The effect of occupational chemical exposure

To examine the effect of occupational chemical exposure on the occurrence of bladder cancer, a study was done in 2015 using data from 1980 to 2014. In 1970, a total of 3080 exposed men and 4051 unexposed men were classified using type-of-job and union records (all the men were members of a particular union). All men were between 25 and 34 years old when the study began. No cases occurred before age 40. Data on smoking (a known strong independent risk factor for bladder cancer) showed that 1019 exposed and 1349 unexposed men were active smokers during the study period. The following 2×2 table summarizes the results:

a)  What kind of study was done here (be as precise as possible)?

b)  Calculate an appropriate relative measure of frequency. Calculate the AR%. If approximately 0.002% of men in the US are exposed to this occupational chemical, what is the PAR%? Briefly interpret all measures you calculate.

c)  Another study that looked specifically at the smoking –> bladder cancer association examined 400 men with bladder cancer and 990 age and race/ethnicity matched men without this condition. The findings of this study are summarized in the table below.

d)  What kind of study was done here?

e)  Calculate an appropriate relative measure of frequency. Assuming that, in the US adult male population, 30% are smokers, what is the PAR%? Compare these values to the occupational chemical exposure values above. Which is the stronger bladder carcinogen, smoking or chemical exposure?

f)  Race/ethnicity is known to be an independent risk factor for bladder cancer.  In this study there were about 50% whites, 30% Blacks, and 20% Asians.  Further, 15% of the study population self-identified as being of Hispanic ethnicity. Given this information, would you need to stratify the crude table shown above and, if so, into how many strata?

g)  In the occupational study, the researchers failed to take smoking into account as a potential confounding variable when they reported their results. Given the other information available from this problem, is this likely to be a major error here? Why or why not? Draw a simple directed acyclic graph that summarizes your answer.

h)  Despite every effort, there were men lost to follow-up.  What would need to be true for loss to follow-up bias to be a major issue?

i)  If in the general US male population, the risk of bladder cancer in men 40-65 is roughly 2.0 per 10,000 per year, is there anything striking about the findings of the occupational study displayed in the table (besides the relative measure of association) that should greatly concern you?