The little blue pill may not have ever been if not for one dutiful nurse.

Viagra, Pfizer’s blockbuster erectile dysfunction drug hit the market in 1998. In 20 years, it’s become ubiquitous: 62 million men all over the world have bought the drug, according to a Pfizer spokesperson. The US military shells out $41.6 million per year for it, and from 2012 onward, the US, Mexico, and Canada spent about $1.4 billion on it annually (although those numbers are projected to fall in the coming years when Pfizer’s patent on the drug expires in 2020).

Despite the drug’s popularity today, the researchers who discovered it weren’t even looking for it. Sildenafil, the active ingredient in Viagra, was originally developed to treat cardiovascular problems. It was meant to dilate the heart’s blood vessels by blocking a particular protein called PDE-5. In animal tests, it seemed to work moderately well: researchers could find evidence that it was impeding PDE-5, and the animals weren’t having any obvious negative side effects. So, it was brought into a phase one clinical trial in the early 1990s, to test whether humans can tolerate a new compound.

All seemed to be going well—except for one weird thing the men enrolled in the study did when nurses went to check on them. “They found a lot of the men were lying on their stomachs,” John LaMattina, who was the head of research and development at Pfizer while this research was ongoing, said on a 2016 episode of the STAT Signal Podcast (listen in around 7:15). “A very observant nurse reported this, saying the men were embarrassed [because] they were getting erections.” It appeared that the blood vessels dilating were not in the heart, but rather the penis (dilating blood vessels is part of the processthat leads to erections).

The sildenafil was working—but in the wrong part of the body. And with that, the so-called “potency pill” was born.

Viagra was approved by the US Food and Drug Administration for use as an erectile dysfunction drug in 1998. About a decade later, researchers began running new clinical trials to see if it could double as a heart drug as originally intended. Sure enough, in 2005, the FDA approved the same compound for a heart condition called pulmonary arterial hypertension, which constricts the blood flow to the lungs and affects both men and women. The drug that makes Viagra work is now also sold under the name Revatio, for pulmonary arterial hypertension.

This isn’t that abnormal. Scientists know how individual parts of the body work, but how all the body’s systems work together is still somewhat of a mystery. Occasionally, this means scientists will know that a drug works, but won’t understand exactly how—and that means new uses for old drugs are found all the time.

Recently, a large-scale clinical trial found that canakinumab (Ilaris) a drug used to treat one form of arthritis made by Novartis, could also be used to treat heart disease after researchers understood that inflammation may play a role in the latter. Before that, in 2008, the compound bimatoprost (Lumagen), which was made by Allergan to treat high eye pressure, was approved to enhance eyelash growth under the name Latisse after people taking it kept reporting the glamorous side effect. And the compound finasteride was first used in the drug Proscar, made by Merck, to treat swollen prostates, but was later approved to treat baldness under the name Propecia when patients reported they noticed hair changes after taking the drug. (Some people taking finasteride have reported some fairly horrific side effects, in some cases long-term.)

In all of these cases, the second application was found after the drug was on the market for a while. As larger populations take a given drug for longer periods of time, there’s a chance for new side effects that didn’t show up even in the largest stage three clinical trials to be uncovered. This is why it’s important to report all side effects—and not just the bad ones—to your doctor, regulatory agencies, or even the drug companies themselves. It could improve or even save lives. In addition, it’s clear that it’s worth researchers’ time to keep investigating a drug that already “works” in case it turns out to have applications for other conditions that scientists didn’t recognize before. The more data, the better.

While drug discovery does involve a lot of research and development, it also involves serendipity and spotting unusual details. If that one nurse hadn’t reported her participants’ reactions to sildenafil, the world wouldn’t have had Viagra. This same kind of luck and vigilance could uncover new uses for existing treatments.