Skip to main content

AdventHealth Uses AI to Balance Capacity Issues Among 9 Hospitals

Analysis  |  By Mandy Roth  
   October 15, 2019

Use of command center AI technology and predictive analytics accelerate patient care and nearly eliminate the need to turn away patients due to capacity issues.

During 2017 nine hospitals in AdventHealth's Central Florida Division collectively turned away 526 patients due to lack of capacity. During the first six months of this year, that number plummeted to just two per month.

The hospitals have not gained additional capacity, nor has demand declined; they now operate more efficiently with real-time performance data visible across the enterprise thanks to AdventHealth's Mission Control command center.

Fueled by AI and predictive analytics, which amass data from multiple sources, a team of more than 50 nurses, EMS and flight dispatchers, transport techs, and other specialists work in the 12,000-square-foot facility with 60 monitors to oversee activities on nine campuses with 3,200 beds and more than 2 million patient visits per year.

Command centers operating at this scale are rare, says Tanya Egbert, senior director of research and strategy for KLAS, the Orem, Utah­­–based research and insights firm focused on healthcare IT. Yet that's changing. Health system executives are keeping an eye on the performance of early adopters, she says, as they document the benefits of greater efficiency, including cost savings, reduced lengths of stay, and the opportunity to improve the patient experience.

Related: AI and Predictive Analytics Enable Tampa General to Save $10 Million, Reduce LOS

AdventHealth's Mission Control executives share an early glimpse into how this type of command center benefits a multihospital system.

The Growing Phenomenon of Command Centers
 

The term "command center" encompasses a range of solutions. In December last year KLAS produced Operational Command Centers 2018, a report about the relatively new phenomenon of command centers, which it characterized as an "emerging market."

The report, which collected data over the previous 12 months, estimates that the vendors operating in this arena have installed less than 170 such solutions, and nearly all involve much smaller scale implementations of patient flow or transfer center software.

Although KLAS did not classify command centers by size, it appears from news reports that perhaps less than a dozen U.S. hospitals or health systems have built NASA-style facilities. "A very low percentage are operating in that top tier," says Egbert, one of the report's authors.

Patients Turned Away Due to Perceived Capacity Issues
 

During 2017, the nine Central Florida hospitals now participating in the Mission Control initiative were successfully operating individually, says Sanjay Pattani, MD, associate chief medical officer for AdventHealth Orlando and executive medical director for Mission Control. However, the system lacked streamlined processes, coordinated functions, or data visibility across the enterprise.

"A lot of processes were manual," says Pattani, "and we relied on phone calls to coordinate care amongst hospitals. Information was not readily accessible or available regarding system capacity. By the time you collected the information, it was already outdated."

Numerous issues were identified, says Penny Porteous, executive director, Mission Control. Some patients were unable to be admitted, nearly half who were admitted through the emergency department had to wait more than an hour for a bed, and once admitted, some experienced delays in treatment.

  • 526 patients were turned away in 2017 "because we did not have the capacity to accommodate them," says Porteous.
     
  • 48% of patients in the emergency department had to wait more than 60 minutes for a transfer to an inpatient bed.
     
  • 56 patients in the system experienced an "excess day" due to inefficiencies in processes. Waiting on a bed, a procedure, or a physician caused delays in treatment or discharge, Porteous explains.
     
  • Utilization at facilities throughout the system varied from 63% to 95%. When an individual hospital reaches 80%­–85% capacity, inefficiencies being to occur, says Porteous. Yet while one hospital in the system might have had capacity issues, another might have had beds available. Lack of visibility across the enterprise made it difficult to conduct transfers between facilities and accelerate patient care, she says.

Technology Harnesses Data to Actively Manage Beds
 

AdventHealth joined forces with GE Healthcare Partners in 2017 to help create the command center. Mission Control went live during November 2018 and moved into its new NASA-style facility on the AdventHealth Orlando campus this past August.

"The concept behind forming a command center was bolstered on the idea that we needed to become more proactive versus our historic reactive nature," says Pattani. Instead of admitting patients to hospitals operating over capacity, "we wanted to actively manage these beds." To do that, AdventHealth decided to employ AI and predictive analytics to harness the massive amount of data generated by the health system.

The solution also uses high-velocity data fusion, a form of AI that processes hundreds of messages each minute—every time something is entered into an EMR or software system connected to Mission Control, explains Jeff Terry, CEO of clinical command centers, GE Healthcare Partners.

'Systems Synchronicity'
 

"A centralized mission control coordinates and establishes what we call 'systems synchronicity,' " says Pattani. "When we have multiple emergencies across our system, our digital sophistication enables us to coordinate [care for these events] simultaneously and in real time by connecting our providers with the right resources to take care of patients in a time-sensitive manner."

This effort has proved to be particularly helpful for the busiest hospital in the system, AdventHealth Orlando. The ED is now able to offer patients a bed at a different location within the system to minimize delays in care.

AI functionality predicts potential stress points and potential high occupancy rates. These forecasts help determine appropriate scheduling for clinical and support staff, as well as needs for other services, including radiology and environmental services.

"AI helps us predict not only where we're going to need help, but exactly what we're going to need to address it," says Pattani.

Fewer Patients Declined; Greater System Efficiencies
 

Mission Control has "streamlined processes and … connected our hospitals so that we're now a lot more aligned and we can view and transition patients across our system," says Pattani.

While officials say that the operation is in its infancy, they have already experienced significant results. AdventHealth shared the following figures with HealthLeaders: [Editor's note: These figures compare the first six months of 2019 against the health system's 2018 performance—not the 2017 figures that prompted AdventHealth to explore a command center solution.]

  • The number of patients turned away each month declined from 17 a month in 2018 to two per month for the first six months of 2019.
     
  • For patients being admitted through the ED, the number of clinical reviews for necessity of care increased from 40% in 2018 to 80% in 2019 due to enhanced efficiencies, says Porteous.
     
  • When comparing Q1 2018 to Q1 2019, Mission Control enabled AdventHealth to double the number of patients transferred from the ED at a hospital with high-demand beds to a facility with greater capacity. These "overcapacity lateral transfers" increased from 120 during the first quarter of 2018 to 241 during the same period in 2019, says Porteous.
     
  • Finally, another performance metric, measuring the time it took from receiving an order to admit an ED patient to the time an available bed was identified, decreased from an average of 91 to 30 minutes. Porteous points out that this measure only involved three of the hospitals.

Reducing Length of Stay and Delivering a More Streamlined Patient Experience
 

"We're now able to visualize in real time what's going on across the system," says Pattani, "but more importantly, we're also able to see what's not happening across the system." In addition to accelerating admissions, the technology also helps identify delays in care for patients who are already in the nine hospitals, he explains. Focusing on this data will escalate care delivery, transition them to appropriate levels of care, and reduce length of stay, he says.

"While patients may not be exposed to the digital sophistication within the command center, they will feel the impact it has on delivering a more streamlined, seamless patient-centered care experience," says Pattani.

Because digital technology is constantly changing, it stimulates continual change for the organizations that use it, says Pattani. Therefore, the capacities of Mission Control will continue to evolve, he says.

"We may not know exactly what we're going to be morphing into," says Pattani, "but we are sure that we're going to be guided by the needs of our community and our patients."

“AI helps us predict not only where we're going to need help, but exactly what we're going to need to address it.”

Mandy Roth is the innovations editor at HealthLeaders.

Photo credit: Courtesy GE Healthcare Partners


KEY TAKEAWAYS

Before AdventHealth's command center, Mission Control, launched in 2018, more than 500 patients were turned away from the health system and others experienced delays in care.
 

While some hospitals had capacity, lack of real-time data visibility across the enterprise made patient transfers difficult.
 

Today, only two patients a month are turned away and the time between receiving orders to admit a patient and finding a bed has dropped from about 90 minutes to 30 minutes


Get the latest on healthcare leadership in your inbox.