How Research Works

Dispelling Myths, Clarifying Concepts, Filling in Gaps in Knowledge…

This week on social media, I learned that the general public really has no idea how some aspects of research work. People either have the idea that research is only radical experimentation on human beings, or a solitary scientist (think someone like Robert Boyle or Isaac Newton) by themselves in a lab working on mice or a petri dish. Research is much broader than that.

I would encourage you to share this widely to help educate the public. I don’t make any money from this. On the contrary, I use valuable time that I should probably be using for my day job and/or side jobs… But it might help calm the world down if everyone understood these things better.

First… scientists rarely work alone.

And it isn’t just that scientists have helper minions. Collaborations are valuable for different expertise, access to different types of experimental methods and/or samples (more on this later). In other words, the idea of one person in charge with no influence from other scientists is quite literally dying.

Research relevant to the human condition takes many forms.

Yes, some studies are done experimentally with human beings. But those are the vast minority. Proof of concept, feasibility, preliminary results, all of that comes from a variety of different kinds of studies before anything is ever put in a human.

In vitro lab studies – Human cell lines (some taken from individuals that volunteer their blood, cell lines that can be purchased from a company) are incredibly valuable. If you want to know how a drug or other substance is going to work, you might test it “in a petri dish” (really in test tubes) using these cells. This approach is incredibly important for making sure something will actually work – lots of things fail at this stage.

Animal studies – before you put anything in a person, you test it in an animal. Mice, rabbits, guinea pigs, primates, etc. You might even obtain cells from animals prior to purchasing them or drawing blood from humans – that’s cheaper. Clearly, animals are not people, and there are some disorders / conditions that are not modeled well in animals. But a lot of approaches go through this level to eliminate things that don’t work.

And then there are studies of humans. But again, the vast majority of these are observational, not interventional. What does that mean? In observational studies, you literally just observe people. You don’t introduce anything – no drug, vaccine, food, nothing. You collect information about their clinical condition and various factors you are interested in. This topic is honestly the focus of multiple hour-long lectures when I teach, so a quick overview paragraph doesn’t really do it justice. But yes, you can learn a great deal from these sorts of studies (cross-sectional, case-control, cohort). Finding the BRCA1/BRCA2 genes that cause breast cancer? Observational, not interventional study. Identifying smoking as a cause of lung cancer? Same. To learn that smoking causes lung cancer and lung disease, no one was forced to smoke cigarettes. Information of how people were already living their lives was used. Electronic medical record analysis that I mentioned in a prior post – same idea. Gasp – the studies that show that autism is caused by genes in some cases – observational studies. And yes, if you design these studies correctly, you can in fact infer cause-effect.

And finally, there are interventional studies in humans. This is where the scientist actually introduces something (a drug, vaccine, educational intervention, dietary modification, etc) to see how disease occurrence changes. It would take a really long time to describe the different phases of clinical trials so I won’t do that, but there is a staging to these things – you aren’t going to launch a brand new drug in a population of thousands before a smaller study is done first for safety.

But the key thing that must be explained here is the concept of informed consent. In any human research study – even the observational ones described above – you must obtain informed consent. Informed consent describes the study, what is being studied, what the intervention is if relevant, and what the benefits and risks are. Every study has risks – even observational studies might have stress or time constraints that might be considered “risks”. If you have been in a research study and you did not sign an informed consent, that scientist was in violation of the law and that is a big problem. I’m dead serious – if you’ve been experimented on and you signed nothing, get a lawyer. But what that also means is that you are aware of risks when you participate in a research study.

Let’s say you are participating in a clinical trial to see if a new drug helps reduce high blood pressure. The new drug will be compared to something. In this case, it would probably be the typical usual blood pressure drug. When you agree to participate in this study, you accept the risk that you might get the “old” drug, and you might not see an improvement, or the new drug might give you a side effect. But the benefit might be that you get better, and that you’ve contributed to science. But what if the condition being studied doesn’t have an existing treatment? Then you get a placebo – a drug that looks just like the new drug, but does nothing. What if you are developing a vaccine against a condition that has no vaccine? Study participants might get a placebo.

Let’s take this a step further… if you want to evaluate whether a new drug has side effects, you have to compare it to a placebo.

I’ve just spent a lot of time describing studies that are relatively uncommon in the grand scheme of all the research that is done in the world. But if you learn nothing else from reading this, understand the essentiality of informed consent and placebos in clinical trials. Oh, and by the way, no one makes you participate in a study. If you don’t like research, don’t volunteer for a study. If you are worried about getting a placebo when you really want the new med / vaccine / treatment, don’t enroll in the clinical trial. If you don’t want to give up your privacy when someone asks you information about various risk factors, clinical conditions, etc, don’t volunteer for an observational study. These things are voluntary! And if do volunteer and they don’t give you a proper informed consent, run for the hills!

Lastly, do not confuse experimental treatment with research study. While experimental treatments may have a research element to them (and should involve some level of informed consent), not all research is an experimental treatment. (It’s like a square is a rectangle, but a rectangle is not a square.) And keep in mind that even when you get a standard treatment done, you are signing a consent. Read it carefully.

Please ask questions!