What exactly is a clinical trial? What the heck is a meta-analysis? Why are people talking about preprints—and do I really need to know?
You’re ahead of the game if you know any of these terms. But there is more to learn. Research is racing along at a dizzying pace as scientists seek to understand and identify solutions to the COVID-19 pandemic. Where knowing a few random terms might have been helpful before the pandemic, now that level of knowledge just skims the surface.
F. Perry Wilson, MD, a Yale nephrologist, teaches a free online course called, "Understanding Medical Research: Your Facebook Friend is Wrong" on the Coursera platform that has attracted 38,000 followers since it was launched in March.
The first thing people need to know, Wilson says, is that research results reported in the media have been interpreted by journalists (who may also cover politics, sports, and other topics)—even science journalists may be looking at the data through a particular filter. “And then that information is filtered through the lens of social media, with your friends, family, and coworkers offering their personal insights,” he says. “This is like a game of telephone, and a lot can change in the process.”
The rapidity at which new information is released can compound the confusion, says Jaimie Meyer, MD, MS, a Yale Medicine infectious diseases specialist. “People who don’t live in the world of scientific research need to know that normally the pace is extremely slow, and it’s slow, in part, because it’s careful,” she says. “But now data is coming in so quickly that you may be reacting to things that aren’t fully vetted, and you have to interpret everything in that light.”
Both doctors agree that it could be helpful for people to know more about medical studies, including the terminology researchers use—and it might even help to look up an actual study in some cases.
We put together a vocabulary to make it easier.
Launching a study
A medical study starts with a question, along with a carefully formatted plan to answer it. It also takes time, input, and, in most cases, money. Here are some basic terms to know.
Hypothesis: A question (and proposed solution/answer) that serves as a starting point for investigation. It could take years to test a hypothesis. Example: Will patients who take a new drug (tested in the study) have significantly lower blood pressure than those who don’t take it?
Placebo: A pill, liquid, or powder, for example, that looks like the drug being studied but is made of inactive ingredients. Placebos are used to ensure that participants can’t tell if they are receiving a medication, since that might affect how they report their symptoms.
Research grant or funding: Money to fund part or all of a study. Money can come from a variety of sources, including academia, the government (such as the National Institutes of Health [NIH]), a philanthropic source, a pharmaceutical company, and even crowdfunding.
“There's no doubt that different funders have different priorities and those priorities are not always exclusively in the best interest of people’s health and well-being,” says Dr. Wilson. For that reason, he advises looking at how a study was funded, what role the funder may have played in the study, and whether the funder contributed to decisions about whether to publish the results.
What you want to evaluate here, Dr. Meyer says, is whether the people analyzing, interpreting, and presenting the results have any real or perceived conflicts of interest that might affect how those tasks are done.
Types of studies
There are different approaches to research, and knowing what kind of study you are looking at can help you put its results into context. For example, studying a new treatment in a lab may provide important information about its potential, but there may still be questions about how that treatment will perform when used to treat people in a clinical setting. Here are some types of studies.
Basic science study (also called “bench research”): A study conducted in a lab that does not involve human subjects; it may involve cells or animals, for example.
Preclinical research: The trial process begins with what’s called “preclinical research.” This is when drugs or products are first tested on cell cultures and animals—not humans. If it passes the preclinical phase, it can move on to clinical trials.
Clinical trial (also known as an interventional study): A research study in which one or more human subjects are assigned to one or more interventions (these could include a treatment first studied in a laboratory). An intervention is a procedure or treatment such as a drug, nutritional supplement, vaccine, behavior (like exercise or diet), device, or anything else that is provided for clinical research purposes.
Phases of clinical trials: Studies that focus on drugs or biological products undergo phases based on definitions developed by the Food and Drug Administration (FDA):
- Phase I: An experimental treatment is given to a handful of patients. Researchers may assess how participants’ bodies react to the treatment, watch for side effects, determine a safe dose, and decide on the best way to give the medication (orally or by infusion).
- Phase II: The drug or treatment is tried on a larger group of patients (the number varies, but it can be up to several hundred). Researchers aim to determine the effect the treatment has on a disease and the rest of the body. Often, these studies are too small to tell if the treatment affects certain outcomes (like death), but they may show an improvement in laboratory values, symptoms, or other markers.
- Phase III: Larger groups (from hundreds to up to tens of thousands) are given the treatment, and researchers compare its effectiveness with placebo or standard treatments. Phase III trials are designed to test whether the treatment improves very important outcomes (like death, or remission of a disease), says Dr. Wilson. Once this is complete, a new drug application can be filed with the FDA, which also reviews all other available research. If the FDA approves it, doctors can use the treatment on patients. The standard approval procedure at the FDA requires two Phase III trials showing efficacy. You can get approval with just one trial if the treatment addresses a rare disease or if there are no existing treatments for the condition, adds Dr. Wilson.
- Phase IV: Once the product reaches the public, researchers look to see what happens to people who get the medication or intervention in the “real world.” This phase provides additional information about long-term risks and benefits, and the best way to prescribe the drug.
Observational study: Researchers observe a group of individuals and measure their outcomes. Participants are not given any special interventions.
Cohort study: A type of observational study defined by similarities in the group being studied (for example a study of older Americans living alone, or a study of smokers). Cohort studies follow that group to observe what outcomes happen to them.
Randomized controlled trial: A study that randomly assigns participants into different groups. It is known as “the gold standard” because it is a way to reduce bias by balancing certain characteristics participants might have (like gender, ethnicity, or income) between groups.
Systematic review: A review of all previous research designed to answer a specific research question.
Meta-analysis: Like a systematic review, a meta-analysis analyzes data from multiple studies that have been done on a particular subject to derive conclusions based on the whole body of research. But a meta-analysis combines the data from those prior studies to arrive at an “average” effect of the intervention.
Single- and double-blind studies: Two different approaches to testing a treatment or other intervention. Participants in a single-blind study don’t know which treatment, if any, they are getting. In a double-blind study, neither the subjects nor the researchers know which treatments participants are receiving.
Conducting a study
To conduct a medical study properly, researchers need to determine what kinds of trial participants to recruit. They also need someone to lead the study. Here are some basic terms.
Principal investigator (PI): The person who acquires funding for the study, designs the study, and carries out the plan for the research. He or she is responsible for all of the oversight of the study. He or she is usually the first or last author on the paper when it is published.
Healthy volunteer/patient volunteer: The former is a study participant volunteer who has no health problems. The latter has a health problem and participates to help researchers better understand, diagnose, treat, or cure that problem.
Eligibility criteria: Key requirements for clinical trial participants that could include such things as age, or health or gender status. A study may have inclusion criteria that a person must meet to participate, and exclusion criteria that bars them from joining the study.
Dr. Meyer notes that elderly people and women who are pregnant or of “childbearing potential” are commonly excluded from clinical trials, which can lead to potential questions about a drug's effectiveness in those populations. “If you fall into one of these groups, you may not know for sure whether a drug tested in a trial is safe or effective in someone like you,” she says.
Control Group: A group of clinical trial participants who are assigned a placebo or an intervention other than the primary one being tested (researchers may test against the existing standard of care or “treatment as usual”) so that researchers can compare the two.
Randomization: The process of assigning clinical trial participants to treatment or control groups randomly, using the element of chance to reduce bias. “Randomization is now usually done by a computer, but one can think of it as a “coin flip,” says Dr. Wilson.
Variable: An attribute that varies among study participants. In a study that involves people in general, participants may have variable attributes in such areas as income, gender, health status, weight, or whether or not the participants are smokers.
Adverse event: A negative change in the health of a study participant. This may or may not have been caused by the intervention given in the study.
Serious adverse events are pre-defined and, if they occur during the course of a study, an oversight committee can pause the study. “This is especially important to understand as multiple clinical trials of COVID-19 vaccines were paused at one point,” says Dr. Meyer, “but this is a normal part of carrying out clinical trials.”
Publication of the study
Usually, it can take a decade from the moment the researcher asks the initial question to the time a fully researched and vetted paper is published, and publication is by no means the final word on the topic. For instance, there may need to be multiple studies of a particular drug before the FDA will approve it and doctors will use it to treat patients. It’s important to keep in mind, especially as scientists work more quickly toward solutions for COVID-19, that some published research is released before it’s fully reviewed, in a form known as preprints.
Peer review: A review of a study by multiple experts who provide criticism and, in most cases, ask for revisions. Reviewers are often anonymous to the authors of the paper and are almost never paid for their services. The peer review cycle may be repeated multiple times.
Preprint: An early draft of a research study that has not yet gone through a formal peer-review process. The use of preprints has exploded during the pandemic, say the doctors. “You’re seeing things that are very raw. Often through the process of rejection and revision, these papers are substantially revised,” says Dr. Meyer.
While preprints can provide the research community with a wealth of data and opportunities for collaboration, lay readers need to be wary, says Dr. Meyer. Dr. Wilson agrees. “I have seen the media use these pre-peer-reviewed studies as the basis for stories. Often, the good stories will say that the study has not yet been peer-reviewed, but even that tends to get lost in the interpretation.”
Rapid review: An approach to reviewing a medical study that follows the principles of a formal peer review process, but that is completed more quickly. Because reviewers are asked to complete reviews on an expedited track, journal editors use this judiciously, applying it only to those papers that have urgent implications for basic science or clinical care.
Impact factor (IF): A number that reflects the yearly average number of citations that articles published in the previous two years in a given journal received. The impact factor is a way to show a journal’s importance, especially when compared to others in the same field. A higher IF generally indicates that a journal has greater importance “One thing people might not be aware of is that there are a ton of journals out there, and there is a huge difference between them in terms of quality, scope, and impact,” says Dr. Meyer. Some journals are not peer-reviewed, and some are frankly biased, she says. “The Journal of the American Medical Association [JAMA] and The Lancet are seen as very high-impact journals, and they will only accept manuscripts and data of the highest quality.”
Open access: An approach to publishing a study that makes it free to anyone to read. The downside is that journals that provide open access may charge researchers high fees to publish a study, says Dr. Wilson. “There is controversy as some of these journals may offer to publish a study without conducting adequate peer reviews—a practice called predatory publishing,” he says.
How to read the actual study
You can go beyond the headlines about new research and read the actual studies. Some journals may charge fees to read them, but there are a variety of ways to get around the paywalls. You can check your local library, which may subscribe to some medical journals. Another option is PubMed, a government database of citations and abstracts that often includes links to the full-text articles on the publishers' websites.(PubMedCentral is an archive for free full-text biomedical and life sciences journal articles.) Yet another approach is to email one of the study’s authors directly—the first or last author listed usually includes his or her email with the abstract, which is free to read.
Abstract: A brief overview of a published paper. “If you really want to understand a study, you should at least read the abstract,” says Dr. Meyer. “It’s a quick read, and it’s raw, not refined by other people’s viewpoints.”
Introduction: Essentially the background, including previous research and the rationale behind the decision to conduct the study. The hypothesis and objectives are usually included here.
Methods: Includes such essential details as the eligibility criteria for the participants (if it’s a clinical study) and how they were recruited, how the study was conducted, and what equipment, instruments, and procedures were used.
Results: Just the raw data, sometimes accompanied by a chart, graph, or table. If something changed during the time the study was being conducted—if some participants left, for instance—this section should provide an explanation.
Conclusion/Discussion: An interpretation of the results that may include opinions from the study’s authors. The strengths and limitations of the study may be included here, as well as recommendations for further research.
When one study contradicts another
Even if you understand how studies work, it can be confusing when one study contradicts another, or experts disagree. But that’s normal, according to the doctors. “Everything we do in science really needs to be interpreted in terms of all of the things coming out, not just the one sexy paper,” says Dr. Meyer. The picture can change long after a study is published, as more studies are completed and doctors put a new treatment that has been studied into practical use, she says.
But, if a piece of research does catch your attention—or you want to know more about vaccines and potential treatments for COVID-19, for example—Dr. Wilson says it still might be useful to do a deep dive and read a few studies.
“If you find something that could make a difference in your own life, I would encourage you to go to the primary source,” he says. “We need to embrace the scientific method now, more than ever. We've got a challenge in front of us that I think we have a good chance of beating, but we can only do that with the best quality science.”