall results? In other words, if interviews had been conducted on 100 % of eligible cases, how different might the observed odds ratio have been from the one given in the journal article's abstract? Is it likely that the 16% non-interviewed cases and the 28% non-interviewed controls would produce serious distortion? What further data would you want to evaluate this distortion, rather than just speculate on it?

  9. Explain the Fontham et al. statement (page 36, column 2, lines 15-16) that colon cancer controls "provided an opportunity to examine the issue of recall bias associated with a recent diagnosis of cancer."

**10. Use the data on "Education" in Table 4 of the journal article to calculate the odds ratio for the association between less than high school education vs. (some college + college + graduate education) and lung cancer in nonsmoking women. Consider "exposed" as women with less than high school education; consider "nonexposed" as the combination of women with some college, college, and graduate education. Use the population controls as the control group.

1. What are the values of cells (A, B, C, and D) of the 2 by 2 table?


2. What is the value of the odds ratio?


3. Can you spot the incorrect number in Table 4?

11. Interpret this odds ratio in a sentence. Can this odds ratio be used as an estimate of the risk ratio or incidence density ratio?

12. The 95% confidence interval around the odds ratio calculated in 10b above is (1.67-3.06). What information does this confidence interval provide?

13. In Table 5 of the journal article, the adjusted odds ratio for lung carcinomas (outcome) and ETS (exposure) obtained with the population controls was 1.20 (0.93- 1.55). Changing only the outcome to adenocarcinoma of the lung yields an adjusted odds ratio of 1.36 (1.02 - 1.84).

a. Explain what is meant by the "adjusted" odds ratio.

b. How do you interpret the findings both in Table 5 and in Figure 1 of the journal article, in which adjusted odds ratios for adenocarcinomas of the lung are most often larger than those for all lung cancer?