Waste Not, Want Not
Tech Projects Puts U S Healthcare Inefficiencies under the Knife
BY NOVID PARSI
PORTRAITS BY TODD WINTERS
Tina Esposito, Advocate Aurora Health, Chicago, Illinois, USA
As if the high cost of private healthcare in the United States wasn't bad enough, much of that expense is avoidable. According to a 2019 study in the Journal of the American Medical Association, about 25 percent of the country's healthcare spending is wasteful. That waste tallies at least US$760 billion per year.
“U.S. healthcare has problems around access, cost and quality—the three things that patients look for,” says Peter Chang, MD, vice president of care transitions, Tampa General Hospital, Tampa, Florida, USA.
—Peter Chang, MD, Tampa General Hospital, Tampa, Florida, USA
Yet some U.S. healthcare organizations are proving that IT projects can help them find efficiencies and deliver better care.
After implementing a cloud-based medication management tool in 2018, LIFE St. Mary, a program in Trevose, Pennsylvania, saw costly emergency department (ED) visits decrease by 16 percent and total hospitalizations decrease by 27 percent. Providers and pharmacists use the tool, called MedWise, to review the medications of high-risk patients and prevent adverse drug events that could send them to the hospital.
In Winston-Salem, North Carolina, Infinite Leap, a healthcare-technology provider, recently implemented various real-time location services (RTLS) projects at Wake Forest Baptist Health. These solutions help Wake Forest Baptist Health's staff locate clinical equipment and use it efficiently. In 2019, Infinite Leap launched a project that provided all inpatients with RTLS badges, allowing staff to better track patient location and status, and to optimize workflows, like patient rounding and discharge processes.
Wake Forest Baptist Health executives estimate that all completed RTLS projects save the health systems US$3.5 million annually and reduce patient wait times by up to 50 percent. But in order for these IT initiatives to deliver genuine benefits, project managers have to bridge the interests of tech specialists and healthcare professionals while making sure medical practitioners embrace change.
Healthcare IT projects cannot deliver solutions for their own sake. They must cut costs and improve care, which means that data specialists and clinicians have to work together to figure out what the projects must do and how they can best do it.
“Analytics have unbelievable potential to provide insight, but they're only as good as the partnership between the data people and the clinical people,” says Tina Esposito, chief health information officer, Advocate Aurora Health, Chicago, Illinois.
—Tina Esposito, Advocate Aurora Health, Chicago, Illinois, USA
TAMPA GENERAL HOSPITAL
That was a lesson Advocate Aurora Health learned early on during a project to develop a predictive analytics tool. The healthcare system's leaders wanted to determine the patients most at risk of having to return to the ED so that its care managers could call them and make sure they were taking care of themselves. Historically, the healthcare system tried to make that determination by looking at the patient populations that had cost the most and had the most ED trips in the past.
The problem was that the past didn't predict the future. The patients who had gone to the hospital the most weren't necessarily the ones who would use it the most in the future. “We had to predict the patients who would be the highest cost, rather than relying on a retrospective look at who was the highest cost,” says Tim Arnold, manager of Advocate Cerner Collaborative at Cerner, a healthcare IT services provider. Mr. Arnold lives in Chicago.
The team asked its healthcare providers for the types of patients that would most benefit from the intervention calls from care managers—so the data specialists could create a predictive tool that would identify them. The clinicians came up with about half a dozen such conditions, including asthma and dementia, and for its pilot project, the team focused on heart failure. The predictive model pulled hundreds of variables to determine the heart failure patients who had the highest risk of going to the hospital within 90 days.
The clinicians and data scientists collaborated closely through regular, formal check-ins and more casual conversations. When the data analysts found that about 20 percent of patients who returned to the hospital did so within 10 days of discharge and 30 percent within two weeks, the clinicians realized the phone-call interventions had to happen as quickly as possible.
Before the project, care managers contacted patients about two weeks after identifying them; now, they contact them just four or five days after they've been identified by the predictive model.
Completed in 2018, the pilot project for the heart-failure cohort resulted in a 23 percent reduction in hospital trips. And Advocate Aurora used the same cross-disciplinary team approach for subsequent predictive analytic initiatives, such as a 2019 project for patients with chronic obstructive pulmonary disease.
PHOTO COURTESY OF ADVOCATE HEALTH
“There's excitement in healthcare around big data and how it can predict diseases or mortality, but the question is: What do you do about that? If you can predict mortality, can you prevent or delay it?” Mr. Arnold says. “That question doesn't always get asked if you have only data scientists and they don't partner closely with the clinicians to find ways to improve the care of patients.”
—Tim Arnold, Cerner, Chicago, Illinois, USA
Healthcare IT projects must rely on interdisciplinary collaboration, says John Grdina, PMP, director of project management, Infinite Leap, Winston-Salem, North Carolina. “You have to deal with folks in clinical and IT departments, so you have to speak a project management language that both of these groups will understand.”
RX FOR CHANGE
On healthcare IT projects, the project management often involves change. “It's easy to come up with a project idea. It's hard to sustain the effort,” Dr. Chang says.
In 2018, Tampa General launched the CareComm project (see “Coordinated Care,” page 27), a command center that uses apps to provide real-time insights into how patients and providers use the facility, in part to maximize staffing efficiencies. “This requires extensive change management to show the stakeholders that what we're doing will save them time,” Dr. Chang says. “When you're asking people to change, you have to communicate that benefit out of the gate.”
The CareComm project team found an effective way to do that: illustrating the benefits the new tech would provide patients. During the user training sessions, the CareComm project team discussed a specific patient whose care was delayed as a result of the hospital's standard process. “Everyone says, ‘That's not acceptable; that's not what we want to deliver,’” Dr. Chang says. The team then demonstrated how that patient's care would improve with the new, more efficient process. “Patient stories are extremely valuable when trying to get people to change,” he says.
On Infinite Leap's RTLS projects, its team likewise shepherds end users through the process of adopting a major change. “Getting people to adapt to what we're doing can be a roadblock, but we have to get them to change their standard processes for the better,” Mr. Grdina says.
—John Grdina, PMP, Infinite Leap, Winston-Salem, North Carolina, USA
The team clears that roadblock by securing executive support from the start, in part by establishing an executive steering committee. “We can tell anyone this is the right project to do, but if we don't have an executive buy-in, it's hard to get the benefits realized,” Mr. Grdina says. “If we have the executive sponsors onboard, it's easier to trickle that support down to the end users.”
Like the CareComm team, Infinite Leap engages users through extensive training that demonstrates both how to use the technology and the benefits it provides. “We don't close a project until everyone who will touch the technology has been adequately trained in it,” Mr. Grdina says.
Having super users helps the adoption process, says Joanna Wyganowska, PMP, vice president of marketing, Infinite Leap, Fargo, North Dakota. Super users are the first to learn the new technology and then pass on that learning to their colleagues. “We have nurses who serve as super users so when other nurses have questions about the technology, it's easier for them to learn from someone in their own field,” she says.
Tabula Rasa conducts both group training sessions and one-on-one sessions, says Brian Adams, CFO, Tabula Rasa HealthCare, Moorestown, New Jersey. “Our implementation phase is heavy on training so that end users are familiar with how to navigate the platform and how to interpret its clinical information so they get the maximum benefit.” PM
Executives at Tampa General Hospital, a 1,000-bed community hospital and academic medical center, knew they had room to improve. “We saw a lot of inefficiencies in our system. There was a lack of coordination and automation around the delivery of healthcare,” says Peter Chang, MD, vice president of care transitions, Tampa General Hospital, Tampa, Florida, USA.
The problem is one that almost every healthcare facility faces. “A hospital is essentially a very large, expensive waiting room. Patients wait for test results, medications and physicians,” Dr. Chang says. “We wanted to take the time period between those events and compress them as much as possible.” Tampa General's emergency department sees about 300 patients a day. “So minutes and even seconds are extremely valuable,” he says.
Tampa General wanted to find IT solutions to achieve two main outcomes: “driving costs as low as they can be and improving quality and access for patients,” Dr. Chang says.
So in 2018, it launched the CareComm project. In partnership with GE Healthcare, Tampa General created a command center that uses tiles, or apps, that provide realtime information about how patients and providers use the facility. That information gets displayed on the center's 40 screens, and the center's workers collaborate with one another to identify efficiencies and communicate them to each other and to hospital staff. They ensure patients arrive and leave the hospital on time, without waiting unnecessarily for care.
For example, Tampa General has five helicopters it dispatches to trauma patients. In the past, making sure the hospital was ready for a trauma patient would require multiple calls among many staffers, resulting in delays. Now, as soon as a helicopter takes off, CareComm team members quickly access real-time information about the intensive care unit and its room availability so that, as soon as the helicopter lands, the patient can go directly to the available room and the waiting team.
With the imaging department's tile, the CareComm team wanted to reduce the amount of time it takes for inpatients to receive imaging. That tile's owner group identified a bottleneck: Some patients arrived at their scheduled MRI procedures without first having had the necessary IV access. So, per the user group's suggestion, the tile delivers an alert to the care team if a patient has been scheduled for an MRI without having IV access.
“We take the most efficient routes with the most efficient care teams to deliver the most efficient care,” Dr. Chang says. With eight screens and 10 team members, CareComm's temporary, 800-square-foot (74-square-meter) command center opened in December 2018, the same month that the team released its first tile. The permanent 8,000-square-foot (743-square-meter) center, with 40 screens and 32 team members, opened in August 2019. The project will deliver 20 tiles by the end of 2020. In its first year, the CareComm project saved 9,000 unnecessary days of patient stays by getting procedures done, and patients discharged, as soon as safely possible. “If we aim to decrease the cost of healthcare, we have to become more efficient,” Dr. Chang says.
Improving Outcomes Three U.S. healthcare organizations found efficiencies and value through IT.
Yale New Haven Health
Location: New Haven, Connecticut
In 2019, Yale New Haven Health finished a project to replace its legacy mainframe-based accounting system with a new platform that merges financial and clinical data. Now, healthcare leaders can gather and analyze data to drive better clinical outcomes and reduce excess costs.
Payoff: The organization saved an estimated US$150 million.
Location: Philadelphia, Pennsylvania
Rolled out in 2018, the Palliative Connect tool uses predictive analytics to inform providers about the patients who would benefit from palliative care consultations. It helps coordinate between a patient's primary care team and palliative specialists.
Payoff: The number of patients discharged to hospice nearly doubled, while inpatient mortality, 30-day readmissions and intensive care unit transfers all declined.
CLOCKWISE FROM TOP LEFT. PHOTOS COURTESY OF YALE NEW HAVEN HEALTH, PENN MEDICINE AND UCHEALTH
Location: Aurora, Colorado
UCHealth's predictive analytics project, launched in 2018, uses variables in patient outcomes before and after elective surgery along with patients’ electronic health records to draw connections between preoperative conditions and outcomes.
Payoff: The organization cut costs by operating more efficiently.