Dynamic deployment: 5 persistent myths busted
The definition of dynamic deployment and its implementation are explained in this mythbusting primer
By Dale Loberger
Dynamic deployment is a topic, due to an evolving understanding, that lacks a universal definition. Unfortunately what people hear can often be very different from what was intended. And for a profession that demands a high degree of precision in how we communicate, it can be very frustrating when we casually converse in terms that lack clear meaning.
To correct some of the ideas that I often see confused in discussions about dynamic deployment, here are five of the most common misconceptions and the current reality behind each statement.
Myth 1: Dynamic Deployment is the same as System Status Management
While dynamic deployment owes a lot to the foundation that Jack Stout created in his ideas of System Status Management developed during the 1980's, they are developing into two distinct levels of implementation strategies today.
System Status Management is only a relatively dynamic option with a fixed number of pre-defined plans. A separate plan is developed for each hour of the day and every day of the week given all of the possible system service levels, which means a total of 168 plans must be created for each possible service level for each week. Available resources are placed based on the hourly plan that has generally taken anticipated seasonal demand into account for that hour.
Some agencies will generalize these ideas and create a limited set of plans that are based strictly on the number of available vehicles and place resources according to a predetermined schedule regardless of the time-of-day or day-of-week and the variations in demand that it entails. This is neither dynamic deployment, nor SSM. The success of this over-simplification, known simply as closest vehicle dispatching is based on the decreased chute-time of having a vehicle already on the road and the luck of having it somewhere near the call simply because it is not with others at some centralized base or station.
Dynamic deployment is based on an understanding of the dynamic nature of demand in near real-time. To properly locate resources for the greatest benefit, we need to know where we will be needed next.
Since people are creatures of habit who congregate with others that are similar to them, they create a consistency in demographics that generalizes the probability of a specific set of medical requests. There are also traffic patterns and common driving habits that, along with road conditions, allow certain trauma calls to be predicted. As a result, statistical algorithms can forecast areas — not specific addresses — where emergency calls are most likely to be generated with surprising accuracy and redundancy.
Even though SSM takes demand into account, it was established as a manual process that generalized planning into discrete hours of the day. True dynamic deployment requires an automated computer process that allows a more regularly updated forecast of demand in conjunction with the current location of all available units. It is, therefore, a continual match between currently available units and the highest probability of call demand on a near real-time basis rather than 60-minute planning increments.
Myth 2: Dynamic deployment means "street corner posting"
The core principle of dynamic deployment is the pre-positioning of resources to be closer to expected calls for service in the very near future, but exactly where units will post in wait of their next assignment is determined by an agency's policy. It is this internal decision that determines where they are willing to allow their medics to wait. Some systems are fine with the actual nearest street corner to the center of an identified hot spot while others are willing to give a few blocks in order to locate their crew in a relatively safe or more comfortable environment. Sometimes that post is actually a purpose-built station. To the system, a post is no more than a location where assets are assigned, but to the crews in the field, a post means much more.
Any posting location simply depends on the administration of the system rather than the strategy itself. The primary objective of dynamic deployment is to position the responding vehicle closer to the next call. Each time an ambulance is closer to that future request for service compared to coming from a centralized point, the strategy has succeeded to some degree. Consequently, the deployment of ambulances in this manner does not dictate that every street corner becomes a legitimate potential posting location.
Myth 3: Dynamic deployment means constantly moving from post to post
The name, dynamic deployment, certainly implies the movement of resources to adjust the coverage to anticipated demand. The constant bouncing from one post to another, however, instead of moving from a post to a call is evidence that something is not working correctly in the implementation of a response strategy.
Common issues can be that the demand is simply not well understood or it may be that there is an extra capacity of resources for an extended period of time. Another possibility is inherent in the strategy employed for moving vehicles between posts.
In order to meet a new plan based on changing demand or a new service level, a domino style movement can cause multiple vehicles to move while accumulating the shortest amount of overall time. The unfortunate side effect of this maximized efficiency can be a recommendation that increases the total number of moves to achieve the new goal in the fastest time.
A leap-frog style movement can achieve the same plan, often in nearly the same amount of time while displacing fewer vehicles in the process. However, these movement recommendations will allow ambulances to pass one another which is an occurrence that some systems try to avoid as a rule. A benefit to allowing these resources to move past one another can be a better balance of workload if the number of previous calls is factored into the consideration for movement to a particular post.
Often, an efficient movement strategy with an overall reduction in the unloaded miles driven, can be achieved by exercising prudent judgement in following automated system recommendations. As a result, the efficiency of a well-managed system can mean not only a reduction in response times, but a reduction in the non-reimbursed operating costs as well. If these costs go up significantly as a result of posting, there is most likely a larger problem with its implementation.
Myth 4: Dynamic deployment sacrifices random calls in outlying areas by focusing on high density populations
Statistically speaking, there is always a greater likelihood of a larger number of calls from a more densely populated area than there is from other rural areas that contain fewer people. It follows then that anticipating demand has a propensity to recommend coverage of the commonly higher demand urban areas; however, that is not a given. While anticipated demand is certainly a driving force for dynamic deployment, it does not need to be the only consideration.
Depending on unique features of the service area, it is prudent to balance pure demand with some basic geographic coverage through a weighting factor. In an exclusively high density area, the need for weighting geographic coverage may not be necessary.
In general, weighting demand more heavily than geography cuts a larger percentage of response times to an absolute bare minimum. But that efficiency is at the cost of other infrequent remote calls.
By increasing the weight of pure geographic coverage, dynamic deployment recommendations can strike a balance in the risk of these singular calls from low-density, remote areas with the typical high-density, high-demand areas. In this case, the aggregate response times are reduced at the percentile level by reducing the unfair burden on remote constituents. The only way to ensure the lowest possible response times for everyone is to increase the overall supply of resources which comes at a significant financial cost. Dynamic deployment balances community needs with the most efficient services from available resources based on a given risk tolerance in order to control all costs.
Myth 5: Dynamic Deployment is about chasing meaningless response times at the expense of crews
One of the most dangerous tasks involved in EMS is driving the ambulance. The use of red lights and sirens is especially risky. A primary benefit of dynamic deployment is the reduction of the distance between the location of the vehicle when it is dispatched and its destination. The fewer miles driven under emergency conditions lowers risk for the crew and the public.
A secondary benefit to positioning crews closer to the call before it is made is a reduction in the time the patient is waiting for service. While only a small percentage of calls may warrant an immediate response, the practice of attending all high-priority calls in a timely manner ensures that you are responding effectively in those cases when seconds literally do count. But even if a fast response is not critical to the clinical outcome, it is typically appreciated in patient satisfaction. An efficient response, therefore, is not strictly about time, as safety can be a measurable benefit as well.
To learn more about dynamic deployment read Dynamic deployment: 5 more persistent myths busted.
- Stout, Jack L. (1983 May 22) System Status Management: The Strategy of Ambulance Placement. JEMS.
- Ingolfsson, A. (2013). EMS Planning and Management. In Operations Research and Health Care Policy (pp. 105-128). Springer New York.
- Dean, Stephen S. Why the Closest Ambulance Cannot be Dispatched in an Urban Emergency Medical Services System. Prehospital and Disaster Medicine, 2008, 23, pp 161-165
- Ward, Michael J. (2012 October 18) How Fire Departments Mangle Ambulance Deployment. EMS1.
- Brophy, John R. (2014). Dynamic Deployment: A Primer for EMS. Createspace. Charleston, South Carolina.
About the author