A Brief Guide to Understanding Medical Studies

Chances are, you routinely come across news stories reporting, often excitedly, on the latest medical studies. Unfortunately, not all of these reports accurately convey the results of a study, and many stories lend credibility to studies that lack true medical significance. This brief guide is designed to help you understand the three main types of medical studies you may encounter in news reports and in Worst Pills, Best Pills News articles.

Randomized, controlled trials (frequently reported as ’clinical trials’)

These studies are considered the “gold standard” in clinical medicine because they are designed to allow investigators to determine whether an intervention (for example, use of a drug or medical device) causes a health benefit or an adverse event. At the beginning of the study, subjects are randomly assigned to receive either an experimental intervention, such as a new drug being developed, or a control intervention, such as a placebo or an already-marketed drug. This random assignment process minimizes the chance that factors other than the intervention, known as “confounders,” will account for subjects’ outcomes at the end of the study and makes it possible to determine whether the study intervention might have caused the outcomes.

For example, investigators wanting to determine whether Drug X cures headaches would randomly assign half of the subjects to Drug X and half to a placebo pill that looks and tastes like Drug X but has no real effects on the body. If this randomization process is done correctly, subjects on Drug X and subjects on the placebo will be similar to each other in age, sex, socioeconomic status and many other characteristics that might otherwise have caused more or fewer headaches to occur in either group. Because these characteristics are systematically “evened out” before the subjects receive either Drug X or a placebo, it becomes less likely that they are responsible for the study outcome.

Randomized, controlled trials are the best way to glean cause-and-effect associations between an intervention and a health outcome. However, a trial must be adequately designed to detect all potential effects on subjects. Many trials, including most conducted by pharmaceutical companies to get drugs approved, are designed to demonstrate only that a product has a health benefit and are generally too small to detect potential adverse effects.

The most serious drug side effects are thus often detected many years after the drug is approved — on the basis of larger clinical trials, adverse event reports to the Food and Drug Administration, or observational studies. This fact is the basis for Public Citizen’s Seven-Year Rule, in which we recommend against the use of any new drug for the first seven years following its approval, unless it offers clear advantages over older therapies.

Observational studies

In observational studies, researchers analyze data on subjects in “real-world” (non-clinical trial) settings, such as hospitals, clinics or the wider community. Investigators do not provide any intervention to subjects in the study and instead rely on data from patients receiving usual care from their physicians. Information on patients’ medical care and health status is obtained from either one-on-one interviews or reviews of large databases containing patient health information. These data are then analyzed to investigate associations between various exposures and diseases or other health outcomes of interest. Perhaps the most well-known observational study is the decades-old Framingham Heart Study, which has been a critical source of information on risk factors for cardiovascular disease.[1]

Observational studies are often the only way to find out whether an exposure, such as to a toxin, is harmful to patients, because clinical trials intentionally exposing patients to potentially dangerous products, such as cigarettes, would be unethical. However, because patients are not evaluated in a controlled environment or randomly assigned to the exposure in question, it is often (though not always) difficult to rule out other causes of the patients’ health outcomes.

Case reports and case series

These are descriptions, usually by physicians, of one (case report) or more (case series) patients in their care who have experienced a disease or set of symptoms that is notable, either because the condition itself is new or rare or because it may be linked to a previously unknown cause, such as a drug. A famous example is the case series describing the first patients displaying symptoms of AIDS, now known to be caused by the human immunodeficiency virus,[2] in the early 1980s. Because they simply describe patients’ clinical courses and lack sufficient data for in-depth analyses, case reports and case series are regarded only as possible “early warning signals” of new diseases or side effects, not as conclusive sources of information in and of themselves.

Conclusion

Not all studies are equal. As a general rule, randomized controlled trials are more robust than either observational studies or case reports/series. However, this is not always the case. You should always discuss with your doctor any recent study you read about in a news story before reaching any conclusions on how it may affect your health or treatment decisions.

References

[1] Framingham Heart Study. About the Framingham Heart Study. https://www.framinghamheartstudy.org/about-fhs/index.php. Accessed July 15, 2014.

[2] Centers for Disease Control and Prevention. Pneumocystis Pneumonia — Los Angeles. MMWR Weekly. June 5, 1981:30(21);1-3. www.cdc.gov/mmwr/preview/mmwrhtml/june_5.htm. Accessed July 15, 2014.