Using EMD for Acute Stroke Identification - Official Academy Position Statement


Stroke could become one of thedefining elements of the driving force behind prioritized EMD. In recent years,thrombolytic therapy with tissue plasminogen activator (t-PA) to break downblood clots that precipitate “brain attacks,” has resulted in new hope forpeople who suffer stroke.


Clinical studies have shown that,if given during the early stages of a stroke, t-PA can indeed improve outcomesfor many stroke victims. The National Institute of Neurological Disorders andStroke (NINDS), t-PA Stroke Study Group, concluded in its 1995 paper, Tissueplasminogen activator for acute ischemic stroke, “Despite an increasedincidence of symptomatic intracerebral hemorrhage, treatment with intravenoust-PA within three hours of the onset of ischemic stroke improved clinicaloutcome at three months.”77 Appropriate t-PA administration improvesthe long-term outcome in a significant number of patients. Overall, t-PAtreatment is beneficial, despite the fact that it does cause seriousintracranial bleeding in some patients. For 11 percent of patients, if they getto hospital rapidly, and are treated by a stroke team using thrombolytic drugs,they’ll go home rather than to a long-term care facility.”78


Earlier interventions lead toimproved patient outcomes (as “time is muscle” during acute MI, “time is braincells” during stroke) and the goal should be a 90-minute time to treatment,rather than the latest acceptable treatment time of

three hours.


This short window for effectiveintervention has led to the widespread realization that, as stated in USAToday, “People having strokes should be treated with the same urgency as thosesuffering heart attacks.”


Patient groups are calling for “...amajor overhaul of outmoded stroke responses nationwide, upgrading stroke to atime-dependent, urgent medical emergency.”79 This publicity has ledto many agencies proposing that for all requests that report stroke-likesymptoms, response be upgraded to an obligatory lights-and-siren mode both tothe scene and during transport of the patient.


We believe that in most situationsthe small time savings of a L&S response to the scene will not alone make asignificant difference in stroke outcomes and that the initial response shouldbe the same as for chest pain in the absence of symptoms which suggest thepatient is arresting.            


For chest pain without thesesymptoms, charlie (not delta) determinants drive responses forthe cardiac age-range patient. We believe the similarities between stroke andacute MI warrant changes in protocol 28: Stroke (CVA), now that there is anintervention that significantly improves the outcome for many stroke patientsif given soon enough.


In stroke, the situation is in manyways analogous to acute MI. Therefore, we believe there should be more paritybetween the stroke and chest pain protocols by upgrading the Stroke dispatchprotocol to drive determinants that are similar to the charlie determinants on the Chest Pain protocol. However,protocol 28: Stroke does not warrant the deltadrivers we use in protocol 10 (chest pain). These delta determinants are driven by priority symptoms thatsuggest the patient is arresting and requires on-scene treatment within two orthree minutes.


If a patient is arresting,effective on-scene intervention (defibrillation) by first responders can makean enormous difference, but the window of opportunity is extremely short.On-scene intervention for stroke patients is still very limited and will likelyremain so until the on-scene use of neuroprotective drugs can be shown to beeffective. In the meantime, the existing charliedeterminants do not imply that stroke (or MI without evidence of arrest) is nottime-sensitive, but rather that it is reasonable to respond to stroke and acuteMI without L&S and to only respond with L&S when there is evidence thatthe chest pain patient is deteriorating or arresting.


A stroke patient should call 9-1-1as soon as symptoms appear, hence a need for increased public awareness andeducation. Pepe, et al., in their 1998 paper, Ensuring the chain of recoveryfor stroke in your community, point out that, “The sheer logistics of reachingand retrieving patients, even in a ‘scoop and run’ mode, leads to significant timelapses, a concept often unappreciated by those unfamiliar with the delivery ofemergency patient care in the out-of-hospital setting.”80 They showthat typical delivery of a patient to the ED will not take place for at least30 to 50 minutes after a stroke is recognized by the patient or bystander.(This breaks down to 15 minutes from caller recognition to EMS-professionalidentification of the stroke at the scene; 10 to 20 minutes on-scene time; and5 to 15 minutes transport time.) The NINDS guidelines: Rapid identification andtreatment of acute stroke,81 recommend that a physician shouldevaluate a stroke patient within 10 minutes of arrival at the ED, and that astroke specialist should be available or notified within 15 minutes. In orderto avoid giving thrombolytic treatment to a patient with a brain hemorrhage, aCT scan must be done. The CT scan should be started within 25 minutes, theresults available within 45 minutes, and treatment (when appropriate) startedwithin 60 minutes.


Hunt, et al.,36 measured that a typicalL&S run saves, on average, 43 seconds over a cold run. While this savingseems surprisingly small, there are other studies that appear to confirm theseresults.11 Even allowing for rural versus metropolitan differences,it seems likely that a L&S response is probably only saving one or twominutes over a cold response time.


In a rapid (15 to 20 minute) “scoopand run” delivery to a facility where the CT turn-around is also fast, an extracouple of minutes for a coldambulance journey to the scene is a small, but significant, portion of the timetaken before the t-PA injection-but the total time is still likely to allowtreatment of patients within a 90-minute window. In a conventional ALS responseand delivery (30 to 50 minutes) to a facility that does not meet the NINDSguidelines for CT scanning, the time lost by a coldresponse will very likely still be less than one percent of the total time andis therefore probably not alone a major source of delay.


While these time constraints do notreally suggest that the “time criticality” of thrombolytic stroke treatmentwarrants the extra risks of a L&S response, we do need a response mode thatconfers the same sense of urgency without the use of lights and sirens, i.e.,one that gets the patient there quickly, but without creating unnecessaryhazards for emergency personnel or civilians. Stroke is indeed a time-criticalmedical emergency, but it is saving hours (through public education) andminutes (through ED compliance with the NINDS time guidelines) that isimportant, not saving seconds that can place others at unnecessary risk.


            In1996, the Academy stated, “It is the temporary position of the Academy that, atthis time, no changes are necessary within protocol 28’s key questions, postdispatch instructions, or determinant codes. However, in light of changingscience, we recommend the addition of a new in additional information, tostate:


‘The adoption of in-hospitaladministration of clot dissolving drug therapies may require special assignmentof units equipped to evaluate patients for this therapy in areas adopting itfor trial and on-going treatment. Based on the current consensus recommendationto provide this treatment within 3 hours of the occurrence of stroke symptoms, the use of lights-and-siren (hot) responses is generally notindicated at the present time unless priority symptoms are present.’”


More recently, at the 1998 Councilof Research meeting the Council recommended several additional modificationsthat emphasize the availability of thrombolytic treatment, but also recognizedthat other elements in the survival chain (such as pre-notification of strokeresponse teams at the receiving hospital and eliminating unnecessary ALSprocedures at the scene) will have greater impact on stroke survival. Further,the Council stood by the 1996 recommendation that L&S responses are notgenerally indicated for stroke when priority symptoms are not present.


Since the 1998 Council of Researchmeeting, the Academy has continued to examine the stroke issue in great detail.For a number of reasons, the Council of Standards has now established anage-dependent triage for stroke and stroke-like symptoms. The new determinantsappear as follows:


C-1      Notalert

C-2      Abnormal breathing

C-3      Speech or movement problems

C-4      Numbness or tingling

C-5      Vision problems

C-6      Sudden onset of severeheadache

C-7      STROKE history

C-8      Breathingnormally  35

B-1      Unknown status (3rd party caller)

A-1      Breathingnormally < 35

The version 10.3 changes have beenapproved by the Council of Research and then were formally ratified by theCouncil of Standards—effective July 20, 1998.


The 10.3 changes will improve theMPDS in several ways. Statistically, the age range given would only fail toupgrade less than three percent of all strokes, and less than one percent ofall ischemic strokes, to the charlieresponse. The recommended ALS capability of the charlieresponse for stroke would increase the sense of urgency when compared tothe BLS capability of the current alpharesponse, and might even improve the patient’s reception at the ED.82Finally, upgrading stroke to an ALS-level response sets the scene for on-siteintervention when new neuroprotective drugs gain approval.


As with all MPDS determinant-drivenresponse modes, local medical control has the final word on who, when, and howfield personnel actually respond. Due to differences in response configurationoptions, available facilities, and local driving conditions, this situation maybe somewhat different among systems. While any delay in the stroke patient’schain of recovery is undesirable, in most circumstances the extra time taken bya cold response to the scene is avery small portion of the total (call-to-treatment) time. Stroke is atime-sensitive medical emergency and should be responded to with a sense ofurgency comparable to that used in response to chest pain suspected to be dueto myocardial infarction.


Therefore in the absence of additional prioritysymptoms, for both Chest Pain and Stroke the Academy recommends dispatching anALS-level response without the use of L&S (see “Response DeterminantMethodology” protocol). As more data on thrombolytic treatment becomesavailable, the precise nature of t-PA’s time criticality will become clearer.For now, it appears that the benefits of the small time savings of an L&Sresponse may not outweigh its additional risks.     


Fig. 6-15.        Priority dispatch evolves to accommodate changes intreatments and accepted standards of care. Stroke management has recentlyundergone radical changes due to the introduction of thromolytic therapies.This position statement presents the rationale for changes to the MPDS firstreleased in version 10.3 (now updated to v11.2). “Using EMD for Acute StrokeIdentification,” by Sinclair, R. and Marler, J. Reprinted from the JNAEMD.83