|
� |
Chapter 8 Exercises
08-exercises.htm 7/1/2010
8.1 Vitamins and
neural tube defects (p. 169). Source: Milunsky et al.,
1989.
8.2
8.3 Cytomegalovirus
(CMV) and restenosis (p. 170).
8.4 Cardiac arrest
during high-intensity activity (p. 170).
8.5 Hepatitis B and
liver cancer (fictitious data). A study of Hepatitis B infection and liver cancer found 65
cases of liver cancer in 65,000 person with Hepatitis B. In a Hepatitis-free
cohort, there are 5 cases in 215,000 individuals.
(A) Calculate the rate ratio of liver
cancer associated with Hepatitis B infection. Interpret this statistic.
(B) Calculate the RD per 100,000 person-years. Interpret the results.
(C) Which of the above statistics
quantifies the strength of the association? Which quantifies the excess
number of cases per unit person-time with exposure?
8.6 Stress
hormones and coronary disease. Psychosocial stressors stimulate the secretion of
catecholamines from the adrenal glands. As part of this stress response,
catecholamines stimulate the sympathetic nervous system, constrict arteries,
accelerates the heart rate, and increase cardiac ejection volumes. These may
have deleterious effect on heart health. In a study was conducted in Evans
County, Georgia in the early 1970s looked at whether high levels of endogenous
catecholamines increase the risk of coronary heart disease (CHD) (Cassel,
1971).� A cohort of 609 CHD-free
white males with high catecholamines levels experienced 76 coronary
incidents. In contrast, a cohort of 487 males with low catecholamines
levels experienced 44 incidents. Calculate the relative risk of coronary
disease associated with high catecholamine levels. Interpret this
finding.
8.7 Old
hat. John Snow (1855) compiled the
following information during his investigation of the
|
Water Source |
Cholera
deaths |
No. of
households |
|
Southwark & Vauxhall Co. |
1263 |
40,046 |
|
Lambeth |
98 |
26,107 |
(A) Calculate the cholera mortality rates per 1000
household by water source.
(B) Calculate the risk difference associated with the
Southward & Vauxhall Company water source.
(C) Interpret your findings.
8.8 Joseph Goldberger. In 1918, Joseph Goldberger (1874 - 1929) wrote "of the 127 households owning cows which supplied milk during the three months preceding the date of the canvass in the spring of 1916, two ... were pellagrous; whereas of 451 not owning such cows, forty-seven ... had one or more cases of pellagra." Let "not owning a cow" be the risk factor. Calculate R1, R0, RD, and RR.
8.9 Zocor. A randomized trial evaluated the
cholesterol-lowering drug Zocor on the risk of heart attack (4S
Study Group, 1994). The treatment group consisting of 2,221 subjects
experienced 111 fatal heart attacks. The placebo group of 2,223 individuals
experienced 189 such incidents.
(A) Arrange the data to form a 2-by-2 cross-tabulation of
counts.
(B) Calculate the RR.
(C) In relative terms, how much did the Zocor decrease the risk of a heart
attack?
8.10 Cell phone
use and auto accidents. One aspect of cell phone use that deserves our attention
is its potential to cause driving accidents. Dreyer (1999) estimated
fatal auto accidents rates of 5, 10, and 12 per 100,000 person-years for light,
moderate, and heavy cell phones use. Using the light users as the reference
category, calculate the effect of moderate- and heavy-use in absolute
terms.
8.11 Strong associations do not necessarily affect large numbers of
people. Review the data in Table 8.6 (p.
165) in the text and explain in plain terms why the absolute effect of smoking
on ischemic (coronary) heart disease is larger than the effect of smoking on
lung cancer occurrence even though smoking is associated with a modest relative
risk for heart disease and a large relative risk for lung cancer.
8.12 In plain English. How would you explain each of these epidemiologic measures to a lay audience? (A) risk (B) relative risk (C) risk difference (D) attributable fraction in the population
8.13 Injuries in farm workers [similar
to exercise 11.4]. McGwin et al. (2000) completed a cohort study in which
farmers were contacted biannually to determine whether an agriculture-related
injury had occurred. Results by ownership status and race were as follows:
|
Group (number) |
Events |
Person-years |
|
Caucasian Owners |
67 |
2047 |
|
Af-Am. Owners |
27 |
821 |
|
Af-Am. Workers � |
37 |
359 |
(A) Calculate the rates of injury per 1000 person-years
within each group.
(B) Calculate the RR of injury in the Af-Am owners compared to the Caucasian owners.
(C) Calculate the RR of injury in the Af-Am workers compared to the Caucasian owners.
(D) Is there an association between race and risk?� Is there an association between farm
ownership and risk?
8.14 Vitamins and neural tube defects, population
preventable fraction.
(A) Take the data from exercise 8.1 and put in into a 2-by-2 table.
(B) Assume data represent a random sample of population
pregnancies. Determine the proportion of neural tube defects in the population
that would be prevented with universal prenatal folic acid use. [Use either
formula 8.17 or formula 8.18.]
8.15 CMV and restenosis, attributable
fraction.
(A) Display the data from exercise 8.3 in the form of a
2-by-2 cross-tabulation.
(B)
Calculate the fraction of restenosis cases in CMV+ cases that is attributable
to the CMV infection. [This question addresses the fraction of exposed cases
due to infection.]
(C) Assume data are a random sample of patients
undergoing angioplasty. What fraction of the restenoses
in this patient population is attributable to CMV infection? [This
question is clear in referring to all patients, not just CMV+ patients.]
8.16. Why
is it incorrect to say that a risk ratio of 10 indicates a high risk of
disease?
8.17 Prevention
of back injuries at work? A cohort study was completed to determine the
effectiveness of back belts in preventing disabling low-back injuries at work
(Wassell� et al., JAMA. 2000 Dec 6;284(21):2727-32) From April 1996 through April 1998, the
investigators interviewed 2939 material handlers who wore a back belt every
work day. Within this group there were 502 reported back injuries.� They also interviewed 2601 material handling employees
who chose not to wear back belts. Of those employees who chose not to wear back
belts, 455 reported back injuries.
(A) Display the data in the form of a
2-by-2 cross-tabulation of counts.
|
|
Back Injury |
No Back Injury |
Total |
|
Used back belts |
|
|
___________��������������������� |
|
Did not use of back
belts |
|
|
___________ |
|
Total |
|
|
|
(B)
Calculate the incidence of back injury among the workers who used back belts.
(C)
Calculate the incidence among the workers who did not use back belts.
(D)
Calculate the relative risk of back injury associated with back belt use.
(E)
Interpret these results in terms a layman would understand.
8.18 Coronary disease interventions. An investigation pooled data from multiple clinical trails to study the effectiveness of two interventions for coronary disease. The article starts with this summary: “Coronary artery bypass graft (CABG) and percutaneous coronary intervention (PCI) are alternative treatments for multivessel coronary disease. Although the procedures have been compared in several randomised trials, the long-term effects on mortality in key clinical subgroups are uncertain.”
a) The following information on overall mortality is reported “…575 of the 3889 patients died in the CABG group compared with 628 of the 3923 patients in the PCI group.” Based on this information, calculate the risks within each group. In addition, calculate the risk difference associated with CABG, i.e., let CABG represent the exposed group. Show work.
b) Based on the data in part a of this question, what would you tell a layperson considering their treatment options for coronary disease? Limit your response to one or two clear, grammatical sentences.
c) This same study addressed whether effectiveness was modified by patient characteristics. One such patient characteristic was diabetes. “Of the 1233 [patients] with diabetes, 143 of the 615 patients assigned to CABG died, compared with 179 of the 618 patients assigned to PCI.” Calculate the risks and risk difference within diabetics.
d) Based on the data in part c of this question, what would you tell a diabetic considering their treatment options?
e) Within non-diabetics, the authors report: “… of the 6561 patients without diabetes, 432 of the 3263 [CABG] patients and 448 of the 3298 [PCI] patients died.” Calculate the risks and risk difference in non-diabetics.
f) Based on the data in part e of this question, what would you tell a non-diabetic considering their treatment options?
8.19 CVS and limb deformities. A cohort study was conducted to evaluate the risks of congenital limb deformities associated with a prenatal screening procedure known as chorionic villus sampling (CVS).� Twenty-nine (29) of the births in women receiving 39,421 CVS procedures had congenital limb deformities. Eighteen (18) congenital limb deformities occurred among the 123,489 births not exposed to CVS.
a) Display the data in 2-by-2 cross-tabulated form.�
|
|
Limb Defect |
No Limb Defect |
TOTAL |
|
CVS + |
|
|
_________ |
|
CVS – |
|
|
_________ |
|
TOTAL |
|
|
_________ |
b) Calculate the incidence proportion (average risk) in the CVS+ group.� Express the incidence in terms of “per 10,000 procedures.”
c) Calculate the incidence in the CVS− group. Again, express the risk “per 10,000 procedures.”
d) Calculate the risk ratio associated with CVS.�
e) Calculate the risk difference of limb defects associated with CVS.
f) Which of the above statistics informs us of the excess risk per 10,000 CVS procedure?