Sensing Falls: Predictive AI Helps Nurses Reclaim Time at Bedside

The following is a guest article by Deepak Gaddipati, Founder & Chief Technology Officer at VirtuSense.

Dowsing rods find underground water. Achy joints forecast bad weather. Upturned leaves predict a storm. Just as these natural guides indicate events, artificial intelligence (AI) powered sensing technology is now used in healthcare to predict patient falls. In doing so, the technology relieves nursing observation burdens, frees up time for patient care, and reduces risk of falls with injury by up to 96 percent.

Patient fall prevention has always been a measurable quality and safety goal for healthcare provider organizations. According to the World Health Organization, prevention strategies should emphasize education, training, creating safer environments, prioritizing fall-related research, and establishing effective policies to reduce risk. But amid the pandemic, rising nursing shortages, and burnout, AI-powered sensing technology for fall prevention must be implemented alongside traditional strategies to relieve nursing burdens while maintaining quality patient observation and care.

Traditional fall prevention methods have limits when it comes to eliminating falls. For example, many organizations employ virtual sitters who observe patients in their room via a video feed. This option only watches for falls versus truly preventing them, and could cost upwards of $1.75 million annually to monitor 100-beds.

This article presents valuable rationale for the use of AI-powered sensing technology for acute-care fall prevention to free up nursing time and reduce healthcare costs.

Innovation in Patient Observation

“Observation is one of nursing’s most important roles alongside documentation, clinical care, and the patient relationship,” according to Thomas Hale, MD, PhD, MS, Chief Medical Officer, VirtuSense. However, there have been challenges with using technology for efficient and effective patient observation. Timeliness is one of these challenges.

Telesitters remotely observe 10 or more rooms at a time and are alerted when a fall is occurring, but by the time staff arrive the patient may already be on the floor. Alarm fatigue is another important concern that has hindered the use of traditional observation technology.

Research indicates that 72 to 99 percent of all hospital alarms are false leading to slow acute care nursing response and patient falls. For example, sensing bed pads have an incredibly high rate of false alarms because the sensor can’t differentiate between a patient shifting in bed or attempting to get up.

More can be done to provide nurses with reliable technology that accurately predicts patient falls before they happen, thereby relieving at least one aspect of the nursing workload. For example, pairing AI technology with infrared sensors and movement data provides a solution that effectively monitors patients without the intrusiveness of a camera in the room.

  • AI sensors trained on millions of hours of data can detect the difference between someone shifting in their bed or intending to exit the bed with 99% accuracy.
  • Alerts are routed to a central console as well as designated smartphones 30–65 seconds before a bed/chair exit occurs for high fall risk patients.
  • AI sensors and analysis are so accurate, nurses aren’t responding to false alarms.

Real World Scenario in Sensing Falls

“Nursing staff take rapid and proactive action to prevent patient falls when sensing technology picks up these imperceptible changes and applies AI algorithms. Being able to compute the data at a repeatability that makes the information useful also produces better information for clinical decision-making,” adds Dr. Hale. For example, one academic medical center reduced falls on a single nursing unit from 4.56 to 0.36 per 1000 patient days. Falls with major injury dropped from just under one fall per month to zero. Monthly spending on fall prevention for the unit also plummeted, by 66 percent.

These results demonstrate the importance of proactive alerting, and how it can improve patient safety while making acute care nursing teams more efficient and effective.

Making a Positive Impact on Nurses Today

The healthcare industry continues to endure a nursing shortage that has been fueled by the COVID-19 pandemic. Many senior nurses have retired, burnout runs rampant, and the cost for travelling nurses has exploded. These challenges intensify as diagnoses, medications, and patient acuity become more complex in the acute care setting.

The only way to address nursing workload burdens and make a positive impact on nursing satisfaction is through technology advancements.

Preventing patient falls is a major safety concern. However, it is a serious challenge amid the current nursing climate. AI-powered sensing technology for fall prevention takes this concern off the nursing team’s shoulders while safeguarding the patient and supporting proactive healthcare.

About Deepak Gaddipati

Deepak Gaddipati, Founder and Chief Technology Officer at VirtuSense is a technology visionary whose passion is to fix healthcare. He acquired peerless expertise in machine vision, deep-learning, and “IoT” while developing the first commercial full-body, automated scanning system that is widely deployed across most U.S. airports. Deepak’s mission behind VirtuSense is personal: in 2009, his grandmother fell while walking to the bank and broke her hip. She died within ten days from the injury. Leveraging his artificial intelligence and machine learning knowledge, he founded VirtuSense to develop a tool that could proactively identify fall-risk in older adults.

 

   

Categories