Why Most Drugs That Enter Clinical Trials Never Reach Patients

A numbers game from the start

Before a new medicine ever reaches a human trial, it has already survived years of laboratory work and animal testing. Even so, most compounds that begin clinical trials in people will fail somewhere along the way. This is not usually a sign that the science was wrong at the outset. It reflects the fact that trials are deliberately built as a series of filters, each one testing something different, and each one capable of stopping a drug in its tracks for a different reason.

Understanding where and why drugs fail tells you far more about the trial system than simply listing what each phase is called.

Phase 1: the safety filter

The earliest human trials involve a small number of healthy volunteers or, for some conditions such as cancer, patients who have exhausted other options. The purpose here is not to find out whether the drug works. It is to find out whether the human body tolerates it, at what dose, and with what side effects. Researchers are watching how the drug is absorbed, broken down and cleared from the body.

Drugs can fail at this stage for reasons that have nothing to do with their eventual usefulness. A compound might be metabolised too quickly to have any effect, or too slowly, building up to toxic levels. Unexpected side effects, even mild ones, can be enough to end development if they suggest a risk that would outweigh any benefit.

Phase 2: the graveyard of good ideas

This is where the largest proportion of drugs fail, and it is worth pausing on why. Phase 2 trials involve a larger group of patients who actually have the condition the drug is meant to treat. For the first time, researchers are testing whether the drug does what it is supposed to do, not just whether it is safe.

Many drugs fail here simply because the biological theory behind them, however plausible in a laboratory dish or an animal model, does not translate into a measurable benefit in real patients. Human disease is often more complex than the models used to study it. A drug that shrinks tumours in mice, or that alters a biological marker in a small early study, may show no meaningful clinical benefit once tested properly in the people it is meant to help. This gap between laboratory promise and clinical reality is the single biggest reason drug development is slow and expensive.

Phase 3: proving it at scale, and finding rare problems

Drugs that clear Phase 2 move into much larger trials, often across multiple hospitals and sometimes multiple countries, comparing the new treatment against an existing standard treatment or a placebo. The scale matters. Some side effects are rare enough that they simply do not show up when only a few hundred people have taken a drug. Only when thousands of people are involved does a pattern become statistically visible.

Phase 3 also tests something Phase 2 often cannot: whether the benefit holds up across a realistically varied population, including people with other health conditions, different ages, and different backgrounds, rather than the narrower group typically recruited for earlier studies. A drug that looked effective in a carefully selected Phase 2 group can lose its apparent benefit, or reveal new risks, once tested this broadly.

Why failure at this stage is expensive and painful

Because Phase 3 trials are large, long and costly, a failure here represents a much bigger loss of time and money than an earlier failure. It is also the stage where the human cost of uncertainty is most visible, since large numbers of patients have been enrolled in good faith, some receiving the experimental treatment and some the comparator, without knowing in advance which will prove better.

This is why trial design includes safeguards such as randomisation, blinding, and independent monitoring committees that can review data as it accumulates and recommend stopping a trial early, either because the drug is clearly harmful or, less often, because the benefit is already overwhelming and it would be unethical to keep giving some patients the older treatment.

Regulatory approval is not the finish line

A drug that succeeds through Phase 3 still has to be assessed by a regulator, in the UK the Medicines and Healthcare products Regulatory Agency, before it can be prescribed. The regulator examines the full trial dataset independently, rather than simply accepting the manufacturer’s conclusions.

Even after approval, monitoring continues. Post-authorisation studies, sometimes called Phase 4, track the drug once it is being used by much larger and more varied populations than any trial could realistically include. Rare side effects, drug interactions and long-term risks can still emerge here, which is why reporting systems for adverse effects remain active for as long as a medicine is on the market.

What this means for how you read health news

When a new treatment is described as promising after early results, it is worth remembering how many stages remain, and how many treatments that looked promising at Phase 1 or Phase 2 never make it further. The trial system is not designed to be pessimistic. It is designed to make sure that by the time a drug reaches a pharmacy, its benefits and risks have been tested against reality rather than theory, at a scale large enough to be trusted.

Anyone wanting to check the current status of a specific trial, or understand the rules governing research involving people in the UK, can consult the Health Research Authority and the MHRA directly, both of which publish guidance and trial registers for public reference.