The Efficacy of Advanced Life Support: A Review of the Liter

[madmedic8522]

ace, thanks! I'm all for making improvements and changes out in the street as long as they benefit the pt and their out come, no matter what the situation is.

i think it gets back to our education and being responsible for it, which is why i always encourage my medic students (I'm a preceptor) to ask about their treatments at the hospital, so they/we can better understand what we are doing out here.

as far as our effectiveness, it gets back to what azcep stated, we need to take our profession more seriously.

madmedic

[Ace844]

ace, thanks! I'm all for making improvements and changes out in the street as long as they benefit the pt and their out come, no matter what the situation is.

i think it gets back to our education and being responsible for it, which is why i always encourage my medic students (I'm a preceptor) to ask about their treatments at the hospital, so they/we can better understand what we are doing out here.

as far as our effectiveness, it gets back to what azcep stated, we need to take our profession more seriously.

madmedic

"mad, & Everyone"

Good points, but my question is this. How do we take what we see as positive 'results, outcomes, etc..' from the anecdotal realm and 'show' that some of what we do 'actually works'? After we accomplish this how do we proactively create effective long-lasting local change as a catalyst and begining point for 'sweeping' postive changes across the board? It seems to me that a well written 'comprehnsive' un-biased study which may show this would be good for everyone, even the Fire Depts who dabble in ems...no?

Thanks,

ACE844

[AZCEP]

Get in good with the IRB?

Hopefully your medical control physician would be willing to help with gathering the data.

Audit of PCR's for the last 2-3 years, looking for specific complaints and treatments.

When you do something that works, keep copies of the documentation so when it happens again you can compare the two.

Just suggestions, but not outside the realm of possibility.

[Ace844]

Get in good with the IRB?

Hopefully your medical control physician would be willing to help with gathering the data.

Audit of PCR's for the last 2-3 years, looking for specific complaints and treatments.

When you do something that works, keep copies of the documentation so when it happens again you can compare the two.

Just suggestions, but not outside the realm of possibility.

Is this something that your system actively practices? If not why? Would it work if you tried it? Just curious.

ACE

[madmedic8522]

ace, sorry i got off the subject!!

it would seem to me that it would have to start at the level of our med con, whether it be their assisting us with a direction to follow or taking over the project all together. i would think that the information base would already be in place, it just needs to be studied!

after that not only a well written comprehensive unbiased study, but the adoption, by different educational programs available to EMS providers, up to and including continuing education.

now my question is: how do we narrow down the cause and effect, to get this started?

only an opinion!

madmedic

[Ridryder 911]

One needs to be aware of the various educational standards as well. Not all M.D.'s are created equal as well. There are so many variables that an efficient study would be hard.

I agree we need to look and examine our system. What we are doing and how we are doing it apparently is not working. Then again let's not fix what is not broken... hence the need for detail studies to find out what is good and bad.

EMS unfortunately was mis-designed. Paramedics placed in metro areas, where really the care was not as needed as in rural areas, where there is no medical care and that at least some medical care could be administered in a timely manner. So is true on helicopters.. they are not needed for larger urban areas with a fight distance of 3 -5 miles.. where they are as needed to a regional trauma center idistance is 35 to 60 miles. The true difference of the Golden Hour...

I am all for patient outcome studies; however one needs to be cautioned when reading and observing any study. Outcome studies has been popular the last few years, and unfortunately there has been many biased studies that are sometimes associated with them as well. Everyone needs to remember who is performing the study, why they are truly performing the study, and whom is sponsoring the study.

Although ethics and scientific statements should be always accountable and credible, it is not always that way.

I hope EMS continues to evaluated, but as well, I hope it is those in EMS that can perform those studies, and not some opinionated biased group. Good or bad. Studies will always be controversial and how we apply their findings is what really matters.

R/r 911

[Ace844]

Hello Everyone,

Here's some more literature on this topic.... I hope you enjoy them...

(1: Ann Emerg Med. 1988 Oct;17(10):1034-41. Related Articles @ Links

Comment in:

· Ann Emerg Med. 1989 Aug;18(8):909-10.

Effect of mobile paramedic units on outcome in patients with myocardial infarction.

Dean NC, Haug PJ, Hawker PJ.

Department of Emergency Medicine, LDS Hospital, Salt Lake City, Utah.)

To investigate the effect of mobile paramedic units on outcome, we prospectively studied for two years all patients with myocardial infarction admitted to the LDS Hospital emergency department who sought aid prior to cardiac arrest. One hundred thirty-four patients who received prehospital care from a mobile paramedic unit were compared with 101 patients who selected another means of initial care. Mortality, occurrence of life-threatening arrhythmias, and change in Killip class at 24 and 48 hours were the outcome variables. Data analysis by multiple logistic regression revealed that outcome was not improved, but a 29-minute median delay in hospital arrival occurred in paramedic-treated patients. Defibrillation was the only beneficial treatment performed by paramedics that could be identified. Current mobile paramedic unit procedures may need to be streamlined to eliminate the delay in hospital arrival resulting from extensive prehospital care.

http://www.iitd.ac.in/tripp/publications/p...ry/mvfiwoco.PDF

http://pdm.medicine.wisc.edu/20-4%20PDFs/Isenberg.pdf

http://www.cochrane-injuries.lshtm.ac.uk/P...INALReport2.pdf

http://www.hta.nhsweb.nhs.uk/execsumm/SUMM217.HTM

http://pdm.medicine.wisc.edu/19-4%20PDFs/Mock.pdf

http://www.medscape.com/viewarticle/448918

http://www.nhtsa.dot.gov/people/injury/ems...ppendices-a.htm

http://bmj.bmjjournals.com/cgi/content/full/324/7346/1135

http://www.who.int/violence_injury_prevent...l/en/index.html

www.aetmis.gouv.qc.ca/site/download. php?3f84607e6bb5c2dbe12e736439956b41

http://pdm.medicine.wisc.edu/birnbaum4.htm

Also, for such an important topic to our profession, you guys sure don't have much to say...why? Isn't this important to you and the group at large??

Out here,

ACE

[Ace844]

(Journal of Emergency Medicine

Volume 31 @ Issue 1 , July 2006, Pages 1-5

doi:10.1016/j.jemermed.2005.08.007

Copyright © 2006 Elsevier Inc. All rights reserved.

Original contribution

How well do paramedics predict admission to the hospital? A prospective study

Saul D. Levine MD⁎, , Christopher B. Colwell MD⁎, Peter T. Pons MD⁎, Craig Gravitz EMT-P, RN†, Jason S. Haukoos MD, MS⁎ and Kevin E. McVaney MD⁎

†Paramedic Division, Denver Health Medical Center, Denver, Colorado

⁎Department of Emergency Medicine, Denver Health Medical Center, Denver, Colorado

Received 2 July 2004; revised 4 April 2005; accepted 1 August 2005. Available online 22 June 2006.)

Abstract

A study was designed to determine whether paramedics accurately predict which patients will require admission to the hospital, and in those requiring admission, whether they will need a ward bed or intensive care unit (ICU) monitoring. This prospective, cross-sectional study of consecutive Emergency Medical Service (EMS) transport patients was conducted at an urban city hospital. Paramedics were asked to predict if the patient they were transporting would require admission to the hospital, and if so, whether that patient would be admitted to a ward bed or require an ICU bed. Predictions were compared to actual patient disposition. During the study period, 1349 patients were transported to our hospital. Questionnaires were submitted in 985 cases (73%) and complete data were available for 952 (97%) of these patients. Paramedics predicted 202 (22%) patients would be admitted to the hospital, of whom 124 (61%) would go the ward and 78 (39%) would require intensive care. The actual overall admission rate was 21%, although the sensitivity of predicting any admission was 62% with a positive prediction value (PPV) of 59%. Further, the paramedics were able to predict admission to intensive care with a sensitivity of 68% and PPV of 50%. It is concluded that paramedics have very limited ability to predict whether transported patients require admission and the level of required care. In our EMS system, the prehospital diversion policies should not be based solely on paramedic determination.

Introduction

The nationwide problem of Emergency Department (ED) and hospital overcrowding has brought the practice of ambulance diversion to the forefront. Diversion has grown increasingly complex and many hospitals now have numerous different categories of diversion, such as ED, intensive care unit (ICU), trauma, obstetric, pediatric ICU, ward, and psychiatric diversions. In order for a particular type of diversion category to be effectively acted upon, an implied assumption is made that when a hospital declares a diversion, the prehospital caregiver is able to accurately triage patients and predict the need for admission and level of care.

Several studies have evaluated the role of prehospital providers in Emergency Medical Services (EMS) systems and the ability of paramedics to determine clinical diagnosis and prognosis (1). Some investigators have found EMS providers to be accurate in their ability to triage patients, whereas others have found accuracy rates to be unacceptable (2, 3, 4, 5, 6, 7, 8 and 9). A number of publications have raised the question of whether paramedics can safely predict if a patient needs transport at all (4, 10 and 11). If field personnel are able to accurately predict which patients will require admission, and particularly the need for intensive care, specific EMS diversion categories may be possible.

The purpose of this study was to determine whether paramedics can accurately determine which patients will require admission to the hospital, and in those who are admitted, whether they will require admission to a ward bed or an ICU.

Methods

Denver Health Medical Center (DHMC) is an urban county hospital and level I trauma center that accepts both medical and trauma EMS transports. The annual ED census at DHMC is approximately 55,000. The overall admission rate from the ED is approximately 18% and, of those, approximately 25% are admitted to the operating room (OR) or ICU. The Denver Health and Hospital Authority is the agency contracted to provide 911 emergency medical service to the City and County of Denver using a two-tier system in which fire serves as the first basic life support tier and paramedics serve as the second advanced life support tier. The Denver Health Paramedic Division serves a population of approximately 550,000 based upon year 2000 census data with 136 paramedics, covering a geographical area of approximately 150 square miles. Paramedic Division ambulances respond to approximately 65,000 requests for emergency medical assistance annually and transport approximately 45,000 patients to Denver area hospitals. At the time of the study, Denver Health employed only Emergency Medical Technician (EMT) paramedics and not EMT basics. The average duration of paramedic experience at the time of the study was 7.4 years, with a range of 0.3 to 28 years.

A prospective, cross-sectional study was performed that surveyed paramedics treating all EMS patients transported to DHMC from June 18 to July 18, 2001. All patients transported to DHMC were included in the study. Patients transported to other hospitals were excluded. The paramedic attending to the patient on each transport was asked to complete a standardized form upon arrival at the hospital. The forms were collected and stored in a locked data collection box in the ED.

Data collected included the EMS trip number, a field diagnosis, and whether or not the paramedic believed the patient would be admitted. If the paramedic did predict admission, the further distinction of “ward” or “ICU” admission or “other” was made. Ward admission was defined as those patients admitted to any inpatient service including medicine, surgery, neurosurgery, orthopedics, obstetrics/gynecology, or psychiatry, and ICU admission was defined as those patients admitted to the medical, surgical, or coronary intensive care unit. Patients who were thought to require a bed with telemetry monitoring, but not intensive care unit monitoring, were classified as ward admissions. The category “other” included the morgue, the psychiatric emergency area, the obstetric screening room, or leaving against medical advice (AMA). For the purposes of the study, admission was defined as actual transfer to an in-hospital bed, not just the decision by the physician that the patient should be admitted.

Data from the study form were entered into a Microsoft Excel spreadsheet (Microsoft Corporation, Redmond, WA) for analysis. Sensitivities, specificities, positive predictive values (PPVs) and negative predictive values (NPVs) with 95% confidence intervals (CIs) were calculated using Stata Version 8 (Stata Corporation, College Station, TX).

The study was approved by our Institutional Review Board and a waiver of informed consent was granted.

Results

According to computer-assisted dispatch records, 1349 patients were transported to DHMC during the 1-month study period. This accounted for 55% of all EMS transports performed in the City and County of Denver by the Paramedic Division during the same time period. Research forms were completed on 985 patients (73%) and complete data were available for 952 (97%) of these. Of the 952 patients, 533 (56%) were men and 847 (89%) were over the age of 17 years. Twenty patients (2%) who were triaged immediately upon arrival from the ED to the psychiatric emergency care area, the obstetric screening room, or were dead on arrival and sent to the morgue were excluded, leaving a total of 932 patients included in the final analysis.

Paramedics predicted 202 (22%; 95% CI: 19%–25%) patients would require admission and 730 (78%; 95% CI: 76%–81%) would be discharged from the ED. Of the 202 patients expected to be admitted, 124 (61%) were felt to be candidates for admission to a ward bed and 78 (39%) were felt to require an ICU bed (Table 1).

Table 1.

Paramedic Predictions of Patient Disposition from the ED vs. the Actual Patient Disposition Actual disposition from ED (n)

ICU admission Ward admission Discharge from ED Total

Paramedic predictions of patient disposition from ED ICU admission 39 24 15 78

Ward admission 8 49 67 124

Discharge from ED 10 65 655 730

Total 57 138 737 932

ED = Emergency Department.

Of 78 patients predicted to go to the ICU, 63 (81%; 95% CI: 70%–89%) were admitted to the hospital. However, only 39 (50%; 95% CI: 38%–62%) were actually admitted to the ICU, whereas 24 (31%; 95% CI: 21%–42%) went to a ward bed, and 15 (19%; 95% CI: 11%–31%) were discharged home.

Of the 124 patients paramedics believed would be admitted to a ward bed, 67 (54%; 95% CI: 45%–63%) were discharged and 8 (7%; 95% CI: 3%–12%) were admitted to the ICU. Of the 730 patients paramedics felt would be discharged, 75 (10%; 95% CI: 8%–13%) were admitted, of whom 10 (13%; 95% CI: 7%–23%) were admitted to the ICU.

The sensitivity, specificity, PPV, and NPV of the paramedics’ triage decisions for patient admission to the hospital (all admissions), ICU, and ward compared with actual outcomes are shown in Table 2, Table 3 and Table 4. Similarly, the sensitivity, specificity, PPV, and NPV of the paramedics’ triage decisions for patient admission to the hospital (all admissions) based upon the nature of the emergency (medical vs. trauma) are shown in Table 5 (29 patients were not included due to incomplete data needed to categorize the nature of their emergency). Analysis of the medical vs. trauma admission data was post hoc and not part of the original study design.

Table 2.

Paramedic Predictions for Admission to the Hospital (ICU plus Ward) vs. Actual Admission to the Hospital Actual disposition % (95% CI)

Admission Discharge Total

Sens. 62% (54–68)

Paramedic predictions of disposition Admission 120 82 202 Spec. 89% (86–91)

Discharge 75 655 730 PPV 59% (52–66)

Total 195 737 932 NPV 90% (87–92)

CI = confidence interval; PPV = positive predictive value; NPV = negative predictive value.

Table 3.

Paramedic Predictions for Admission to the Intensive Care Unit vs. Actual Admission to the ICU Actual disposition % (95% CI)

ICU admission Non-ICU admission Total

Sens. 68% (55–80)

Paramedic predictions of disposition ICU admission 39 39 78 Spec. 96% (94–97)

Non-ICU admission 18 836 854 PPV 50% (39–62)

Total 57 875 932 NPV 98% (97–99)

CI = confidence interval; PPV = positive predictive value; NPV = negative predictive value.

Table 4.

Paramedic Predictions for Admission to the Ward (Compared to All Other Outcomes: ICU Admission and Discharge from the ED) vs. Actual Admission to the Ward Actual disposition % (95% CI)

Requires ward admission All other dispositions⁎ Total

Sens. 36% (28–44)

Paramedic predictions of disposition Requires ward admission 49 75 124 Spec. 91% (88–93)

All other dispositions⁎ 89 719 808 PPV 40% (31–49)

Total 138 794 932 NPV 89% (87–91)

CI = confidence interval; PPV = positive predictive value; NPV = negative predictive value.

⁎ Admit to ICU or discharge from ED.

Table 5.

Paramedic Predictions for Admission to the Hospital (ICU and Ward) vs. Actual Admission Stratified by Nature of the Admission (Medical or Trauma)⁎ Actual disposition % (95% CI)

Requires admission† Discharge Total

Medical Sens. 53% (43–62)

Paramedic predictions of disposition Requires admission† 60 40 100 Spec. 86% (81–89)

Discharge 54 236 290 PPV 60% (50–70)

Total 114 276 390 NPV 81% (76–86)

Trauma Sens. 71% (60–80)

Paramedic predictions of disposition Requires admission† 61 41 102 Spec. 91% (88–93)

Discharge 25 406 431 PPV 60% (50–69)

Total 86 447 533 NPV 94% (92–96)

CI = confidence interval; PPV = positive predictive value; NPV = negative predictive value.

⁎ Twenty-nine patients were not included as there were insufficient data to categorize the nature of their emergency.

† Admit to ICU or ward bed.

Discussion

The problem of ED overcrowding is well documented (12, 13, 14 and 15). One result of overcrowding is for hospitals to declare a diversion status in an effort to direct patients being transported by ambulance to other hospitals. This places a greater burden on EMS crews to determine which patients are appropriate for particular or specific hospitals. Some categories of hospital diversion appear straight forward (ED divert) whereas others are less well defined (ICU, ward, psychiatry, CT, etc.). Diversions also have been shown to increase transit times and distances traveled (8).

Novel approaches to hospital diversion could be implemented if prehospital caregivers were capable of accurately categorizing patients in terms of level of needed care. Studying prehospital triage decisions has been problematic, however, due to patient variability and the lack of objective triage protocols. Richards and Ferrall demonstrated certain patient characteristics and chief complaints associated with increased admission rates, which potentially make hospital admission somewhat more predictable (4). These researchers attempted to evaluate paramedics in terms of their “gut feeling.” Our study also asked paramedics to subjectively triage patients.

Hauswald found paramedics are unable to determine which patients do not need transport (10). Similar studies have also demonstrated a poor agreement between predicted triage and actual disposition or diagnosis (6, 7, 9 and 16). Other research, however, demonstrates excellent paramedic diagnostic accuracy by chief complaint or organ system (2 and 3).

The paramedics in our study were asked to predict whether or not patients they were transporting to the ED required admission to the hospital and, if so, the level of required care at the hospital. This study demonstrates that paramedics have a modest ability to predict the need for admission to an ICU and limited ability to determine the need for admission to a non-ICU bed. Of note, the paramedics were somewhat better able to predict these bed needs for trauma patients when compared with non-trauma patients. Only half of the patients in our study with a paramedic anticipated ICU requirement ended up actually admitted to the intensive care unit. In addition, there were 10 patients the paramedic thought would be discharged home who were ultimately admitted to the ICU.

There are several limitations to our study. Selection bias may have resulted because 27% of EMS transports to the hospital did not have a data collection form completed and because we did not capture EMS transports to other receiving hospitals.

The decision to admit a patient is based upon numerous historical and clinical factors that are not routinely available to paramedics. In addition, the decision to admit is not one that paramedics routinely participate in or have formal training about, thus limiting their ability to make accurate predictions. Variable paramedic experience, particularly history of working in an ED, may have affected their ability to make an accurate admission prediction.

Further bias, in the form of incorporation bias, may have resulted if paramedics discussed the cases with physicians, nurses or technicians in the ED, as well as their direct observation of continued care before the admission prediction card was submitted. We used the Emergency Physician’s decision (in conjunction with consultants in some cases) as the criterion standard to which the paramedic prediction was compared. Discrepancies in the patient’s disposition between different physicians or consultants may have also biased the results.

Conclusion

Based on our investigation, paramedics have limited ability to predict whether transported patients need admission to the hospital and, more specifically, whether they require intensive care or ward admission. This has important implications about the limitations of diversion strategies that rely on paramedic prediction, and whether prehospital diversion policies should be based on paramedic determination. Further work to develop, assess and utilize hospital diversion guidelines and categories as well as the ability of prehospital providers to accurately implement those categories is needed (17).

References

1 K.W. Neely, M.E.R. Drake and J.C. Moorhead et al., Multiple options and unique pathways a new direction for EMS?, Ann Emerg Med 30 (1997), pp. 797–799. SummaryPlus | Full Text + Links | PDF (303 K)

2 J.J. Schaider, J.C. Riccio and R.J. Rydman et al., Paramedic diagnostic accuracy for patients complaining of chest pain or shortness of breath, Prehospital Disaster Med 10 (1995), pp. 245–250. Abstract-MEDLINE

3 R. Sahni, J.J. Meneqazzi and V.N. Mosesso Jr, Paramedic evaluation of clinical indicators of cervical spine injury, Prehosp Emerg Care 1 (1997), pp. 16–18. Abstract-MEDLINE

4 J.R. Richards and S.J. Ferrall, Triage ability of emergency medical services providers and patient disposition a prospective study, Prehospital Disaster Med 14 (1999), pp. 174–179. Abstract-MEDLINE

5 K. Qazi, J.A. Kempf and N.C. Christopher et al., Paramedic judgment of the need of trauma team activation for pediatric patients, Acad Emerg Med 5 (1998), pp. 1002–1007. Abstract-MEDLINE | Abstract-EMBASE

6 J.P. Santoro, P. Smith and T.J. Mader et al., Accuracy of field diagnosis by paramedics [abstract], Acad Emerg Med 5 (1998), p. 390.

7 S.M. Sasser, M. Brokaw and T.H. Blackwell, Paramedic vs. emergency physician decisions regarding the need for emergency department evaluation [abstract], Acad Emerg Med 5 (1998), p. 391.

8 K.W. Neely, R.L. Norton and G.P. Young, The effect of hospital resource unavailability and ambulance diversions on the EMS system, Prehospital Disaster Med 9 (1994), pp. 172–176.

9 J.E. Pointer, M.A. Levitt and J.C. Young et al., Can paramedics using guidelines accurately triage patients?, Ann Emerg Med 38 (2001), pp. 268–277. Abstract | PDF (99 K)

10 M. Hauswald, Can paramedics safely decide which patients do not need ambulance transport or emergency department care?, Prehosp Emerg Care 6 (2002), pp. 383–386. Abstract

11 S. Silvestri, S.G. Rothrock and D. Kennedy et al., Can paramedics accurately identify patients who do not require emergency department care?, Prehosp Emerg Care 6 (2002), pp. 387–390. Abstract

12 G.M. O’Brien, M.D. Stein and S. Zierler et al., Use of the ED as a regular source of care associated factors beyond lack of health care insurance, Ann Emerg Med 30 (1997), pp. 286–291.

13 R.A. Hayward, A.M. Bernard and H.E. Freeman et al., Regular source of ambulatory care and access to health services, Am J Public Health 81 (1991), pp. 434–438. Abstract-EMBASE | Abstract-MEDLINE

14 K. Grumbach, D. Keane and A. Bindman, Primary care and public emergency department overcrowding, Am J Public Health 83 (1993), pp. 372–378. Abstract-EMBASE | Abstract-MEDLINE

15 R.A. Lowe, G.P. Young and B. Reinke et al., Indigent health care in emergency medicine an academic perspective, Ann Emerg Med 20 (1991), pp. 790–794. Abstract

16 W.G. Baxt, C.C. Berry and M.D. Epperson et al., The failure of prehospital prediction rules to classify trauma patients accurately, Ann Emerg Med 18 (1989), pp. 1–8. Abstract

17 J.P. Campbell, V.A. Maxey and W.A. Watson, Hawthorne effect implications for prehospital research, Ann Intern Med 26 (1995), pp. 590–594. SummaryPlus | Full Text + Links | PDF (508 K)

Original Contributions is coordinated by John Marx, md, of Carolinas Medical Center, Charlotte, North Carolina

Reprint Address: Saul D. Levine, md, Department of Emergency Medicine, University of California, San Diego, 200 W. Arbor Drive, Box 8676, San Diego, CA 92103-8676

[Ace844]

(American Heart Journal

Volume 151 @ Issue 6 , June 2006, Pages 1255.e1-1255.e5

doi:10.1016/j.ahj.2006.03.014

Copyright © 2006 Elsevier Inc. All rights reserved.

Clinical Investigation

Feasibility and benefit of prehospital diagnosis, triage, and therapy by paramedics only in patients who are candidates for primary angioplasty for acute myocardial infarction

Arnoud W.J. van 't Hof MD, PhD, a, b, , Saman Rasoul MDa, b, Henri van de Wetering Ma-ANPa, b, Nicolette Ernst MD, PhDa, b, Harry Suryapranata MD, PhDa, b, Jan C.A. Hoorntje MD, PhDa, b, Jan-Henk E. Dambrink MD, PhDa, b, Marcel Gosselink MD, PhDa, b, Felix Zijlstra MD, PhDa, b, Jan Paul Ottervanger MD, PhDa, b, Menko-Jan de Boer MD, PhDa, b and on behalf of the On-TIME study groupa

aIsala Klinieken, Zwolle, The Netherlands

bAmbulance dienst regio IJssel Vecht, The Netherlands

Received 13 October 2005; accepted 20 March 2006. Available online 15 June 2006.)

Background

Despite data showing that time to treatment is very important in ST-elevation myocardial infarct patients, unacceptable long delays to reperfusion remain present in daily life practice. We sought to evaluate the feasibility and effect of improving logistics by early infarct diagnosis in the ambulance and immediate triage to a percutaneous coronary intervention (PCI) center performed by paramedics only without interference of a physician.

Methods

In the On-TIME study, 209 patients were included after prehospital infarct diagnosis and triage in the ambulance (ambulance group, n = 209). Infarct diagnosis was made by highly trained paramedics with the help of a computerized electrocardiographic algorithm. The accuracy of diagnosis, time to treatment, left ventricular function, and clinical outcome were compared with the patients who were diagnosed and triaged at a referral non-PCI center (referred group, n = 258). Left ventricular function was assessed before discharge using a nuclear technique.

Results

Acute myocardial infarction was accurately diagnosed in 95% of patients in the ambulance group, as compared with 99% in the referred group (P = .01). The percentage of patients in whom pharmacologic pretreatment (heparin, aspirin, tirofiban, or placebo) was initiated in the ambulance within 90 minutes after the onset of symptoms was 59% in the ambulance group versus 43% in the referred group (P < .01). A left ventricular ejection fraction of <40% was present in 25% in the ambulance group, as compared with 38% in the referred group (P = .013). After multivariate analysis, ambulance triage was independently associated with a left ventricular ejection fraction >40% and a favorable long-term clinical outcome.

Conclusions

Early, prehospital infarct diagnosis, triage, and therapy in the ambulance with direct transportation to the nearest PCI center, performed by trained paramedics only, is feasible in 95% of patients. Ambulance triage resulted in earlier diagnosis and initiation of therapy and was independently associated with a better left ventricular function and clinical outcome, as compared with triage and transportation from a referral non-PCI center.

Article Outline

Background

Patients and methods

Statistical analysis

Results

Univariate and multivariate analysis

Discussion

Prehospital care

Mechanism of benefit

Limitations

Conclusion

Appendix A. The On-TIME study group

References

Background

Primary coronary angioplasty has been shown to be a very effective reperfusion modality in patients with acute myocardial infarction (MI),1 and 2 even when additional transport is necessary to a percutaneous coronary intervention (PCI) center.3 However, time to reperfusion is often considerably longer when the data from registries are reported instead of the data from randomized trials. In a large registry in the United States, only 5% of patients had a door-to-balloon time <60 minutes.4 Although door-to-balloon times decreased from 90 to 70 minutes in the most recent Euro Heart Survey analysis on acute coronary syndromes,5 many patients still have unacceptable long times to treatment. Some of the delay might be prevented by prehospital infarct diagnosis and triage, selecting patients who are candidates for primary angioplasty who can immediately be transported to a PCI center. However, it is unclear whether this can be reliably performed by paramedics without interference of a physician.

This study evaluates the feasibility and benefit of prehospital infarct diagnosis and triage in the ambulance by paramedics only and compares outcome with triage at a referral non-PCI center.

Patients and methods

The design, inclusion and exclusion criteria, and main findings of the On-TIME study have been described previously.6 In this study, 209 patients were included after prehospital infarct diagnosis and triage in the ambulance (ambulance group, n = 209). The accuracy of diagnosis, time to treatment, the quality of reperfusion, left ventricular function, and clinical outcome was compared with the patients who were diagnosed and triaged at a referral non-PCI center (referred group, n = 258). The presence of an ambulance equipped with 12-lead electrocardiogram (ECG) diagnostic facilities determined whether a patient was transported to a non-PCI center first or immediately referred to the PCI center.

Emergency transportation was performed after a telephone call, done either by the ambulance driver (ambulance group) or the referring physician (referred group), with the aim to prepare the arrival of the patient directly at the catheterization laboratory. Recruitment and randomization in the ambulance was initiated only after a period of training in prehospital infarct diagnosis and care for at least 6 months. All paramedics had at least 2 years of training at a (cardiac) intensive care unit and work strictly according to board certified protocols. Computer diagnosis in the ambulance was made based upon a fixed algorithm, which has been previously described.7 and 8 Before transportation, all patients received 500 mg of aspirin, 5000 IU of unfractionated heparin intravenously, and study medication (tirofiban or placebo). Post PCI, all patients were treated with Clopidogrel (300 mg loading dose followed by 75 mg daily for 1 month), aspirin, β-blockade, statin, and angiotensin-converting enzyme inhibition.

Time from symptom onset to diagnosis was defined as the time from the onset of symptoms to the time of the diagnostic ECG, either made in the ambulance or at the non-PCI center. Total ischemic time was defined as the time from symptom onset to first balloon inflation.

All angiographic and electrocardiographic parameters were analyzed by an independent core laboratory (Diagram BV, Zwolle, The Netherlands) and scored by observers who were unaware of randomization and outcome data. A correct infarct diagnosis was defined as prolonged chest pain with typical evolutionary electrocardiographic changes, coupled with an unstable coronary lesion on the angiogram and a rise in creatine kinase of >3 times the upper limit of normal. Prevented MI was defined as a correct infarct diagnosis without a rise in creatine kinase of >3 times the upper limit of normal. A false-positive infarct diagnosis was made in those patients who did not meet the criteria for a correct or prevented MI. Left ventricular ejection fraction (LVEF) was assessed in the patients recruited in the Zwolle area only and was measured with a radionuclide technique before discharge, as previously described.1 The data on LVEF were gathered by a specialist in nuclear medicine, who was blinded to the clinical data. Clinical outcome, being the incidence of death or recurrent MI, was assessed at 1-year follow-up.

Statistical analysis

Statistical analysis was performed with the SPSS 10.0 statistical package (SPSS, Chicago, IL). All noncontinuous angiographic variables were analyzed using the χ2 or Fisher exact test. Continuous variables were analyzed using analysis of variance or Mann-Whitney U test. The continuous variables time from symptom onset to diagnosis and total ischemic time were dichotomized, based upon the median value (94 minutes and 188 minutes, respectively). Key outcome parameters were the percentage of patients with a correct infarct diagnosis, LVEF before discharge, and the 1-year incidence of death or recurrent MI. Risk stratification was based upon the previously described TIMI risk criteria.9 To assess independent predictors of left ventricular function and clinical outcome, multivariate analysis was performed using stepwise logistic regression. All parameters that were significantly different between the groups (sex and distance to the PCI center) and all baseline characteristics associated with left ventricular function or clinical outcome with a P ≤ .10 were entered into the model. The cutoff value of 40% for determination of a poor left ventricular function was based upon a value less than the 25% percentile (38%).

Results

A correct infarct diagnosis was present in 95% of patients triaged by ambulance personnel, as compared with 99% triaged at a referral center (P = .01). The 14 patients with a false-positive infarct diagnosis either had severe aortic stenosis (n = 3), left ventricular hypertrophy with strain (n = 1), previous MI with persistent ST elevation (n = 2), atrial fibrillation with early repolarization (n= 2), pericarditis (n = 2), gastrointestinal disease (n = 2), or others (n = 2).

Baseline characteristics of the ambulance and the referred group are described in Table I and were not significantly different, except for sex and distance to the PCI center (45 vs 24 km, P < .001). The percentage of patients who were diagnosed within 90 minutes or underwent balloon inflation within 3 hours after the onset of symptoms was significantly higher in patients after triage in the ambulance (Table I). Both initial perfusion of the infarct-related vessel and final myocardial perfusion as assessed by the myocardial blush grade was better in the patients triaged in the ambulance (Table I).

Table I.

Baseline, electrocardiographic and angiographic characteristics Referred (n = 258) Ambulance (n = 209) P

Baseline

Age (y ± SD) 62 ± 11 61 ± 11 .88

Male sex 77% 85% .03

Diabetes 9% 13% .17

Hypertension 29% 26% .37

Smoking⁎ 66% 64% .68

Anterior MI 46% 46% .90

Previous MI 10% 7% .24

Previous CABG 1.6% 2.4% .52

Previous PCI 4.3% 5.8% .46

Killip class >1† 16% 19% .39

TIMI risk score >3‡ 42% 46% .38

SO to diagnosis <90 min 43% 59% .001

Ischemic time <3 h 29% 52% .001

Electrocardiographic

Cum ST elevation (mm) 10 ± 7 11 ± 8 0.24

Cum ST deviation (mm) 15 ± 9 15 ± 9 0.81

Angiographic

Single-vessel disease 41% 49% .11

Pre PCI

TIMI 2, 3 35% 44% .04

Post PCI

TIMI 3 91% 91% .97

MBG 3 47% 59% .02

CABG, Coronary artery bypass grafting; Cuon, cumulative; MBG, myocardial blush grade as previously defined10; SO, symptom onset.

⁎ Current or previous smoking.

† Defined as systolic blood pressure <100 mm Hg or heart rate >100 per minute.

‡ TIMI risk score as described by Morrow et al.8

Prevented MI was present in 15% in the ambulance group and in 10% in the referred group (P = .08). Left ventricular ejection fraction was measured before discharge in 318 (75%) of the 426 patients with a confirmed diagnosis of acute MI, recruited in the Zwolle area. The LVEF was 46% ± 10% in the ambulance group, as compared with 44% ± 11% in the referred group (P = .17). An LVEF <40% was present in 25% in the ambulance group, as compared with 38% in the referred group (P = .013).

Univariate and multivariate analysis

Left ventricular function. Univariate predictors of a poor left ventricular function were male sex, anterior infarct location, the presence of diabetes, an initial heart rate >100 beat/min, a TIMI risk score >2, and nonambulance triage. After multivariate analysis, only male sex, anterior infarct location, and nonambulance triage were independently associated with a poor left ventricular function (Table II).

Table II.

Predictors of Left Ventricular Function and Clinical Outcome Variables OR 95% CI P

Left Ventricular Ejection Fraction < 40%

Univariate

Male gender 1.8 0.9-3.6 .08

Diabetes 2.1 1.0-4.4 .04

Anterior infarct location 6.3 3.7-10.8 <.01

Heart rate > 100/min 3.1 1.0-9.3 .03

TIMI risk score > 2 5.9 3.2-10.8 <.01

Distance < 41 km .8 0.5-1.4 .54

Ambulance triage .5 0.3-0.8 .01

Single vessel disease 1.0 0.7-1.8 .71

Multivariate

Male gender 2.6 1.2-5.5 .02

Anterior infarct location 3.7 1.6-8.9 <.01

Ambulance triage 0.4 0.2-0.8 .01

Death or recurrent myocardial infarction at 1-y follow-up

Univariate

Age (per y) 1.05 1.0-1.1 .02

Male gender 1.8 0.6-5.2 .29

Hypertension 2.7 1.3-5.5 .01

Anterior infarct location 1.9 0.9-3.2 .08

Heart rate > 100/min 3.0 1.2-7.9 .03

TIMI risk score > 2 2.2 1.1-4.5 .03

Single vessel disease 0.4 0.2-0.9 .03

Distance < 41 km 0.7 0.3-1.4 .29

Ambulance triage 0.3 0.1-0.7 .01

Multivariate

Ambulance triage 0.3 0.1-0.9 .03

Hypertension 2.5 1.1-5.5 .03

Single vessel disease 0.4 0.2-1.0 .05

Anterior infarct location 2.3 1.0-5.3 .05

Clinical outcome. At 30-day follow-up, 1.0% of patients with a correct diagnosis of acute MI had died in the ambulance group, as compared with 3.2% in the referred patients (P = .2). At 1-year follow-up, total mortality and the combined incidence of death or recurrent MI was significantly lower in the ambulance group (2.1% vs 6.0%, P = .04 and 3.6% vs 10.5%, P = .006, respectively). After multivariate analysis, ambulance triage remained an independent predictor of survival free from death or MI at 1-year follow-up (Table II).

Discussion

This study showed that correct infarct diagnosis, triage, and therapy can be performed by highly trained paramedics only, without interference of a physician, in 95% of patients. Furthermore, it shows that ambulance triage is an independent predictor of an LVEF >40% and a favorable clinical outcome, as compared with triage at a referral non-PCI center. The mechanism of the beneficial effect seems to be the combination of earlier diagnosis, earlier initiation of pharmacologic pretreatment, and a better initial and final myocardial perfusion. It showed that further streamlining of logistics, in which an unnecessary visit of a non-PCI center is prevented, saves time and is associated with a better left ventricular function and clinical outcome.

Prehospital care

The most recent version of the American College of Cardiology/American Heart Association guidelines for the treatment of patients with ST-elevation acute MI state that primary angioplasty is the preferred reperfusion strategy if it can be performed within 90 minutes after first patient contact.11 However, it was recently shown that only 5% of transferred patients in the United States have a door-to-balloon time <90 minutes.4 In addition, in Europe, every day door-to-balloon times are considerably longer than those reported in randomized controlled trials.12 This long delay prevents further development of primary angioplasty programs and is often stated as an excuse for treating the patients with thrombolysis or a combined pharmacoinvasive strategy, which, so far, has not been supported by evidence from randomized controlled trials. Many obstacles have been associated with the lack of development of a prehospital diagnosis and care program of patients with an acute MI. The fear of a false-positive infarct identification when diagnosis is made without interference of a physician is one of them. This study shows that the combination of trained paramedics, together with a validated computerized ECG algorithm, results in a correct diagnosis in 95% of cases.

Mechanism of benefit

It is unlikely that the reduced total ischemic time alone may sufficiently explain the beneficial effect on left ventricular function. The moment of time saving is important: saving 30 minutes in the very early phase of MI is expected to result in a larger benefit, as compared with saving 30 minutes later on, the so-called golden hour of reperfusion therapy.13 and 14 Most patients triaged in the ambulance were diagnosed within 90 minutes and treated within 180 minutes after the onset of symptoms. This very early diagnosis and therapy is probably related to the better myocardial blush grade after PCI in this group.15 In addition, a prehospital diagnosis gives the opportunity for early initiation of antithrombotic and antiplatelet pretreatment during transportation. Agents such as aspirin, heparin, and glycoprotein IIb/IIIa blockers have shown to improve initial patency of the infarct-related vessel.16 and 17 The very early initiation of these agents in the ambulance group might explain the higher initial patency rate in this group, resulting in a higher rate of prevented MI and might contribute to the better left ventricular function and clinical outcome of these patients. This study confirms previous findings from our group, which showed that extra transportation delay is associated with a worsening of the LVEF.18 The difference in outcome cannot be explained by a difference in medication after discharge, as shown in Table III.

Table III.

Medication at 30-day follow-up Referred (n = 239) (%) Ambulance (n = 202) (%) P

Aspirin 93 89 .17

Oral anticoagulation 10 10 .90

β-Blocker 89 84 .13

Calcium antagonist 6 4 .24

Nitrates 17 15 .51

ACE inhibitor 53 57 .43

Statin 88 85 .33

Clopidogrel⁎ 81 76 .19

ACE, angiotensin-converting enzyme.

⁎ Use at discharge.

Limitations

This study is a post hoc analysis of patients recruited in the On-TIME trial and not a randomized comparison between prehospital triage in the ambulance versus triage at the referral center; however, randomization would be unethical because this would deliberately prolong time to treatment in 1 arm. The presence of an ambulance equipped with 12-lead ECG diagnostic facilities determined whether a patient was transported to a non-PCI center first or immediately referred to the PCI center. Inclusion and exclusion criteria were the same for both study groups. However, multivariate analysis might correct differences in baseline characteristics between the groups, but this statistical correction might not overcome the problem of undetectable confounders and is less reliable in a relatively small-sized trial with a low incidence of the outcome parameter of interest. Left ventricular ejection fraction was routinely performed at discharge in the patients recruited in the Zwolle area only (prespecified LVEF substudy). Patient characteristics of these Zwolle patients, however, did not differ significantly from the patients recruited at other PCI centers.

Conclusion

Prehospital infarct diagnosis, triage, and therapy in the ambulance is feasible without physician interference when performed by highly trained paramedics using a validated computerized ECG software in 95% of patients. In addition, ambulance triage was an independent predictor of a left ventricular function >40% and was independently associated with a favorable clinical outcome. Therefore, all efforts should be made to implement prehospital infarct diagnosis, triage, and therapy in the care of patients with an acute MI and to further improve cooperation with ambulance personnel in this regard.

References

1 F. Zijlstra, M.J. de Boer and J.C.A. Hoorntje et al., A comparison of immediate coronary angioplasty with intravenous streptokinase in acute myocardial infarction, N Engl J Med 328 (1993), pp. 680–684. Abstract-MEDLINE | Abstract-EMBASE | Full Text via CrossRef

2 E.C. Keeley, J.A. Boura and C.L. Grines, Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review of 23 randomised trials, Lancet 361 (2003), pp. 13–20. SummaryPlus | Full Text + Links | PDF (409 K) | Full Text via CrossRef

3 M. Dalby, A. Bouzamondo and P. Lechat et al., Transfer for primary angioplasty versus immediate thrombolysis in acute myocardial infarction: a meta-analysis, Circulation 108 (2003), pp. 1809–1814. Abstract-MEDLINE | Abstract-Elsevier BIOBASE | Full Text via CrossRef

4 B.K. Nallamothu, E.R. Bates and J. Herrin et al., NRMI Investigators. Times to treatment in transfer patients undergoing primary percutaneous coronary intervention in the United States: National Registry of Myocardial Infarction (NRMI)–3/4 analysis, Circulation 111 (2005), pp. 761–767. Abstract-MEDLINE | Abstract-EMBASE | Abstract-Elsevier BIOBASE | Full Text via CrossRef

5 S. Behar, Main findings of the EHS-ACS II. Results of the Euro Heart Survey in Acute Coronary Syndromes 2 Presented at the meeting of the European Society of Cardiology, Stockholm (2005).

6 A.W.J. van 't Hof, N. Ernst and M.J. de Boer et al., On-TIME study group. Facilitation of primary coronary angioplasty by early start of a glycoprotein 2b/3a inhibitor: results of the ongoing tirofiban in myocardial infarction evaluation (On-TIME) trial, Eur Heart J 25 (2004), pp. 837–846. Abstract-EMBASE | Abstract-MEDLINE

7 E.W. Grijseels, M.J. Bouten and T. Lenderink et al., Pre-hospital thrombolytic therapy with either alteplase or streptokinase. Practical applications, complications and long-term results in 529 patients, Eur Heart J 16 (1995), pp. 1833–1838. Abstract-EMBASE | Abstract-MEDLINE

8 M.N.S.K.J. Ernst, M.J. de Boer and A.W.J. van 't Hof et al., Prehospital triage for angiography-guided therapy for acute myocardial infarction, Neth Heart J 12 (2004), pp. 51–56.

9 D.A. Morrow, E.M. Antman and A. Charlesworth et al., TIMI risk score for ST-elevation myocardial infarction: a convenient, bedside, clinical score for risk assessment at presentation: an intravenous nPA for treatment of infarcting myocardium early II trial substudy, Circulation 102 (2000), pp. 2031–2037. Abstract-MEDLINE | Abstract-EMBASE | Abstract-Elsevier BIOBASE

10 A.W.J. Van 't Hof, A. Liem, H. Suryapranata and on behalf of the Zwolle myocardial infarction study group, Angiographic assessment of myocardial reperfusion in patients treated with primary angioplasty for acute myocardial infarction: myocardial blush grade, Circulation 97 (1998), pp. 2302–2306. Abstract-Elsevier BIOBASE | Abstract-MEDLINE

11 E.M. Antman, D.T. Anbe and P.W. Armstrong et al., ACC/AHA Guidelines for the management of patients with ST-elevation myocardial infarction—executive summary. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to revise the 1999 guidelines for the management of patients with acute myocardial infarction), J Am Coll Cardiol 44 (2004), pp. 671–719. SummaryPlus | Full Text + Links | PDF (1089 K)

12 D. Hasdai, S. Behar and L. Wallentin et al., A prospective survey of the characteristics, treatments and outcomes of patients with acute coronary syndromes in Europe and the Mediterranean basin; the Euro Heart Survey of Acute Coronary Syndromes (Euro Heart Survey ACS), Eur Heart J 23 (2002), pp. 1190–1201. Abstract-EMBASE | Abstract-MEDLINE

13 Fibrinolytic Therapy Trialists (FTT) collaborative group, Indications for fibrinolytic therapy in suspected acute myocardial infarction: collaborative overview of early mortality and major morbidity results from all randomised trials of more than 1000 patients, Lancet 343 (1994), pp. 311–322.

14 G. De Luca, H. Suryapranata and J.P. Ottervanger et al., Time delay to treatment and mortality in primary angioplasty for acute myocardial infarction: every minute of delay counts, Circulation 109 (2004), pp. 1223–1225. Abstract-Elsevier BIOBASE | Abstract-EMBASE | Abstract-MEDLINE | Full Text via CrossRef

15 G. De Luca, A.W.J. van 't Hof and M.J. de Boer et al., Time-to-treatment significantly affects the extent of ST-segment resolution and myocardial blush in patients with acute myocardial infarction treated by primary angioplasty, Eur Heart J 25 (2004), pp. 1009–1013. Abstract-EMBASE | Abstract-MEDLINE

16 F. Zijlstra, N. Ernst and M.J. de Boer et al., Influence of prehospital administration of aspirin and heparin on initial patency of the infarct-related artery in patients with acute ST elevation myocardial infarction, J Am Coll Cardiol 39 (2002), pp. 1733–1737. SummaryPlus | Full Text + Links | PDF (81 K)

17 G. Montalescot, M. Borentain and L. Payot et al., Early vs late administration of glycoprotein IIb/IIIa inhibitors in primary percutaneous coronary intervention of acute ST-segment elevation myocardial infarction: a meta-analysis, JAMA 292 (2004), pp. 362–366. Abstract-Elsevier BIOBASE | Abstract-EMBASE | Abstract-MEDLINE | Full Text via CrossRef

18 A.L. Liem, A.W.J. van 't Hof and J.C.A. Hoorntje et al., Influence of treatment delay on infarct size and clinical outcome in patients with acute myocardial infarction treated with primary angioplasty, J Am Coll Cardiol 32 (1998), pp. 629–633. SummaryPlus | Full Text + Links | PDF (64 K)

Appendix A. The On-TIME study group

Steering committee

MJ de Boer, E Boersma, AJ van Boven, R Buirma (Nonvoting member), J Dille, AWJ van 't Hof, and RJ de Winter.

Ambulance coordinators

F Hollak (Ambulance Dienst Regio IJssel Vecht), F de Pooter (Ambulance Dienst Regio Noord West Veluwe).

Referral center coordinators

T Bouwmeester (Winschoten), R Brons (Meppel), R Dijkgraaf (Harderwijk), W Jap (Apeldoorn), MJ de Leeuw (Assen), A Mosterd (Amersfoort), C Oei (Heerenveen), and J Saelman (Hoogeveen).

PCI center coordinators

The Netherlands: JM ten Berg (Nieuwegein), AJ van Boven (Groningen), JHE Dambrink (Zwolle), and RJ de Winter (Amsterdam).

Italy: S Petronio (Pisa).

[Ace844]

(Evaluating an Emergency Medical Services—Initiated Nontransport System

Michael J. Schmidt A1 @ Daniel Handel A1, Christopher J. Lindsell A1, A2, Lindsey Collett A3, Paul Gallo A3, Donald Locasto A1

A1 Department of Emergency Medicine, University of Cincinnati College of Medicine, Cincinnati, OH

A2 Institute for the Study of Health, Cincinnati, Ohio

A3 City of Reading Fire Department, Reading, OH)

Abstract:

Objective. To evaluate both factors predicting nontransport and mortality rates in an emergency medical services system with a nontransport policy. Methods. We reviewed data from 1,581 transported and nontransported patients from October 2001 to July 2003. Patients who refused transport against medical advice were excluded. Extracted data included demographics, run characteristics, chief complaint, and clinical impression. Transported and nontransported patients were compared using Mann–Whitney U or chi-square tests. Logistic regression identified factors predictive of nontransport. A Social Security Death Index search determined 30-day mortality.

Results. A total of 1,501 runs involving 1,059 patients were included. Median age was 60 years (range, 0–97 years). A total of 427 (40.4%) were male; 107 (10.2%) were nonwhite. Older patients were more likely to be transported (odds ratio, 1.03; confidence interval, 1.02–1.03). Race, frequency of calls, mutual aid, or time of day did not significantly influence probability of transport. Patients with cardiovascular, respiratory, and gastrointestinal complaints were more likely to be transported than those with other conditions (P < 0.005); patients with endocrine, trauma, and miscellaneous complaints were less likely to be transported (P < 0.003). Patients with renal, obstetrics/gynecology, and hema matology/oncology were complaints all transported. Mortality was 4.9% (confidence interval, 3.9%–6.2%) for transported patients and 1.0% for those not transported (confidence interval, 0.2%–3.7%).

Conclusions. Age is a determinant when deciding on transporting patients. Patients with complaints with potentially higher acuity were transported most often. Only two nontransported patients died within 30 days, although it is unknown whether initial transport would have changed their mortality. Our data suggest that emergency medical services–initiated nontransport is influenced only by age and chief complaint and may not result in significant mortality.