How Bioequivalence Studies Are Conducted: Step-by-Step Process

When a generic drug hits the pharmacy shelf, you might assume it’s just a cheaper copy of the brand-name version. But behind that simple label is a rigorous scientific process designed to prove it works exactly the same way in your body. This process is called a bioequivalence study. It’s not guesswork. It’s not theory. It’s a tightly controlled clinical experiment that measures how your body absorbs and uses the drug-down to the last nanogram per milliliter of blood.

Why Bioequivalence Studies Exist

Before the 1980s, companies had to run full clinical trials to prove a generic drug was safe and effective. That meant millions of dollars and years of time. The U.S. Hatch-Waxman Act of 1984 changed everything. It created a shortcut: if you could prove your generic drug behaved the same in the body as the original, you didn’t need to repeat expensive clinical trials. The same rule applies across the European Union, Japan, Canada, and most other countries. The goal? Safe, affordable access to medicines. The FDA estimates generic drugs saved the U.S. healthcare system over $1.6 trillion between 2010 and 2019. But that savings only works if the generics are truly equivalent. That’s where bioequivalence studies come in.

The Core Principle: Same Amount, Same Speed

A generic drug doesn’t need to look like the brand-name version. It can have different fillers, colors, or shapes. But it must deliver the same active ingredient into your bloodstream at the same rate and to the same extent. That’s the definition of bioequivalence. The two key numbers that matter are:

• Cmax: The highest concentration of the drug in your blood after taking it. This tells you how fast the drug gets absorbed.
• AUC (Area Under the Curve): The total amount of drug your body is exposed to over time. This tells you how much of the drug gets absorbed overall.

If these two values for the generic drug fall within 80% to 125% of the brand-name drug’s values, regulators consider them bioequivalent. For drugs with a narrow therapeutic index-where small differences can cause serious side effects-this range tightens to 90% to 111%.

The Standard Study Design: Crossover Trials

Most bioequivalence studies use a crossover design. That means each volunteer takes both the generic drug and the brand-name drug, but in a random order. One group gets the generic first, then the brand after a break. The other group gets the brand first, then the generic. This design cuts down on individual differences. If one person naturally absorbs drugs slowly, they’ll show that pattern for both versions, making the comparison fairer.

Typically, 24 to 32 healthy adults participate. Sometimes more-up to 100-if the drug behaves very differently from person to person. These volunteers are carefully screened. No smokers. No major health issues. No medications that could interfere. They’re kept in a controlled clinical unit for the duration of the study.

The Washout Period: Letting the Drug Clear Out

After taking the first drug, volunteers don’t just jump right into the second. They wait. This is called the washout period. It’s usually five times the drug’s elimination half-life-the time it takes for half the drug to leave the body. For a drug that clears in 8 hours, that’s 40 hours. For a slow-clearing drug like some antidepressants or anticonvulsants, it can be weeks. Get this wrong, and leftover drug from the first dose skews the results. One contract research organization reported a study costing $250,000 and three months of delay because the washout period was too short for a drug with a 72-hour half-life.

How Blood Samples Are Taken

Before the drug is given, a baseline blood sample is drawn. Then, after dosing, blood is taken repeatedly-usually at least seven times. The schedule isn’t random. It’s designed to capture the full absorption and elimination curve:

• Just before dosing (time zero)
• One sample just before the expected peak concentration (Cmax)
• Two samples around the peak
• Three or more samples during the elimination phase

Sampling continues until the area under the curve (AUC) reaches at least 80% of the total possible exposure (AUC∞). For many drugs, that means collecting samples for 3 to 5 half-lives. Blood is spun down to get plasma or serum, then frozen until analysis.

The Lab Work: LC-MS/MS and Precision

Measuring drug levels in blood isn’t like testing sugar in urine. The amounts are tiny-often nanograms per milliliter. That’s why labs use liquid chromatography-tandem mass spectrometry (LC-MS/MS). This method can detect a single molecule of a drug among millions of others in blood. The method must be validated to strict standards: accuracy within ±15% (±20% at the lowest detectable level), precision under 15% variability. If the lab’s method isn’t solid, the whole study fails. BioAgilytix’s 2023 white paper found that 22% of bioequivalence studies face delays due to analytical method problems, costing an average of $187,000 per delay.

Statistical Analysis: The 90% Confidence Interval

The raw data-Cmax and AUC values for each volunteer and each drug-is log-transformed. Why? Because drug concentrations in blood don’t follow a normal distribution; they’re skewed. Log transformation makes the math work.

Then, a statistical model called ANOVA is used. It accounts for differences between sequences (who took which drug first), periods (first vs. second dosing), and individual variability. The key output? The 90% confidence interval for the ratio of the geometric mean of the generic to the brand-name drug.

If that interval falls entirely between 80% and 125% for both Cmax and AUC, the study passes. No exceptions. No rounding. No “close enough.”

When the Standard Doesn’t Work

Not all drugs fit neatly into this model. Some are highly variable-meaning different people absorb them very differently. For these, regulators allow more complex designs:

• Replicate crossover designs: Each subject takes the drug multiple times (e.g., test-reference-reference-test). This gives more data on variability.
• Reference-scaled average bioequivalence: Used by the FDA for drugs with high variability. The acceptance range expands slightly based on how variable the brand-name drug is.
• Parallel studies: Two separate groups, one gets the generic, the other the brand. Used only for drugs with half-lives longer than two weeks.
• Pharmacodynamic or clinical endpoint studies: For topical creams or inhalers, where measuring blood levels doesn’t reflect what’s happening at the site of action. Instead, researchers measure skin redness, lung function, or other direct effects.

The FDA’s 2020 guidance on topical products, for example, now requires clinical endpoint data-not just blood levels-to approve generic steroid creams.

What Happens Before the Study Even Starts

A bioequivalence study doesn’t begin in the clinic. It begins months earlier with:

• Reference product selection: Only one batch of the brand-name drug is used. Regulators require it to be from an intermediate dissolution profile-meaning not the fastest or slowest dissolving batch, but the middle one.
• Test product preparation: The generic must be made at commercial scale. At least 10% of a full production batch or 100,000 units, whichever is larger.
• Dissolution testing: The generic and brand are tested in simulated stomach fluids across pH levels 1.2 to 6.8. The dissolution profiles must match with an f2 similarity factor above 50. If they don’t, regulators will question whether the drug will behave the same in the body.
• Pilot studies: Many companies run small-scale versions first. Dr. Jennifer Bright, former director of the FDA’s Office of Generic Drugs, said pilot studies cut pivotal study failure rates from 35% to under 10%.

Common Pitfalls and Why Studies Fail

The FDA’s 2022 Bioequivalence Study Tips document breaks down why studies fail:

• 45% fail due to inadequate washout periods
• 30% fail because sampling times were poorly planned
• 25% fail because the statistical analysis was wrong

Other issues include using the wrong reference product, poor analytical method validation, or not accounting for food effects when the drug is supposed to be taken with meals. Even small protocol deviations-like a volunteer eating a snack 10 minutes late-can trigger a rejection.

Real-World Outcomes: Successes and Failures

Teva’s generic version of Januvio (sitagliptin) got approved in 2021 after a single successful study with 36 subjects. That’s efficient. Alembic Pharmaceuticals, on the other hand, had its generic version of Trulicity (dulaglutide) rejected in 2022 because Cmax values were inconsistent across multiple studies. The brand-name drug has a complex delivery system-injectable, long-acting. The generic couldn’t match its absorption profile.

In 2022, the FDA approved 936 generic drugs based on bioequivalence data-98% of all generic approvals that year. That’s the power of this system. But behind every approval is a team of scientists, statisticians, and clinical staff working for months to get those numbers right.

What’s Next for Bioequivalence?

The field is evolving. Modeling and simulation tools like PBPK (physiologically based pharmacokinetic modeling) are being used more often to predict how a drug will behave without running full studies. The FDA approved 27% of generics in 2022 using biowaivers-essentially skipping human studies for simple, highly soluble drugs based on BCS classification.

There’s also growing pressure to expand these approaches to complex products: inhalers, injectables, and topical gels. The EMA and FDA are working on new guidance. The goal isn’t to weaken standards-it’s to make them smarter. Faster. Less wasteful.

But for now, the gold standard remains the same: real people. Real blood samples. Real numbers. And a 90% confidence interval that must fall between 80% and 125%.

That’s how we know your generic pill won’t just look different-it will work the same.