Before a drug makes it to the FDA, the company has to show how it works in animals. Scientists run tests on different animals, and in the end bring that data to the FDA in the form of an Investigational New Drug application. If the FDA signs off, the company starts testing the drug in humans.
The FDA isn't gathering that data or running the trials. That's the responsibility of the drugmaker. The agency is just there to review it.
That's when phase one begins. During this trial, drugmakers administer the drug to a group of healthy humans who don't have the disease. They're just determining if the drug is safe. Researchers can see if the drug is toxic at higher levels and figure out if there are any major side effects that would become a problem in later trials.
Like phase one, the second trial tests whether a drug is safe, but this time the drug is given to people who have the disease, rather than healthy people.
This trial is "randomized." That means half of the participants receive the drug and the other half gets a sugar pill called a placebo. And while this test is mostly about safety, having the control group helps the researchers begin to pick up clues about whether the drug is working as designed.
It is at this stage that the drug-approval process could end if the Trump administration picks a commissioner in favor of approving drugs only after they've been proven safe and not necessarily effective.
Based on the current clinical-trial process, there's just a 30% chance of a drug moving from phase two to phase three, according to data from biotech trade organization BIO.
Now that a drugmaker is pretty sure a drug is safe, it moves to phase three. It's in this stage where researchers find out if the substance actually does what they hope it does. Working with the FDA, the drugmaker designs a larger trial that homes in on the ideal dosage, patient population, and other factors that could make or break approval.
Once all the data is collected and analyzed, the company can submit a new drug application. It includes data from the trials, the preclinical information, and details about how the treatment is manufactured. If the FDA accepts that application, the agency has 10 months and, in some priority cases, six months to make a decision.
During that review period, the FDA may hold an advisory committee meeting, where an independent group vets the drugmaker's data. From there, the group votes on whether the drug is safe, effective, and if the benefits outweigh the side effects. The FDA takes the committee's view into consideration but is not bound to follow its advice.
This is where the FDA explains why the drug didn't get approval. The drugmaker then decides whether to correct the mistakes or collect any missing information before reapplying.
Once a drug is approved, that's not the end of the FDA's responsibilities.
The FDA and the drugmaker still collect data on side effects. For example, in July 2015, the agency changed the label on a commonly used class of antibiotics to show potentially permanent side effects. The agency also has some oversight over how the drugs are marketed, making sure the ads don't mislead.
And should a drug start causing serious problems that didn't show up in the original trials, the FDA can take it off the market.