Lung Cancer and Smoking Case Study
Lung Cancer and Smoking Case Study
Here’s some information to assist you with questions on the last case study. Be sure to include all equations with your answers.
#4. Where else (besides a hospital) could cases be found? Where else might you find controls for this study (besides in a hospital)? Use Ch 6 to assist you in answering these.
#7. This question is asking if hospitalized patients (w/o lung cancer) are closely matched to the general population (w/o lung cancer). Explain your answer. (Hint: Look at the percentage you get for controls who smoked. Is this exposure level something you’d see [even back then] in the general population?)
#8. This question is mainly getting at this: Even if not as high as the cases (lung cancer pts), if the controls have a very high exposure level (smoking), then how might this affect the results for this case-control study (Odds Ratio)?
#9. Here’s the equation for proportion of cases who smoked: 1522/1530 x 100 = 99.48% Do the same for proportion of controls who smoked.
#11. The odds of smoking for the cases is: 1522/8 = 190.25:1 Do the same for the odds of smoking for the controls.
#12. Divide the odds to get the OR. You’re also asked to do the ‘cross-product ratio’. This is just another term for the equation you’ve used for the Odds Ratio from Ch. 6. Set up your 2×2 and see if you come up with the same OR from that equation. Post both equations and answers for this.
#14. You’ll end up with 4 Odds Ratios on this one. One for each dosage of cigarettes (1-14; 15-24; 25+; and ALL). These are NOT age categories; they are dosage (# of cigarettes smoked/day). Set up your 2×2 table for each category. Your ‘no exposure’ row (0 cigarettes) will be the same for all the 2×2 tables.
#15. Look at the 4 different ORs and note any differences by dosage. Is there a dose-response relationship between exposure and disease? #16. This question relates to things you learned in Ch. 10 that could affect the results of a study (like certain types of bias). Name some of the factors.
#18. The mortality rate in the cohort section is the ‘incidence rate’. The person-years is the population size. So, for the mortality rate for those who smoked 1-14 cigarettes (Table 3), you would take 23 / 38,600 x 1000. This gives you a mortality rate of 0.60/1000 person-years. #18. Rate Ratio is another term for Relative Risk. To find this, you would take the mortality rate among the exposed and divide by the mortality rate among the nonexposed. For the 1-14 cigarette group, this would be 0.60 divided by 0.07 (mortality rate among the nonexposed). #18. Rate Difference is the same as Risk Difference. For the 1-14 cigarette group, you would find this by subtracting 0.07 from 0.60 (incidence among exposed – incidence among the nonexposed).
#19. This proportion is called ‘attributable risk percent’ or the more recognized (from your book ~ Ch 9) term of ‘etiologic fraction’. Since there are two different equations you could use for EF, you would either use the mortality rate or the RR for ‘All Smokers’ from #18 in your equation.
#20. You’ll take the number of deaths among All Smokers (from Table 3) and multiply that by your answer for #19 (EF%). This is the number of deaths that could have been averted if no one had smoked.
#21. Since RR is the strongest measure of association, use those numbers to determine which disease has a stronger association with smoking. Use the AR% (EF) as your second reason. Explain your answer.
#22. Population attributable risk percent is the same as ‘population etiologic fraction’ (from Ch. 9). You would use the mortality rate for ‘All’ for these calculations (given in Table 4). There will be two equations for your answer (Lung Cancer & CVD) and don’t forget to answer the compare/differ questions, as well.
#23. Multiply your answers from #22 times the Mortality Rate for ‘ALL’ from Table 4. You’ll do this for both lung cancer and CVD. Your answers will be followed by “lung cancer (or CVD) deaths per 1000 person-years”.
#24. Use the RRs from Table 5 to help you in answering this question. Note the differences in RRs among Current Smokers, Ex-Smokers (by years since quitting), and Non-Smokers (never smoked). Then discuss what this implies for public health and preventive medicine. #26. Answer the first question as to which study design has the largest sample size, costs more, and takes longer to complete. The remaining factors are answered as advantage/disadvantage.