Healthcare Analytics, Population Health Management, Healthcare Big Data


Big Data Visualization Improves ED Patient Flow, Satisfaction

Workflow improvements and big data technology combine to produce "air traffic control for patients" at SwedishAmerican Hospital.

By Jennifer Bresnick

- “If you can’t measure it, you can’t manage it” is a familiar and powerful adage for data scientists and statisticians across many different industries, but it’s a message that the healthcare system is just starting to take to heart.  

Big data visualization and patient satisfaction

As industry reforms, changing payment models, and the increasing impact of patient satisfaction scores begin to put pressure on providers to become more efficient, the ability to analyze, visualize, and act upon insights from new sources of big data is becoming a critical competency.

Meaningful workflow adjustments and cultural change are not always easy to enact, however, and many hospitals struggle with developing the process improvement strategies that will give them the competitive edge they need to survive turbulent financial times.

But at SwedishAmerican Hospital in Illinois, a division of UW Health, change is a very good thing for patients and providers alike. A targeted initiative to improve patient flow between the emergency department and the inpatient setting has resulted in a significantly quicker discharge process, explain Chief Nursing Officer Ann Gantzer and Director of Emergency Services Chad Thompson.

By investing in big data visualization technology and some simple front-end process improvements, the hospital has been able to trim hours off the patient discharge process, improve patient satisfaction scores, and reduce the time it takes to get emergency patients admitted.

READ MORE: Healthcare Analytics Help Mercy Uncover Cost, Quality Insights

Defining the scope of the problem

“We were having some log jams in our emergency department,” said Gantzer in an interview with “Chad and his group did as much as they could for process improvement and efficiencies within the ED, but it became apparent that it was definitely a mismatch between supply and demand.”

Poor communication between the emergency department and the inpatient units left providers with a fuzzy sense of when they would be able to move their patients out of the ED, Thompson said, and slow discharges from the inpatient wards left providers on both sides of the equation juggling frustrating traffic jams in the 333-bed hospital.

The solution required a two-pronged approach involving a technology upgrade and a focus on the human component.  On the health IT side, SwedishAmerican decided to supplement its existing MEDHOST emergency department information system (EDIS) with a new patient flow offering from the vendor.

“Patient Flow HD gave us transparency into what was happening in the inpatient units,” Thompson said.
“We are now able to see bed occupancy rates, identify potential discharges, and figure out what people were waiting for.  That helped us overcome the various barriers for that discharge so that we could proactively start planning our process.”

READ MORE: How Do Artificial Intelligence, Machine Learning Differ in Healthcare?

The visualization tool helps physicians and nursing staff across the organization identify barriers and implement solutions when quality metrics or timeframe goals aren’t being met.

“It helps let the inpatient managers know when ED volume is significantly increasing,” explained Gantzer.  “They can make necessary changes in their unity because they’ll know that they need this many cardiac beds or this much room for a particular surgery.  The individual units can react so much more quickly now that they know that the demand is there.” 

“One of my peers said, ‘This looks like air traffic control for patients,’” she added.  “That is truly how our administrators look at this tool.”

Combining health IT improvement with organizational change

Technology is just one piece of the puzzle – and arguably not even the most important one, in this case.  Gantzer and Thompson credit their success to the fact that they paid close attention to the underlying workflows that may have been hindering the efficient transfer of patients.

READ MORE: Data Warehouse, ERP Tools Top Wish List for Value-Based Care

“Before implementation, we took the opportunity to enhance our discharge process from the inpatient side, and did an in-depth process improvement analysis and flow study of how we could execute a discharge differently,” said Thompson.  “That was an important first step.  All too often in healthcare, we buy a program, make it live, and then we find ways to work around it.  That’s not very effective.”

“It was so important to do that process work before we had the technology help us,” agreed Gantzer.  “It’s definitely true that a lot of times, we apply technology and then force-fit it into the workflow.  We chose to do things differently this time because we had been guilty of doing what a system lets us do instead of using the system to support us.  It made a huge difference to do this the right way round.”

“We started with largest med/surg unit in the hospital, because if we could figure out how to facilitate the discharge process in that particular unit, we could adapt it anywhere else,” she continued.  “We had a process team of clinicians.  We also involved nurses, clinical pharmacy, social workers and case managers, and process improvement experts.  We also had help from some nursing students taking part in a leadership program.”

The team gave stopwatches to the nursing students, who timed exactly how long it took to accomplish every step of the discharge process.  Gantzer and Thompson then reviewed each task to separate critical actions from time wasters.

“We found that it took six hours from the moment of the order until the patient went out the door,” said Gantzer.  “But four hours of that was non-value-added time.  So it actually only takes two hours to discharge a patient, which is where we set our benchmarks.  That’s what we strive to meet.”

“We ended up eliminating almost 50 percent of the steps we were going through to cut that time down appropriately.  By looking at patterns in demand from the ED and the demand from our surgery volume, we determined that we needed every floor to discharge fifteen percent of their patients by 11:00 in the morning to ensure that there would be enough room for everyone.  That comes out to about two patients each.”

Small workflow changes can produce big results

The goal was an attainable one, but it required cooperation and buy-in from physicians and nurses.  A few changes to the daily care routine caused little disruption but produced major results.

“All the physicians really had to do in order to accomplish this goal is switch around when they see which patients,” Gantzer said.  “Obviously, the ones that need the most critical attention get seen first.  But after that, they would see the patients who may be less critical, but who are still going to be in the hospital for a while.  And they’d save their discharge patients for the end of the day.”

“So we flip-flopped that.  Now they go from critical patients to some of their discharge patients to the stable patients staying in the hospital.  We didn’t get a lot of resistance to that, because once they understood what the problem was and how this change was going to help, they realized it would work to everyone’s benefit.”

The nursing staff followed suit, putting patient discharge at the top of their task list and leaving less time-sensitive daily work for slightly later in the day.

“They understood the concept of prioritizing these tasks.  It isn’t a foreign concept.  After all, if a pre-op patient had to be ready for surgery by 9:30, they would make sure that patient met their deadline.  So we just took the same approach to discharge.”

“It’s only two or three discharges per floor.  So in the morning huddle, the leadership identifies which patients should be able to meet that time frame, and the nursing staff just approaches that as one of their priorities.  It wasn’t really a major adjustment.”

A clear and immediate value proposition helped to ensure widespread buy-in, added Thompson.  “Nursing staff really bought in when we showed them the four hours of wasted work time and introduced the cleaned up processes so that they could execute a discharge faster,” he said.  “They had been kind of suffering in silence, in terms of not voicing all of their frustrations that they had with the process, so they view this as a significant win.”

Nurses and physicians weren’t the only ones who had to be involved in the project, Thompson pointed out.  The housekeeping staff, vital to patient safety and smooth turnover yet often overlooked for their contributions, also had to keep pace with the new order of patient care.  

“Once you discharge a patient, you can’t get a new patient into the room until the space is cleaned and completely turned over,” said Thompson.  “So the housekeeping staff started changing their workflow, too.  If Unit A had a significant number of discharges and Unit B had fewer, then the housekeepers working on those other units would go to Unit A and help.”

“The housekeeping staff have always been major contributors to the team, but with this visualization and this data, they’re able to show off their impact.  There’s a sense of pride for everyone who’s playing a part in helping the admission turnaround time go down.  That really helps with securing buy-in.”

Lower wait times, higher patient satisfaction scores

Nurses, physicians, and housekeepers all stood to gain from cutting wasted hours from their days, but they are not the only ones with a stake in the game.  High quality, person-centered care is the goal for SwedishAmerican, and keeping patients as happy as possible during a stressful and difficult event like a hospital admission is critical for overall success.

Patients forced to wait for care are often understandably negative about their experiences, and may voice these sour opinions on satisfaction surveys or online review sites.  These public consumer opinions not only drive business to or from a healthcare organization, but also affect CMS quality ratings and may impact value-based reimbursement.

“We know that our patient satisfaction scores specific to discharge instructions and nursing responsiveness in our pilot unit improved dramatically,” stated Gantzer.  “When we rolled it out across the organization, those same domains also increased.”

In addition to cutting discharge times by close to two-thirds, the hospital has also decreased the time it takes to get admitted from the emergency department.

“Last month our admission turnaround time for admitted patients was 58 minutes on average,” said Thompson.  “We’re also starting to look at that data and break it down by percentile to see how many of the patients get admitted in less than an hour. We’re meeting our goal of having 75 to 80 percent of the patients placed in under an hour.”

“If we rewind the clock back to four or five years ago, we were waiting well over 200 minutes, on average, for a routine admission placement.  So that’s a significant improvement.”

A collaborative spirit and a clear value proposition are essential for achieving these results, Gantzer asserted.  “You have to realize that this is everyone’s problem.  Patient throughput is not just an emergency department issue.  Everyone has to work together to get these things done in a timely manner.  Once you start stressing the fact that these changes can be really good for the patient, everyone seems to get on board pretty quickly.”

Thompson suggests a long-term commitment to dismantling data siloes and process barriers that prevent communication and coordinated effort.  “Patient flow can seem like such a huge, monolithic problem that people often feel defeated by it,” he acknowledged. “But don’t be overwhelmed.  Just start in small bits.  Work for savings one or two minutes at a time, and then continue to chip away at the process.  You’ll end up seeing really great results if you keep at it.”


Join 25,000 of your peers

Register for free to get access to all our articles, webcasts, white papers and exclusive interviews.

Our privacy policy

no, thanks

Continue to site...