In public health research, we often learn about an ‘association’ or ‘correlation’ between a risk factor/s and a health outcome, but to infer causality generally requires a review of the existing body of evidence and a critique of the study. Correlation is necessary, but not alone sufficient to make causal inferences with significant confidence. The Bradford-Hill Criteria is one tool that researchers use to evaluate an association/correlation to help determine if one can infer a causal relationship. Often when a new study is published, the media reports the findings without addressing the study’s results within the context of the existing body of data/evidence, and reports associations that may or may not be causal.
Prior to completing part two of this assignment, make sure to complete your weekly readings, as well as review the following article: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2706236/ .
1) Research the APU Library for a recent (within the last 5 years) epidemiological study. The study must be from a peer-reviewed journal, and use an epidemiologic study design of either a case-control or prospective or retrospective cohort design. Do NOT present a study that is a review. The study should report on an association between a particular risk factor and a health outcome in humans. You can also use Google Scholar or a web-search to locate a scholarly article, however you must locate the entire study article (not just an abstract). Attach the study article you located to your post as a PDF. If you need assistance, contact classroom support.
2) Very briefly summarize the following: study purpose, methods/study design, and findings. Explain the correlation / association reported. Include the Relative Risk / Odds Ratio reported.
Does the study have internal and/or external validity and why? Choose at least four (4) of the BH Criteria to discuss with respect to the correlation / association reported in your study.
This is the book used for the course. It is online
Epidemiology for Public Health Practice: By Robert H. Friis, Thomas Sellers
Copy and paste the following questions into a Word Document and include your answers beneath the question. Write concisely and completely. You may use outside sources but you must cite and reference any that you use. All formatting should follow APA format.
1) Define and give the formula for relative risk. How does it differ from an odds ratio?
2) A cohort study was conducted to study the association between coffee drinkers and anxiety in a population-based sample of adults. Among 20,000 coffee drinkers, 500 developed anxiety. Among the 50,000 non-coffee drinkers, 250 cases of anxiety were observed. Develop a table to determine the relative risk of anxiety associated with coffee use. What is the relative risk (RR)? What is the etiologic fraction? Show your work including your 2×2 table.
3) Compare and contrast absolute vs relative effects. What are the measurements for each and when is each used?
4) You are reviewing the results of a cohort study of vitamin D supplements and prostate cancer. The study had an alpha level of 0.05 and 95% confidence interval (CI). The study findings report a relative risk (RR) of 0.75, P value = 0.20, 95% CI (1.45 – .60).
Interpret these results: comment on the RR, P value, and CI.
5) Compare and contrast intervention studies (controlled clinical trials vs. community interventions). What are advantages and disadvantages of controlled experimental vs. quasi-experimental designs?
You can use a scholarly journal article but it cannot be a review study. Please attach the article and provide the reference in APA format