masthead
 

Chapter 13

Epidemiology in Clinical Settings
  1. Predictive value depends on disease prevalence, but sensitivity and specificity do not. What might cause the sensitivity and specificity of a test to vary from one population to another?

  2. Suppose that you wished to conduct a prospective cohort study to evaluate the benefits of prostate-specific antigen testing as a screening tool for prostate cancer. What outcome would most interest you? What biases would affect the study results? Would these biases also affect the results of a randomized trial?

  3. Because everyone eventually dies, why would we not say that the case fatality rate among patients with any disease is 100%?

  4. Under what conditions might one find that the baseline difference in a variable in a clinical trial is “statistically significant” but, nevertheless, not confounding? Under what conditions might we find that the baseline difference is not “statistically significant” but, nevertheless, is confounding?

  5. The Alzheimer’s disease cooperative trial manifested confounding by MMSE score. If the trial were repeated, would you expect that this same risk factor would be confounding again?

  6. Equipoise is a state of genuine uncertainty as to which of two treatments is better. Ethicists consider equipoise to be an ethical requirement for conducting a randomized therapeutic trial: if the researcher is already of the view that one treatment is better than the other, it would be unethical for that researcher to assign patients to the treatment that he or she believes is inferior. Under what conditions can equipoise be achieved in a placebo controlled trial?


Website Terms and Conditions and Privacy Policy
Please send comments or suggestions about this Website to custserv.us@oup.com        
cover