Review Questions for Chapter 16

Last update: 5/7/2010

�16.1 Introduction

  1. Why is it necessary to base preventive measures on knowledge of causal mechanisms? 
  2. When is it necessary to go forward with preventive measures before knowledge of causal mechanisms are complete?
  3. Define causal inference.
  4. Is there such thing as ultimate proof in empirical sciences? Explain.
  5. True or false? Despite difficulties in interpretation, much of what we know about human health and disease comes from observational studies.
  6. Describe the two types of decisions in epidemiology. Identify how they differ. 
  7. Provide an example in which the discovery of an effective preventive measure predated the discovery of the causal mechanism.
  8. Describe The Problem of Induction.
  9. True or false? According to refutationist philosophy, the value of a scientific hypothesis depends on the degree to which it can be disproved.
  10. Why can you never prove that all swans are white?
  11. Describe the asymmetry of positive and negative proof. 

�16.2 Advisory Committee to the Surgeon General on Smoking and Health

  1. When was the Surgeon General's Report on Smoking and Health published?
  2. True of false? Association = causation
  3. True of false? Statistical methods alone cannot establish proof of a causal relation.
  4. List the five main criteria for causation presented in the The Surgeon General's Report on Smoking and Health. 

�16.3 Hill's Framework

  1. What is the most direct measure of the strength of an association?
  2. Why do strong associations provide better evidence than weak associations?
  3. Are weak associations indicative of a non-causal relationship?
  4. If multiple studies consistently show the same results, is this proof of causality?
  5. True of false? Establishing the proper temporal relation between cause and effect is the only mandatory causal element of Hill’s framework. 
  6. Describe plausible temporal relations between these factors: (A) deficient care-giving, (B) environmental lead exposure, (C) intellectual impairment in children (D) pica. 
  7. Provide an example of an inverted U-shaped dose-response relationship.
  8. How does the element of coherence differ from that of plausibility? [tough one]
  9. What types of experiments are considered by epidemiologists? 
  10. Early in the AIDS epidemic, before HIV was discovered (circa 1983), epidemiologists realized groups at high risk of HIV groups shard characteristics with groups at high risk of Hepatitis B. This suggested the diseases were spread by similar mechanisms. Which of Hills casual elements is being addressed by this argument. (Note: A competing hypothesis at the time was that HIV was caused by environmental toxins, such as those in amyl nitrate "poppers.") 
  11. A study on micro-nutrient and Alzheimer's disease found progressively lower risks of brain atrophy with increasing  levels of folic acid consumption (Snowdon et al., 2000). Which of Hill's causal criteria is addressed by this statement?
  12. List the short labels we use to discusses the elements of Hill's causal inference framework.
  13. The association between oral contraceptive use and clotting diseases has been the subject of considerable epidemiologic debate. Below are statements that bolster causal arguments on this subject. In each instance, select the single best criterion addressed by the statement.  
    Criterion    : (a) Strength (b) Consistency (c) Specificity (d) Temporality (e) Biological Gradient (f) Plausibility
      Statements: 
        (1) The rate of cardiovascular disease increases progressively with higher doses of estrogen in oral contraceptive formulations.
        (2) Lab studies show that oral contraceptive hormones increases the “stickiness” (“coagulability”) of blood .
        (3) Studies from the 1970s suggest early formulations triple the risk of cardiovascular death in women of childbearing age.
        (4) Nearly all studies show a positive association between oral contraceptive use and cardiovascular disease.
        (5) The risk of cardiovascular disease increases soon after oral contraceptive use begins.
  14. Match the element of the causal inference framework with its description. 
    Criterion: (a) strength (b) consistency (c) specificity (d) temporality (e) biological gradient (f) plausibility (g) coherence (h) experimentation (i) analogy
    Brief descriptions:
    ___ Requires that exposure precedes disease onset by a reasonable amount of time
    ___ Evidence from clinical trials and studies in the lab are used in support of the epidemiologic information.
    ___ Increases in the level, intensity, duration, or total amount of exposure leads to progressive increases in risk
    ___ Large RRs provide more firm evidence than small ones.
    ___ Diverse methods of study carried out in different populations under a variety of circumstances by different investigators provide similar conclusions
    ___ The cause leads to one specific disease mechanism, and that the mechanism disease come from this single cause
    ___ Evidence "sticks together" as a whole 
    ___ Similarity in some respect but not in others of an otherwise different phenomenon. 
    ___ Association is based on established biological fact and theory.

Key