OrbitaASSIST AI-Powered Bedside Virtual Assistant: Interview with Nick White, Orbita Executive VP of Patient Care

Boston-headquartered Orbita, a healthcare innovation company offering conversational AI platform and virtual health assistant solutions, has released the OrbitaASSIST, a voice-powered, AI-driven bedside virtual assistant. Patients use a smart speaker to make requests, and the system intelligently triages and sends requests to various care team members.

OrbitaASSIST can be used as a replacement for or adjunct to existing nurse call systems. Patients often find it frustrating to use these traditional systems to communicate with care staff who are nearly always busy and need to focus on multiple priorities. The solution prioritizes requests based on their urgency or nature. Each request is routed directly to the care team member who is the most appropriate person to respond.

The system was tested over an 18-month period in pilots at various care environments including a leading teaching hospital in Sydney, Australia. In the 8,000 patient requests made through the system, Orbita reports that the median response time to patient calls for help fell by as much as 70%. Research also showed a reduction in complications such as patient falls. Additionally, users gave the system high ratings, with nurses feeling more confident in responding to patient requests, and patients reporting they would like to use the system again in the future.

Medgadget had the
opportunity to talk to Nick White, Orbita
Executive Vice President of Patient Care Solutions and OrbitaASSIST co-creator

Conn Hastings, Medgadget: Please give us an overview of the limitations of current nurse call systems.

Nick
White, Orbita:
Within
care facilities, there is growing desire to provide patients faster and better
access to information, care, and services. These organizations have an
incentive to ensure that each patient’s stay is not only of the highest quality,
but also delivers each patient an individualized experience that is as pleasant
and as cost-effective as possible. 

OrbitaASSIST was developed to
address a significant challenge facing in-facility care settings where patients
struggle to communicate their needs to busy staff.

Traditional
nurse call systems (including push buttons and pull strings) do not effectively
prioritize frequent and varying requests made by patients, let alone provide
reassurance that someone has heard and will respond to their need. These outdated,
binary systems fail to adequately support patients and their care teams,
resulting in reduced patient satisfaction, inefficiency of clinical staff and
care delivery, and higher risk of unintended outcomes.

Two-way
communication systems, such as the addition of a speaker to a call button, allow
the patient to convey a message by calling and when answered speaking, most
often to a care team member located at a central nurse station. This requires a
person to answer the call, and they may be tasked with multiple activities and not
be in a position to get a care team member to respond to the patient in a
timely fashion. As a “voice” system, this offers some improvements, however it does
not deliver the full range of benefits of a voice-powered, AI-driven virtual
health assistant and digital workflow at the bedside.

Medgadget: How does OrbitaASSIST improve on these existing systems, in terms of enhancing patient safety and well-being?

Nick
White:
Patient
experience scores are critical to a hospital’s success. In organizations using OrbitaASSIST,
we’ve seen an uplift in patient experience scores. We’ve also seen improved
nurse response times. Why is this important? Previous studies have linked
released time to care for nurses with reductions in mortality rates and patient
lengths of stay as well as nurse burn-out rates. We are also targeting improvements
in the number of hospital acquired complications, such as falls and pressure
sores.

With OrbitaASSIST, a patient
expresses a need via a request made through a smart speaker by simply
speaking; for example, “Ask the nurse for a blanket,” or “Tell the nurse I have
fallen.”  

Artificial intelligence (AI) and machine learning
operate on the backend to triage, prioritize, and intelligently route requests
to appropriate care team members who view requests on mobile, tablet, or
desktop devices. The patient hears a natural language response such as “someone
will bring you a blanket soon,” or “we’ve raised an alarm with the nurse.”

The request is routed
to specific care team members who receive a notification on a smart device,
which includes the nature of the request and suggested response time, allowing
them to prioritize their actions before responding. The concept is to have
the person most suited to respond to the request alerted and able to respond in
a timeframe that is most appropriate (i.e.; a fall needs an immediate response
by a clinician vs. a request for a pillow which can be fulfilled by a
non-clinical team member with less urgency.)

Amazing stories have come
out of our pilot sites including a spinal ward at the Prince of Wales Hospital.
There, patients saw reduced response times when requesting help for pain and the
facility reported reduced risks of falling.

OrbitaASSIST can
offer entertainment features such as music, news or podcasts. With this, we’ve seen
improvements in patient moods as we seek to as to combat the loneliness and
isolation that arises with lengthy hospitalizations for traumatic injuries.

Measurable improvements and highly positive feedback have come from patients, nurses and other care team members using
OrbitaASSIST. Here’s a list.

  • Patient response times are up to
    70% lower when compared with industry benchmarks.
  • Patients complaints about nurse
    response times dropped significantly and, in some instances, were eliminated
    (100% of patients surveyed said they’d like to use it if hospitalized again).
  • Nurses find the system extremely easy to use — and 87% of nurses
    report feeling more confident in responding to patient requests. 
  • Care facilities saw a reduction in
    falls and other hospital acquired complications.
  • Patients share that the
    entertainment aspects of the device transformed their stay in hospital 

Medgadget: Does the system require training to recognize the requests of specific patients, or identify the most appropriate staff member to deal with each request? How do in the in-built AI and machine learning systems function in the system?

Nick
White:
To help get this innovation to market and start
making a difference for patients and nurses, the first evolution of
OrbitaASSIST is built on a technology stack that includes Amazon (AWS and
Alexa), the Orbita conversational AI platform, and the ServiceNow to enable the
digital workflow.

OrbitaASSIST benefits
from the inherent capabilities of these vendors to ensure a secure,
healthcare-centric, enterprise-grade platform optimized for maximum
extensibility and seamless integration capabilities
with existing
infrastructures.

In designing OrbitaASSIST,
we’ve taken into account the feedback from a team of nurses with hundreds
of years of combined nursing experience. Their input has been critical to the
design of the conversational capabilities and natural language processing.

In additional, working with the care teams has helped us to define
key areas of flexibility that hospitals need to support different modes of care
within their facilities. Each unit has the ability to modify the business rules
such as level of prioritization, suggested response times, escalation and
routing, enabling them to configure OrbitaASSIST to match how their care system
runs.

The OrbitaASSIST team has
worked closely with our technology partners to fine-tune the solution to
optimize the accuracy of voice recognition. We set a critically high bar in lab
tests so that we would meet the needs of patients and care teams in a clinical
setting.

OrbitaASSIST has
received 27 awards including a Good Design Award, the national iAward for
Research and Development, a 2019 Webby Technical Achievement in Voice, a Gold
Innovation Spike, the AMY’s Trail Blazing Product of the Year, and a Prime
Award for Excellence in Patient Support.

Medgadget: Do you think that OrbitaASSIST could be used for residents in their own homes, to request something of their caregiver or alert their clinician to an emergency?

Nick
White:
We are currently working with community-based care
teams to have OrbitaASSIST tailored for a residential setting. We also have
another product that is successfully deployed into the home environment called
OrbitaCONNECT. OrbitaCONNECT can be used to support management of individual
care pathways, and the power of the two solutions combined will change the face
of care at home.

Today, OrbitaASSIST is live
in hospitals and aged care facilities, focused on supporting patients and
nurses. However, the potential for support in the home is clear, and as our
clients look to optimize care delivery and adherence in the home, the use of
voice enabled digital workflow and the broader set of Orbita solutions have a
significant role to play.

Medgadget: When and where will the system roll out? Do you see this type of technology being commonplace in the future?

Nick White: The system has been tested and is now being used
successfully at a number of sites in Australia. Orbita is working closely with several additional customers
looking to implement
conversational AI solutions in the global market. Orbita will support these clients as they move forward with
their offerings by accelerating the development of enhanced features that
enable them to capture greater value as they scale those deployments.  

Because it is now offered by Orbita rather than Deloitte, the OrbitaASSIST solution is available for
pilots not only in Australia
and New Zealand, but also North America and Europe.

We believe this technology will be commonplace in the future not only in hospitals, but also in skilled care facilities, senior living environments, and community-based care organizations.

Link: Orbita homepage…