Drug Recall Authority: How the FDA Legally Removes Unsafe Medications

When a medication turns out to be dangerous, people assume the FDA just pulls it off the shelves. But that’s not how it works. The FDA doesn’t have the power to force a drug recall. Instead, it asks manufacturers to do it voluntarily. And in nearly every case, they do.

How the FDA Actually Removes Unsafe Drugs

The legal foundation for drug safety in the U.S. comes from the Federal Food, Drug, and Cosmetic Act (FD&C Act) of 1938. Under this law, the FDA can inspect factories, review clinical data, and monitor side effects after a drug hits the market. But when something goes wrong - like contamination, incorrect dosing, or hidden toxins - the agency can’t issue a direct order to remove the drug. It can only request a recall.

This might sound like a loophole, but the system works because manufacturers have everything to lose by refusing. If a company ignores an FDA request, the agency can take legal action: file for a court injunction to stop production, seize inventory, or even shut down facilities. For most companies, that’s a far worse outcome than a voluntary recall. So they comply - quickly.

In 2022, out of 4,312 drug recalls in the U.S., only three required FDA enforcement. That’s 99.7% voluntary. The FDA doesn’t need to force recalls because the consequences of saying no are too high.

The Three Levels of Drug Recalls

Not all recalls are created equal. The FDA classifies them into three categories based on how dangerous the drug is:

• Class I recalls are the most serious. They involve drugs that could cause serious injury or death. Examples include pills with lethal contamination, like NDMA in blood pressure meds, or insulin with incorrect concentrations. These make up just 2.1% of all recalls but demand immediate action.
• Class II recalls are more common - about 69% of all recalls. These involve drugs that might cause temporary or reversible health problems. Think wrong label instructions, missing active ingredients, or packaging errors that could lead to misuse. Not life-threatening, but still risky.
• Class III recalls are the least urgent. They’re about minor issues like incorrect labeling, mismatched batch numbers, or packaging flaws that don’t affect safety. These rarely harm patients, but the FDA still requires them to be fixed.

The classification determines how far the recall goes. A Class I recall might require hospitals to notify every patient who received the drug. A Class III might only need a notice to distributors.

How a Recall Starts

Recalls don’t happen out of nowhere. They usually begin in one of two ways:

• Manufacturers find the problem themselves. Companies are required to run annual stability tests on every drug they make. If a batch fails - say, it starts breaking down too fast or shows unexpected contaminants - they must report it to the FDA and start a recall.
• The FDA spots the issue. Through its MedWatch program, the agency collects over 1.2 million reports each year from doctors, pharmacists, and patients about bad reactions. If patterns emerge - like multiple cases of liver damage linked to a specific batch - the FDA investigates. If evidence mounts, they contact the manufacturer and ask for a recall.

Once a recall is initiated, the manufacturer must submit a “Recall Strategy” to the FDA within 24 hours. This includes how far the recall will reach (retail? hospitals? patients?), how they’ll notify people, and how they’ll collect returned products.

Why Devices Are Different

Here’s where things get confusing: the FDA can force a recall on medical devices. Under 21 CFR 810, if a device poses a serious risk - like a faulty pacemaker or a broken ventilator - the agency can legally order the manufacturer to pull it from the market. No request. No negotiation. Just a command.

Why the difference? It comes down to history. When Congress passed the Medical Device Amendments in 1976, they gave the FDA stronger powers for devices because they were seen as more inherently risky. Drugs, on the other hand, were governed by the older 1938 law - which never gave the FDA that kind of authority.

Experts argue this creates a dangerous gap. A contaminated pill and a broken insulin pump can both kill. But one can be pulled by court order. The other can’t.

The Valsartan Recall: When the System Struggled

In 2018, a cancer-causing chemical called NDMA was found in valsartan, a common blood pressure drug. The FDA issued its first public alert on June 8. By June 29, U.S. manufacturers had recalled all affected lots. On the surface, it looked like a success.

But the truth was messier. The NDMA came from a raw ingredient made in China. Chinese suppliers didn’t respond to FDA requests for 17 days. The delay meant patients kept getting contaminated pills longer than they should have. And because the FDA couldn’t force a recall, they had to wait - and warn the public repeatedly.

This case exposed a major weakness: global supply chains. Many drugs today are made with ingredients from multiple countries. If one supplier ignores the FDA, the entire recall can stall.

What Hospitals and Pharmacies Do

When a recall hits, hospitals and pharmacies don’t just wait for a notice. They have to act fast. The American Society of Health-System Pharmacists (ASHP) requires every pharmacy to have a recall plan. That includes:

• Assigning someone to check FDA alerts daily
• Tracking lot numbers in their inventory system
• Notifying patients who received recalled drugs
• Documenting every step for audits

But it’s not easy. A 2022 survey found that 68% of hospital pharmacies struggled to match recalled lots with patient records because manufacturers use inconsistent labeling. In one case, a Class I recall was delayed by 3.7 days because the pharmacy couldn’t identify which patients got the bad batch.

That’s why companies like Recall Masters and Recall Index now offer software to track recalls in real time. Over 70% of U.S. hospitals use these services - not because they want to, but because they have to.

Is the System Broken?

Some experts say yes. Dr. Sidney Wolfe from Public Citizen testified in 2019 that the FDA’s inability to mandate recalls creates dangerous delays. He pointed to the valsartan case and others where patients were exposed to harm for weeks while regulators waited.

Others argue the system works. Janet Woodcock, former FDA deputy commissioner, said 99.7% of recalls happen within 10 days of FDA notification. The voluntary model has been stable for decades.

The real issue isn’t the system - it’s the lack of modernization. The FD&C Act was written in 1938. The FDA’s recall powers haven’t changed since then, even as drugs became more complex, supply chains became global, and risks grew.

In 2022, Congress considered the FD&C Modernization Act, which would have given the FDA mandatory recall power for drugs. But the provision was removed during committee review. Industry groups like PhRMA spent millions lobbying against it, arguing voluntary recalls are “99.98% effective.”

Meanwhile, new threats are emerging. Biologics - drugs made from living cells - are harder to test and more prone to contamination. The FDA says it’s only a matter of time before a major recall is delayed because of a foreign supplier who won’t cooperate.

What’s Next?

The PREVENT Pandemics Act, introduced in late 2023, includes a provision (Section 3103) that would finally give the FDA the power to order mandatory drug recalls. If passed, it would be the biggest change to drug safety law in nearly a century.

But it’s still uncertain. PhRMA continues to oppose it. And without political pressure, Congress may leave things as they are.

For now, the system relies on trust. Trust that manufacturers will act fast. Trust that global suppliers will respond. Trust that hospitals can track down every pill. It’s worked - so far. But when a life is on the line, trust isn’t enough.