Skip to content

Reducing alarm fatigue: Packard Children’s researchers apply big data to a big problem

AlarmIf you've ever had a too-sensitive smoke detector in your kitchen, you've experienced alarm fatigue. You make toast or boil water and the smoke detector screeches, perhaps causing you to tell it in exasperated tones that nothing is on fire. You might ignore the racket or remove the detector's batteries, putting you at risk if there's ever a real fire.

Alarm fatigue is a big problem in hospitals, too, where automated monitors track patients' vital signs. The monitors' alarms sound frequently, but the vast majority of alarms don't indicate true crises, and all the false alarms desensitize doctors and nurses to the beeping and can slow their responses to real emergencies.

A new paper in the Journal of Hospital Medicine from a team of researchers at Lucile Packard Children's Hospital Stanford proposes a fix that may sound obvious, but would change current practice: Match the heart-rate and respiratory-rate thresholds that trigger alarms to what's actually unsafe for hospitalized children. The "normal" ranges now used for monitors are somewhat arbitrary; they're based either on measurements from small numbers of healthy kids or on physicians' consensus opinions rather than measurements from real hospital patients.

To determine safe ranges, electronic medical records are a treasure trove because they contain thousands of real vital-sign measurements. The Stanford team, led by Veena Goel, MD, analyzed the heart- and breathing-rate data from 16 months of Packard Children's records and calculated the 5th and 95th percentiles for the measures, cutoffs chosen as reasonable thresholds for abnormally low or high values. They broke the data down by age range and, to avoid having the results skewed by critically ill children, excluded data from patients who spent time in the intensive care units.

The researchers compared their calculated "normal" ranges to those from another recent study that analyzed real vital-sign data, and also to the NIH-endorsed reference ranges now in use. The two data-based studies are in good agreement, but the differences between the "reference" and data-driven figures are striking. In every age group, the current heart-rate reference range is slower, and the reference range for breathing rates is narrower, than in real life. No wonder there are a lot of false alarms.

The team also checked whether their parameters would have correctly identified a year's worth of real patient emergencies. They concluded that their heart-rate and breathing-rate thresholds are at least as safe as the reference ranges now in use, and that the data-based thresholds should be further tested in a prospective study.

"Data-driven vital alarm sign limits have the potential to decrease false monitor alarms, alarm-generated noise and alarm fatigue," the researchers wrote, adding that their work could be the first step in the development of an even better technique: "Ultimately, using a patient's own physiologic data to define highly personalized vital sign parameter limits represents a truly precision approach, and could revolutionize the way hospitalized patients are monitored."

Previously: Hacking for the win: Stanford students design new technology for nurses in need, Monitoring patients' vital signs using a touch-free video system and A new view of patient data: Using electronic medical records to guide treatment
Photo by debsilver

Popular posts

Category:
Genetics
Sex biology redefined: Genes don’t indicate binary sexes

The scenario many of us learned in school is that two X chromosomes make someone female, and an X and a Y chromosome make someone male. These are simplistic ways of thinking about what is scientifically very complex.