Without evidence of benefit, an intervention should not be presumed to be beneficial or safe.

- Rogue Medic

Utilization of warning lights and siren based on hospital time-critical interventions


Also posted over at Paramedicine 101 (now at EMS Blogs), and at Research Blogging. Go check out the rest of the excellent material at these sites.

This study was the topic for discussion on the EMS Research Podcast.[1]

Go listen to the podcast.

To simplify the title –

Does the use of lights and sirens get the patient to the hospital in time for life-saving treatment?

The routine use of lights and siren (L&S) by emergency medical services (EMS) personnel has been a longstanding tradition, but with evidence mounting concerning its risks, many are now questioning their utility.1–4 [2]

This is not just appropriate, but essential.

We have too many treatments/procedures that are based on nothing more than superstition, tradition, and/or wishful thinking. We need to evaluate what we do in as unbiased a way as possible to find out if there is any benefit to any patient, rather than just blindly continue with each standard of care myth-based intervention.

Morbidity and mortality from collisions involving emergency vehicles is a major public health hazard.[2]

Traffic fatality is always one of the top causes of line of duty death in EMS. If a patient is unstable, crashing on the way to the hospital is definitely not a good idea. Is there any benefit from the risk of L&S driving?

Roughly 70% of fatal ambulance crashes occur during utilization of warning L&S.14[2]

As the EMS providers wrote the chart, there was a questionnaire to confirm if the times documented were accurate. If EMS personnel subjectively felt documented times were not accurate, the chart was excluded from the study.[2]

The time of travel in the control group was recorded by two medical students and one EMS fellow traveling in their personal vehicles from the location of the 9-1-1 response to the hospital. They drove during the same day of week and time of day as did the original call. They were instructed to obey all traffic laws and speed limits. All time was recorded in minutes. Any significant time delay due to weather patterns was noted and excluded from analysis.[2]

From the paper, it is not clear where they did this study. I have worked for all of the hospitals, except one. That is the University of Medicine and Dentistry of New Jersey – Robert Wood Johnson Medical School in New Brunswick, NJ. The demographics listed are not consistent with any of the other hospitals listed. RWJ is the most suburban of the hospitals and that is something that should have a bearing on the way we assess the applicability of this study to individual systems. This is a variable to consider in the way traffic affects transport times.

A total of 112 charts were used in this analysis. The average difference in time with versus without L&S was -2.62 minutes (95% CI = -2.60– -2.63 minutes (min), paired t-test p-value <0.0001; signed rank p-value <0.0001) such that patient transport with no L&S took on average of 2.62 minutes longer than when using L&S.[2]

95% CI = -2.60– -2.63 minutes?

That is a surprisingly narrow confidence interval.

The average transport time in minutes with L&S is 14.5 ±7.9 min (1SD) (range = 1–36 min). The average transport time without L&S is 17.1± 8.3 min (range = 1–40 min). The time difference ranged from 24 min faster with L&S to 16 min slower with L&S.[2]

Here is another point that raises questions that are not answered in the paper.

Did one of the ambulances crash?

Did one of the ambulances get lost?

Or, should I ask, Did two of the ambulances crash/get lost/whatever?

All we know is that there were 2 transports that took dramatically longer with L&S than without L&S.The major roadway connecting the university hospital with neighboring towns was frequently under construction. Although this factor could account for prolonged times in the lights and sirens group, it also could have equally affected the control group.[2]

And it could explain the two extra-long L&S transports.

Huber regression estimates no significant effect of time with L&S on the difference between the two mean transport times, with an increase of 0.02 minutes (95% CI = -0.06–0.10) in the difference due to a 10-minute addition in transport time with L&S. This finding is contrary to the expectation of L&S being even more useful for longer distances.[2]

In other words, the longer trips did not result in significantly more time saved using L&S.

3 1/2 seconds to 6 seconds (0.06 to 0.10 minutes) for every additional ten minutes of transport time. Travelling at 60 MPH (Miles Per Hour) for 10 minutes, this would save less time, than increasing the speed to 61 MPH. Travelling at 30 MPH (Miles Per Hour) for 10 minutes, this would save less time, than increasing the speed to 31 MPH.

If the ambulance increases speed from 60 to 61 MPH, it is going to be barely noticeable in the back.

If the ambulance increases speed from 60 to 61 MPH, it is going to be barely noticeable in the back.

If the ambulance turns on the Lights & Sirens, it is going to be very noticeable in the back.

Where is the benefit that justifies the increased risk?

The complete logs of interventions provided to the study patients were evaluated. Of the 112 patients transported with L&S, 108 (96.4%) were treated with PIs only. Five (4.5%) patients transported with L&S also received time-critical HI.[2]

PIs are Prehospital Interventions – treatments that can be provided by the paramedics (ALS or Advanced Life Support personnel).

HIs are Hospital Interventions – treatments that cannot be provided by paramedics. Fibrinolytics, neurosurgical evacuation, cardiac catheteriztion, and transvenous pacing in this study.

In other words, they were racing to the hospital, to have treatments that could have been provided by the paramedics.

However, there are times when it may be more appropriate to have something done in the more controlled setting of the hospital, rather than on scene or in the ambulance.

It is also possible that the medical command physician ordered that the paramedic not provide a treatment that is within the paramedic’s scope of practice. This can be for a treatment that is only permitted with medical command contact or a treatment that is permitted on standing orders, but that the medical command physician specifically ordered be withheld until the patient is at the hospital.

What about certain procedures that are often unsuccessful due to operator error, such as transcutaneous pacing or cardioversion. Even in the hospital, it is not unusual for some operator error to be involved when using these procedures to treat unstable patients.

The last patient was diagnosed with an unstable, third-degree heart block and required immediate transvenous pacemaker placement secondary to ineffective capture with a transcutaneous pacemaker.[2]

Immediate? This was not done within the time saved by L&S transport, so hardly immediate.

Ineffective capture?

Is that the same as complete lack of capture?

As in pronouncing a patient dead because of ineffective cardiac output as demonstrated by being pulseless, apneic, and asystlic?

As I frequently like to point out –

Failure to capture with a transcutaneous pacemaker is frequently operator error.

Tom Bouthillet of EMS 12 Lead was not on the podcast, but he has made similar statements about transcutaneous pacing.[3]

No HI was administered within the first 2.62 minutes of arrival. All five patients were admitted to a critical care unit and the average length of stay in the hospital was 10 days. No deaths occurred in the group who received HI.[2]

Was any time saved that made any difference in outcome?

We do not know, but this study did not provide evidence to support L&S transport.

Were any treatments provided any sooner?

We do not know, but this study did not provide evidence to support L&S transport.

it is possible that patients with more serious illnesses had lights and siren compared with those who were less critical. Since only 7% of patients during this time interval did not have L&S, it is unlikely that this influenced the results.[2]

93% of these ALS patients require treatment EMS cannot provide?


Only 5 patients did and none of them needed these treatments in the amount of time saved by L&S.

93% of these ALS patients are unstable?

That has not been my experience in any of the systems where I have worked. If anything, even the reverse is too high.

7% of ALS patients being unstable is too high.

So why all the commotion?

Because a mentality exists in the system that L&S result in improved patient care,[2]

We need to expose these myths for what they are – superstitions.

Go listen to the podcast.


[1] Driving with Lights and Sirens: EMS Research Episode 8
EMS Research Podcast

[2] Utilization of warning lights and siren based on hospital time-critical interventions.
Marques-Baptista A, Ohman-Strickland P, Baldino KT, Prasto M, Merlin MA.
Prehosp Disaster Med. 2010 Jul-Aug;25(4):335-9.
PMID: 20845321 [PubMed – indexed for MEDLINE]

The full text of the paper is available as a free PDF download from Prehospital Disaster Medicine from that issue’s index. – PDM has moved, so these links do not work.

The download link is in the page number – page 335. – PDM has moved, so these links do not work.

[3] Transcutaneous Pacing (TCP) – The Problem Of False Capture
EMS 12 Lead

Marques-Baptista A, Ohman-Strickland P, Baldino KT, Prasto M, & Merlin MA (2010). Utilization of warning lights and siren based on hospital time-critical interventions. Prehospital and disaster medicine : the official journal of the National Association of EMS Physicians and the World Association for Emergency and Disaster Medicine in association with the Acute Care Foundation, 25 (4), 335-9 PMID: 20845321



  1. You’re asking the titans of EMS, specifically of the cult of high performance EMS to not only look at, but also apply valid research?

    But, what about the work of the venerable Archbishop of HPEMS, Jack Stout? What about removing data that doesn’t fit their model to make their model work? What about using epinephrine? What about using helicopters because the car looks bad?

    What; was that Heretic known as Bledsoe right in that the small amount of data used can yield accurate predictions? That epi doesn’t work even though we’ve always done it? That helicopters and their angels don’t always make things better?

  2. Great timing. I just posted somewhere else in the past week on this same topic (more or less). I feel that if EMS personnel were made to defend their use of RLS as if in the court room every time they “run hot”, we would see the use of RLS drop to next to nothing. Now if we could just convince people of the same thing while en route to the scene.

    Hey, I’m a dreamer