The clinical trial was structured to statistically compare alarm data collected before and after the BEAMS nurse call interface was implemented and nurses were alerted to patient alarms by the system.
To begin with, BEAMS units were programmed on the alarm tones for all patient applied medical equipment so alarms could be reliably detected and distinguished from ambient noise. The BEAMS technology uses proprietary signal processing and deep machine learning to memorize alarm tones based by equipment types. BEAMS firmware updates can be upgraded securely and wirelessly over the network without any human intervention.
Eight programmed BEAMS units were then installed in eight rooms on a ward and connected to the Hospital’s Wi-Fi network. Baseline data was collected for eight weeks on when and where alarms were recorded and how long it took for the alarms to be addressed by nursing staff. In the second phase, all eight units were connected to the existing nurse call system with no changes to existing infrastructure. Again, data was collected for an additional eight weeks to record the improvement in the response times to alarms in each room now that BEAMS could trigger a nurse call.
BEAMS has a default 40 second latency built in before any action is taken to minimize the alarm fatigue caused by spurious or accidental equipment alarms that occur from time to time.
The BEAMS dashboard captures and stores the meta-data recorded for each event which includes: unit serial number, date & timestamp, alarm type, duration and location. BEAMS devices listen for patient alarms and then securely record them over Wi-Fi to a repository stored either in the Cloud or on a local encrypted server within the customer’s chain of custody.