Ending Ambulance Diversion

"Central Hospital from Medic 22, en route with one BLS trauma on board. ETA seven minutes."

"Medic 22 from Central, negative. We are on BLS divert."

"Medic 22, copy that." New traffic: "Southside Hospital from Medic 22, en route with one BLS trauma on board, ETA 22 minutes."

Hospital diversion—the practice of effectively "closing" an emergency department to area EMS agencies—occurs on a daily basis throughout the United States. While it tends to occur more frequently in urban areas with higher populations, diversion is no stranger to suburban and even rural hospitals.

In recent years, as hospital leaders started analyzing their surge capacity, a stark reality stood out. With emergency departments running at or near capacity on a regular basis, how could the hospitals possibly step up to a terrorist act or pandemic?

In spring of 2007, current Airlift Northwest Executive Director Chris Martin was well into her 23-year career as administrative director for emergency services at Harborview Medical Center in Seattle, WA. She recalls sitting with a handful of people at a Starbucks coffee shop in Seattle's University District looking to brainstorm surge capacity ideas.

"What we quickly realized was, how could we do surge when we can't even do our normal daily load?" Martin recalls. The group shifted to brainstorming how emergency department (ED) throughput could be improved, which led to the idea of becoming efficient enough to end ambulance diversion.

Coincidentally, concerns from the area EMS community to the Regional EMS & Trauma Council about diversions at all 18 hospitals in Seattle and King County were increasing. Some complaints included transporting units being diverted with critical patients on board and families ending up at different hospitals than their patients. The timing was ripe for a new way of doing business.

Overview of Providers and Service Area

Harborview, located in downtown Seattle, is a primary destination hospital for emergency medical systems serving nearly 2 million people in Seattle and surrounding King County. It is the base hospital for Seattle Fire Department and King County Medic One EMS operations. It is the only Level I trauma center in the state of Washington, and also serves as the primary Level I trauma destination for Alaska and Idaho.

Within King County, multiple fire-based agencies, private companies and one primary air provider (see Table I) transport ALS and BLS patients to 18 hospitals (see Table II). In 2009, more than 18,000 ALS patients and nearly 78,000 BLS patients transported by EMS agencies comprised approximately 20% of the total number of ED patients seen in these hospitals.

Diversion Consequences to EMS Agencies

Hospitals can call divert (close to EMS) at a number of status levels (see Table III ), and EMS agencies are not always given a clear understanding of why an ED closed. All they know is the hospital is on divert. To them it means the ED is not available, when in reality it may only be a BLS divert, or an ALS divert, or some particular subspecialty care that is not available.

Seattle Fire Department Assistant Chief A.D. Vickery says diversion impacts his agency's ability to function at peak performance. "Our EMS system in Seattle is fire-based, utilizing all SFD apparatus in a layered response system, with the closest fire apparatus dispatched to the scene," he says. Vickery says all Seattle firefighters are EMT-certified as a condition of employment. Because all on-duty apparatus is available for any type of call, delays encountered on an EMS call could impact response to a subsequent EMS call, structure fire, motor vehicle accident or some other emergency. "Any delay in patient delivery impacts the entire fire/rescue/EMS response system negatively, delaying our service delivery," Vickery concludes.

American Medical Response (AMR) is responsible for the majority of BLS patient transports in King County. AMR's operations manager for King and Snohomish Counties, Kaylee Garrett, says diversion affects her crews' abilities to provide the best service to their patients and also impacts AMR's ability to meet its contractual obligations for emergency and interfacility transports.

Garrett was concerned about making ambulance crews find the "next best" destination for their patients while en route. AMR crews are also limited to a 20-minute window from arrival at a hospital until they are back in service. ED backups and diversion to another hospital that is backing up due to someone else's divert status makes that standard hard to meet. When the crews are backed up at hospitals or finding alternatives they are not available for the next emergency call or interfacility transport.

According to Bob Berschauer, AMR's director of government relations for Northwest/Great Lakes Region, "In years past, we have had issues with trying to find a receiving facility. Our ambulances have been unable to rapidly transfer patients, which results in a loss of available units to respond to requests for service." He says that while diversion is not currently as big an issue as it has been in the past, "Everyone knows it can become an issue again, so that is why it is so important to work on this even when it is not presently a problem."

Diversion Consequences to Hospitals

Depending on the hospital, decisions to divert ambulances are not limited to emergency department staff, but are often initiated by bed-management staff, administrators and even hospital executives.

In King County, hospitals have become accustomed to using diversion as a management tool for running their EDs. But hospitals and EMS agencies had difficulty tracking diversion frequency and consequences. Existing data collection tools were limited and not very reliable.

"Some hospitals think going on divert reduces workload," says David J. Carlbom, MD, associate director of emergency services at Harborview. "But I can't find any evidence in the literature to support that." He says ED overload is more affected by inpatient activities elsewhere in the hospital than by sheer numbers coming in the ED door.

An immediate consequence is the ripple effect of one hospital going on divert on other hospitals in its region. On Seattle's First Hill, there are three hospital EDs within walking distance of each other. Get out of the urban core, and the distances grow.

Jennifer Graves, RN, MS, ARNP, and nurse executive at Swedish Medical Center's Ballard campus, sees diversion as failing to deliver exceptional patient care. "Our primary focus at Swedish is to deliver exceptional patient care, and our teams realized some time ago that closing our doors and diverting patients away during a very vulnerable time was certainly not contributing to the health and well-being of our community," she says.

Turning away patients is harmful to the patients, to the care delivery system and to the hospitals' bottom line. According to Carlbom, patients entering the ED via medic or ambulance transport tend to be sicker and are more likely to have insurance or some other means of paying for their care.

Because there are so many choices in King County, hospital destination decisions are made in concert with the patient's wishes. While patient choice is not allowed to get in the way of the best hospital choice, BLS patients like to be taken where they've been treated well in the past, or where their primary care physician practices. Diversion can get in the way of these decisions and waste healthcare dollars, says Carlbom. "When a patient can't get to his or her home hospital, no patient history is readily available, tests get repeated, wasting time and money, and the patient's outcome can be impacted," he says.

Driving Forces

In 2008, limited data on diversion and examples of its consequences were presented at a summit with all King County hospitals represented. Hospitals agreed in theory to attempt a zero-diversion goal, but with nobody in overall charge and voluntary compliance, little progress was made. The analogy used was citizen CPR classes where you are taught that if a person is not singled out to call 9-1-1 and report back, the call may not get made. A new model to add responsibility needed to be crafted.

Harborview, in conjunction with King County Public Health and the Central Region EMS and Trauma Council, followed up on this first effort by initiating the King County Diversion Project designed to create a perpetual zero-diversion system status for itself and the other 17 hospitals in the county. Funding was secured with federal grant money administered by the King County Healthcare Coalition.

Initial Steps

Martin, Graves and Carlbom all agree that the best first step taken was bringing Clark Hartley on board to manage the project. Hartley brought years of experience as a hospital-based paramedic at Johns Hopkins Hospital in Baltimore, MD. Having worked for the office of the chief operating officer, he knew hospital and prehospital operations intimately, which would be invaluable for bringing both worlds together in his new job as manager of the King County Diversion Project.

Martin says having a point person like Hartley full time adds energy, effort and visibility to keep the project on people's minds. Going back to the CPR analogy, Hartley is the guy who calls for help and stays in faces (politely) to make sure everyone is doing their part.

 Hartley began with a search for data. He found there was inadequate data available for hospital administrators to define the scope and source of diversion problems. Prior to 2008, King County relied on an outdated home-grown data capture system that provided no historical data for comparison or analysis.

In 2008, Harborview, using grant funding, purchased the WATrac system on behalf of all the hospitals in King County to collect data on hospital diversion status. Although not as flexible as users would have liked, it did capture more useful data than they could get before. From late 2008 to early 2009, the frequency and accuracy of hospital reporting dropped dramatically due, in part, to heavy turnover of personnel and lack of state resources. In addition, lower patient counts reduced the need to divert, which contributed to reduced reporting efforts.

An early part of the Diversion Project was to encourage hospitals to follow through with initial and ongoing training to ensure usage and upgrade accuracy on data that was entered. Having more comprehensive and accurate data collected enhances its usefulness to administrators. Having more real-time updates improves level-loading of patients. “Giving EMS providers access to timely information enables them to make well-informed decisions in the field, which is critical to getting the right patient to the right place the first time,” says Hartley.

King County Diversion Project Elements

Hartley researched no-diversion systems around the country to find best practices. Boston, which closely matches King County in population and number of hospitals, had success with a city-wide zero-diversion program that spread outward. Today, Massachusetts is a widely published zero-diversion state.

Las Vegas leaders lobbied the state legislature to enact laws requiring hospitals to off-load patients in 30 minutes and a technologic method of reporting for accountability. Salt Lake City has worked to eliminate diversion by shifting the focus on improving patient throughput. San Diego is using better communications between hospitals to reduce diversion. They also use technology for real-time situational awareness.

Common terms were defined so everyone used the same words in the same way. "Once we got everyone using the same terminology, it became easier to accomplish things," says AMR's Garrett.

Hartley organized a steering committee from throughout the region with representatives of all the hospital and prehospital partners in the county (see Table IV). Only executive-level people with the authority to make decisions and commit resources were sought. He selected the representatives himself because "As the new guy, I had no personal biases that could get in the way."

Hartley believes earlier efforts to move toward zero divert were undermined because there was no centralized approach with a single leader and global participation. Carlbom says Hartley's method of bringing everyone to the table was an interesting approach. "We all had lots of little projects, but Clark pulled them together into one large, comprehensive project where everyone could see all efforts underway," he says. "It was apparent that an EMS-driven plan could hurt hospitals and a hospital-driven plan could alienate EMS. Getting everyone at the table meant that instead of all of us working in our small silos, we could focus on integrating all our work to the greater good."

Initially, the committee met monthly as an advisory group to refine plans and make recommendations, then quarterly as plans began to be put into play. Hartley expressed pride and respect for the committee members. Nobody he solicited said no, and these busy people maintained an 80% attendance rate over 18 months.

Hartley met with leadership at every hospital and the medical director at every agency, as well as fire and EMS leadership, recognizing that while hospitals are making the decisions of whether to stay open, the consequences of closing are very much a prehospital issue, too.

Process to Achieve Zero Diversion

Primary reasons hospitals reported “closed” to ambulance services:

• Boarding in the ED
• Catheter lab down
• ED saturation
• Hospital saturation
• No acute care beds
• No critical care beds
• No telemetry beds
• Number of ambulances waiting.

One fact that became clear from the beginning was that diversion was a hospital problem, not an EMS problem. Eliminating diversion would require communication, cooperation and better patient throughput anywhere a patient may be sent or kept in the hospital.

"The big 'ah-ha' I've seen across the county is the realization that diversion is not an isolated emergency department issue, but a system phenomenon that occurs when there are defects or bottlenecks in any number of related areas downstream," says Swedish Hospital's Graves. "As throughput in all of our operational departments improves, there is a positive correlation with our emergency department cycle times and hours on divert."  

This means getting everyone on board, from the ED to labs and admitting to radiology to the hospitalists, Graves says. "The ED physician may need to admit the patient. The patient wants to get to his room as soon as possible. Let's get everyone working to make it happen and eliminate the waits and delays."

Eileen Bulger, MD, FACS, director of emergency services at Harborview, echoes Graves' observations. "The ED backs up when there are no patient beds in the hospital and they board in the ED," she says. "There are several benchmarks for patient flow in the ED that are in line with the national standards promoted for emergency medicine. If there is nowhere for the patient to go, then meeting the time benchmarks becomes challenging."

Changing the culture of how diversion is used was important, says Hartley. "Diversion was a management tool and often a crutch," he says. "People relied on it so much they became accustomed to using it for daily operations." Hospitals took a hard look at how they used diversion, who had the authority to put the ED on divert, and under what circumstances. New policies were drafted and implemented by some hospitals that moved the decision to call divert higher on the management chain.

In the Swedish Hospital system, says Graves, one of their keys to success was "enabling those in the trenches to escalate issues in a coordinated way to ensure that appropriate resources were marshaled in an effort to prevent diversion, and removing authorization of divert from front-line caregivers and putting the responsibility for this on the nursing director and medical officer of the day." The goal was to make calling divert a second option rather than a first.

Better level loading of the hospital EDs by BLS transport agencies is being helped by technology with the FirstWatch Early Warning Surveillance System. Two of the three central dispatch centers in the county have collaboratively tied their computer-aided dispatch systems together to form a regional dashboard. They can see, in real time, all ambulance and medic units at or en route to all 18 hospital EDs. The screen shows the unit designation, identifying it as an ALS or BLS unit, average transport unit wait time at that hospital, and longest wait time. Dispatch centers can help route transporting units to the closest, best hospital with the shortest wait time. Patient care is optimized, unit out-of-service times are reduced, and EDs are loaded more evenly. Once funding is secured, the third dispatch center has discussed plans to implement this technology.

"The majority of ambulance patients are BLS-level in acuity. So, working to make sure all hospitals are receiving a fair share of BLS patients instead of concentrating them into the hospitals with high trauma loads will keep beds available for the higher acuity patients overall," says AMR's Berschauer.

One potential roadblock to achieving level loading that cannot be overlooked is patient choice. It factors into customer satisfaction and, for private companies, ultimately profitability. “Level loading is a very sensitive subject. It required a high level of situational awareness, both at the service provider and regional advisory levels,” says Hartley.

"This must be balanced with the patients who have a definite demand as to what hospital they want to go to," Berschauer continues. "Since by definition a BLS patient is usually not in emergent danger, but still in need of urgent treatment, honoring destination requests is considered by many to be a patient right."

Data-Driven Change

Hartley says it is impossible to overstate the importance of having reliable data and using it to improve functions. With this project, compliance rates and accuracy from the hospitals using the WATrac system have improved dramatically. Administrators have data they can trust to measure benefits gained through operational process redesign.

Collaboratively sharing hospital data to all partners has unleashed the power of transparency, according to Hartley. Hospitals began to see trends and patterns emerge. He began hearing questions like, "Why was the other hospital able to cope and we weren't?" "When we (or a neighbor) went on divert, was the cause internal or external?"

Swedish Hospital's Ballard campus is one of four EDs in the Swedish system in King County. Graves says having good data to review has given them some good-natured internal competition. There's a healthy spirit of, "if they can do it, we can do it" among the different campuses.

Current Participation Levels

When Hartley began having data analyzed in 2009, hospitals in King County were averaging more than 3% of cumulative hours on divert. In 2009, as data became more reliable, hospitals realized their own efforts had begun to pay off as diversion hours dropped. Hartley’s contacts in Boston and Las Vegas were optimistic that no-diversion could be achieved in Seattle/King County, based on their own experience of reaching no-diversion starting with much higher levels of diversion activity.

Through Hartley's outreach with data, concerted efforts in every hospital, and the power of transparency, 2010 data showed a high of 3.42% of cumulative hours on divert one month and a low of 0.93% of hours in another month. From January through November 2010, 14 hospitals achieved zero divert in one or more months. Nine hospitals have achieved and maintained zero divert every month.

Due to its unique situation as the base hospital for two large EMS systems and the premier regional destination for the sickest patients, Harborview has yet to reach the zero-divert benchmark, but Bulger is optimistic. "There is a lot of support within to move to a no-divert policy," she says. "Because of our high census, Harborview has the greatest challenge with this, but it is our goal to establish a process to handle the high volume and avoid divert. Bottom line, I think we are moving in the no-divert direction and just need to get buy-in from the entire hospital and have a mechanism within the system to move patients who do not need Level 1 trauma care to other hospitals in our system when we are too full."

Initial Impacts/Benefits to EMS

Seattle Fire Assistant Chief Vickery has noticed the progress made so far. "We see benefits in all facets of our EMS delivery system with the implementation of a no-divert policy," he says. "Foremost is efficient patient care. Additionally, our fire units are available on a more frequent basis to respond to emergencies when patient delivery and throughput are enhanced. Time/fuel/maintenance and personnel costs all are minimized with a no-divert policy."

Garrett says she is not seeing her crews take the brunt of being redirected as much as they were, but the reality is that they still have to find places to take patients and must be comfortable talking to them about changing their choice of hospitals.

Berschauer notes that "A combination of reduced hospital patient census and a real focus on the part of hospital staff to improve throughput in all hospital departments has alleviated those issues at the present time."

Next Steps

The regional EMS & Trauma Council is conducting a three-month no-diversion pilot scheduled from March 1 through May 31 this year. It is expected that all 18 hospitals will participate. Results will be carefully studied to see what needs adjusting in order for them all to make zero-divert permanent. Results will be reported and reviewed with numerous stakeholders, including the Washington State Department of Health, the state hospital association, all hospitals and all EMS agencies.

Participants in this project all see zero-divert throughout the state as a distinct possibility. Rural areas with limited hospital choices are already essentially there due to the absence of effective alternatives.

Berschauer is optimistic. "I think the goal is feasible," he says.  "There will always be legitimate outliers that affect a hospital’s ability to provide service.  The object of a no-diversion policy is to keep the ability to divert to a tightly defined and very rare event, instead of a more common occurrence with undefined criteria."
Harborview's Carlbom believes the biggest challenge will be trying to convince people it will be OK. "We have to just do it," he says. "We're so close now, at less than 2% in King County. People have to take the plunge. It'll be scary at first, but will be well worth it."

Ed Mund began his fire and EMS career in 1989. He currently serves with Riverside Fire Authority, a fire-based ALS agency in Centralia, WA. His writing and photos have been published in several industry publications. Contact him at ems@medstrat.com.