WHY THE GLASGOW COMA SCALE HAS GOT TO GO

Whenever my colleagues or nurses ask me what GCS a patient has, I just make something up that sounds about right. Judging by their smirks I know they think that they know I didn’t check properly. The truth is far worse. I don’t know the Glasgow Coma Scale. 

Despite med school and my 10 years as a doctor having worked thousands of patients where the GCS is relevant I still haven’t managed to learn the damn thing.  I guess it in so many ways never made sense to me, and furthermore I can’t think of one single instance where it affected my clinical decision-making.

So I’m delighted every time I run into someone who also thinks the GCS is rubbish and has no place in emergency medicine. Maybe I have a friend in one Dr Steven Green. He  wrote a great editorial in a recent issue of Annals of Emergency Medicine that outlines the problems with the Glasgow Coma Scale.

Below are some of the things he touches on. I have linked to some of his references.

It was never intended for use in trauma or emergency medicine

When Teasdale and Jennet invented the GCS back in 1974 it was  designed as a tool for repeated bedside assessment of various neurological functions in patients in a neurosurgical ward. It was never intended to be used as tool in trauma or emergency medicine. And it was most definitely not supposed to be used the way we use it. The sub scales were never meant to added up to a total score.

The inventors even publicly disowned the GCS in the form we know it . Back in 1978 they wrote ‘We have never recommended using the GCS alone, either as a means of monitoring coma, or to assess the severity of brain damage or predict outcome.’

It is not as reliable or precise as you think

For the GCS to work as intended it needs give reproducible results that allows reliable monitoring over time and allows for being communicated between professionals. It does not.

One study compared GCS assessments between pairs of emergency physiscians. Only in 38% of the cases the GCS scores were the same and in 33% of cases the scores varied with more than two points. The precision of the GCS is greatly overstated.

Because of it’s seemingly high level of detail it gives us a false impression of having created order out of chaos. It is a numeric illusion.

The subscales aren’t comparable and don’t add up

The GCS tells us to sum up the scores from the three different scales (motor, vocal, pain). That implies that the three scales are somehow interrelated and furthermore have the same clinical magnitude. It should be obvious to anyone that that is highly unlikely.

A most glaring example, is how the same GCS can predict different mortality depending on the inherit components. Studies have compared mortality with variation in subscale scores that result in the same GCS. This is a good illustration of the futility of adding the subscale scores

A GCS of 4 with the components 1+1+2 (E+V+M) predicts a mortality rate of 48%. But if we calculate the same GCS of 4 with the components 1+1+2 it will suddenly predict a mortality rate of 27%.

This is another aspect to which the authors objected in vain.

 The GCS is not consistently teached or remembered

Due to its complexity and the way it is defined the GCS is hard to teach consistently and even harder to remember. When tested on the GCS even neurosurgeons get it right only 56% of the time. Military physicians calculate the GCS correctly only 15% of the time.

It doesn’t really do anything for you.

The GCS does not reliably predict brain injury, need for intervention or prognosis. There is actually no evidence it yields any benefit above unstructured assessment.

Here you might argue that ‘it does, it really does, it helps me decide when, for example, to intubate a patient or not‘.

Does it really? When was the last time you intubated a patient as a result of a calculated GCS? We (me and my man Dr Green) thought so. You intubated as a result of a fuzzy, unstructured assessment and at best used a retrocalculated GCS to vindicate that decision.

Simpler scales perform just as well

Studies show that using any of the GCS subscales or other scales such as the AVPU work just as well.

Read the editorial here. Dr Green named his editorial

‘Cheerio, Laddie! Bidding Farewell to the Glasgow Coma Scale’

I find the title a bit weak and ambiguous. Wavering even. What is he trying to tell us? What is his message? Is he pro- or con-GCS?

This entry was posted in Emergency Medicine, Neurology, Trauma. Bookmark the permalink.

21 Responses to WHY THE GLASGOW COMA SCALE HAS GOT TO GO

  1. Vince D says:

    Love the blog! I came across it on LITFL this week and I’m working through your old posts now.
    As an EMT working in the US, I too have found almost no use for the numerical portion of the GCS. On the plus side however, its structure provides a slightly more detailed picture of the patient than just AVPU, so when I’m calling in report to a hospital, I like to run through the parameters without trying to regurgitate the values. “Patient’s eyes open to pain, speaks but cannot form coherent sentences, and withdraws from pain.” I know what that patient looks like, and so does the nurse I’m speaking to.
    Later, when I’m filling out my paperwork that requires documentation of the GCS every 5 min, I’ll finally get around to adding up the numbers.

    • admin says:

      Hi Vince,
      You are right, of course. The GCS does have some merit. In my view, just not in the immediate care phase of things. Verbalising your observations, like you suggested, rather than phoning in a number (that the person in the other end is not likely to make sense of) is a great way to work around the big issue with the GCS.

      And THANKS for reaching out. You might well be our first reader! So thanks for that.
      We will keep watching medicalapproach.
      Keep it going. It is looking good and it´s the greatest way to learn!
      And please send any feedback, suggestions, brutally honest criticism you might have. We would love to hear it from someone working in the big country where EMTS have come a long way! /D

  2. Josh says:

    Wonderful editorial. I’m just getting into medical blogs and i’m glad i came across this one if for no other reason that to have read this piece. I’m a Neurointensivist in the US, dealing almost exclusively with neurosurgical and critical neurologic patients. I think the GCS is essentially worthless, even in my patient population, for all of the reasons you mention above. There are far better scoring systems out there if you want to communicate a patient’s condition (i.e. the FOUR score). Even better, take 20 seconds to actually describe what you see! It’s priceless.

  3. Eldon says:

    What do you propose to replace it with? I’m not going to say whether I think the GCS is adequate or not but until somehting else comes along to replace it I think it works fine. Like any tool when used properly it does work. Just because a person is either too lazy or incompetent to learn how to use a tool doesn’t make the tool bad, just useless in that person’s hands. I’ve been in emergency medicine for close to 30 years in both the military and civilian environments and have used the GCS many times to describe my patients and to receive information concerning a patient. It’s not perfect but until you have an adequate or better replacement that can be used in the field then I suggest you learn how to use it properly, especially when it concerns the health and welfare of your patients.

    • admin says:

      Hi Eldon,
      Thanks for your reply. Fair points. Clearly, knowing and applying the GCS properly, you’re not the problem. The problem is the rest of us incompetent lazy bastards (who you work with) who don’t know the GCS, don’t use it properly or who happen to think it’s useless. What do you think of Dr Greens editorial and references? Clearly there is a problem with the GCS? As for better alternatives, in emergency care the AVPU or the simplified motor scales seem to be all I need. /Admin

  4. Charles says:

    Ok, I’m not a doctor or anything of that nature. I am an EMT and a Paramedic Student. The GCS, as with any other grading scale is a tool, and simply that. It should not take the place of an actual assessment. Instead of getting wrapped up in what number your patient falls in, why not treat them for how they present, and not how a number says they should present. Use the GCS as a documentation guide, and nothing more. Lets put the focus back on true, hands on, patient assessments.

  5. Brian Williams says:

    I think you may be on to something I am a paramedic in the metro Detroit and Ann Arbor Michigan and we look to the doctors that study these things and give us pointers because we have to make these decisions in less than optimal settings. We have a very progressive medical director Dr. Bob, he is really demanding when it comes to us being trained in advanced skills like rapid sequence intubation and other things that he feels will help us to troubleshoot any dificulty for the best patient outcome. We don’t have the benifit of medical school and when you say we should not be using the GCS it makes me wonder what we could use. We have the AVPU scale and it works, but nothing else really provides the easy to use accurate scale that can convey the the information required in a set of numbers. I am asking you and your man Dr. Green to present a new scale that will address the issues you brought up, we pre hospital EMS workers in Ann Arbor Michigan would love to help in any way we can. We have many studies and trials of different things and any advancement in rapid patient assessment would be well received. I say “don’t curse our darkness without shining some light”, give us the tool you imagined and change our world. Thank you.

    • admin says:

      Brian,
      Thanks for your input. I´m NOT proposing yet another tool. I am suggesting we get rid of one tool I happen to think, for acute medicine purposes, is flawed. Perhaps we have the tools we need. We have the AVPU, we have the FOUR score Josh mentions above, the scandinavian RLS-scale and a few more. But most importantly, we have the fuzzy logic we base our management on, that can´t ever be replaced by numbers on a scale. The GCS doesn’t add anything to that. I like how some of the commentators above emphasise a communicating short description of what they see.

  6. josh says:

    Pardon me if i am mistaken, but isn’t their an error in the following paragraph:

    “A GCS of 4 with the components 1+1+2 (E+V+M) predicts a mortality rate of 48%. But if we calculate the same GCS of 4 with the components 1+1+2 it will suddenly predict a mortality rate of 27%.”

    How was 48% and 27% calculated off the same 1+1+2? I believe it should have read another way adding to 4?

    Great post though!

  7. Billy Warwick says:

    I have to agree with your opion on the GCS. I have found in my 13 years in EMS and as a Paramedic instructor that I tend to treat the patient and the hell with the GCS. I have intubated people with GCS of 3 and some with GCS of 10 so I have to agree it needs to be set aside and find a better alternative. BW NREMTP

  8. Pingback: THE GCS STILL DOES NOT BELONG IN EMERGENCY MEDICINE | ScanCrit.com

  9. Seth Trueger says:

    @EMchatter tweeted this and I love it– and you reminded me of the best part of Green’s editorial: “A remarkable insight into the scale’s complexity was the embarrassing 2003 realization that one fourth of British hospitals were actually using the original 12-point form of the GCS rather than the current 13-point version, perhaps for decades without anyone noticing and correcting the error.” (PMID 14525875)

  10. Pingback: ¿Sirve para algo la Escala de Coma de Glasgow? | Reanimación.net

  11. Pingback: ¿Sirve para algo la Escala de Coma de Glasgow? | Medicina General y Salud Ocupacional

  12. Pingback: #EMDutch Review – 2 | EMDutch

  13. Pingback: CASE #6 UPDATE – FIND THE BLEEDING, STOP THE BLEEDING « Rural Doctors Net

  14. Austin says:

    I am glad I ran across this post. We are using GCS to assess dialysis pts for medical necessity. I have numerous questions as to the validity of the GCS assessing altered mental status. Anything under 15 would be a “blanket” AMS when the person can be alert and oriented, but unable to move extremities due to conditions that may or may not be related to a brain injury. Would you give these people a 15 on the GCS when they are alert and oriented, but cannot move an extremity? I am a paramedic and I am trying to instruct EMT’s/Paramedics on proper use of the GCS, but I am finding I have more questions than answers. There is too much grey area in the GCS.

  15. Pingback: NEURO – ¿Para qué sirve la Escala de Coma de Glasgow? |

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>