Split-Second Destination Decisions

This past Sunday night about 2245 hours, a Detroit police officer was shot in the head while responding to a domestic violence call. The Detroit Free Press,�in an article identifying the shooter, reported that the incident happened at an apartment complex in the vicinity of the 10000 block of Joy Road near Wyoming Avenue on Detroit�s west side. According to Channel 4 News in Detroit, Rapid Response EMS was dispatched and arrived on scene in less than 60 seconds. The officer was transported, with a police escort, to a level two trauma center,� as reported by another local news source,�with the patient being handed over within 22 minutes of the original dispatch. Now, Detroit Police Chief�James�Craig is asking why the injured�officer “wasn’t taken to the closest hospital.”

These are the facts as I have been able to glean them from multiple news reports and summaries. I have no inside knowledge of this particular incident or even any great understanding of Detroit in general, but I believe there are several interesting questions worth a larger discussion here from the perspective of a complete outsider. For those who may have more inside information of this situation, I will point out that I have no interest in any past conflicts that this particular EMS provider has had with the Detroit Fire Department or a memo now resurfacing from an incident last October specifying that injured “Detroit firefighters will be transported by Detroit EMS only.” That is a totally separate matter that relates potentially to medical care, not destination decisions.

A FOX 2 news article made a statement that they are “still looking into why a critically wounded Detroit police officer was taken to a hospital in Dearborn– when there were two hospitals that were closer.�At least one of the hospitals that was passed up is better equipped to deal with a gunshot wound to the head [emphasis added].”�Another article, updated during my research today,�has since made a correction stating, “This story has been updated to clarify that an ambulance driver [sic] did not pass any hospitals while transporting a wounded Detroit police officer.”

The question we in EMS are often forced to answer is what facility is “closest,” however that question does not always have a static answer from every incident. One of the first articles I read on this case printed a thumbnail map (since removed) similar to the image on the left (which you can enlarge by clicking on it.) What immediately struck me was that the shooting occurred at the center of a triangle formed by the three “closest hospitals.”

If we consider distance to be “as the crow flies,” or perhaps more appropriately, “how the medical helicopter flies.” We will get one set of distances and travel times. Here, Henry Ford to the east appears closest in straight-line distance followed by Sinai-Grace to the north and finally Beaumont in Dearborn to the southwest. If we consider road miles of the shortest path, the order changes with Sinai-Grace at 4.8 miles, Henry Ford at 5.6 miles and lastly again, Beaumont at 6.8 miles. However, if distance is measured in drive-time, specific values change (according to my tests using Google Maps) depending on the amount and direction of flow of traffic. In all of my time tests, Sinai-Grace came in dead last due to the number of local street segments traversed and I suspect a large number of traffic signals. These typically narrower streets and signaled intersections are not only slower to travel, but more dangerous when traveled using red lights and sirens (some studies will show this is especially true with a police escort.)

The travel times in my daytime investigation during a typical work week varied with the other two destination hospitals and probably would still be different from a late drive on a Sunday evening. Of some significance in comparing the “best routes” is the number and direction of turns. As a general rule, right turns are safer than left turns which must cross opposing lanes of travel. The other consideration is the speed limit of the roadways. While I assume the ambulance was travelling above the posted rate going code 3, it is the faster roads that are built to a higher level of safety and will more easily accommodate higher speeds with fewer traffic control devices (lights or stop signs.) The route to Beaumont had the highest number of miles on restricted access highways that have the highest speed limits in any city. Consequently, this may have been a very good choice based on actual travel-time as well as safety considerations.

The call was still a judgement one and I will not defend one or the other as the best choice given my lack of knowledge in Detroit, but I will defer to the judgement of crews that travel these streets regularly both as emergent and routine traffic.

The other consideration in this call was the trauma rating of the hospital. As I understand it, both Henry Ford and Sinai-Grace are level 1 while Beaumont is only a level 2 facility. Given the severity of the wound, some deference would likely be given to the better equipped hospitals. However, the real difference between these levels is typically whether there is a teaching and research program available. The surgical capabilities should actually not be significantly different.

With drive times so close to being similar, I can sit comfortably in the safety of my arm chair typing that the choice of Henry Ford would have been quite practical; however, I may well have made a different choice myself as I place myself behind the wheel (as I will be doing tonight.) That immediate “split-second” decision of east versus south west is much more difficult in the moment. And this is exactly the type of situation where I would be grateful for the input from the MARVLIS in-vehicle client that sorts destinations choices by distance and provides an optimal path based on time-of-day with turn-by-turn driving directions.

 

Advertisement

Still Solving Problems in Lexington

An awful lot can happen in five years. I know that my own understanding of EMS deployment has deepened a great deal in that time. It was that long ago that I wrote a post about The Cost of Saving Money using Lexington County, SC, as an example. The county EMS Director, Brian Hood, and the now-retired county GIS Manager, Jack Maguire, made a huge statement about how EMS and GIS can work together and achieve incredible results. At that time, Lexington County EMS credited technology with giving them an advantage that helped them plan and respond better.  Even though they were experiencing an average annual growth rate in calls-for-service of about 7-1/2 percent, they had gone over 4 years without adding a single new truck to their fleet. The close relationship EMS had developed with their GIS group also benefited everyone by improving the quality of their street data for all county users. I have repeated this story over the years but when I revisited them recently for a follow-up, I was amazed to learn how much we had both matured.

Chief Hood began by stating that ten years ago their average response time was 11 minutes. Since then, growth in demand for services has continued to range anywhere between 3.5 and 11 percent annually. Still, they have not added a new ambulance to their fleet, but through continual improvement they have that same average response time of 11 minutes today. Their goal is 12 minutes at the 90th percentile. However, pending legislation in the state of South Carolina known as R.61-7 may require times at the 95th percentile for Advanced Life Support (ALS) response. Guaranteeing service at that level can be a daunting challenge for any manager. The response of Chief Hood was to develop a process to address the demands as well as the realities of his agency. At the core of that process is MARVLIS Deployment Planner (a tool for automating system status management) and MARVLIS Deployment Monitor (a live view of current resources and demand with real-time recommendations.) These tools give the Chief and his staff the information they need to know for scheduling and dynamically deploying resources. “If you took these tools away from me, I could not do my job,” said Hood. “History absolutely repeats itself and this system is frighteningly accurate.”

In addition to facing increasing demands and tighter response times, Lexington is facing a lack of paramedic resources the same as many other areas of the country. It is recognized that sending ALS level resources to every call can be expensive and even wasteful of these limited resources when record reviews show that 70 percent of responses only require a Basic Life Support (BLS) level of care. The new solution they have just begun testing is a tiered approach where calls are being triaged based on nearly 200 determinate descriptors to categorize the initial response level. To prevent dispatching high acuity resources to low priority calls, it is not always the closest unit that is assigned to a call by dispatchers. The lowest categories of Alpha and Bravo level are only sent BLS providers in a vehicle that could otherwise provide ALS care. Rather than requiring an ambulance intercept in the event an upgrade of care is required, command staff will arrive in a quick response vehicle to supplement the care available and effectively transform that ambulance into a full ALS unit.

They are also looking at improving provider safety by questioning the use of lights and sirens on most calls. Just as calls can be categorized for the level of responders, they can be categorized for “cold” and “hot” responses that can limit the dependance on lights and sirens. This is still very much a work in process, but key to making it successful will be in the support of county commissioners. The goal of arriving on scene to the highest priority calls on-time 95 percent of the time will mean that other calls designated in the lowest priority responses will take longer. It’s just common sense that decisions must be made when a system has a defined budget with limited resources to get an important job done. The vision to see the larger picture and to achieve the greatest good for all who are involved is the hallmark of real leadership. Problems never really go away, the list just keeps changing and they keep solving them.

HP-EMS Profile: MedStar Mobile Healthcare

As I was going to be in Dallas for the Fire Rescue conference, I decided to go a little early and pay a visit to MedStar Mobile Healthcare (the renown “birthplace of Mobile Integrated Healthcare”) just over in Fort Worth, Texas.  For anyone who may not have been paying attention to the industry during the last few years, community paramedicine has become a hot topic at conferences for EMS systems that are looking to fill a gap in the healthcare needs of the community.  Significant savings can be realized just in reducing transport demand, especially by “loyal EMS customers”, but additional cost avoidance is available to the hospital in preventing re-admittances.  If you are looking for additional information about implementing a similar program, Matt Zavadsky, director of public affairs at MedStar Mobile Healthcare, has written an excellent description of Community Paramedicine and why it’s the future of our profession.

There is really no doubt that EMS as a practice is changing. However, Paramedics and EMTs will always be critical in responding to emergency calls for service, but MedStar has helped show that they can also be effective in using their skills far beyond that traditional role. While it was the MedStar reputation for innovation in delivering high performance EMS related services that enticed me to visit, I was really most impressed by the back-end systems that keep the care providers on the road and doing their job effectively. Community Care Paramedics like Jimmy Aycox, pictured here with his Panasonic Toughbook, rely on the MARVLIS Client software not only for accurate routing information but also patient details presented from the CAD for filling out patient care reports.

But what makes it all work in the field actually starts in the dispatch center, whether the calls are emergent or scheduled.  Technology is a critical piece used to find the right resource and route the closest paramedics to the right call.  In many routing systems, the travel impedance (the factor that tries to model the real-life movement of a vehicle) is based simply on speed limits to calculate the time required to move from one intersection to another. These systems are static and do not account for various traffic patterns throughout the day or any seasonal variations such as school being in or out of session.  Then there is also the issue of planned road closures or closures due to accidents that can also significantly affect navigation. In this news story about MedStar, the problem with traffic and road closures is highlighted along with their response in employing new technology to account for these issues. During my visit, System Status Controller, Stacey Sokulsky told me that their “older GPS technology could be up to 2 minutes off [in predicting drive times], but I have not seen MARVLIS be off by more than 10 or 15 seconds.” This can make a big difference in selecting which vehicle to dispatch.

Having the right tools makes the job much easier and allows progressive systems like MedStar Mobile Healthcare to do more outside of the traditional role and thinking. Thanks for letting me get a peek at the heart of your system.

The Cost of Saving Money

There are two fundamental ways to save money: either cut your budget and make do with less resources or invest in process efficiency to cut future expenses while continuing to provide at least the same level of service.  These are important considerations as the costs of doing business clearly continue to increase, whether we consider the expenditures on goods used in providing a service or the price of fuel used to deliver that service.  At the same time, the ability to effectively raise the price of the delivered service in order to recoup those additional expenditures is not typically possible.  This paradox leaves many ambulance services in a quandry.  If your decision is to continue operating at a diminshed capacity to reduce spending, there is probably little advice I can offer.  But if cutting service is not a prudent long-term option, then we can look at how an agency looks to improve performance.

A good case study may be the Lexington County (SC) EMS, a service directed by Brian Hood responding to roughly 30,000 calls per year.  As a growing county outside one of the largest cities in South Carolina, Lexington has seen their call volume grow at an annual rate of about 7.5%.  To keep pace with this growth in demand, they would likely have needed to add one new vehicle per year at a cost exceeding $3M for an ambulance, crew, station, equipment, etc.

Additionally, one of the ongoing, and rapidly growing expenses, for any service delivery organization is the rising cost of fuel.  Controlling mileage can often be a great option for managing expenses.  Of particular interest to some services is also the incurring of fines for exceeding response times.  This is a particularly good area to control costs for affected services since this type of expenditure does not lead to any revenue nor does the payment itself provide any service.  But to achieve these goals, performance must improve.

High Performance EMS is about doing things better – specifically providing advanced pre-hospital care with a focus on higher economic efficiency.  The result of better performance is patient satisfaction with cost savings.

For Lexington County, the more efficient posting of ambulances based on predicted demand and time-based routing using MARVLIS has allowed the service to actually improve response in the face of rising demands while foregoing the acquisition of additional resources and the commitment to ongoing costs associated with them.  Over the 4 year period since implementing MARVLIS, the projected cost savings of $3M per year is compounded by the avoidance of recurring staff costs and therefore totals closer to$16M for that period.  The initial investment required to make that savings was less than $400K in hardware and software systems.  As a result, satisfaction has improved for both the patients (who experience quicker service), and also staff (who endure fewer post moves while being closer to incoming calls.)  This is not just a promise or hope of what might happen, but a real world experience of leveraging county GIS services and an experienced High Performance EMS consultant, Bradshaw Consulting Services.  Watch for a more detailed article soon from Esri Press.