Like every health system, Allina Health based in Minneapolis, Minnesota, relies on surgical procedures as a core financial driver.
In 2018, as part of an overall health system initiative to improve patient flow and increase capacity, Allina sought to optimize its surgical services by eliminating as much waste and inefficiency in every surgical process – from the time surgery was indicated in clinic through the day of surgery and discharge from PACU.
“Our hypothesis was that there were surgical cases we were losing as a result of the flow of an operative day being inconsistent and unpredictable, so we wanted to utilize standard work across our ORs to ensure our patients and surgeon experienced as predictable surgery day as possible,” explained Bill Evans, vice president of surgical services and orthopedics at Allina Health.
“A key part of reimagining surgical services was to create a system-wide surgery scheduling policy to fill OR time more effectively,” he continued. “However, even after standardizing our scheduling processes to maximize OR utilization, we still believed significant opportunity for improvement existed.”
For example, surgeons looking for slots outside their assigned block times, as well as newer surgeons seeking to establish enough volume to justify a block assignment, were hampered by manual scheduling processes that hadn’t materially changed in 30 years. Those processes also frustrated both OR and clinic schedulers, leading to low job satisfaction.
“After improving every part of the process chain and addressing OR utilization through policy, we remained confident that we had not solved the access problems our surgeons were telling us they were experiencing,” Evans noted.
“Our robot surgeons, in particular, were telling us they could not get access despite a low utilization rate,” he continued. “It became clear we needed a new approach to surgery scheduling, and we believed we needed a solution that was purpose-built to address the unique challenges of OR scheduling – and designed around surgeons.”
Allina Health simply could not afford to wait for its EHR vendor or other IT systems to eventually roll something like this out down the road, he added.
Before making any proposal, Allina Health worked with health IT vendor Qventus to carefully study its situation, system, and current and future needs.
“They met with all stakeholders to be sure they understood the pain points and challenges,” Evans recalled. “Ultimately, we selected Qventus to enable us to automate our operating room scheduling and maximize our surgical program with the support of machine learning AI.
“The perioperative solution Qventus proposed would automate and make uniform OR scheduling and yield greater efficiencies, while simultaneously relieving the burden on overworked staff,” he continued. “The Qventus Perioperative Solution automates every step of the scheduling process and eliminates the manual processes that cause scheduling bottlenecks and mismatches.”
The system’s AI-powered software combines pattern recognition and predictive capabilities with principles of behavioral science to address Allina’s serious scheduling problems.
There were a few key differences between Qventus’ technology and that of other vendors that stood out to Evans and his team.
“Predictions alone don’t drive value, they have to drive action,” he said. “That’s where Qventus applies their behavioral science expertise. When they nudge block owners for early release, they show them ‘What’s in it for me.’
“Surgeons don’t lose much by releasing time early, but their upside is improving their block utilization – and priority for being notified of available time in the future,” he continued. “Furthermore, the system doesn’t just predict that a block will be merely underutilized; it actually predicts specific unused times so that it’s easy for block owners to act on.”
Allina Health also was looking to grow case volumes.
“Some solutions use machine learning to ‘overbook’ their ORs to fit in more cases, which introduces significant risk and frustration,” Evans contended. “Moreover, other tools simply bombard practices with open time alerts when time opens up.
“We wanted to build a case mix that helps achieve our strategic goals: optimize site of care, grow strategic service lines, build relationships with key practices, etc.,” he continued. “The machine learning helps us attract cases that align with our priorities.”
When practices – including independents – search for time, they personalize how open time results are ranked based on practice patterns and health system priorities, just like how Netflix learns viewing patterns and makes recommendations.
Allina Health’s system doesn’t just wait for cases to come in. When time opens up, the machine learning matches available times to surgeons who would be a strong fit and proactively offers them time, much like how Amazon sends a product recommendation based on purchase patterns.
“In short, the ‘push/pull’ capabilities I described, on an automated, machine learning platform, was exactly what we were looking for to help solve our access issues,” Evans said.
MEETING THE CHALLENGE
Allina Health phased in the system over several months at its flagship hospital, Abbott Northwestern, starting with its robotic surgeons, who had been most vocal in their complaints about access issues.
The second phase included all block holders, and the third phase represented all surgeons who had privileges at the hospital but did not hold block time.
“We chose to make usage of the system voluntary rather than mandatory, and intentionally positioned it as a tool to help the surgeons, even though its use will also enable us to optimize and grow our surgical services,” Evans explained. “Currently, more than two of three elective cases are scheduled through Qventus.
“Our phased approach was highly successful from the beginning,” he continued. “In the first days, we were already exceeding our performance goals, and the tool was so intuitive we started seeing cases being added on from surgeons who weren’t planned for Phase I but were seeing how it could help their practice.”
Once staff worked with all block holders in Phase II, they were thrilled with the results in terms of surgeon uptake, but also getting unneeded block time freed up sooner so other surgeons could use it.
Allina Health’s results since launching Qventus in mid-2022 have been impressive:
“We have added 3.5 cases per OR per month, on average,” Evans reported. “The additional cases means greater revenue, shorter waiting lists for patients and busier and more satisfied surgeons.
“We’ve had a 36% increase in cases per surgical robot per month,” he continued. “Previously, it was difficult to schedule robotic ORs efficiently with the result that a robot-equipped OR was not always available when needed and non-robotic procedures were sometimes performed in robot-equipped rooms.”
100-plus hours of OR block time have been released early per month. This is of great benefit because the earlier the time is released, the more likely it is that it can be filled by another surgeon. The Perioperative Solution automatically prompts surgeons to release blocks of time they are unlikely to use, which creates greater efficiencies throughout the system.
“Two of three elective cases are automatically scheduled via the Perioperative Solution,” Evans noted. “This level of automation relieves the burden on our schedulers and on surgeons’ schedulers. It frees them to perform more important tasks.
“Anecdotally, we are hearing from our clinics that we have made it so easy to schedule that we are the default solution they always use first,” he added.
The Perioperative Solution has exceeded Allina Health’s expectations from a quantitative results perspective from the very beginning of its use.
“But also vitally important, it has provided our scheduling teams with an innovative tool that makes their jobs easier,” Evans said. “It frees up time for them to focus on more complex scheduling cases and providing higher touch service for our teams, rather than requiring mundane low-value tasks, such as dealing with back-and-forth phone calls and faxes.
“I am convinced we will see additional unplanned savings as a result of reduced turnover in our scheduling staff,” he added.
ADVICE FOR OTHERS
“Automation powered by artificial intelligence and machine learning is the future of healthcare,” Evans stated. “It’s the obvious and necessary answer to staffing shortages, rising labor costs and shrinking surgical revenues.
“Many health systems are waiting for their EHR or other IT systems to rollout these types of solutions,” he concluded. “However, waiting comes at a significant opportunity cost. These automation solutions have rapid ROI and are purpose-built to solve these unique problems; the EHR and other systems are not.”