Jeff Clawson, M.D.

Jeff Clawson, M.D.

Brett Patterson

Brett Patterson

Ask Doc

By Jeff Clawson, M.D. and Brett Patterson

The following question is a summary of inquiries regarding an experimental, 2-question algorithm, fostered from outside the Academy, intended to speed time-to-compressions in cardiac arrest cases.


Dear Dr. Clawson:

Has the Academy considered adopting the 2-question algorithm recommended in the recently published Emergency Medical Service Dispatch Cardiopulmonary Resuscitation Prearrival Instructions to Improve Survival From Out-of-Hospital Cardiac Arrest: A Scientific Statement From the American Heart Association? I hear this approach is being used in Seattle and Arizona in an effort to speed time-to-compressions in cardiac arrest cases. It advocates starting compressions immediately for any patient who is not conscious and not breathing normally. Is this a safe and effective approach?


Excellent question! The Academy has been aware of this approach for some time and our Medical Council of Standards has carefully reviewed and considered it. However, we cannot endorse it presently for a variety of reasons—including whether it is safe and effective.

As you are aware, a strong focus on early and uninterrupted compressions has brought attention to the early recognition of cardiac arrest by dispatchers, specifically with regard to those patients in the early stages of arrest when agonal respirations may produce caller confusion about a patient’s status of breathing.

Currently in the MPDS®, the recognition of absent or uncertain breathing by the caller prompts an immediate, ECHO-level response and subsequent Pre-Arrival CPR Instructions. If the caller is unsure or ambiguous when asked if the patient is breathing, the protocol directs an immediate response without delay or any further evaluation. Further interrogation about breathing is simply not appropriate. Only if the caller reports breathing, but the trained EMD remains suspicious about the potential of agonal breathing, should the Agonal Breathing Diagnostic Tool be used. Importantly, using this tool to verify breathing takes only 10–20 seconds.

After hearing of the work done by Dr. Tom Rea and Dr. Ben Bobrow at the ECCU conference last September, the Academy reviewed a cross section of cardiac arrest calls from several centers and interviewed QA and training personnel to determine if significant delays in getting hands-on-chest were occurring because of difficulty recognizing absent or agonal breathing. What we found was compelling and it has inspired changes, although not a change in wording of the Case Entry breathing question. Here is a synopsis of what we found and what the Academy is doing:

First, it was evident from the call review and the attitudes of the people we interviewed that some EMDs were misinterpreting the protocol’s intent with regard to the Agonal Breathing Diagnostic Tool (ABDxT). Rather than using the tool when the EMD was suspicious about a caller’s positive answer to the breathing question, EMDs were using the tool to verify uncertain or ambiguous answers from the caller. In other words, they were trying to validate that the person was not breathing, rather than trying to disprove a positive answer in doubt. The latter is why the tool was developed. Uncertain breathing, by rule, should be considered not breathing and requires no further diagnostic.

In summary:

Don’t use the ABDxT if caller unsure Use the ABDxT if you are unsure (not convinced)

We believe this delaying phenomenon has become cultural in some centers and the issue is being addressed through curriculum changes, instructor updates, continuing dispatch education, a pending Journal article, and changes to the protocol (reinforced by mandatory protocol update education), as follows:

The initial wording of the diagnostic tool, which instructs EMDs when to use the tool, has been changed from:

“When the patient is unconscious or not alert and is breathing abnormally or irregularly…”

To the following:

“When the patient is unconscious and breathing reported by the caller is questionable, or when mandated by the protocol…” (1)

Additionally, a new Rule (2), two Axioms (3 & 4), and an UNCERTAIN BREATHING definition(5) have been added to limit the use of the tool and encourage prompt compressions.

In addition to these Case Entry enhancements, several other changes have been implemented to shorten time-to-compressions in v.12.2, including a direct pathway from Case Entry to PAIs, bypassing Protocol 9 in the cardset; removal of the mouth/airway check in PAIs (v12.2); removal of the PAI breathing check in PAIs when the patient is reported as non-breathing/INEFFECTIVE breathing/UNCERTAIN BREATHING at Case Entry; and a new “Multiple Rescuers” pathway that gets hands on chest by the second rescuer before more detailed instructions are provided to the caller.

The affect of these changes on hands-on-chest time is expected to be substantial and is currently being studied.

In essence, we believe that the current standard of cardiac arrest detection in the MPDS is sound, and efforts to correct apparent, misguided attitudes about the use of the breathing diagnostic tool are well under way. We are also confident that the changes to v13.0 of the MPDS will provide additional time savings.

With regard to abandoning the current approach and adopting the suggested “Is s/he breathing normally?” approach across the board, we have several concerns:

1.Cardiac arrest is only a small portion of the unconscious patient call volume. From a London Ambulance Service data set of MPDS Determinant Codes with cardiac arrest outcome percentages, we know that out of 5,108 cases coded in the ECHO-level of Protocol 9: Cardiac Arrest (this includes 6 specific ECHO codes describing various presentations of ineffective breathing), 2,150 (42%) suffered cardiac arrests.

This is compared with 20,447 patients within the Chief Complaint of Protocol 31: Unconscious that were coded in the ECHO and top DELTA levels (still unconscious), in which only 491 (2.3%) had a CA outcome. One very interesting comparison is found in Protocol 31, looking at the top three codes where the patient is still unconscious—E-1, D-1, and D-2: (see table on page 8).

It appears that the protocol is quite accurate in separating Agonal/Ineffective Breathing states from Effective ones (even if abnormal).

2.There are Chief Complaints other than Protocol 31 where the patient may present unconscious but breathing abnormally: Protocols 2, 3, 4, 7, 8, 11, 12, 13, 15, 17, 23, 27, 29, 30, 33, 34, and 37. Several of these, (2, 3, 4, 7, 8, 15, 23, 27, 29, 34), by definition, have potential safety issues which, at Case Entry, are yet to be determined. Failure to identify certain things present or lurking at the scene through Key Questioning, requiring intervention through DLS instructions, as is routinely done by correct protocol use, could and, predictably, will harm patients, laypersons, and responding rescuers. A myriad of safety issues are currently addressed through Key Questioning on Chief-Complaint-specific protocols in the MPDS prior to hands-on bystander intervention.

3.There are many situations in which a patient may present as unconscious with abnormal breathing, but not be in cardiac arrest, and in desperate need of airway opening and management. Simply starting 600 compressions (or compressions only) is a death sentence, as many of these patients are comatose and will not respond or “fight back” as has been anecdotally claimed by some. Without an open airway, even passive breathing cannot be effective.

4.On medical cases in which the Chief Complaint is Unconsciousness, the very first question after Case Entry’s “Is s/he breathing?” is “Is her/his breathing completely normal?” We think this is a much better way of erring on the side of the patient because, if the answer to the first breathing query is “no” or “I’m not sure,” there is no reason to ask about normalcy, which is a judgment call by the caller—the correct Chief Complaint is identified and CPR commences. If the patient is breathing and the normalcy question becomes necessary, the time taken is only 3 or 4 seconds and an airway and second breathing assessment is always provided for patients that remain unconscious.

5.A risk/benefit ratio has been cited to support immediate and continuous compressions for unconscious patients with abnormal breathing. Proponents cite a 2010 study conducted in Seattle that looked at outcomes of patients who survived their unnecessary CPR experience. The Academy believes there are important limitations to this study and that further research is necessary before concluding that the clinical risk of unnecessary CPR is indeed minimal and inconsequential.

•Of the 247 patients assessed in the study, 5 (2%) suffered a fracture. On the surface, this seems an insignificant number, unless you are one of the five. However, the MPDS is used in over 3,500 communication centers worldwide and literally affects millions of patients. A standard practice of performing unnecessary compressions on roughly 50% of unconscious, abnormally breathing patients has the potential to cause tens of thousands of fractures, and millions over a relatively short period of time. This hardly equates to “Do no harm.”

•The patients assessed for side effects were alive. It is simply unknown whether or not the deceased patients were alive prior to their CPR experience. The side effects ascertained were related to physical injuries like broken ribs and visceral organ injury in survivors, not a failure to open airways, provide rescue breathing, or other more insidious internal compromises that may have resulted in, or hastened, death. There are a number of medical conditions where the compressions process would definitely make things worse, not better, for these patients—while not technically “injuring” them. This type of inappropriate treatment/lack of correct treatment has never been studied, and is reminiscent of the old question, “All those who are dead, please raise your hands.”

The MPDS has been very carefully evolved over the past 30-plus years to care for a very wide range of patients, and many lessons have been learned. One of those lessons has been to very carefully vet Proposals For Change (PFCs) for impact on protocol outside of the intended target because the complexity of dispatch protocol is such that one part almost certainly affects another. In short, while cardiac arrest survival is and always has been a top priority, it cannot be our only focus. And while focusing on improving outcomes for this relatively small fraction of our patients, we must remain very mindful of how we are affecting the rest of our patient population in never forgetting the 1st Law of Medical Dispatch: “First, do no harm."