Not that long ago, the dispatch center was just a room located in a basement or the back of an office, staffed by a person “trained” to answer a telephone and push a microphone button, with a data set generally gathered on a paper log with a pencil. Times have really changed.
One needs look no further than the “Chain of Survival” to understand the importance of today’s dispatch center. Immediate activation of response agencies, early CPR, rapid defibrillation and early and effective ALS initiation all emphasize the need for minimizing time and accurate decision-making. The success of each of these important components is directly impacted by trained dispatchers and the technological tools at their disposal.
One of the axioms of EMS response is that if it doesn’t go right at dispatch, there is little chance for the patient. In the modern world of fewer unit hours, increased demand for service, higher expectations and stressed revenue sources, that axiom is now: If it doesn’t go right at dispatch, there is little chance for the system.
The dispatch center is no longer really a dispatch center. About 10 years ago, it could have been termed a control center, but even that has changed. Now, with responsibilities that include protocols for “hear and treat” and “see and treat,” and interfacing on a regular basis with the other components of the healthcare community, it truly can be termed a “clinical hub.”
To be effective and efficient in this role, accurate data is needed–because data drives outcomes. This article will provide an understanding of the diverse data sources and the uses of data in this new world. It all begins here.
Ensuring That the First, First Responder Is First
The first point at which accurate data on any emergency can be captured in most systems is when outside communications are received at the initial answering point. However, response time in most systems starts when the call is received at the EMS dispatch center. I mention this now because this is “lost” data, and, more importantly, a lost opportunity.
With immediate activation of response emphasized as the first link in the Chain of Survival, obtaining this data is extremely important for analytical purposes. Ironically, technology may not be the limiting factor, because these centers often employ sophisticated Computer Aided Dispatch (CAD) hardware and software programs. Rather, there seems to be a reticence by many initial answering points to provide that data because of the concern that it will disclose a significant delay in call processing. This data becomes increasingly important, however, as the focus on response times transitions from outputs, where arbitrary response times are measured and outliers penalized, to clinically significant, evidence-supported outcomes.
“Okay, Tell Me Exactly What Happened”
After capturing the address of the emergency and the call-back number, the highly trained Emergency Medical Dispatcher (EMD) starts the clinical assessment of the patient in a matter of seconds. Similar to the paramedic’s field protocols, the protocols used by the EMD are medically approved and founded in clinical evidence. They require the EMD to interrogate the caller for demographics, characteristics and the patient’s general problem, and then ask further specific systematized questions to determine the acuity level of the patient and the proper system response needed. This is all driven by data!
Without a doubt, every EMT and paramedic reading this article has responded to a “shortness of breath” call with red lights and siren, only to arrive and find a non-acute patient breathing relatively normally. That can lead them to question the reliability of the data, if, in fact, dispatch protocols are so data-driven. It is an interesting question with an explainable answer.
The Medical Priority Dispatch System (MPDS) is the mostly widely used protocol. It is also the most studied and has a large evidence base to support its clinical foundation. Everything MPDS does is based on data.
First, it is important to understand that the MPDS is designed to over-triage patients because the EMD is not actually with the patient and can’t see the patient. This will probably change as technology evolves, but for now it is the reality. Hopefully most people will agree that it is better to over-respond to the patient not in respiratory distress than to under-respond to the patient who was actually breathing agonally and is found to be in cardiac arrest.
Second, the published peer-reviewed evidence shows that the MPDS process really does work. Each action of the EMD is captured as a data point within the MPDS software, resulting in the assignment of a patient condition, or “determinants.” Currently there are more than 300 determinants in the system.
After the MPDS system was introduced at the London Ambulance Service, it showed a 200% increase in the identification of cardiac arrest over the next three years.1 Similar studies have shown the accurate identification of chest pain, seizure and stroke patients.
Equally important for EMS systems today is the use of data by the EMD to identify the non-acute patient. This is not only feasible using a structured dispatch protocol system, but peer-reviewed research has proven that it can be done–and is safe. In fact, in one study it was found that 99% of the calls triaged as “alpha,” the lowest acuity category, did not meet any higher acuity criteria.2
This data is extremely important for the dispatch center. Not only does accurate triage save valuable ALS resources for the critical call, but accurate triage of these calls by the EMD can provide that caller with better, more appropriate resources outside the traditional 9-1-1 system. For example, by linking the MPDS dispatch software to nurse triage software, the Louisville EMS system and MedStar Mobile Healthcare in Fort Worth, Texas, have safely implemented telephone triage, or “hear and treat.”
This type of dispatch system is also a key for community paramedicine. Without data driving the process, the community paramedic, or advanced paramedics as they are called in England, would not be as successful as they are now. With the rapid spread of community paramedicine to the United States, it is imperative that the dispatch system, and clinical hub concept, be founded on dispatch data, and that data be clean, unbiased and accurate.
The CAD: Data, Data, Everywhere
The pencils and paper logs are long gone, hopefully, replaced by all kinds of new CAD technologies. In fact, in many dispatch centers, it is hard to find a piece of paper anywhere. It is all software-driven, with complex interfaces linking the call-taking to the CAD and the CAD to the field. With these programs, there truly is data, data, everywhere, and the chances of the data being accurate are almost 100%, as long as it’s entered accurately–which often requires no more than a push of a button.
More can now be learned sooner to speed the response with the right resource. In the past, most dispatch programs just captured times as follows: Call Received; Unit Alert; Unit En Route; At Scene; En Route to Hospital; At Hospital; and In Service.
Of course, there were variations, but those were the basics. The data was then converted into reports as follows, with the first leading the way: Response Time; Response Time Exceptions; Dispatch Processing Times; Unit En Route Times; and Percent of No Transport.
Those reports and others allowed system managers to attempt to identify outliers and hopefully, make positive changes in a timely manner. That timely manner often actually meant providing feedback a month or two later. With data driving care, this would all be a little late.
A supervisor involved in strategic deployment strategies, or as it is more commonly known (and often despised), system status management, faced even more obstacles. Hand-entered data, maps on walls, and acetate with erasable markers were the tools of the day. It is no wonder the skeptics had the advantage. By the time a system status plan could be developed, it was outdated.
Fast-forward to 2014 and the world of real-time data. With a plethora of CAD systems and software applications, the opportunity for care to be driven by data instantaneously, and in a “predictive” manner, is almost limitless. Again, it all begins at dispatch–the hub.
Many communications centers, such as those serving Louisville; Reno, Nev.; Fort Worth, Texas; and Richmond, Va., use FirstWatch software to monitor operational indicators such as dispatcher call-processing times to provide 9-1-1 center supervisors and dispatchers with real-time feedback. Information captured by a dispatcher or call-taker and entered into the CAD can be monitored automatically by FirstWatch in real time along with software like ProQA, which provides structured dispatch protocols.
With this combination, even the “unstructured” information that a dispatcher enters into the “notes” field can be monitored for key words or phrases. Some centers link their phone, CAD and ProQA data systems to interface with FirstWatch from first call ring throughout the call-taking/dispatch process. Many EMS systems then link that call center data to ePCR data, and some, like Sedgwick County EMS in Kansas, even link that to hospital data, for measurement of performance outcomes.
Regardless of the evidence, the focus on some semblance of a response time standard will never be eliminated. The key will be to ensure that there is sufficient coverage to meet a standard and at the same time maintain resources to treat the vast majority of the patients accessing the system in an economically challenged environment. With real-time data collection, interpretation and feedback, this has become much easier.
Understanding the possibilities is a first step, beginning when the 9-1-1 call is received. With the appropriate interface, both the address and the telephone number automatically populate the call-taker’s screen. Within 30 seconds the call-taker uses MPDS to process the call as life-threatening and, using another interface, transfers the data to the CAD, which then automatically populates the dispatch software.
At this point, based on available software applications and real-time data, the dispatcher’s decision becomes one based on education rather than assumption. Many EMS systems use Automatic Vehicle Location (AVL) software that continually transmits vehicle location and availability to the dispatch center using another interface. At a glance, the dispatcher knows the options. Other programs, specifically mapping, also may be available to show coverage, or lack of coverage, the best route for the vehicle, any obstacles that might be encountered (think bridges, trains, road work), and, based on time of day, traffic.
The decision is made and the rest is a push of a button. The crew is notified, the time automatically stamped, and even the route automatically recorded. Another push of the button by the crew on arrival at the scene and an accurate response time is recorded. Dispatch then looks at the real-time map again, determines both the location of the coverage gaps and the likelihood of where the next life-threatening emergency will occur, and moves the remaining available units. These actions take only seconds, and the coverage plan remains optimized.
Operationally, the collection of data like this quickly opens new doors for supervisors and managers. Every aspect of the operation can be monitored and performance measured. Although it’s true that the focus of data collection can be on the individual dispatcher, in a performance-based EMS agency, the collection and use of data is best served when it is used to make system analysis easier and system changes faster. Understanding the complex nature of a dispatch center and the almost overwhelming responsibilities involved with the data are inextricably linked to better clinical care.
Clinically, real-time dispatch center data also saves lives. Using mapping, or GIS, the locations of past cardiac arrests can be plotted along with the time of day and response times. This leads to interfacing with coverage maps, AVL and the ultimate timely response. Although evidence has proven that the eight-minute response time standard is arbitrary, the same evidence shows that interventions at the four- to five-minute mark after a cardiac arrest markedly improve survival. This is data that saves lives.
In addition, by combining the MPDS system with dispatch notes, medical directors and clinical supervisors can be automatically notified of specific clinical events, including cardiac arrests, chest pain or high-risk pregnancy. Combining this dispatch data with the data collected on board the unit, such as 12-lead EKG or specific PCR documentation, can create quick and effective quality improvement loops.
Expanding the Clinical Hub
Data can also play a part in driving clinical care through the use of non-traditional applications within the clinical-hub concept. The following are but two examples.
The threat of anthrax terrorism following the events on Sept. 11, 2001, along with the SARS outbreak experienced by Toronto EMS in 2003 and the more recent panic when it was believed that H1N1 would create the biggest influenza epidemic since 1919, clearly demonstrated the need for syndromic surveillance. The U.S. is unique in that it has real-time detection technology developed by FirstWatch that uses trigger alerts based on accumulated MPDS data to detect possible disease outbreaks.
FirstWatch data, which uses secure processes and meets HIPAA requirements, is presented on “dashboards” so users can instantaneously see the status of any dataset. FirstWatch is also set by many systems to alert the comm center of the early signs of a chemical, biologic, radioactive or nuclear attack, or for naturally occurring events like epidemics.
FirstWatch can also send an automated page or email, or both, to provide advance warning of an event. When I was the chief executive of the Richmond (Va.) Ambulance Authority, for example, FirstWatch predicted a flu outbreak in the city two days before the public health director received his notification. Automatic alerting software such as this reduces workload and increases awareness by automating key notifications for sentinel or situational awareness events, such as the explosion of a bomb or suspicious clusters of patients that could be an indicator of an epidemic.
Where the U.S. lags is in the registry of automatic external defibrillators (AEDs). For instance, in Denmark there is a National AED Network that has placed an AED for every 1,100 people, and the appropriate dispatch center knows where each is placed and can direct the responder to the closest location. While AEDs clearly play a part in the chain of survival, where they are can be a mystery without a registry. The time has come to be aggressive, and the place to link the registry is in the clinical hub.
Driving the Chain
With the changes in dispatch technology over the past 10 years, one can only speculate about the future. Whatever the direction, the concept of the chain of survival and the need for speed will remain as new methodologies, medications and techniques are discovered. For these new elements to reach the patient, the dispatch clinician will need the best data available to initiate all links in the chain. It all starts at dispatch.
References
1. Heward A, Damiani M, Hartley-Sharpe C. Does the use of the Advanced Medical Priority Dispatch System affect cardiac arrest detection? Emerg. Med. J. 2004;21(1):115—118.
2. Hinchey P, Myers B, Zalkin J, et al. Low acuity EMS dispatch criteria can reliably identify patients without high-acuity illness or injury. Prehosp Emerg Care. 2007;11(1):42—48.
