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

- Rogue Medic

Some HEMS Crash Data Charts

All of the charts were updated at 15:27 EDT 10/16/08. My explanation is after the first chart and in footnote [5].

I am way behind in responding to comments on the last post, Helicopters and Bad Science. This is a partial response. Dave wanted to know about the relative risks of ground transport vs. HEMS. This is a good question. The charts here may help to put that in some perspective. The risk of a crash when traveling by ground is much lower than by HEMS.

Dave also suggested that HEMS provides expertise. Maybe. There are excellent HEMS crews, but as with anything else, there are some you would not want treating anyone you care about.

I wanted to find another way of looking at some of the data on HEMS (Helicopter Emergency Medical Services) crashes and fatalities. Here are some charts I made.[1]

An apology for some poor labeling. LODD means Line Of Duty Death. It is a shorthand that have become so accustomed to using that I forgot to define it. The chart above is comparing all EMS fatalities, HEMS crash fatalities, and ground EMS crash fatalities. The way I had it labeled was that it was comparing crashes. When it comes to ground EMS, I don’t even want to try to count them. Looking at the fatalities is more accurate, since they are much harder to ignore.

I did not include anything that mentioned a plane. I did include anything that mentioned an ambulance crash. The totals include all sorts of other causes of death. The site[2] may not present an accurate listing, but it seemed to be objective on the presentation of the information and more comprehensive than the other sites I looked at. I did not include total or ground EMS for 2008, but we do have some HEMS numbers. HEMS so far in 2008 – 13 crashes, 24 dead, 9 injured. Add a paramedic, who was struck by a rotor while loading patients into a helicopter.[3] 10/15/08 A crew of 3 and an 18 month old patient. The crash resulted in a fire, which is one of the factors that has been cited as increasing the likelihood of fatality. “Thirty-nine percent of all deaths occurred in crashes with postcrash fires; 76% of crashes with postcrash fire were fatal compared with 29% of other crashes.”[4]

As of 10/16/08 I updated the chart to add the 5 extra fatalities for 2008. I updated all of the charts to reflect the changes described in footnote [5]. There is no joy in the new data supporting the claim that HEMS is abused. Real people have died. This brings only sadness. While I am a bit cross-eyed from changing all of the charts, those who died, their families, and friends are having an immeasurably worse day and would be glad to only experience my minor inconvenience.

This chart and the next two charts are from the NTSB Special Investigation Report on EMS Operations, but reflect adjustments based on what appears to be more complete data from HAI[5]

This chart differs from the first chart for at least a couple of reasons. This is from the Helicopter Association International AMSA.[6] This is limited to helicopter data. The reporting for either, or both, organizations may contain inaccuracies.

Same source[2] and Bureau of Labor Statistics.[7]

The division of the number of all aviation fatalities by 10 is purely to have the data appear on a similar scale. Where it shows 35.2 as the first number, read it as 352. EMS aviation fatalities are dramatically lower than those for all aviation. EMS aviation fatalities are a part of all aviation. This division was to demonstrate that the fatalities from all aviation have been trending down, while those from EMS aviation have been trending up much more. All aviation numbers are down by about half over the time period of the graph, while EMS aviation is up by a factor of about 4.

A couple of quotes from an article[8] examining HEMS crashes –

It is estimated that there are 650 helicopters in the United States dedicated to air medical service,1 and the present series revealed an average of 12 helicopter EMS crashes per year in recent years. Thus, 1 medical helicopter in 4 is likely to crash during 15 years of service (12X15/650=28%), which suggests that investments in prevention have the potential to be highly cost effective.

Thirty-nine percent of all helicopter EMS crashes result in 1 or more deaths.

The odd thing is that patients keep surviving these crashes. Not all crashes happen with a patient on board. If helicopters are being utilized appropriately – not as substitutes for ground paramedics, who are not trusted to make decisions anyway – then the patients should not be healthy enough to survive these crashes.

The reality is that, in many places, medical directors have a baseless fear of paramedics making decisions.

Not completely baseless, there are bad medics out there. The bad medics are the fault of the medical directors not screening medic candidates appropriately and not providing appropriately aggressive oversight once the paramedics are working.

If you doubt this, how do you explain the systems that are full of excellent paramedics? Are they some variation of cancer clusters – competence clusters? No. They are the result of competent medical direction.

Why won’t Maryland adopt competent medical direction at the state level?

Medical direction that demands competence from EMS providers. Medical direction that provides Maryland residents with the kind of competent providers available in some other places. Some of the local medical directors do seem to be trying to change this, if I read the comments correctly.


^ 1 The charting software I used is at NCES Create a Graph website.
This is from the National Center for Educational Statistics. The charting software is easy enough for a medic to use without help from a little kid, but probably would have been much faster with a little kid doing this. I created these charts by entering the data from the listed sources.

^ 2 Dave’s EMS
Line of Duty Deaths page
See the addendum to [5] for an explanation of the higher fatality numbers than Dave has listed.

^ 3 Arizona DPS Medic Victim Of Blade Strike During Rescue
From Aero-News Network

^ 4 Baker SP, Grabowski JG, Dodd RS, Shanahan DF, Lamb MW, Li GH.
EMS helicopter crashes: what influences fatal outcome?
Ann Emerg Med. 2006 Apr;47(4):351-6. Epub 2006 Jan 19.
PMID: 16546620 [PubMed – indexed for MEDLINE]

^ 5 Aviation Special Investigation Report
Special Investigation Report on Emergency Medical Services Operations
National Transportation Safety Board
Free Abstract . . . . . . Free Full Text PDF

Addendum reflected in the charts. This is directly from the NTSB paper cited here:

* Salt Lake City, Utah (FTW03FA082). On January 10, 2003, an EMS helicopter crashed into terrain while maneuvering in dense fog on an aborted mission to pick up a patient. The pilot and flight paramedic were killed, and the flight nurse was seriously injured.
* Redwood Valley, California (LAX04FA076). On December 23, 2003, an EMS helicopter was en route to pick up a patient when it collided with mountainous terrain while operating in high winds and heavy rain. The pilot, flight nurse, and paramedic were killed.

NTSB fatality numbers for the year are listed as 3, but they describe at least 5 fatalities. I have changed the data on the charts to reflect this. I changed all data to reflect the Helicopter Association International Air Medical Service Accidents 1991 – 2004 data. This is the source [6] directly below. They include individual details of all of the HEMS crashes. I do not believe that they include fixed wing data, if I am wrong, please let me know. Also, the 29 fatalities for 2004 does not reflect the HAI number of 18. Are these fixed wing data. Is HAI missing something? 10/23/08 a Med Flight Air Ambulance crashed killing 5. This was a Lear Jet, so I adjusted those numbers down. That still leaves a difference of 5. If someone has an explanation for this, please let me know. In the mean time I will keep looking. I have updated the charts by increasing the numbers with the belief that the higher NTSB numbers reflect fixed wing data.

^ 6 Helicopter Association International
Air Medical Service Accidents 1991 – 2004
Free PDF
I originally listed a different paper and a different link. They were not bad references, they were just different material than what I obtained the data from. This PDF is the correct source.
Main page for Helicopter Association International.
Link page for crash data and other statistics from HAI.
Their search page information and the search page intself – U.S. Civil Helicopter Accident Database.

^ 7 Injuries, Illnesses, and Fatalities
Bureau of Labor Statistics
A bunch of Free Full Text and PDFs of workforce fatality data.

^ 8 This is the same source as footnote [4]
Baker SP, Grabowski JG, Dodd RS, Shanahan DF, Lamb MW, Li GH.
EMS helicopter crashes: what influences fatal outcome?
Ann Emerg Med. 2006 Apr;47(4):351-6. Epub 2006 Jan 19.
PMID: 16546620 [PubMed – indexed for MEDLINE]



  1. I believe one of the charts- the one comparing the deaths due to ground EMS crashes versus air ambulance crashes- is inaccurate. According to the NTSB, the fatal EMS helicopter crashes were for the year 2003 were:6/7/2003 1 dead, 1 serious, 1 uninjured; Salt Lake City, UT1/28/2003 1 dead (repositioning flight with only the pilot on board); West Chicago, IL1/10/2003 2 dead; Salt Lake City, UTTherefore I question whether the aviation crash numbers are under reported. The ground ambulance crash numbers seem to be in line with what has been reported by injury prevention researchers (including a group from the CDC).

  2. You can add another helicopter, 3 crew and a patient to the death toll tonight.When will the madness end?

  3. As always, a really interesting (and thought provoking) post. Definitely a lot to think about here. The NTSB report is now on the list of things to read….And don’t forget to say a prayer for the crew and patient killed Wednesday night just outside of Chigago.

  4. The Erstwhile Medic,I have looked at the information, compared different sources, and come to the conclusion that I need to redo all of the charts. They have now been updated. They also include the 4 more killed last night. According to the Helicopter Association International Air Medical Service Accidents numbers, I had put the wrong link in the footnote, it is actually 9 fatalities for 2003.01/10/03 – 4 fatal.01/29/03 – 1 fatal.06/07/03 – 1 fatal.12/23/03 – 3 fatal.I expect that the numbers are not completely accurate, but I like that 2 of the sources list some details of the specific crashes, so the numbers can be independently evaluated. I do not want to just be pointing to big numbers, but accurate numbers, as much as possible given my resources.

  5. Anonymous,I do not like this kind of evidence to support my position, but maybe people will make some serious adjustments in the criteria that affect the decision to fly, rather than drive.

  6. Walt Trachim,Thank you for the kind words. The NTSB has some interesting reading, but the more I look at on the Helicopter Association International site, the more I like it.In a business that is supposed to be about saving lives, we have a lot of trouble going about it in the right way.

  7. Sounds like a Number Needed to Treat calculation for helicopter transports is in order here. I’m betting the data isn’t available yet.Also sounds like some of the folks in charge need to do some math. Anytime someone whips out the “how many deaths are acceptable” argument in the press I start thinking “true believer”.Save the beliefs for church. The hospital/ambulance/scene is a place for science.Helicopters have their place if transport times are long or air medics are more experienced than ground medics AND the patient is unstable or potentially unstable. Otherwise ground transport is more efficient.

  8. Ted,I think we mostly agree. The problem is trying to get all of the research evaluated in a way that selects the good studies and ignores the poorly controlled ones, even if they have large numbers.The places that use poorly trained medics with poor oversight, should not be the examples we use, but it is not so easy to identify that in the studies. One example of the variability of quality in EMS is made pretty obvious in the intubation research. Some places almost never miss, while others seem to aim for the place that smells like chili. Aggressive oversight should prevent this. The feedback should be easy – waveform capnography.When it comes to trauma, there are many reasons for not getting feedback. HIPAA is one that many use, but that is a misunderstanding. The focus on transferring a live patient misleads us into believing that what we are doing is the reason the patient looks good at the time of transfer. Similarly, we try to get pulses back in cardiac arrest and are less focused on the patients who leave the hospital alive. I believe that most of what we do (drugs and procedures) to get a pulse back, actually decreases their chances of walking out of the hospital. We focus on making the vital signs better, instead of not making the patient worse.Better vital signs are not necessarily better for the patient.There is a ton of research on trauma triage, so I will try to get something posted on that this week.

  9. RM,I agree wholeheartedly. More research, demanding excellence in training and performance, and focussing on outcomes instead of immediate results (we’re getting there! Hi-dose epi and MAST pants are out of the protocols!) are essential for the advancement of EMS.But keep in mind that exactly half of the services out there will be below average. The Least Common Denominator exists and cannot be removed. Raised maybe, but not removed. You HAVE to write protocols with the LCD in mind–that’s who they’re for. And sometimes that means “get this patient away from that moron at Hydrocephalic EMS and into the hands of a chopper jockey”.I’ll freely admit that this is a short-term solution while you remediate Hydrocephalic EMS, but it’s a place to start.

  10. Ted,It is a place to start. In Pennsylvania they have new protocols coming out Nov, 1. They have been trying to push things forward in several areas. I have been pleasantly surprised with several changes, but they still need a lot of work, especially on the LCD factor.Other places seem to be content putting people on the street, with the expectation that these people will not be able to recognize what they have. This makes no sense. They are presented with various bogeymen, but no way to determine what is stable, potentially unstable, and clearly unstable. One example of this is the treatment of tension pneumothorax. How many awake and alert patients get harpooned because of diminished breath sounds? Too many.

  11. Good post. Thanks for all the great info. I would like to see some data that is weighted for the explosion in numbers of Air Medical programs in recent years (fueled in part by an increase in medicare reimbursement for these services that, I’m told, makes them profitable at about a call per day). I am sure the problem will not look quite as severe. This does not excuse the issues, but it will offer some needed perspective.Also, I think that it is important to note that the helicopters that are crashing are flying single engine aircraft with one pilot, no NVG (Night Vision Goggles), no TAWS (Terrain Avoidance Warning System), and aren’t IFR (Instrument Flight Rules, as opposed to Visual Flight Rules) capable. I am no expert in this area, but this is what I’ve been told by those experts whom I trust.The point: I suspect that there are industry-proven best-practices that can make HEMS very safe. I don’t think much is going to change without some increased federal regulation.

  12. Ben,From footnote 4. Susan Baker, et al. go into some detail on the flight characteristics that lead to crashes. It is from the Annals of Emergency Medicine, so not available without a subscription. If you send me an email, I can send you some pdfs. Susan Baker has written a lot on aircraft safety, so there are some insights from the aviation perspective. Helicopter Association International has a lot of information. They seem to be the group pushing hardest for self-regulation. Many others talk about self-regulation, but it only seems to be a delaying tactic. HAI seems to be serious. They are also an excellent source of information that appears to be unbiased. I will be posting more on HEMS and will try to post on the financial considerations.

  13. Thanks. I’ll send you an email.On the issues of self-regulation, it is my perception that the industry has been “self-regulating” for years, and that maybe this isn’t working so well.

  14. Ted,

    One thing I forgot to comment on. That half of the people are below average sounds right, but it is misleading.

    For example, you have 10 EMS organizations to keep the math simple. 9 of them are consistently excellent, but one organization waltzes to the beat of its own dysfunctional ADD drummer. The organizations are of the same size. Some of the people in the less-than-ideal organization may be good providers, but most are not. When you average the quality of all of the providers, more than 90 percent should be competent, but the less-than-stellar ones drag the average down significantly. Almost 10 percent would be below average and more than 90 percent above this inappropriately low average and far above the ridiculously low Lowest Commom Denominator.

    Similarly, eliminating a small percentage of dangerous people should lead to a tremendous improvement in the average quality of care and should eliminate almost all of the avoidable bad outcomes.

    Outside of a class in statistics, most distribution does not follow a Bell Curve pattern.

  15. “I think that it is important to note that the helicopters that are crashing are flying single engine aircraft with one pilot”Actually there was a study done that said that single engine helicopters were at no greater risk of a crash than dual engine helicopters. However, I do agree that two pilot operations should be the standard, just as they are for commercial airlines. That would more or less necessitate the use of twin engine helicopters though because of the added weight.”I would like to see some data that is weighted for the explosion in numbers of Air Medical programs in recent years (fueled in part by an increase in medicare reimbursement for these services that, I’m told, makes them profitable at about a call per day). I am sure the problem will not look quite as severe. This does not excuse the issues, but it will offer some needed perspective.”Look at Part 121 aviation (more commonly known as airlines) which had until recently seen astronomical growth yet there was only one fatal crash in the past several years. Even with exploding numbers of flights, the number of flights that ended with an explosion dropped. That is what is achieved when an industry is forced to adopt the mindset that safety is paramount and the cost of doing business is not measured in human lives. While I do agree that it is important to view the crash rate in perspective- the recommended method is number of crashes per X hours flown (with X normally being either 10,000 or 100,000 depending upon the scale of the problem) rather than the number of helicopters- the fact remains that these same issues continue to plague the industry indicates that greater oversight is needed because the industry is, not to put to blunt of a point on it, standing aroud with their heads up their collective asses whilst they continue to kill our colleagues and occasionally their patients. In the cold light of the day after a crash, when one is standing over a smoldering pile of wreckage that once was a beautiful flying machine it does not matter one iota whether there is one medical helicopter flying or 10,000 flying. We owe nothing less to our fallen brothers and sisters to make sure they did not die in vain and hold those in part responsible for their deaths (read as: the administration of the aeromedical operations who encourage or demand flights into hazard conditions) no less accountable than a drunk driver who knowingly and wantonly gets behind the wheel despite the fact he is putting lives at grave risk for little to no benefit to anyone except for fulfilling his own selfish desires.

  16. The Erstwhile Medic,

    You wrote –

    We owe nothing less to our fallen brothers and sisters to make sure they did not die in vain and hold those in part responsible for their deaths (read as: the administration of the aeromedical operations who encourage or demand flights into hazard conditions) no less accountable than a drunk driver who knowingly and wantonly gets behind the wheel despite the fact he is putting lives at grave risk for little to no benefit to anyone except for fulfilling his own selfish desires.

    An excellent point that does change the perspective. If we hold bar tenders liable for allowing patrons to drive while impaired by alcohol, how can administrators avoid responsibility for pressuring crews to fly in circumstances where impaired would be an improvement?

    The drunk bar patron may choose to spend his money elsewhere after the bar tender takes the drunk’s keys. After that night, the drunk may go somewhere that does not act as responsibly.

    The flight crew may be putting their jobs/personnel records on the line if they do not do as the administrator suggests. If they do fly, they may be putting their lives and the lives of patients on the line.

    Getting away with flying when it is inappropriate may only lead to discounting the risk even more on future flights.

  17. “If we hold bar tenders liable for allowing patrons to drive while impaired by alcohol, how can administrators avoid responsibility for pressuring crews to fly in circumstances where impaired would be an improvement?”The most frequent way they skirt responsibility and culpibility is because of the way the FAR (the regulations pertaining to aviation) are writing. The ultimate and final responsibility for the safe operation of any flight is at the feet of the pilot in command, regardless of pressures- real or perceived- from other sources among other things. Technically someone could have a gun at the head of the pilot, order him to do something that results in a stall and a crash and ultimately it would still be listed as something along the lines of: “The NTSB determines the probable cause of this accident to be failure of the pilot in command to maintain adequate airspeed resulting in a stall and subsquent uncontrolled descent into terrain. A contibuting factor was the asshole in the copilot’s seat holding a Smith and Wesson.” However, that being said, I have seen commercial operations- mostly on-demand cargo companies- held to account in NTSB reports for improperly influencing the decision making of their pilots, usually cited as a contributing factor under the guise of something to the tune of “inappropriate corporate culture”, etc. To my knowledge, the NTSB has never cited the corporate culture issue in a specific air ambulance crash, but then again we have never seen a plenary review of the subject with all stakeholders involved. As much as I criticize the air ambulance industry, I would gladly sit down with any group on either side and help to hammer out the needed changes in a way that keeps helicopters flying to aid those who truly need them AND keeps me from having to send condolences to the family of another friend or colleague. The two are not mutually exclusive and what is really needed- before the FAA really bends the industry over the table and dilates their lower colons as a group- is that a multipartisan meeting needs to be held with the following criteriaNo politics- so the Maryland panel is outNo bullshit- unless you have hard evidence, keep quiet about your opinion because this is science, not a meeting of religious zealots. No excuses-if you want to keep flying you’re going to have to invest in safety one way or the other. It will be much cheaper and easier to do it voluntarily rather than at the ever so pleasant behest of the Federal Aviation Administration or an administrative court judge. “Getting away with flying when it is inappropriate may only lead to discounting the risk even more on future flights.”As one of my EMS instructors was fond of saying: “Relying on experience alone, without truly learning from the mistakes inherent in that experience, simply results in making the same mistakes with an ever increasing level of confidence.” I personally love to cite that bit of advice whenever some pilot with more flight time with me (which is pretty much all of them given that I am still working on obtaining my pilot’s license) says I MUST be wrong simply because they “have __,000 hours of flight time!”. Usually my retort to them is followed by a request that they sign a release granting me access to their autopsy records in case they crash in a state or county that won’t hand them over as a courtesy to me as a researcher.

  18. The Erstwhile Medic,We can hope that someone forces a few alternative view points on the Maryland panel. It is amazing that Dr. Bass and Dr. Scalea conclude that others are not objective, just because others disagree with them. I suppose that is a pretty good definition of bias.

  19. Arguably, that is probably the most disgusting thing about the whole Maryland panel: the fact that Dr. Scalea- a brilliant surgeon and someone I hold in rather high regard (on everything but this)- would be so boneheaded and dense as to refuse to see that just because it’s Maryland doesn’t mean the laws of time, distance and cause and effect do not bend to fit his view of the situation. Also, the state of Maryland itself criticized the system and more or less proved that the system is horrendously flawed. Honestly, I think Bass is simply trying to save his job at this point more than anything else. If this panel does not address things in a productive manner and another helicopter goes down because of his inaction, if he thinks the state government and everyone else is biased against him now, all he will have to do is wait. As for Dr. Scalea, I think he is honestly just trying to cling to the ideals of Dr. Cowley. He is amongst the brightest and most talented trauma experts out there (on par with Ken Mattox, et al), but he is painting himself into a corner where he will eventually have his reputation (or memory, depending upon whether he dies before this is all resolved) permanently marred by his comments on this matter. It is much the same way that people who are true students of EMS, EM and trauma care tend to view Dr. Cowley in very mixed light because while he was a major reason for the development of a current system he was also a tyrant who handled dissent regarding his opinions about as well as the SS tolerated it. So I guess if Dr. Scalea wants to go down in history as this generation’s R. Adams Cowley, he is doing a good job of working towards that goal. As much as I disagree with him on this topic, it saddens me a great deal for such a damn fine physician (and a relatively nice guy too) to be inexorably heading down such a path.

  20. The Erstwhile Medic,I don’t know Dr. Scalea or Dr. Bass. I only know what I have read, both in the quotes attributed to them and what they have written themselves. I am not impressed, at least not in a positive way. Based on that, they seem to be exemplars of the problems with medicine. Too much of a focus on the anecdote at the expense of science, at the expense of medicine.

  21. I have met Dr. Scalea on a couple of occasions, and he is a decent person. To put it mildly, I was kind of shocked to hear the blatant disregard he had for the research on the subject and his disdain for the questions of his colleagues of all ilks. I’ve also met Dr. Bass and likewise, he seems to be a reasonably intelligent and nice guy, but I think of him in much the same way I do President Bush: a nice guy to sit and talk and have a beer with, but not someone you want in a position of authority because of his professional attitude and stances on important issues. Personally I think the biggest problem is his case is that he has been left in a position of absolute authority for far too long and it has gone to his head.

  22. The Baltimore Sun published a good review of the years medevac crashes:


    And of course the first comment is from a Maryland EMT-B that seems to be characteristically lacking in assessment skills.

  23. The Erstwhile Medic,Well, I do not expect to be invited to share an adult beverage with either doctor. It is an old adage, that power corrupts.

  24. Yeah, I doubt Bass or Scalea would ever sit down and talk with me given what I have said publicly regarding their stances on this, but I could be wrong. However, I would not be surprised if myself, along with the likes of Ken Mattox and a couple of others, are burned in effigy at the opening of the meeting Maryland panel. In fact, I would actually welcome being in such company….

  25. The Erstwhile Medic,It is interesting that Dr. Bledsoe will be on this panel. If anyone were to burn me, I would definitely prefer that it be in effigy, rather than in the flesh.

  26. I think they included Bledsoe solely to give the impression that they aren’t totally ignoring the issue, which is what they are actually doing. At least we will be able to have a “man on the inside” so we can find out some of the defenses the panel is using and can turn it back against them at a later date.

  27. Let’s hope that Dr. Bledsoe’s contributions will be more than just nominal. I know he is capable of contributing a lot, but are the rest capable of paying attention to what he has to say?

  28. Anonymous,Thank you for the link. It is a significant part of my most recent post – Update on the Maryland Crash and Unnecessary Helicopter Transport.

    I agree about the comment on the article. There is a second similar comment.

  29. That is the big question: will the rest of the members actually heed any of his advice.

Speak Your Mind