Anatomy of a beep: A medical device giant and an avant-garde musician set out to redesign a heart monitor’s chirps

WASHINGTON — As deals struck by health care behemoths go, this was one of the stranger ones. On one side, you had a medical device giant, with a phalanx of PR professionals carefully guarding the company’s image. On the other, you had a consultant who didn’t sound much like a consultant:

“I am synthetic life form ‘Yoko K.,’ assembled in the US with components made in Japan,” one of her websites explained. “I am designed to assume the role of an ‘electronic musician.’ I am one of many secret agents sent to this time to plant magical thinking in people through the use of ‘pre-22nd century nostalgia Mars pop music.’”

In other words, Yoko K. Sen is an ambient electronic musician, born in Japan but transplanted to the United States, where she’s layered her breathy, machine-modulated vocals over ethereal blooms of synth at galleries, in concert halls, and on award-winning albums. In recent years, though, she’s created a new, more corporate niche for herself: revamping the soundscape in hospitals. Medtronic had hired her, late in 2017, for a related project, to help design the beeps patients would hear from their cardiac monitors at home.

The beginning of one of Sen’s compositions, called “Lou.” 

We live in an era of constant redesigns, and health care — with its cheerless institutional bent and vomit-like color palette — has proven especially ripe for reimagining. Fashion icons have taken on the hospital gown. Architects have gone after the hospital room, and celebrity chefs have barged into the cafeteria. One bigwig San Francisco designer even tried to rebrand death. You name it, there’s probably someone out there working to give it a makeover. And sometimes, these aesthetic changes might just save lives. As the New York Times reported in 2014, when patients moved into homier hospital rooms as a test, they not only felt more comfortable but also requested less pain medication.

Behind all of these initiatives is the idea of refocusing health care on the patient. To do that, the thinking goes, you often need to fix problems so glaring they’re easy to overlook. Medical sound is a perfect example. What, in the 21st century, could be more ubiquitous than a beep? We’re so surrounded by electronic noises that we hardly notice them — and in a hospital, where that digital symphony becomes even more feverish, an unnoticed alarm can mean a dead patient. So redesigning those chimes requires a delicate balance: Attention-getting but not startling, easy to differentiate but simple enough to learn.

That was the task facing Sen, 38, when Medtronic asked her company, Sen Sound, to come up with around 10 tones for their new remote heart monitor. This medical music was for the home, not the hospital, but many of the same constraints applied. Nearly a million Americans have bits of Medtronic hardware nestled inside their chests to keep tabs on — and, sometimes, intervene in — their heart rhythms. Those implantable devices come with a plastic doohickey that sits at the bedside as a kind of cardiac satellite, taking in data on how well a person’s blood is being pumped and then beaming the information out to a webpage for their physician to check.

That bedside monitor was the instrument on which Sen’s newest piece was to be played. It was a strange assignment, applying the lessons of dreamy electronica to everyday electronics. “There’s not a lot of sound design expertise in the medical industry per se — certainly not the kind of expertise you would normally associate with writing music, or creating little melodies,” said Michael Wiklund, general manager of the human factors engineering practice at the global safety consulting firm UL.

But there was a precedent. A cruise-ship-pianist-turned-anesthesiologist, an opera-loving clinician, and a small flock of psycho-acousticians — to name a few — had spent decades working on similar compositions, with oversight from a kind of United Nations of medical sound. They’d all been struggling with the essential, but oft-overlooked, questions that now faced Sen: Just how much could you improve something as limited as a beep? And if you did, how much of a difference would it make?

For Sen, those questions began with an illness. She first got sick around 2012, and by 2014, her condition had worsened enough for her to spend the year undergoing a frenzy of poking and scoping and testing to figure out what was wrong. She prefers not to talk about the diagnosis itself, but the sonic disorientation she felt during weekly hospital visits has become an essential part of Sen Sound’s pitch.

Her career had been built on listening — placing a beat just so, shifting tones to create an otherworldly shudder — and now she was surrounded by a thoughtless hellhole of sound: beeps, footfalls, rattles, screams, beeps, and more beeps.

“I remember I was hooked to several different monitors,” she said. “One of them was beeping so loud, nonstop, and I asked a nurse who came in, ‘Excuse me, this thing keeps beeping, is this OK?’ And she said, ‘Yeah, that thing just beeps.’ That gave me such a lasting impression. ‘Yeah, that thing just beeps.’”

She didn’t know it at the time, but that unnecessary beeping would open up a sort of crossroads for Sen. She’d had moments of self-reinvention before. When she was a little kid, her family moved every year or so — Nagano, Niigata, Tokyo — following her father’s jobs as a forest ecologist. She’d lived in six or seven different cities by the time she was 10. She just thought it was normal to move periodically, building your social world afresh.

Her entry into electronic music was just as precipitous. By then, she was living in Washington, studying international affairs at George Washington University. Music had been an obsession for her while she lived in Japan. She’d started classical piano at age 3, threatening to throw a tantrum when her teacher said she was too little. Afterward, she picked up instruments and groups the way other kids amass detentions — clarinet in a wind symphony, drums in a rock band, voice in a jazz combo. But in college, all that had been reduced to voice lessons and nocturnal trips to the dance floors of D.C., then throbbing with the steady thump of late-’90s house.

Then, one morning toward the end of her degree, she woke up with a strange feeling — “like a déjà-vu, almost,” she said, “like a tingly sensation.” As subtle as it was, she knew what she had to do. “Up until that day I never wrote a piece of music, I never touched music-making software,” she said. “I just woke up and I was like, ‘Oh I’m making electronic music today.’”

She called an acquaintance who had a studio setup, asked if she could borrow it, and laid down a track with beat and baseline, lush chords and vocals on top. Later, she found out that that was also the day her voice teacher had died suddenly. She sometimes thinks of that unexpected urge — it still occasionally overtakes her — as an instructor’s parting gift.

The tingly sensation had led to soundscapes, which had led to albums, which had led to artist’s residencies, which had led to new collaborations. Her illness interrupted all of that. When, finally, her health began to improve, she felt giddy at the simple fact of getting through a day of work again. In 2015, she and a friend were selected for a two-week residency at Ideo, a firm bent on redesigning design itself (and, as it happens, the same company where an exec wanted to repackage death).

Beforehand, Sen had been skeptical of design thinking — “I thought it’s some kind of a hipster thing” — but she came out a convert. “I started to talk to some designers,” she said. “’Do you guys use this design thinking for, like, sound in hospitals?’ They were like, ‘No, not really.’”

If they’d been avid readers of the medico-acoustical literature, they would have known better. There was, in fact, a long history of sound design for hospitals, almost Shakespearian in its rivalries and alliances and rhythmic obsessions.

In the ’80s, just as synth-pop was oozing into prominence, anesthesiologists decided that the electronic racket inside operating rooms was getting out of hand. They were dealing with a plague of beeps, and they weren’t the only ones. A psycho-acoustician in Cambridge, England, Roy Patterson, had recently applied himself to the alarms in trains, planes, and fire stations — and so he was asked to try to restore the sanity and safety inside hospitals next.

What he proposed were seven pairs of sounds, rounded so they wouldn’t startle, with snippets of silence to allow a clinician to think and react. He had a specific ditty for problems with oxygenation, ventilation, heart function, artificial perfusion, drug administration, and temperature — plus a general alarm, for good measure — and each melody came in two moods. If played slowly and once through, it meant “caution.” Sped up and extended in length, it meant “emergency.” It was like a small musical language, with only a handful of words.

Two pairs of Roy Patterson’s sample sounds: A cardiovascular “caution” alarm and “emergency” alarm, followed by an artificial perfusion “caution” alarm and “emergency” alarm. 

Some doctors, though, were quick to dismiss those sounds as unsound. The whole project had begun because all of these machines were indistinguishable in their high-pitched bleating. Now, some said, Patterson’s samples were just as confusing.

One musician-turned-clinician named Dr. Frank Block Jr. became so riled up that he planned a kind of counterattack — half prank, half study — at the 1990 American Society of Anesthesiologists annual meeting. During an after-dinner panel discussion in front of a room full of colleagues, he played pop songs that might work as alarms for the scenarios Patterson had had in mind. For a cardiac alarm, he’d chosen “I Left My Heart in San Francisco.” Oxygenation would be “Love is Blue.” Issues with drug administration — think about an IV that’s dripping down — would trigger “Raindrops Are Falling on My Head.”

He even wrote a paper about how recognizable these alarms became to the audience once they’d heard each song and been told its hidden medical meaning. Age was not a factor, he hypothesized, because of the recent rise of oldies stations. “When I proposed this, a lot of people came running up and said, ‘Are you kidding?’ Quite honestly, I did not know the answer. … It was absolutely on the line between genius and madness,” he said.

But the border between genius and madness wasn’t really the International Organization for Standardization’s cup of tea. Nor, it turns out, was Patterson’s suite.

The group is a bit like the U.N. for industry, bringing together bodies from 162 nations to set international norms on everything from food safety to anti-bribery systems. What these international standardizers did like in Patterson’s work was the idea of urgency encoded in rhythm, and they used the same trick when they joined forces with the International Electrotechnical Commission to solve the problem of medical alarms once and for all.

They settled on two patterns: three beats for semi-urgent alarms, and five beats for dire ones. And they asked Block and a few other musically inclined members of the working group to write some tunes to go with them.

The composers knew, right off the bat, that the assignment was impossible. With so little rhythmic latitude, these tones just wouldn’t be distinctive enough for a harried nurse or doctor to tell them apart. Even so, Block sat down at the grand piano in his living room in Little Rock, Ark., and began composing show-stoppers for IEC Standard 60601-1-8. Again, he was omnivorous in his inspirations. A careful listener might have heard the old NBC chime in the alarm for ventilation, and in the artificial perfusion warning, the chant of the Winkies from “The Wizard of Oz.”

Two pairs of the sounds proposed by Block and his colleagues: A cardiovascular medium-urgency alarm and high-urgency alarm, followed by the NBC-inspired medium- and high-urgency alarms for ventilation. 

As uncatchy as those melodies were, they made it into the standard, and when the first edition came out in 2003, it gave device companies the choice to use those compositions or write their own within haiku-like constraints. The U.S. Food and Drug Administration recognizes the standard, too, as a kind of regulatory shortcut.

“Quite honestly, most of the companies went with a fixed pitch. They said, ‘We don’t like your tunes, we’re not musicians, we don’t have composers on staff, we’re just going to use a fixed pitch.’ So that’s what’s happening today,” said Block, who is now a professor of anesthesiology at the Augusta University/University of Georgia Medical Partnership, and who has publicly apologized for his role in derailing the Patterson sounds. “I literally have had in the OR a cardiac monitor and an anesthesia machine and an infusion pump, and they were all going bahm bahm bahm, they were all making identical sounds, and they were all compliant with the current edition of the 60601-1-8.”

It isn’t just him. There has been plenty of research into the mysteries of medical sound. These musician-scientists now know that softer alarms, with the right harmonics, can cut through background noise like a soprano through an orchestra. They know that Block’s melodies are, as he predicted, hard to recognize, even with training. But that research often can’t be heard in the beeps of medical devices. As Dr. Joseph Schlesinger, assistant professor of anesthesiology and critical care medicine at Vanderbilt, explained, “Now, we have alarms that are easily confused and difficult to learn and don’t really tell us what’s wrong.”

Sen’s agreement with Medtronic came about the way agreements often do in the age of Silicon-Valley-style entrepreneurship: a stint at an incubator, a mentor who knew someone at a foundation, a talk at Stanford Medicine X, a tweet seen by the right people.

She had spent 198 Swiss francs on her very own copy of IEC Standard 60601-1-8. She had made sound maps of hospitals. She’d given workshops in which health care executives wore blindfolds, to deepen their listening. She’d noticed that the fixed pitches played by a cardiac monitor and a bed fall alarm, when taken together, formed a tritone — “the Devil in music.” But this would be among the first designs that Sen Sound had to deliver.

Instead of creating sounds for different failures that could result in patient death, as Block had done, she needed to compose a Suite for Bedside Monitor, in which each movement had only half a second to tell patients about the transmission of their cardiac data. One jauntier piece might have been titled “Properly Sent to Physician Portal.” Another, on the more serious side, might be “Wi-Fi Trouble.”

The company hadn’t received any complaints specifically about the monitor’s current soundscape, but, as Jenny Ramseth, a director of customer experience at Medtronic put it, “I mean, nobody was complimenting our tones, either.” The team worried that those alarms, programmed purely to get attention, might be too alarming.

In some ways, the impulse was a good one. As an administrator of a 7,000-person Facebook support group for those with implantable cardiac defibrillators, Jasmine Wylie knows that many patients leave the hospital with little education about the device inside them and its accessories. Many freak out when they hear their own chest beeping — the defibrillator signaling it needs a new battery, say — or when they see their bedside monitor flashing. Some head to the emergency room. Others call 911. So she views any tool that improves the patient’s understanding of these machines as an improvement.

Ironically, though, to those with cardiac devices, there’s also something tone-deaf about the whole musical project. To patient advocate Hugo Campos, it’s as if Ford had invented the Model T and then spent 100 years refining its wooden wheels. After all, both he and Wylie were only vaguely aware that their Medtronic bedside monitors could emit sounds in the first place. What they really want — and what Campos has spent a decade asking Medtronic for — is the ability to see exactly what is going on inside their own heart without having to go through their doctors. “This thing collects tons of data,” said Wylie. “I want to be able to access it, and not wait 72 hours until my clinic opens.”

That wasn’t a problem Sen could solve. Her domain was built of buzzes and beeps — and she spent much of last winter exploring every corner of it, in the company of a Medtronic scientist named Stephen Nelson. The firm wouldn’t make Nelson available for an interview, so he could not comment on being one half of the Lennon and McCartney for Medtronic bedside monitors. But on a humid day in June, Sen sat down at a mini-keyboard to reconstruct parts of their back and forth.

She began by playing, from her laptop, a few of the sounds emitted by the current iteration of the bedside monitor. One was a buzzy, two-toned bleep, which she heard as the last two notes in a G-minor scale. Another was a long note, followed by two lower chirps. The third was just a series of static beeps.

When she had sent out the first to a few of her Facebook friends — a kind of informal customer survey — the reactions were mixed. Some weren’t bothered. Others found it startling or annoying. One person compared it to nails on a chalkboard.

Sen plays a few of the sounds emitted by the current iteration of the bedside monitor.

Sen admits that, while her work as an electronic musician involved both melody and timbre, she’s biased toward the latter, and so that’s where she began. But she quickly realized it was a dead end. Changing the timbre would mean completely redesigning the buzzer within the bedside monitor. “In a sense, you are given an instrument to play and you’re not really allowed to say, ‘This is a trumpet, can I play the flute?’” Sen said.

If she was working with the electronic equivalent of a trumpet, she thought she might try plunking on a digital mute — but that, too, got mixed reactions from her Facebook focus group, and she eventually put it aside.

Instead, she began playing, Block-style, with melody. Just as Block had been limited in the rhythms he could use, so Sen was limited to the frequencies that would be audible to elderly patients. She and Nelson started by laying out the possibilities. “Steve — I remember it was Christmas week — did such hard work to play each note … and then measured the decibel level,” said Sen. “So together we figured out, OK, we get to use this note, not this note.”

With that list of usable notes, Sen drove to a studio in Annandale, Va. The place was owned by a saxophonist friend of hers, so close he was like family. She had married her husband here among the recording equipment, and now, in these rooms carefully calibrated for shaping sound, she and her friend programmed a keyboard so that it could play only the permitted tones.

“He is sort of a maestro of melodies and harmonies,” she said. “I told him just play, have fun, and I recorded everything.” He began to play bursts of notes she hadn’t thought of, standard arpeggios, second inversions, and she emerged with a whole album’s worth of beeps and melodies.

But in the end, these, too, got discarded. “There was a concern that it’s fun and nice, but the goal is really to be clear, and for the tones to be communicating certain scenarios and messages,” she explained.

Those ideas — simplicity, directness — are central in the world of medical sound design, although the current fashion is moving away from beeps alone. Some device makers have been adding speech to sound, with naloxone and epinephrine injectors telling users, Siri-style, what to do.

Meanwhile, a psychologist named Judy Edworthy — “the godmother of alarm design,” Sen called her — has been working on her own suite of sounds. She was a postdoc of Patterson’s in the ’80s and is now at the University of Plymouth in Britain, where she has developed “auditory icons” to replace the alarm melodies proposed in 60601-1-8. Problems with breathing are signaled by a whooshing, a medication issue by the rattling of pills, cardiovascular trouble by the lub-dubbing of a heart. They’re easy to learn, not too irritating, and harmonically rich enough to cut through the noise of an OR or ICU. When paired with a rhythm of beeps to indicate urgency, they seem to meet all the criteria.

“These sounds are almost certainly going to go into the next international standard on alarms,” Block said.

Sen’s proposals weren’t quite so literal. She’d lowered the frequency of the tones and simplified the patterns — more ornament than melody, more “oops” than “red alert.” They were attention-getting but comfortable, she hoped, less grating than the original sounds, but retaining a bit of bite.

Sen demonstrates different ways she tried playing with beeps, to make them more comfortable and meaningful for patients. 

Nelson gave her feedback. She wrote new variations. More feedback came, and more details were tweaked. As much work as she put in, Sen knew that noises often come to us in a blur. Those urgent enough to surface in our consciousness are only there for an instant. They get a blip of thought and then they disappear. That would probably be true for these delicately crafted beeps as well, even in the relative quiet of the home.

Patients won’t hear the final versions for a while — the monitor is still in development, and still has regulatory hurdles to jump. But when they do, chances are they won’t hear the months of calculations and improvisations that went into them. They’ll probably just hear a set of beeps.

That is fine with Sen. As she put it, “The ultimate goal of a sound designer is, when people don’t notice a sound, but … say, ‘I don’t know what it is, but something about this product I like.’”