Question

13.5 A researcher wants to study the association between heavy coffee consumption and urinary bladder cancer. The variable X represents the variable "heavy coffee consumption." X = 1 if a subject drinks coffee heavily in the year 2001, and X = 0 if the subject is not a heavy coffee drinker in 2001. The variable Y represents the occurrence of urinary bladder cancer by the year 2006. Y = 1 if a subject has urinary bladder cancer by 2006, and Y = 0 is a subject does not have urinary bladder cancer by 2006. Suppose that Pr[Y = 1|X = 1] = 0.03 and Pr[Y = 1|X = 0] = 0.01125. Also, Pr[X=1|Y = 1]= 0.667and Pr[X=1|Y = 0]= 0.1816.

(a) The researcher performs a study in which she randomly chooses 100 subjects who are heavy coffee drinkers and 100 subjects who are not heavy coffee drinkers. She then follows these subjects for the next five years and determines whether or not they develop urinary bladder cancer. What type of epidemiologic study is this?

(c) The researcher classifies the subjects according to their coffee consumption and urinary bladder cancer status. Note that 6 of the 200 subjects died within the five years, and so they were left out of the table.

	Bladder cancer	No bladder cancer
Heavy coffee consumption	2	94
Normal coffee consumption	1	97

Compute the estimated relative risk of bladder cancer for the heavy coffee consumption group relative to the normal coffee consumption group.

(d) Of the 6 subjects who died, 4 were in the heavy coffee consumption group and 2 were in the normal coffee consumption group. Further, 2 of the 4 subjects in the heavy coffee consumption group died of urinary bladder cancer whereas neither of the 2 subjects in the normal coffee consumption group died of urinary bladder cancer. Using this additional information, repeat (c).

Answer 1

(a) this is case control  epidemiological study ( where sample population is divided into similar sizes exposed and unexposed and then cases and noncases are observed)

(b) cases control

exposed 2(a)     94(b)  

unexposed 1 (c) 97(d)

Relative Risk ,

  

= 2.04

(d)

cases control

exposed 2+2(a)     94+2(b)  

unexposed 1 (c) 97+2(d)

Relative Risk ,

  

=4