Event Factors
Event factors are those associated with the cardiac arrest itself -- rhythm, witnessed collapse, location of collapse, time of the event, whether a bystander initiates CPR, the cause of the cardiac arrest (ischemic versus cardiomyopathy versus primary arrhythmia), use of an on-scene AED, the presence of emesis, symptoms before collapse, collapse before or after EMT arrival, presence of agonal breathing, activities of the person before collapse (emotional stress, exercise, smoking, and drinking), environmental factors (heat, air quality, ozone level), and position of collapse, decision to begin resuscitation.
| Event Factors | Strength of Association with Outcome |
| Cardiac rhythm | ++++ |
| Witnessed collapse | ++++ |
| Locaction of collapse | ++ |
| Time of event | 0 |
| Bystander CPR | ++++ |
| Cause of cardiac arrest | ++ |
| Use of on-scene AED | ++ |
| Emesis | + |
| Symptoms before collapse | Unknown |
| Collapse before EMS arrival | ++ |
| Agonal breathing | +++ |
| Activities preceding collapse | Unknown |
| Environmental factors | 0 |
| Position of collapse | Unknown |
| Decision to begin resuscitation | Unknown |
Rhythm (strong association with survival):
Rhythm is an extremely strong predictor of successful resuscitation. (Figure 5.1) Ventricular fibrillation has the best chance of survival. Other rhythms such as pulseless electrical activity and asystole have almost no chance of survival.
Witnessed collapse (strong association with survival):
Witnessed collapse is defined as someone at the scene seeing or hearing the collapse. Witnessing the collapse is another extremely strong predictor of survival. In Fig. 5.2 we see the effect of witnessing on survival for the rhythms of VF, pulseless electrical activity and asystole. It is likely that many (perhaps all) of the unwitnessed cases that survive are due to the collapse occurring minutes or even seconds before the patient is discovered. When someone walks into a room and sees a victim collapsed there is of course no way to tell by sight if the collapse occurred seconds or an hour before discovery. Witnessed cases of VF have the highest likelihood of successful resuscitation and form the denominator for all cross community comparison as recommended in the Utstein guidelines. Another problem with using unwitnessed rhythms in the denominator when calculating likelihood of survival is the fact that VF may be the initiating rhythm but after 20 minutes or so will degenerate into asystole. Thus, seeing unwitnessed cases of asystole tells nothing about the causal rhythm.
Location of collapse (moderate association with survival):
Though communities themselves are powerful predictors of survival this section actually refers to where the collapse occurs within a community. Patients who collapse in public locations are more likely to survive compared to those who collapse at home. For witnessed cardiac arrests with VF, collapse in public locations has a survival rate of 47% compared to 28% for collapse at home Part of this is explained by the higher rate of witnessed collapse in public locations as well as the higher rates of CPR in public sites compared to home settings. (Figure 5.3) It is also possible that persons out and about in public are less likely to have significant co-morbidity compared to persons who may be confined to home because of illness. Patients in nursing homes with witnessed VF cardiac arrest have a lower likelihood of survival, 21%, undoubtedly because of their higher co-morbidity compared to persons not requiring nursing home care.
Time of event (no association with survival):
The time of day of the collapse has no relation to likelihood of survival. Furthermore there is no relation of outcome to month of year or day of week. There are some studies showing relationship of time to the cardiac event itself. For example, more myocardial infarctions appear to occur in the early morning hours. But once an arrest occurs there is no relation of time, day, or month to survival.
Bystander initiated CPR (strong association with survival):
There is a large literature on the benefit of bystander CPR. Though not every study agrees the vast majority show an approximate doubling in the likelihood of survival if bystanders begin CPR prior to EMS arrival. It is usually easy to ascertain whether bystander CPR occurs prior to EMS arrival but it is often difficult to precisely measure when the bystander CPR begins. The cardiac arrest registry in King County attempts to reconstruct this time from information on the incident form and dispatch tape recordings. Admittedly this is labor intensive. For communities with less registry resources it is possible to approximate the time from collapse to bystander CPR by using half of the first-in unit response time (time from 911 call to arrival at scene).
Cause of cardiac arrest -- ischemic versus cardiomyopathy versus primary arrhythmia (moderate association with survival):
This is an association that is hard to quantify but I believe a real one, nevertheless. Cardiomyopathies are less likely to present in VF compared to ischemic heart disease associated cardiac arrest and thus have a lower probability of survival. Primary arrhythmias usually present with VF and may be easier to resuscitate compared to ischemic arrests. Some ischemic associated cardiac arrests may have a large myocardial infarction and be impossible to resuscitate. Thus the cause of the cardiac arrest is related to the likely success of resuscitation.
Use of an on-scene AED (moderate association with survival):
An on-scene AED has the potential to provide defibrillatory shocks quickly. It would certainly be a fortuitous if a person collapsed at a location that had an AED on site. In some ways this is analogous to a bystander at the scene knowing how to do CPR. Of course there are lots of conditional factors. Someone has to recognize the collapse, the patient has to be in VF, the bystander has to remember the presence and location of the AED, the AED has to be attached and operated properly. Despite these challenges AEDs offer potential in improving outcomes. The large Public Access Defibrillation (PAD) Trial demonstrated an improved survival in locations with AEDs on site.
Presence of emesis (weak association with survival):
It is assumed that emesis (vomiting) associated with cardiac arrest decreases the likelihood of successful resuscitation owing to obstruction of the airway and aspiration of the stomach contents into the airways and lungs. A study by Reed Simons compared outcomes from cardiac arrest with the presence or absence of emesis and found a negative association with emesis. In other words, if the patient vomited before or during the resuscitation there was a lower likelihood of survival. Simons reported emesis in 32% of 1000 cardiac arrests in King County. Two-thirds of the vomiting episodes occurred before EMS arrival. Patients with emesis had a survival rate of 13% compared to a survival of 18% among patients without emesis. This difference was statistically significant.
Symptoms before collapse (unknown association with survival):
Symptoms such as chest pain or difficulty breathing prior to collapse suggests that the event was likely caused by ischemia or cardiomyopathy. In contrast cardiac arrest with no symptoms is likely the result of primary arrhythmia –namely VF. Since the cause of the event is related to outcome (see above) it is likely that the presence of symptoms is also related to outcomes. However this relationship is complicated by whether the collapse occurs before or after EMS arrival (see below). Patients with symptoms may collapse before or after EMS arrival whereas those with no symptoms invariably collapse before EMS arrival.
Collapse before or after EMS (moderate association with survival):
Cardiac arrest that occurs after the arrival of EMS has a higher likelihood of survival compared to arrest before arrival. This is logical since the personnel are present and can provide CPR and defibrillation almost instantly. For patients with witnessed VF who collapse before arrival of EMS, survival in King County is approximately 35%. One might think that survival for patients who collapse after arrival of paramedics and are in VF would approach 100% but the actual survival is approximately 60%. The reason for the discrepancy probably lies in the fact that patients who collapse after arrival of EMS have serious and ongoing symptoms and thus the reason for the call. These symptoms likely indicate an acute coronary syndrome or worsening of cardiomyopathy. Since the underlying condition may have progressed to the point of fatality, even instant care may not be sufficient. For example, the person may be having a myocardial infarction that is so large that the outcome is death regardless of the timing of care. Patients who collapse before the arrival of EMS are a heterogeneous mixture and include patients with no symptoms and those with symptoms who delay sufficiently until collapse occurs. In either case there is lower likelihood of survival compared to the collapse after arrival group because of the delay in treatment.
Agonal breathing (moderate association with survival):
The presence of agonal breaths (abnormal breaths occurring immediately after cardiac arrest and lasting up to several minutes) is strongly associated with survival because it is a surrogate for a brief period of cardiac arrest. In a study from King County, Jill Clark reported agonal breaths in approximately 40% of all cardiac arrests. Survival in the agonal breaths group was 27% compared to 9% when agonal breaths were not present. The presence of agonal breaths was strongly associated with the underlying rhythm – 56% of VF arrests had agonal breaths compared to 34% for other rhythms. Patients whose arrest was witnessed had agonal breathing 55% of the time compared to 16% for unwitnessed arrests. Agonal breaths occur only in the first several minutes of arrest and therefore its presence suggests that the event happened very recently. Agonal breathing is an underappreciated phenomenon. Most citizen CPR training programs make no mention of its occurrence and many emergency dispatchers are not specially trained to recognize it. This is regrettable since a caller who reports “abnormal breathing” may mislead the dispatcher into thinking the patient is not in cardiac arrest and therefore not offer telephone CPR instructions.
A variant of the agonal breathing occurrence is seizure activity associated with cardiac arrest. With cardiac arrest the brain is deprived of oxygen and this can lead to a seizure. The exact incidence of this phenomenon is not known. If I had to guess I’d say it occurs in 5-10% of cardiac arrests. The presence of a seizure may mislead a bystander into thinking the patient has epilepsy which in turn may lead to delays in calling 911 or delay in dispatchers recognizing the true nature of the problem. It is not known if seizure activity at the time of arrest is associated with a better or worse outcome.
Activities preceding collapse -- emotional stress, exercise, smoking, drinking (unknown association with survival):
As noted in the previous chapter vigorous exercise transiently increases the risk of sudden death but there is no evidence that exercisers who have cardiac arrest are more or less likely to be resuscitated compared to non-exercisers. Of course one would have to control for co-morbidity in the two groups – presumably exercisers have less co-morbidity.
It is possible that other activities such as sex, smoking, caffeine or alcohol consumption, emotional stress, anger, or even high-fat foods may trigger VF. While plausible there are no good data to clinch these relationships. However, the question for us is, once the arrest occurs, whether the presence or absence of these triggers or activities alters the likelihood of resuscitation. Conceivably some of the triggers might impede the resuscitation. The nicotine from smoke could make the heart more resistant to defibrillation. Would EMS personnel provide different therapy if the patient had been smoking or exercising or been in a knock-down argument? Probably not, though it is possible that an immediate trigger might be neutralized by some therapeutic factor. The science on this matter is simple too meager.
Environmental factors -- heat, air quality, ozone level (no association with survival):
Several studies have found an association with the environment and cardiac arrest survival. A 2007 study from the University of Washington reported that long-term exposure to air pollution led to an increase in cardiovascular events in women. This was not a study specifically of cardiac arrest but presumably cardiovascular events and arrests would correlate. Another study found an association with air pollution and EMS calls for complaint of chest pain. Though these environmental factors may play a role in triggering the event I suspect they have a small or no role to play in the likelihood of resuscitation.
Position of collapse (unknown association with survival):
It is possible that the position of collapse may be a determinant of outcome. This might occur in two ways. The position of collapse might delay the start of bystander CPR. For example, the patient could be wedged between the toilet and the bathtub, or crumpled on a stairway, or too heavy to move to the floor, or positioned in one of countless ways that preclude rapid initiation of CPR. A two-year study of 404 cardiac arrests in King County examined the reasons why telephone CPR was not delivered. Altogether 166 patients did not receive bystander CPR either from a trained bystander or as a result of dispatcher-assisted telephone CPR. Among these 166 cardiac arrests there were 8 situations (5%) in which the patient could not be moved into the proper position and thus CPR could not be performed. Another possible way in which position may inhibit successful resuscitation is if the person collapses with his or her head flexed leading to an obstructed airway. In this case the bystander attempting CPR might not appreciate the importance of straightening the head and neck and may have a difficult time providing mouth-to-mouth ventilation. Or if the person collapses with a flexed neck there may be no opportunity for air to move back and forth during agonal respirations. These abnormal respirations do not achieve normal ventilation but they may provide a limited amount of ventilation. A flexed neck will prevent even this suboptimal ventilation.
Decision to begin resuscitation (unknown association with survival):
There are no rules that can tell a rescuer whether the person in cardiac arrest has a good chance, a poor chance, or no chance of successful resuscitation. When the EMTs and paramedics arrive at the scene they must make a split decision to begin or not begin resuscitation. Even though EMTs and paramedics are required to start resuscitation on all patients in cardiac arrest there are clearly situations where resuscitation should not be attempted. I once had the occasion to review the do-not-attempt resuscitation medical protocols for each of the counties in Washington State. I was struck by the variation in these protocols. Most counties require the paramedic to contact the medical control doctor but some allowed the paramedic at the scene to make the decision. Some counties require findings of absolute incompatibility with life (decapitation, rigor mortis) not to begin resuscitation and others allow more subjective signs (cool temperature, lividity – blotchy skin due to pooling of blood in the dependent portions of the body) to be a basis for withholding resuscitation. With this variation in Washington State there undoubtedly exists similar differences throughout the county. The point here is that the decision not to begin resuscitation varies from system to system and may even vary among EMS personnel within an agency. These decisions can affect the denominator of attempted resuscitations and thus have an influence on the survival rate. A medical director of an unnamed Midwestern EMS program told me that his paramedics took virtually every person in cardiac arrest they encountered to the hospital, invariable with ongoing CPR. As a result of this policy his community had a very low survival rate. Enhancing or diminishing the denominator with policies about who to resuscitate will affect the community survival rate but of course will do little to save more patients in cardiac arrest. It does, however, make cross community comparisons very challenging since the criteria for when to initiate resuscitation differs. This is yet another reason to calculate community survival rates based on witnessed VF since patients in VF always have an attempted resuscitation. The only exception to the rule of always attempting resuscitation for someone in VF is if the person has a valid DNR.
Another subtle way in which the decision to resuscitate (in this case the decision to defibrillate) can vary is the specific automated external defibrillator (AED) used by a particular EMS system. Commonly used AEDs in the United States are manufactured by four companies and each has its own proprietary system for VF detection. One might think that VF detection algorithms would be standardized to a national or industry standard but such is not the case. Thus the specific criteria for VF detection can differ from community to community based solely on the type of AED used. Such a possibility is mostly speculative and a study directly comparing different AEDs would be very challenging to conduct. Nevertheless the possibility exists that subtle differences in VF detection algorithms can influence what types of VF are shockable and this in turn could influence the likelihood of survival.