Clinical trials are structured experiments that test whether a new drug, device, or treatment is safe and effective in humans. They follow a step-by-step process, moving from small safety tests in a handful of volunteers to large-scale studies involving thousands of patients, before a treatment can reach the market. The entire journey from first human dose to approval typically takes years, and roughly 90% of drug candidates fail somewhere along the way.
The Four Main Phases
Clinical trials progress through distinct phases, each with a different goal and scale. A treatment must clear one phase before advancing to the next.
Phase 1 is the first time a drug is tested in people. Researchers recruit fewer than 50 healthy volunteers and give them small doses to find out how the body absorbs, processes, and eliminates the drug. The primary goal is establishing a safe dose range and identifying side effects, not treating a disease.
Phase 2 shifts focus to patients who actually have the condition the drug is meant to treat. Between 5 and 100 patients participate, and researchers test different dosages to see which one works best while continuing to monitor safety. This phase also looks at how the drug interacts with other medications. Between 2016 and 2020, only about 29 to 34% of drugs passed Phase 2, making it the biggest bottleneck in drug development.
Phase 3 is the large-scale confirmation study. Anywhere from 300 to 3,000 patients enroll across multiple hospitals and clinics. The drug is compared head-to-head against either a placebo or the current standard treatment. Researchers track effectiveness, side effects, and adverse reactions in detail. This is the data submitted to the FDA when seeking approval. Success rates here are considerably higher, around 70 to 73%.
Phase 4 happens after a drug is already on the market. These post-approval studies follow patients over long periods to catch rare side effects or drug interactions that shorter trials couldn’t detect.
How Bias Is Prevented
The gold standard trial design is called a randomized, double-blind, placebo-controlled study. Each of those words describes a specific layer of protection against bias.
Randomization means patients are assigned to either the treatment group or the control group by chance, like flipping a coin. This prevents researchers from consciously or unconsciously steering healthier patients into the treatment group. It also balances out factors nobody thought to measure, like genetics or lifestyle habits, so the two groups start on roughly equal footing.
Blinding (also called masking) means participants don’t know whether they’re receiving the real treatment or a placebo. In a double-blind study, the doctors administering the treatment don’t know either. This matters because expectations influence outcomes. A patient who knows they’re getting the real drug might report feeling better regardless, and a doctor who knows might interpret symptoms more favorably. Triple-blind studies go one step further, keeping even the team analyzing the data in the dark until the study ends.
The control group, often receiving a placebo or an existing treatment, provides the baseline. Without it, there’s no way to tell whether improvements came from the drug or simply from the passage of time, the placebo effect, or routine medical attention.
Who Watches Out for Participants
Before a clinical trial can begin, it must be reviewed and approved by an Institutional Review Board (IRB). An IRB is an independent committee that evaluates whether the study design is ethical, the risks are reasonable compared to potential benefits, and participants will be adequately protected. IRBs don’t just approve trials at the start. They review ongoing studies yearly and have the authority to require changes to a trial’s design or shut it down entirely if safety concerns emerge.
For large Phase 3 trials, an additional layer of oversight called a Data and Safety Monitoring Board (DSMB) reviews results while the trial is still running. DSMBs are independent groups of doctors, statisticians, and research experts who periodically examine incoming data. They can stop a trial early for several reasons: if participants are experiencing unexpected severe side effects, if the treatment is clearly working and it would be unethical to keep the control group on a placebo, or if the treatment is clearly not working and continuing would expose patients to harm with no benefit.
What Informed Consent Actually Covers
Federal regulations require that every participant receive specific information before enrolling. This includes a description of any foreseeable risks or discomforts, the potential benefits to the participant or to others, and a clear statement that participation is entirely voluntary. You can refuse to join without any penalty, and you can withdraw at any time without losing access to benefits or care you’d otherwise receive.
When relevant, the consent process must also address risks that are currently unforeseeable, since new drugs by definition have incomplete safety profiles. If you decide to leave a study partway through, the consent form should explain what that process looks like and any consequences of stopping treatment abruptly. Informed consent isn’t a single signature on a form. It’s an ongoing conversation, and you can ask questions at any point during the trial.
Who Pays for What
Costs in a clinical trial fall into two categories. Research costs, like the study drug itself, lab tests done purely for the trial, extra imaging scans, and additional doctor visits beyond what you’d normally have, are typically covered by the trial’s sponsor (usually a pharmaceutical company or research institution). You generally won’t be charged for these.
Patient care costs are the routine medical expenses you’d have whether or not you were in a trial: regular doctor visits, hospital stays, standard treatments, and imaging or lab work that’s part of normal care. These are usually billed to your health insurance. It’s worth asking your insurance plan directly whether it covers routine patient care costs for clinical trial participants, since policies vary.
Travel, lodging, meals, parking, and childcare are sometimes reimbursed by the trial, but not always. Ask the research team upfront whether the study offers financial assistance or can connect you with support organizations that help cover those expenses.
How to Find a Trial
ClinicalTrials.gov is the largest public database of clinical studies. You can search by condition, treatment, or keyword, and filter results by the trial’s current recruitment status, phase, and location. Each listing includes the study’s inclusion and exclusion criteria, which are the specific requirements you’d need to meet to enroll. These might specify an age range, a particular stage of disease, previous treatments you’ve had (or haven’t had), or certain lab values.
Some tools built on top of this database use a questionnaire approach: you answer a series of questions about your health, and the system narrows down trials you’re likely eligible for, rather than asking you to read through hundreds of criteria lists yourself. Your doctor or a patient advocacy organization for your condition can also help identify trials that match your situation.
Why Most Trials Don’t Lead to Approved Drugs
The overall failure rate in clinical development is around 90%. Phase 2 is where the majority of candidates stall, largely because a drug that appeared safe in a small group of healthy volunteers turns out to be ineffective, insufficiently effective, or unexpectedly harmful in patients with the actual disease. Phase 3 failures, while less common, are more costly since these trials involve thousands of patients across multiple sites over several years.
Failure doesn’t always mean the drug is dangerous. Sometimes the effect is real but too small to justify the side effects. Sometimes the trial was designed in a way that couldn’t detect a meaningful difference. And sometimes a competing treatment reaches the market first, making the research question irrelevant. Each failed trial still generates safety and efficacy data that informs future research.
Efforts to Make Trials More Representative
Historically, clinical trials have enrolled populations that don’t reflect the diversity of people who will eventually use the treatment. This matters because drugs can work differently across demographic groups due to differences in genetics, metabolism, body composition, and other factors. The FDA now requires sponsors of certain clinical studies to submit Diversity Action Plans describing how they will improve enrollment of underrepresented populations. These plans must outline the specific steps researchers will take, and sponsors can request a waiver only through a formal process with FDA review. The goal is to ensure that by the time a drug reaches the market, there’s evidence it works across the range of people who will actually take it.