To Reduce Hospital Noise, Researchers Create Alarms That Whistle and Sing

In 2012, Yoko Sen was in an emergency room, tethered to a machine bleating relentlessly in her ear.

She was “freaked out,” she said, and felt helpless.

When a nurse returned to the room, Ms. Sen asked if it was O.K. the device was screaming.

“Yeah, this thing just beeps,” she recalled the nurse saying.

Ms. Sen, an electronic musician, was stunned. How could something “so loud and so jarring” be considered normal?

“The fear of not knowing amplified the feeling of anxiety,” she said.

And how, she wondered, could clinicians withstand the clangor?

As she lay there, she said, a cardiac monitor rang out in a tone close to the musical note of C, clashing with a distant device wailing in a high-pitched F sharp, creating what’s called the devil’s interval, a dissonance so chilling that medieval churches forbade it.

Hospitals today can be sonic hellscapes, which studies have shown regularly exceed levels set by the World Health Organization: droning IV pumps, ding-donging nurse call buttons, voices crackling on loudspeakers, ringing telephones, beeping elevators, buzzing ID scanners, clattering carts, coughing, screaming, vomiting.

Then there are the alarms. A single patient might trigger hundreds each day, challenging caregivers to figure out which machine is beeping, and what is wrong with the patient, if anything. (Studies have shown that as many as 99 percent of alarms are false.)

The proliferation of pinging and bleeping can contribute to patient delirium and staff burnout. And because caregivers know that many devices are crying wolf, they might be less responsive or apathetic, a potentially fatal safety issue known as alarm fatigue.

From 2005 to 2008, more than 500 patients in the United States had adverse outcomes, mostly death, because an alarm was ignored, a device was silenced or mismanaged in some way, according to the Food and Drug Administration, which tracks adverse events involving medical devices.

‘The cruelest absence of care’

“You don’t need to have alarms scream at you,” said Judy Edworthy, a professor of applied psychology at the University of Plymouth, in Britain.

But, she said, “people take a lot of convincing” that alarms don’t need to be so startling.

For device manufacturers, sound is often an afterthought in the design, Dr. Edworthy said, and they are worried about being sued if a machine had failed to cry out.

So, without an enforceable, universal standard, alarms have run riot.

They are also using sounds based on an outdated set of international safety standards, which have, paradoxically, perpetuated the din.

Dr. Edworthy, who has been called the godmother of alarms, is leading a passionate group of specialists, including Ms. Sen, who now works with device manufacturers and hospitals to incorporate the needs of patients and clinicians, and Elif Ozcan, who leads the Critical Alarms Lab in the Netherlands.

Together, this group is developing tones that replace the anodyne blare of the current alarms with signals that mimic electronic dance music and or a heartbeat.

They are working to make alarms quieter, combining audible alarms with visual cues like interactive screens that look like paintings, and working to develop a new standard that is likely to go into effect early next year.

“Unnecessary noise is the cruelest absence of care,” Ms. Sen said to a room full of medical professionals at a conference last year about end-of-life management. The words came from the mother of modern nursing, Florence Nightingale, who worked in the Crimean War in the 19th century.

An international standard that perpetuates the din

Deep in the rule book for safety and performance of medical devices is IEC 60601-1-8, which sets the standards for medical device alarm sounds. The particulars of the code were hashed out over many years by a joint working group, assembled by the International Electrotechnical Commission, a nonprofit based in Switzerland that publishes guidelines for electronic and technical equipment used by hospitals.

Among other specifications, the standard sets forth tones for six critical functions: cardiovascular, drug administration, ventilation, oxygen, temperature and artificial perfusion (the flow of blood and oxygen), also known as “the six ways people die.”

At one point, the popular melody “I Left My Heart in San Francisco” was floated as a possible signal for cardiac problems, but ultimately, it did not make the cut.

“The songs are not supposed to be the Billboard top 100,” said Dr. Frank Block Jr., an anesthesiologist and musician, who was on the committee that passed the 2006 standard that is still largely in place.

Among the tones that were approved was a tune reminiscent of the old NBC chime, meant to mimic rising and falling lungs, Dr. Block said.

And the sound for mechanical blood flow and oxygenation was modeled after the “yo-EE-oh” of the Witch’s guards from “The Wizard of Oz,” a musical tritone known as the devil’s interval.

The sound for drug infusions was intended to mimic drops falling and “splashing” up, represented by a jazz chord called an inverted ninth.

But each ditty has the same rhythm and the same number of pulses, making them difficult to tell apart and difficult to learn. And they were never tested. Dr. Block later issued a public apology on behalf of the committee for approving the sounds.

“We did the best we could,” he said recently, “but the sounds were basically terrible.”

Now, Dr. Edworthy is spearheading the creation of a “revolutionary” set of tones, Dr. Block said.

Audio technology has changed drastically since the eight tones were created, said Dr. Edworthy, who has created sonic alerts for nuclear plants and train systems.

“It’s now possible to produce pretty much any sound you want from a medical device,” she said. “Of course, that’s a new set of problems.”

The proposed sounds, called auditory icons, are representative of their functions, like the crumpling paper sound that your computer makes when you throw files in the trash. In this case, the sounds represent critical organ functions and imitate the lub-dub sound of a heartbeat, or a rattling pill bottle for a drug infusion, or a whistling teakettle for temperature.

“We’ve amassed a load of data demonstrating that these sounds work very well,” said Dr. Edworthy, who is collaborating with other researchers, including Dr. Joseph Schlesinger an associate professor at Vanderbilt University, to test how quickly clinicians are able to learn and respond to the sounds, how easily they can be identified, and how loud they need to be.

She has presented her findings to the current committee, which has been described as a “United Nations of medical sound,” and includes representatives from medical device companies and from countries with differing philosophical and cultural perspectives on alarms.

“You’re asking people to make changes that are going to cost millions of dollars, and some just don’t want to,” she added.

That said, the strength of the standard varies between countries, which can adopt all, parts or none of the written guidelines. In the United States, Dr. Block said, the Food and Drug Administration usually follows the standards, but it may add further requirements.

But the bottom line is that no device manufacturer wants a dead patient tethered to one of its machines.

Dr. Edworthy said she was confident that the new sounds would be adopted, provided politics don’t get in the way.

Alarms that do more than beep

At the Critical Alarms Lab, Dr. Ozcan recorded rattling pill bottles and running water to effect Dr. Edworthy’s concepts.

Dr. Ozcan, who has had practice translating vast quantities of data into audio cues for the European Space Agency’s mission control dashboards, said her group at the lab was developing devices to hush the intensive care unit, which can be louder than a vacuum cleaner, and challenging conventional device design, possibly even making alarms “beautiful,” she said.

One of her group’s projects, called CareTunes, is a speculative, even quixotic, melodic design.

The device transcribes a patient’s physiological condition into songs that sound a bit like chill electronic dance music. (Ms. Sen was an artistic adviser to the project.)

The melody is derived from a patient’s vital signs: drums for the heartbeat, guitar for oxygen saturation and piano for blood pressure. When a patient is stable, the tune is harmonious, but it becomes dissonant when a patient’s status changes for the worse, ideally grabbing a caregiver’s attention.

The device would not replace a “code blue,” Dr. Ozcan said, but it could potentially reduce the number of beeps, as caregivers would be alerted that a patient was veering into a danger zone before an alarm is triggered.

The challenge, said Dr. Ozcan, is balancing the needs of patients and clinicians, who would have to learn and integrate new devices into their work flow.

Dr. Ozcan said she was hopeful that the research done at her lab could be applied in other settings, such as air traffic control rooms, or would be relevant for research on how sound influences health in general, especially in work environments.

“We owe it to the community and health care,” she said.

Yoko Sen has since recovered from her illness, but the bleating monitors are still “the soundtrack of my life,” she said.

Through her start-up Sen Sound based in Washington, she has collaborated with medical device engineers to create new tones for home heart monitors, and with interior designers to build a so-called tranquillity room, where clinicians can relax, making them less likely to slam doors or talk loudly.

During a person’s last moments, her eyes might be closed, his nose covered by a ventilator, her food ingested by tube. Unless someone is holding her hand, she might feel nothing.

“For patients who die in the I.C.U., that sound of the alarm might be last sound they hear,” she said.

As part of a project with OpenIdeo, Ms. Sen interviewed hundreds of people about the last sound they would want to hear during their final moments.

Many people said they wanted to hear sounds from nature, like the ocean, or voices of family members.

No one, she said, mentioned bleating alarms.