Chapter 6: Limits to Epidemiology PDF

Preview

Summary

Textbook Chapter 6 Notes...

Description

PROBLEMS AND LIMITS OF EPIDEMIOLOGY #1) Problems with Studying Humans ● Intervention study = it is impossible to control the behavior of human beings under such strict circumstances ● Cohort study = the people who voluntarily choose to do something are likely to be healthier/less healthy than their fellow participants already ● Case-control = people have a difficult time remembering specific details, or may be embarrassed sharing them; controls could differ from the cases in unexpected ways, leading to inconclusive results #2) Sources of Error A study might lead to a wrong conclusion if… ● The reported result is merely a random variation and the association is merely due to chance ● Confounding variables = factors associated with the exposure that may independently affect the risk of developing the disease ● Bias ○ Selection bias = when the control group is insufficiently similar to the treatment group, or there is a systematic difference between people who choose—or are chosen—to participate in a study and those who do not ○ Cohort studies typically occur over long periods of time → people may drop out or be untraceable when it comes time for reporting ○ Reporting/recall bias = a common problem in case-control studies, wherein the case group and the control group systematically report data differently, even if they had the same exposure; e.g. dietary intake reports are notoriously unreliable due to peoples’ tendency to over/underestimate their consumption levels #3) Proving Cause and Effect ● For the most part, epidemiologic studies cannot prove cause and effect ↴ ○ Why epidemiologists usually speak of “risk factors” instead of “causes” ● A study with a large number of subjects is more likely to yield a valid result than a small study ● The stronger the association measured between exposure and disease, the more likely there is a true cause-and-effect relationship ● Dose-response relationship = the relationship between the dose of some agent, or the extent of some exposure, and a physiological response; dose-response effect means that the effect increases with the dose ● Epidemiologic evidence is more convincing if there is a known biological explanation for an association between an exposure and a disease ● A good indication that an epidemiologic result is valid is that it is consistent with other investigations #4) Ethics in Epidemiology ● Most epidemiologic studies are observational and have little potential for harm ↴ ● However, exceptions do occur, especially in the conduct of intervention studies ● All institutions that receive federal funds must… ○ Inform every research subject of the purpose of the study, as well as its benefits and risks ○ Subjects must freely consent ○ Studies must be approved in advance by an institutional review board, who will approve a study if it… ■ Is well-designed ■ Benefits outweigh the risks

● ●

■ Subjects are truly given the opportunity for informed consent Even with today’s strict ethical guidelines, a number of controversial issues still come up Conflict of interest = a situation in which the concerns or aims of two different parties are incompatible

#5) Conflicts of Interest in Drug Trials ● Conflicts of interest often affect reported results ● Food and Drug Administration (FDA) reviews and approves/disapproves of all drugs; their stamp of approval typically means that things are safe to consume ↴ ● However… ○ Harmful side effects (to a drug) may be missed in a clinical trial because they are rare and/or the number of subjects studied may be too small for them to be noticed ↴ ○ Yet, a company may purposefully suppress negative info about a drug during the approval process ■ In fact, evidence now shows that companies may purposefully bias their studies in ways that make the drugs appear safer and more effective than they actually are ● Pharmaceutical industry uses many tricks to prejudice the conclusions of clinical trials, which include… ○ Testing a new drug in a clinical trial against a placebo. Why? Results may be misleading if older, well-established drugs are already in use for the same condition. The new drug will, inevitably, be more expensive than the old one, making it beneficial for the company, but not for the patients, who receive nothing besides another option to solve a problem. ○ Using financial incentives on physician-researchers to come up with results favorable to the companies ■ Industry-sponsored research was nearly four times as likely to be favorable towards the company’s product than research supported by a nonprofit organization ○ Publication bias = companies deliberately choosing to include only positive results in their published findings ■ There have been measures taken to try and prevent this, but a 2017 review of compliance found that most companies were not reporting when they needed to...