In a time of inflation, talent shortages, supply chain challenges and shrinking capital, healthcare decision-makers must be more adept than ever at stretching dollars. And those headwinds are spurring some health system leaders to blaze new trails with emergency medical service providers in the world of artificial intelligence.

"AI and predictive modeling are driving everything we do, and those tools are helping our hospital partners to significantly increase their patient throughput," said Larry Richardson, a former paramedic and chief business officer at AmeriPro Health, a 911 EMS and medical transportation provider based in Atlanta. AmeriPro operates across nine states and has grown more than 100% year over year since launching in 2018.

"We're a data company that happens to run ambulances," he continued. "We know where we need to be and have our resources lined up long before we need to be there, creating much higher odds that we can optimize the level of pre-hospital advanced treatment."

THE PROBLEM

The problem one AmeriPro Health hospital partner faced was a combination of familiar headaches almost every healthcare CEO in the country has had to deal with over the last few years: overcrowding, bottlenecks that were slowing patient throughput, excess dead times with underutilized units, and the lack of readily available and actionable real-time data needed to overcome these challenges, Richardson explained.

"It was a perfect storm of cascading events hampering daily operations and threatening to restrain long-term growth prospects," he said. "Those were the immediate problems inside the facility, but there was also a ripple effect that extended to the broader community.

"It was not uncommon to see a line of our ambulances outside waiting for several hours due to bed shortages," he continued. "Those units were now temporarily out of service, unable to respond to other emergencies that may be occurring in the surrounding area. That's just one example that demonstrates why optimizing patient flow is so critical."

In healthcare, every second counts, he added.

"You can have a team of world-class professionals, the latest cutting-edge technology, and a facility that's second to none in your market, but if you don't have an effective logistics network that ensures patients are getting the right care in the right place at the right moment, then you'll never be able to solve your patient flow problem," he said. "Thankfully, this health system was ahead of the curve in identifying these challenges and showed the necessary resolve to quickly address them."

PROPOSAL

AmeriPro Health's proprietary AI platform would create and develop predictive models that enable it to identify the recurring bottlenecks and provide clarity for decision-makers on where they would need to go and fill in the gaps.

"If you understand what's coming ahead of time with daily patient flow, equipment demand, peak utilization times and staffing needs, then you will inevitably be able to reduce the burden on doctors and nurses while also improving patient outcomes and overall satisfaction," Richardson stated. "Our AI system would allow us to withdraw real-time and historical hospital data to develop an initial algorithm that provides direction within minutes.

"Everything centered around fixing the bottlenecks and creating a cohesive, uniform strategy," he noted. "The hospital's C-suite had one vision of what patient throughput should be. Mid-level managers and directors had another way of looking at it. So, we had to align all perspectives into a working game plan that tied everything together based on the data being collected."

As time goes on, the predictive value that was generated would be further optimized by the ever-widening pool of real-time data coming in, enabling patient flow managers to adjust accordingly and adapt much faster to sudden-changing circumstances, he added.

MEETING THE CHALLENGE

The technology AmeriPro Health deployed on behalf of this health system represented the next chapter for the AI program AmeriPro Health already was using within its internal network.

"As an EMS transport provider, patient care is at the heart of all we do," Richardson said. "But we're also a data company. Through our AI platform, we know where we need to be and have our resources lined up long before we need to be there, creating much higher odds that we can optimize the level of pre-hospital advanced treatment.

"Building on that foundation of success, the next logical step was engaging one of our larger hospital partners, who quickly signed off on the pilot program after we explained to them how this same technology could be used for their own benefit," he continued.

"Our AI tool is a layer. We extract the raw data, run it through a Microsoft platform, and then inject it into AI to give us a predictive model, which provides all of the guidance you need on daily volume/peak times, recommended staffing, and potential bottlenecks, before they actually occur."

Once the data came in, EMS staff made recommendations on how the hospital could pivot to help optimize patient flow. For example, EMS staff used an online-ordering profile through their computer-generated dispatch that's allowing the hospital to drive down its transportation costs by transferring accurate data back and forth between AmeriPro and the hospital's EHR system.

"The data also revealed how the hospital could modify and find new efficiencies for its discharge procedure," Richardson noted. "With the hospital's blessing, AmeriPro Health went into the facility and staffed a discharge lounge where we could monitor patients and then facilitate all of the transfers from the lounge so hospital staff can spend their time and attention on filling beds for other patients.

"As our AI program evolves, our goal is to be able to go to a hospital, layer this information on top of their EHR, and drill down to specific logistics needs and acuity behavior," he said. "If someone is experiencing chest pains, we can use the data to predict how long they'll need to stay and ultimately when they will get discharged and where they need to go."

RESULTS

Within a few months of the AI program being implemented, AmeriPro shaved a full day off of the hospital's average length of stay.

"Those results led decision makers to quickly replicate the program across the other hospitals in its system, which will create a projected 4,700 additional admissions per year across the network," Richardson reported. "We're currently developing strategies for the other campuses and will roll them out in the coming months.

"We also decreased our ambulance response times drastically and with the staffing crisis our technology is helping us significantly optimize our resources to our maximum potential," he continued. "Prior to our program being implemented, the majority of our hospital partners' discharges were occurring between 3 p.m. and 7 p.m. Now, 11 to 12 months later, the ambulance discharges were done by 10:30 a.m."

The math there is very simple: More capacity in the ED earlier in the day, allowing a greater volume of patient inflow, he added. This also directly resulted in increased coverage for the community with more ambulances providing better access to healthcare rather than holding on a wall at the hospital, he said.

ADVICE FOR OTHERS

"Data is the currency of truth," Richardson advised. "So, it's imperative for healthcare leaders to align themselves with a logistics partner who is on top of the technology trends, understands what the hospital's mission is and how to effectively leverage AI so resources and assets can be fully optimized.

"One of the key lessons learned through this experience is how important it is for an ambulance provider to have a seat at the table when decisions are being made on how to move patients," he continued. "The potential savings in dollars and time are simply too great to ignore. We're not a vendor, we're a strategic partner in logistics."

AmeriPro Health was able to add value for its hospital partner, in this case, because the hospital embraced the technology and was able to see this issue through the lens of hospital operations being a shared responsibility that cuts across multiple parties – all working toward a common goal.

"There will always be uncertainty in our industry, but by using AI and other tech tools to better predict what may happen next, health systems will be better positioned to adapt, and thrive, no matter the crisis or sudden shift that will inevitably arise," he concluded.

Follow Bill's HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.