Pharma Built a $6 Billion Belief Engine. Now It Can't Turn It Off.

How pharmaceutical advertising accidentally broke the clinical trial system.

In 1996, painkillers in U.S. clinical trials beat sugar pills by approximately 27%. By 2013, that gap had collapsed to around 9%.

The drugs didn’t get worse. The sugar pills got better.

A 2015 meta-analysis published in the journal “Pain” looked at 84 clinical trials of neuropathic pain drugs from 1990 to 2013. Placebo responses climbed steadily, reaching an average 30% reduction in pain by the end of the study period. Drug responses stayed flat.

And this is only happening in one country. European trials showed no increase. Asian trials showed no increase. Only American ones did. American exceptionalism has conquered many frontiers. But this is a new one.

The Sugar Pill That Ate Clinical Research

FDA drug approval works like this. You give some patients the real drug, others get an inert pill, you measure the difference. If the drug wins by a statistically significant margin, it gets approved. If it doesn’t, it’s back to the drawing board. This is an oversimplification, but when discussing placebos, this is really the nuts and bolts of it all. And Congress codified all of this into law in 1962.

One assumption holds the whole thing together: The placebo arm is stable. The sugar pill is fixed. Known. You’re measuring the real drug against a constant.

In this case, though, the constant is moving. Researchers estimate that more than 90% of potential drugs for neuropathic and cancer pain have failed in advanced clinical trials over the past decade. Not because the drugs don’t reduce pain, but because they can’t reduce it enough beyond what the placebo already delivers.

Defenders of the current system argue, reasonably, that any drug that can’t beat a sugar pill shouldn’t be on the market. The problem isn’t the standard. The problem is that the sugar pill keeps getting better, and only in the country that spends more than any other convincing its citizens that pills work. The U.S. has, in effect, built the world’s most expensive placebo.

Ask Your Doctor if Advertising Is Right for You

The United States and New Zealand are the only two countries on earth that allow pharmaceutical companies to advertise prescription drugs directly to consumers. Every other country has banned the practice.

In 1997, the FDA loosened its disclosure rules for television ads, allowing pharma companies to mention only “major risks” instead of a full list of side effects. Direct-to-consumer (DTC) ad spending climbed from $1.3 billion in 1996 to $6 billion by 2016. The FDA’s own enforcement went the other direction: Violation letters dropped from more than 130 a year in the late 1990s to three in 2023.

What did $6 billion a year buy? Americans now use prescription drugs at higher rates than a generation ago. Physicians are 17 times more likely to prescribe a drug a patient requests by name. And France and Canada’s independent drug assessment systems rated 68% of top-selling U.S. drugs “low added benefit.”

To recap: Americans see the ads, ask their doctors for the drugs, doctors write the script, and most of the drugs aren’t particularly all that good for what they’re trying to treat. The system works perfectly, if you’re the one selling the ads.

On the mechanical side, U.S. trials also got bigger during that timeframe. In 1990, the average neuropathic pain trial enrolled fewer than 50 patients and lasted four weeks. By 2013, that number had grown to more than 700 patients over 12 weeks.

European and Asian trials didn’t show the same inflation. Longer trials with more patients, more touch points with nurses and clinicians, and more clinical staging all track with larger placebo responses. Canadian neuroscientist and Professor of Pain Studies Jeffrey Mogil put it plainly: “The data suggest that longer and larger trials are associated with bigger placebo responses.” It’s a theory you can’t test, about a problem you can’t measure, in the only country where both are true.

Put those two trends together. The only country where placebo responses are climbing is also the only market where pharmaceutical companies spend billions a year telling consumers drugs will make them feel better. The team behind the Canadian assessment outline DTC advertising as one plausible explanation, while noting it can’t be tested directly with their data. (New Zealand, the only other DTC country, has too small a clinical trial market to test the hypothesis.)

How to Unbuild a Placebo

So the placebo problem is getting worse, and it’s concentrated in the United States. What is the industry doing about it?

Not asking why Americans believe in pills so much. Instead, they’re attempting to engineer that belief out of clinical trials.

Researchers developed something called the Placebo-Control Reminder Script, or PCRS: a coached speech read to clinical trial participants before the trial begins, designed to discourage them from reporting benefits if they received a placebo. This is a real thing that real scientists developed and published in a real journal, which is a sentence that shouldn’t need saying, but here we are.

The researchers’ goal was measurement accuracy, not manipulation. But look at the result: They ran an all-placebo trial, published in “Neuropsychopharmacology.” Every participant got a sugar pill. Half got the coaching script. Half didn’t. The coached group reported less improvement in their depression symptoms.

The coaching... worked, I guess? It made people report feeling less better.

Here’s the other approach. Research on the genetics of placebo responses found that a gene variant is linked to how strongly a person responds to a placebo. People who carry two copies of the variant have three to four times higher dopamine in the prefrontal cortex and respond to placebos at far higher rates. That’s about a quarter of the population. One in four humans is, from the pharma industry’s perspective, too hopeful to be used in a study.

A biotech startup called Biometheus began developing genetic tests to identify these individuals and exclude them from drug trials before they could “ruin” the data with their inconvenient biology.

The science is real, though still early stage: A 2018 study found that previously identified placebo-associated gene variants did not predict outcomes in a Phase III inflammatory disease trial. A Pfizer executive told the Washington Post in 2016 that a usable genetic screen was “still a few years off.” That was a decade ago.

To sum, the pharmaceutical industry built a $6 billion-a-year advertising machine that taught Americans to believe in pills. The placebo response rose. And the industry’s answer is coaching scripts that train people not to feel better and genetic tests that screen out people whose biology responds too strongly to hope.

Faith at the Pharmacy, Skepticism in the Lab

The pharmaceutical industry spends $6 to $8 billion a year on DTC advertising to build consumer belief in the drugs it markets. That belief drives interest, which leads to prescriptions, which drives revenue, which funds the next round of advertising.

Then the same industry spends $1 to $2 billion per drug on clinical trials trying to prove those drugs work. And the primary obstacle to proving they work, at least in the U.S., is the belief that the advertising created. The sugar pill cohort keeps getting stronger because the patients in that cohort are Americans who’ve been watching pharmaceutical commercials since 1997, who walk into a clinical trial already half-treated by three decades of consumer marketing.

The industry wants faith in the pharmacy, but skepticism in the lab. It wants consumers who believe in medicine and trial patients who don’t. It’s spending billions on both sides of that contradiction. The left hand spends billions building belief. The right hand spends billions burning it down. Both hands bill by the hour.

This isn’t just a pain drug problem. It shows up in antidepressants too. The average drug-placebo gap on the Hamilton Depression Rating Scale shrank from six points in 1982 to three points by 2008. For most approved antidepressants, fewer than half of the efficacy trials submitted to the FDA found the drug superior to placebo. Not most trials. Not a majority. Fewer than half. If you ever wanted proof that marketing works, here it is: it works so well it broke the science.

Chronic pain. Depression. Anxiety. The conditions where placebos hit hardest are the ones where patients suffer most and have the fewest alternatives. Where your expectation of relief can look like actual relief on a clinical scale. Rising placebo response doesn’t slow down drugs for conditions you can measure with a blood draw or a scan. It slows down drugs for the people whose suffering is the hardest to measure and easiest to dismiss.

Every failed trial adds years to a pipeline that already takes 10 to 15 years from discovery to approval. The people waiting at the end of that pipeline are real, and their pain is not a measurement artifact.

The control isn’t controlled. The baseline is culturally contaminated. Drugs that would have cleared the bar in 1996 fail today in the U.S. but would still pass in Europe, where patients haven’t spent three decades marinating in pharmaceutical commercials.

The industry’s response is not to reconsider the advertising. It’s to coach the hope out. Screen the believers. Redesign the trial. Do everything except turn off the television.