Improving EMS Deployment Performance

I work regularly with agencies that are looking to improve aspects of their operations. Some casual readers may be surprised to know that the focus of those discussions is not always about cutting response times. While response is a simple and common measure, it clearly does not evaluate EMS well and certainly fails to encapsulate many of its complex needs and values. Still, I feel the necessity to address the time objective briefly before going on to other important aspects.  

Of the multitude of scholarly articles written on the subject of EMS response times, many are often cited to me as not supporting the idea of a correlation between time to patient contact and their medical outcome. However, this sort of summary is often overly generalized as a large proportion of the articles have evaluated outcomes of marginal variations in response times greater than 8 minutes and only consider a  patient’s mortality as an outcome. Of those studies that include response times of less than 5 minutes, a positive correlation is usually indicated. Some articles even find value in measures closer to the fabled 8-minute “standard.” In her article entitled “Do Emergency Medical System Response Times Matter For Health Outcomes?” the author, Elizabeth Ty Wilde, states that, “we ?nd that distance is positively and significantly correlated with mortality. Without accounting for the endogeneity of response times, there is no statistically significant relationship between response time and mortality. However, the instrumental variable estimates of the impact of response times on outcomes show that, on average, a minute increase in response times increases mortality by between 8% (measured 1 day after the initial incident) and 17% (measured 90 days after the initial incident).” Further, it is not just the cardiac arrests, strokes, anaphylaxis, and significant trauma calls where time is a primary determinant of outcome, but this measure is arguably the very basis of EMS as an essential service in any community. This political reality, however, is fodder for another post altogether, but sufficient at the moment to show at least probable cause for its continued measurement among a full set of system vitals.

“Reduced response times should never come from driving faster, but rather by driving shorter distances.”

The point of dynamic deployment is not simply to cut response times, but also to make responses safer. Reduced response times should never come from driving faster, but rather by driving shorter distances. For any ambulance call, there are two potential segments of travel. The response phase where a unit is enroute to the patient and the secondary transport phase where a patient is delivered to the “most appropriate facility” for definitive care. It is in the response traffic to a scene when we are most likely to be driving using red lights and sirens. If that distance is shortened by routinely prepositioning the ambulance closer to that next call, the response will automatically be safer for the crew and the public at large. Then, although the absolute distance from the incident scene to the closest facility is fixed, the best route is not always the same during each hour-of-the-day or day-of-the-week given traffic, weather, or other active incidents. The ability to communicate current traffic conditions for selecting the best route is increasingly available and utilized even by commuters who often travel the same route at the same times of the day. Better information on traffic conditions makes for more efficient and safer route selection.

Cutting mileage not only makes for safer and shorter travels, but can reduce the overall mileage of a fleet as well. And since it is the response segments that are cut most often, it is the “unreimbursed” (or “unloaded”) miles that often experience the most significant reduction. There are also the associated costs of fuel and maintenance that can contribute to the savings depending on the deployment model being used prior to the use of an effective posting plan. The knowledge that not all posting strategies are equal must be presumed. The value of any particular dynamic posting plan is in the details of its technology and the integration of its implementation. Jack Stout, the father of dynamic deployment, wrote back in 1983 that “system status management isn’t ‘good’ or ‘bad’ – it is inevitable. Your plan may be simple and stupid, complex and stupid, simple, yet effective, or possibly even complex and even more effective.” Compared to crews sitting in quarters with a vehicle turned off and hooked up to a shore line, the maintenance costs will undoubtedly increase with a dynamic change. Perhaps, it is more appropriate though to state that the realized costs are actually “shifted.” The fuel consumption of an ambulance idling at a post is clearly higher than the example of waiting at a station, but what is the value of the minutes saved in turn-out time and initial travel to an incident for the patient waiting in distress? Opponents will often try to argue the savings down to seconds, but the reality of a well-deployed system is the common elimination of around one minute in chute time from a typical 90 seconds or more to less than 30 seconds in addition to the time gained by moving the starting point minutes down the road from the traditional station. For services that pay a financial penalty for late calls, the savings can be easily measured as in systems that survey the customer satisfaction of their patients.

Shortening time of responses by minutes also contributes to lowering the amount of time that an ambulance is “busy” as measured through a Unit Hour Utilization (UHU) score which Stout defines as the amount of the total time that “a fully equipped and manned ambulance is (available) on the street” compared to the proportion of minutes it is engaged in serving a patient. EMS agencies typically target a UHU of between 0.30 and 0.50 with a focus toward lowering the effective ratio to decrease crew fatigue. Another significant contributor to crew dissatisfaction is the number of post moves requested during a shift. While some agencies incorporate all drive-times (including non-emergency post moves) into the busy ratio of UHU, it is a unique measurable contribution that deserves a separate indication. There are several strategies for minimizing the number of post moves experienced during a shift and can include the consolidation of hours during a plan to eliminate any possible move recommendations at the top of intervening hours and also building a plan that maintains the relative importance of posts during any hour to limit post selection changes to only the higher end of the service levels. These simple changes, while having some minimal negative impact on other performance measures, can decrease unnecessary crew stress as a counter-balance.

“there are an infinite number of iterations on the spectrum between these two extreme ends”

To many readers, dynamic deployment may seem to be an extreme alternative to static station posting. While that can be true, there are an infinite number of iterations on the spectrum between these two extreme ends. Where any service lands along that spectrum will impact three basic competing performance measures of the quality of care delivered to patients, the economic efficiency of a system, and the well-being of their clinical care providers. Moving along the spectrum from the highly intensive strategy of maintaining ready resources at fixed stations equally spread throughout a jurisdiction to the plan for managing a scarce set of resources through a changing deployment pattern will distort the triangle that represents how effectively these competing objectives are met. Approaching the station strategy, represented at the left end of the diagram above, crews are often happy with a fixed assignment. And as long as enough resources are maintained, patients will experience effective care from providers stationed close by any location. This comes at a significant cost to the system to maintain adequate providers at all locations. Conversely, when resources are limited, choices must be made on where they are staged which can raise the anxiety of crews subject to movement on demands of the service. But when posts are forecast correctly, the availability of immediate care does not suffer and the system overall saves money.

Forecasting demand is not a simple affair, but yet another aspect of deployment that must be managed over time and changed with changing demand patterns. At a minimum, demand is recognized to change significantly with school being in or out or session necessitating at least two unique plans across the year.

Few systems truly do well at either extreme in this diagram, but compromise somewhere in the middle often seeking either to balance all aspects, or more likely, to favor particular concerns at any given time. As achievements are made in one aspect or another, then previously neglected objectives rise in importance. There is nothing static about any sort of dynamic deployment plan. System status must be continuously managed and continuously changed to continue to achieve current objectives. Processes for managing that deployment and the assignment of units to calls must also be managed through a strategic protocol.