After a night shift at Abbott Northwestern Hospital, Todd Ostlund would go home and switch off his phone ringer and anything else mimicking the many, many alarms a nurse hears while on duty.

And still, on many nights, he'd be roused from sleep by a "beep, beep, beep" in his dreams.

"I'm never having peace," he said.

Medical device alarms play critical roles in a hospital — to signal trouble with a patient's vital signs or medical equipment, and to draw caregivers to the bedside in time to help. But too often, alarms have been nuisances — set off by patient movements that cause their pulse to spike briefly, or by a momentary kink in an IV line, or patients simply scratching their noses and bumping the blood oxygen monitors on their fingers.

The resulting problem, known as "alarm fatigue," can prevent nurses from responding to real patient emergencies, with fatal results. The Joint Commission, the nation's hospital accrediting body, attributed 80 deaths and 13 serious injuries to alarm-related failures in a recent four-year period, and in 2013 required hospitals to commit to preventing alarm fatigue.

Now, a study from one of Abbott's intensive care units suggests that less, indeed, can be more.

The unit switched the default settings of its pulse-rate alarms from overly conservative thresholds of high and low beats per minute, and to avoid duplication with alarms that go off for related cardiac issues.

The number of pulse rate alarms dropped 76 percent within six months — but Abbott's nursing leaders found no resulting incidents of patient emergencies being missed. If anything, nurses responded even faster to the remaining alarms because they were less likely to be false.

The goal is to avoid any alarm "that isn't going to prompt an urgent response," said Dr. William Dickey, a hospitalist at Abbott who tends to patients throughout the hospital. "It's the difference between bad and absolutely urgent."

Ostlund noticed the change on his nights in that unit. There were far fewer alarms breaking his concentration while he tried to complete a medical chart, or pulling him from one patient to another who wasn't in imminent need of help.

"It relieves some of the stress of being a nurse," Ostlund said. The seven-year veteran added that he has seen colleagues leave the profession in part due to the raucous environment of a hospital.

Mayo's results

Efforts to ratchet back alarms reflect the ebb and flow that emerges in many aspects of medicine. High-resolution CT imaging machines, for example, were broadly used until doctors discovered that their high cost and low-level radiation exposure could outweigh their diagnostic benefits in certain cases.

"Just because there's a new gizmo on the market doesn't mean we need to purchase it," said Dale Pfrimmer, nurse administrator for Mayo Clinic's thoracic and vascular division in Rochester.

Pfrimmer oversaw a similar study in two Mayo critical care units, and said a reduction in alarms over three months did not result in more emergencies such as Code Blues, when unresponsive patients needed resuscitation.

Such emergencies actually declined, though that might have been coincidental. But the alarms that went off did get the staff's attention, he said. "You really focus when you do hear those alarms, because now they mean something."

Missed emergencies aren't the only danger. Nurses distracted by noise can lose concentration and make mistakes, or can fall behind in their work if they err on the safe side and check every alarm immediately. Patients in ICU care also can have their sleep disrupted.

Relatives tune out, too

The risks, and the concept of "alarm fatigue," gained broad national attention after the 2007 death of a woman at UMass Memorial Medical Center in Boston. Nurses failed to respond to an alarm indicating that her monitor's battery needed replacing, and, when her monitor lost power, didn't know she was having a fatal cardiac arrest.

Alarm incidents have been reported in Minnesota, though they reportedly involved poor staffing and staff training rather than noise fatigue. A baby died in a high-profile 1997 case when poorly trained medical foster care providers turned off an oximeter that could have alerted them to her distress.

The family's attorney in that case, Chris Messerly, has come to appreciate the issue because his father has been in intensive care.

"I was thinking these same things as I was sitting in there every day for the last week," he said. "Sometimes [alarms] would go off and no one would come in for, like, 10 minutes."

Frequent alarms do upset visiting families, especially if they sound severe but caregivers don't respond immediately, said Dickey, the Abbott hospitalist. On the other hand, some relatives come to understand the issue so well that they start turning off alarms on their own, which can also create hazards, he said.

Abbott has expanded the alarms project to its neuro ICU. In addition to changing the pulse rate parameters of the EKG alarms, the hospital trained nurses in intensive care to clean the monitor lead sites on patients' skin each day and change the leads because there is research suggesting this reduces false alarms.

Next up are reviews about the frequency of alarms for IV lines. These sometimes go off even when there are temporary kinks preventing the flow of liquid, not real blockages preventing blood, fluids or medicine from getting to patients.

Oximeters, which measure blood oxygen levels, are particularly sensitive to false alarms because they are loosely attached to a patient's finger. Most blare alarms in four seconds, when waiting 10 or 15 seconds might differentiate between false alarms and real problems.

Then maybe Ostlund and other nurses will get some sleep at home. Ostlund was so happy about the alarms project in the ICU that he e-mailed thanks to his supervisors.

"You've got to imagine a constant din," he said. "Now, it's just not there anymore."

Jeremy Olson • 612-673-7744