The plan to improve a community’s survival rate consists of 10 specific but diverse steps. Four of the steps are relatively easy and do not require much in the way of resources. These may be considered the low-hanging fruit. There are six steps that are more difficult and require either modest to considerable equipment or resources—the higher-hanging fruit. It is unrealistic to presume that all 10 local steps can be implemented in any given community; an EMS director and medical director have to decide what is doable in their community. Furthermore, what we consider easy steps may be very challenging in some communities and vice versa. All the recommended steps are cognizant of the fact that real change requires addressing quantitative as well as qualitative factors. The total picture must entail the chain of survival as well as the frame of survival. This guide closes with some thoughts on how to implement change.
A disclaimer: The recommendations put forward in this section cannot simply be used as a template for any emergency medical system. Every community has its own constellation of resources, history, culture and personalities. Indeed this variety, which makes every EMS system unique, is a strength, offering a crucible for new ideas and new programs, which is why every community can become a source of innovation as well as a testing ground for new ideas. For some communities, implementing even one of the recommendations here may prove challenging. For others, they may implement several steps in the first year.
By the end of the two-day Resuscitation Academy (RA) training course, the attendees are fired up to return home and begin to make changes. It is clear that they are bursting with ideas big and small for their home communities. What should they tackle first? What will give the biggest bang for the buck? The RA faculty tells them to pick the low-hanging fruit first. Our advice is to reach for the largest, tastiest, juiciest, and closest pear before climbing the tree. Get some success under your belt and keep plugging away, small step by small step until there is a culture of change, allowing one to pick some of the higher-hanging fruit.
The steps that can achieve the quick results are: 1) to establish a cardiac arrest registry, 2) to begin a program in dispatcher-assisted CPR, 3) to begin a program in high-performance CPR, and 4) to begin rapid dispatch. These steps are neither complicated nor costly, but they are not without challenges.
Three of them require ongoing quality improvement (QI) if they are to reach their potential. High-performance CPR, dispatcher-assisted CPR, and rapid dispatch all require continuous maintenance and nurturing. To do otherwise would be like planting a vineyard and assuming it would do fine without watering and pruning. Programs without ongoing QI and ongoing training will result at best in mediocre and lackluster performance and at worst in no improvement at all.
Step 1: Establish a Cardiac Arrest Registry
A cardiac arrest registry is the first step to improving survival. It is the essence of measurement. One of the mantras at the Resuscitation Academy is “measure, improve, measure, improve. ...” encapsulating the concept of documenting cardiac arrest events (measuring) and then implementing changes for improvement. In turn, continued measurement will determine if the improvement has had an effect and will identify further steps the improvement has had an effect and will identify further steps for improvement. And so on. ... This is the most important mantra of the Academy since it pithily describes the bedrock upon which all programmatic change springs forth.
A registry is a means of taking the entire EMS system’s temperature. If a cardiac arrest is well managed, it’s more than likely that all other conditions will be well managed, too. In this sense, cardiac arrest stands for the whole system. A registry measures more than whether the patient lives or dies, but all aspects related to the care. Was bystander CPR performed? Did the dispatcher provide telephone CPR instructions? How good was the EMT CPR? Were there unacceptable pauses in CPR? Did the paramedics intubate successfully? Given enough cardiac arrests, a profile begins to emerge of where the system is succeeding and where it is failing. This information then informs the specific elements that need improvement.
The cardiac arrest registry’s efforts must be viewed as a core function, and the registry itself must not be threatened with funding cuts or elimination during lean times. It must have sufficient resources and the full support of the medical and administrative directors. Necessary resources include staff time for gathering information from run reports (electronic or paper), dispatch center reports, AED recordings, hospital records, and ideally death certificates. Clearly, a small community will not have the volume of events to justify full-time dedicated staff, but several small communities can join together to establish a registry at the county or regional level. Investigators from Emory University have established, with funding from the Centers for Disease Control and Prevention, a national cardiac arrest registry: The Cardiac Arrest Registry to Enhance Survival (myCARES.net). The registry is open to EMS systems throughout the nation.
As of 2013, there were 50 communities from 17 states participating, plus six entire statewide EMS programs. The registry entails having the EMS system and local hospitals submit data via a Web-based system. CARES overcomes a major obstacle in most well-intentioned registries, namely obtaining outcome data from hospitals. Did the patient live or die and what was the neurological condition on discharge? The CARES project is based on voluntary participation, and all the participants receive summaries of their own community as well as a national summary.
CARES can be customized for the needs of the local community. CARES also provides templates so communities can review their statistics sliced and diced in any way they wish. The main template is the Utstein reporting template, which provides the survival (discharged alive) rate for witnessed cases of VF in which the collapse occurs before the arrival of EMS personnel. For agencies participating in CARES, the Utstein template is automatically generated. Utstein survival report for reporting cardiac arrest survival is also known as the Utstein Template.
The main metric is discharged alive survival from bystander-witnessed cases of VF of cardiac etiology. The survival is expressed as a percentage of all cases meeting this definition. When possible, the neurological status of the survivors should also be determined from the hospital record. CPC scores of 1 and 2 indicated good to moderately good neurological outcomes, and CPR scores of 3 and 4 indicate poor to terrible outcomes. To maintain a cardiac arrest registry (whether in CARES or as a free-standing registry) at its basic level, probably a quarter-time person is needed for a community of one million to gather incident data and obtain follow-up information from hospitals. If the tasks associated with maintaining the cardiac arrest registry are combined with those of collecting and managing data for high-performance CPR and the dispatcher-assisted CPR program, there will be enough work for one half-time employee. This estimate assumes that the EMTs and paramedics are assisting in the data collection such as forwarding run reports and defibrillator downloads. It also assumes that the dispatch center is providing CAD reports and recordings on CPR calls.
The registry should collect information on all cardiac arrests for which EMS care has been provided—in other words, when resuscitation was attempted. The major emphasis, however, should be on cardiac arrests in which ventricular fibrillation (VF) was the presenting rhythm. For communities with limited resources, restricting the registry to cases of VF or witnessed VF is a reasonable measure. Implicit in the concept of a cardiac arrest registry is the assumption that time intervals will be measured accurately. The most important time intervals are those between the patients collapse and the start of CPR, and between the collapse and the first shock. Admittedly it is not usually possible to know the exact time of collapse. Therefore, the first accurate time is the time the 9-1-1 call is answered, and this should be the precise moment the EMS clock starts ticking. For cases involving bystander CPR, the initiation of CPR can be arbitrarily defined as having occurred halfway between the time of the call to 9-1-1 and the time of the first-in units’ arrival. Ideally, all the systems AEDs will be synchronized automatically or manually to an accurate clock.
It is important to be realistic about what a cardiac arrest registry includes. To be fully functional; a registry must have the following three elements:
- Full capture of all arrests meeting the case definition. The case definition we use in King County for an event is a cardiac arrest in which EMS personnel initiate or continue CPR. Patients who receive AED shocks and subsequently do not require EMS CPR are also considered cases. Trauma cases are excluded from the registry unless the trauma involves a low-speed motor vehicle accident or other situation in which the cardiac arrest may have preceded the trauma. Patients who are dead on arrival or have a “do not resuscitate (DNR)” order and those who did not receive EMS CPR do not qualify under the “case” definition. But we do include cases in which EMS CPR is started and then stopped after a DNR order is clarified. However, since these cases are invariably non-VF cases, they do not affect the VF survival rate.
- Measurement of critical variables: Witnessed collapse, collapse before EMS arrival, first rhythm obtained, shockable rhythm, bystander CPR, telephone CPR, time of call to dispatch center, time of EMS CPR, estimated time of bystander CPR, time of first compression for dispatch-assisted CPR, and time of first defibrillation.
- Measurement of outcome: death at scene, death in hospital, discharge alive (ideally with a determination of neurological outcome). To obtain critical information on the outcome of all patients admitted to the hospital, a good working relationship with area hospitals is essential. A registry is part of ongoing quality improvement and is considered protected information in most states (and consent from the patient to release medical information is not required). The completeness of the registry can of course vary from the bare minimum of information to hundreds of variables. In King County, we have a registry that is comprehensive and serves as the basis for many studies. The 300 variables we collect from CAD reports, incident reports, defibrillation downloads, voice recordings, hospital records, autopsy reports, and death certificates would be considered excessive for routine quality improvement. A good basic registry can be achieved with 14 event and three outcome variables.
The list of variables in the preceding page exhibits a minimal data set.
For communities not participating in CARES, information about a free-standing registry may be obtained by contacting the Resuscitation Academy. This free-standing registry, called CATS (Cardiac Arrest Tracking System), includes 14 event variables and three outcome variables. It comes with a data dictionary and is a simple-to-use registry which collects only essential variables. Though CATS is not linked to a national registry, it will allow any EMS system to monitor its own performance in managing cardiac arrest and to track changes in survival over time. CATS uses an Access data base in which to record information, and thus, should be readily available within most organizations. CATS can display the data in an Utstein Template and can be exported to Microsoft Excel to allow for customized displays.
Step 2: Implement Dispatcher-Assisted CPR with Ongoing Training & QI
Most dispatch centers claim to have dispatcher-assisted CPR protocols in place, but in practice they don’t offer CPR instructions very often. (Synonymous terms are telephone CPR, dispatcher CPR, and dispatcher-assisted telephone CPR.) Admittedly, it is difficult and stressful for dispatchers to determine the presence of cardiac arrest and provide CPR instructions; it is far easier simply to reassure the caller that help is on the way. But the center whose culture supports its dispatchers to assertively offer callers CPR instructions over the telephone is a center that has the chain of survival firmly in its grasp. This kind of culture can exist only if someone has responsibility for teaching dispatcher-assisted CPR, monitoring the program, and watching it like a hawk. Someone has to listen to recorded information from all cardiac arrest calls and give feedback to individual dispatchers, as well as to the entire staff. It is just as important to review the calls in which instructions were provided (how could it be done faster, better?) as it is to review the calls in which cardiac arrest was not recognized (how can we do a better job identifying cardiac arrest?).
The adjective “assertive” describes a useful mindset for dispatchers when fielding possible cardiac arrest calls. A take-charge attitude that moves ahead with CPR instructions, when there is reasonable likelihood that cardiac arrest is present, is the attitude needed for this program to succeed. If the dispatcher is overly cautious or holds back in the face of uncertainty, the instructions will seldom be given or there will be considerable delay in their implementation.
One element of any successful dispatcher-assisted CPR program is training, which should include continuing education. Dispatchers in King County receive an initial 40 hours of training in emergency medical dispatching, and are thereafter required to complete eight hours of continuing education every year. Special emphasis is placed on recognizing cardiac arrest and delivering CPR instructions.
Recently, a five-year randomized clinical trial in King and Thurston (Wash.) counties and London, England, looked at whether dispatcher-assisted CPR achieved better survival with standard CPR (mouth-to-mouth with compressions) instructions than with chest compressions-only instructions. The trial found no difference overall in survival, but there were non-statistical improvements in survival and neurological recovery with chest compression only. As a result we now provide chest compressions-only instructions for all adult cardiac arrests. (Chest compressions-only instructions for adults is also recommended by the American Heart Association.) The dispatchers provide standard CPR instructions (mouth-to-mouth combined with chest compression) for cardiac arrest in children and infants (fortunately rare events) and when there is an obvious respiratory cause of arrest such as drowning, hanging or smoke inhalation.
Prior to delivering these instructions, the dispatcher has determined the likely presence of cardiac arrest. This is achieved by asking two critical screening questions. Note that the instructions call for chest compression only CPR. If respiratory arrest is the suspected cause of cardiac arrest then ventilation instructions are provided. There are slightly different protocols for children and infants as well as protocols for choking patients.
In King County, there is an expectation that “every call is a cardiac arrest until proven otherwise”. (We think of this as a step specific mantra for dispatcher-assisted CPR). Though only 1% of the calls will actually be a cardiac arrest, nevertheless this expectation primes the dispatcher to always ask the two screening questions (unless the caller is the patient) as quickly as possible. (View Dispatcher All Caller Interview, p. 10.) These two questions must be asked of every caller who isn’t the patient:
• Is the patient conscious (awake)?
• Is the patient breathing normally?
If the answer is “no” to both, the dispatcher immediately begins instructions. (View Adult CPR Instruction, p. 11.) This gives the second-step specific mantra “No, no, go,” which is short for “No not conscious; no not breathing normally; go and begin CPR.)
Dispatchers learn the significance of agonal respirations and how to recognize them. It is particularly important that they offer CPR instructions when there are agonal respirations, since these patients are the ones most likely to be resuscitated and discharged from the hospital. Agonal breathing is present in approximately 60% of patients with VF cardiac arrest. Yet the presence of agonal breathing will often confuse the caller and/or dispatcher into thinking the patient is not in cardiac arrest.
Despite our considerable training on how to recognize agonal breathing, it remains a challenge. When asked, “Is the patient breathing normally?” the caller often responds with “a little” or “sometimes” or “I’m not sure” or “I think so.” The dispatcher is seeking a “yes” or “no” response to the question and instead receives an ambiguous reply. When the dispatcher asks the caller to describe the breathing, the replies are varied and include snoring, slow, grunting, groaning and gurgling. Often, the agonal breathing can be heard in the background, but if not, the dispatcher may ask the caller to bring the phone to the patient in order to better hear the breathing. In terms of training and motivation, the director of a dispatch center should do whatever is necessary to ensure that the center achieves a 50 percent rate of offering dispatcher-assisted CPR instructions in cases of cardiac arrest. This is not an unrealistic target. Achieving such a target will require a sponsor—someone who takes charge of the desired change, has the authority to mandate it, and establishes training, professional expectations, and ongoing audits to see that the change is fully implemented. Once dispatchers realize how vital they are to the chain of resuscitation, and especially when they see concrete evidence of their success, they will become the staunchest advocates of dispatcher-assisted CPR.
As with high-performance CPR, the quality of dispatcher-assisted CPR can be measured. At a minimum; every call of cardiac arrest must be reviewed with the following elements measured:
• Was cardiac arrest recognized?
• Were the two basic questions asked:
1. Is the patient conscious (awake)?
2. Is the patient breathing normally?
• Were agonal respirations (if present) recognized?
• Were telephone CPR instructions offered?
We believe ongoing QI of dispatcher-assisted CPR should include the following attainable goals:
- Recognition of cardiac arrest by the dispatcher in 75% of all cardiac arrests treated by EMS.
- Recognition of cardiac arrest within one minute (on average).
- Provision of dispatcher-assisted CPR in 50% of all cardiac arrest calls treated by EMS (excluding the calls in which bystanders are performing CPR at the time of the call).
- First compression started within two minutes (on average).
Feedback must be provided to the dispatcher following every event.
The American Heart Association in 2012 issued a Scientific Statement strongly endorsing dispatcher-assisted CPR, including the importance of asking the two identifying questions, special training in the recognition of agonal respirations, and a vigorous ongoing QI program.
Step 3: Implement High-performance CPR with Ongoing Training & QI
Recent studies demonstrate the connection between quality CPR and survival from cardiac arrest. Not only is the time interval from collapse to onset of CPR predictive of survival, but also the quality of the CPR is just as important. The better the CPR, the better the outcome. Since 2005, when we trained every EMT in our system how to perform high-performance CPR, we have seen a dramatic increase in survival. Resuscitations that go on for 50 or 60 minutes with a pharmacy of medications and 10–20 defibrillatory shocks and a patient who survives—with excellent neurological recovery—used to be the exception, but now seems commonplace. It is as though the onset of high-performance CPR suspends death and gives a better opportunity for the defibrillatory shocks and medications to work their magic.
High-performance CPR is as much a construct as a measurable skill. The construct says that letter-perfect CPR is the goal of all resuscitations. This skill can be achieved in training, as well as through review of real events. We believe training on manikins with “paper strip recorders” is the best teaching tool, as it provides instant feedback (i.e., tangible and quantifiable) about the quality of CPR.
Last, and perhaps most important, is that there is an ongoing QI program that provides feedback with specifics about CPR performance to EMTs following every cardiac arrest. A QI program should, among other things, measure the percentage of time chest compressions are performed during each two-minute interval between rhythm assessments. Well-trained EMTs should be able to provide chest compressions for at least 90% of the available time. Most defibrillators allow digital downloads following the resuscitation, including precise measurements of CPR percentage and quality.
Another aspect to high-performance CPR is the choreography between the EMTs and the paramedics. Other terms used to describe this flawless team performance are the “dance of resuscitation,” the “CPR ballet” and the “pit-stop approach to CPR.”
Observing well-trained rescuers engaged in high-performance CPR it is indeed like watching a well-choreographed dance. The term “pit-stop” refers to the pre-defined role of each rescuer and the very minimal waste of time. Like a professional race car pit crew, each member of the team knows exactly what to do and does it with the minimal wastage of time and effort. This choreography means the members switch or rotate roles with minimal interruptions—current protocols call for rhythm analysis every two minutes. Thus, the chest compressor and the ventilator can switch roles every two minutes. With sufficient personnel at the scene, one EMT can start compression, the second EMT can attach the AED pads, the third EMT can provide ventilation, while the fourth EMT can feel the femoral pulse (in order to define the location of the artery and determine if a shock leads to a perfusing rhythm). A fifth EMT, if present, could be the “captain” of CPR and provide direction to the crew (typically the person providing overall direction is the one who operates the automated external defibrillator). Paramedics should optimally intubate and place the IV with no interruption in chest compression. Clearly if there are fewer rescuers the responsibilities must be aggregated.
In our system, we think of the EMT crew as owning CPR, meaning that they are responsible for the quality and directing assignments. The paramedics own advanced life support, meaning they are responsible for intubation, starting an IV, and administering medications. Upon arrival, paramedics become the overall team captains, but they know to delegate CPR to the EMTs. The EMT team not only keeps track of the quality of CPR, but also keeps track of the timing of interventions. Since there is a rhythm analysis every two minutes, an EMT is the official timekeeper, literally using a stopwatch. (In King County we go slightly beyond two minutes of CPR in order to end with 30 chest compressions prior to every rhythm analysis.)
For the EMTs and paramedics to understand why high-performance CPR is so critical, part of our training includes instruction in the science of CPR. Peter Kudenchuk, MD, has developed a compelling 30-minute video on the science of CPR and why letter-perfect CPR is so important. This video is part of every King County EMT’s training.
Step 4: Implement Rapid Dispatch
With rapid dispatch, the closest EMT-staffed vehicle is dispatched within seconds, when specific medical emergencies are reported to the 9-1-1 dispatcher. The dispatch should occur even while additional information is being gathered from the caller. The quick arrival of at least an EMT vehicle, allows the EMTs to perform CPR and deliver the first defibrillatory shock. On the other hand, it may be immediately clear that both EMTs and paramedics are required (such as a report of ongoing CPR) and thus, both vehicles can be rapidly dispatched. If additional information from the caller suggests that paramedics will not be needed after all, the dispatcher can call off the paramedic unit with a code green message.
The symptoms or complaints that should trigger a rapid dispatch are: unconscious or suspected cardiac arrest, chest pain, difficulty breathing, stroke symptoms, ongoing seizure, and significant trauma. A community should carefully measure its current time interval from first ring into the alarm center to specific dispatch of the first responding unit (EMT or paramedic unit). This time interval is variously labeled but will be called “dispatch time” here. The National Fire Protection Association (NFPA) sets a dispatch time standard of 60 seconds for critical events. Rapid dispatch can do much better than that. The rapid dispatch target time in King County dispatch centers is 15 seconds or less, especially when the address is auto populated into the dispatcher CAD (computer aided dispatch) system. The first mention of a critical symptom mandates an immediate dispatch. Dispatch centers that place priority on a rigid, predefined protocol over rapid dispatch are doing their communities a disservice, and in the case of cardiac arrest, are delaying life-saving therapy.
The concept of rapid dispatch applies to tiered-response EMS systems as well as single response systems. Many EMS systems’ dispatch protocols require full information before even a single rescue vehicle can be sent. That may be an acceptable procedure for the majority of calls, but speed is of the essence in a life-or-death situation, and in those cases usual dispatching protocols must be short-circuited. In Seattle and King County, we train dispatchers to use rapid dispatch when they hear certain key words and phrases from callers—either the short list of symptoms in the preceding paragraph or words such as “collapsed,” “unconscious,” “can’t breathe” and “heart attack.” We also urge the dispatcher to use common sense and immediately send EMTs whenever a caller otherwise conveys the likelihood of a critical event. In King County rapid dispatch is used in approximately 30% of EMS calls. We believe rapid dispatch saves 30–60 seconds in dispatch time for the most critical medical emergencies.
Given the fact that survival falls about 10% for every minute of delay in CPR and defibrillation, rapid dispatch can add 5–10% to a community’s survival rate. All of this can happen with no additional staffing or resources. (Not bad. Now you see why this is one of the low-hanging fruits. View Community Scorecard, p. 14, for more on this.)
One more point to be made: EMS dispatch centers (whether stand-alone centers or part of larger combined centers such as fire and police) must have protocols authorized by medical directors. Medical expertise is necessary to provide pre-arrival instructions as well as determine the urgency of the caller’s complaints, and thus, how quickly units must be dispatched. The logical person to fill this roll is the EMS medical director. Unfortunately many dispatch centers still do not involve the EMS medical director in the writing or approval of protocols for medical emergencies, a bizarre disconnect between the patient and the EMS system.