chbare Posted May 24, 2012 Share Posted May 24, 2012 Dwayne, how often do humans blow the interpretation? Look at all the discussion here. We certainly do not have consensus. Also, why go with one in 50? Just food for thought, but we are willing to accept p values of three, four and nearly five percent when it comes to literature. However, I cannot deny that automated interpretations can get it wrong, but a machine will never get tired at 03:00 after several calls and a machine is very good at using a consistent approach to anslyse the ECG. The information can be helpful. Determining axis and detecting certain anomalies such as digoxin effect, limb lead reversal and fascicular blocks can be much easier if the machine picks up on the presence of said anomalies. I am certainly not advocating for throwing your brain away but don't throw away free information if it is potentially helpful. 1 Link to comment Share on other sites More sharing options...
paramedicmike Posted May 24, 2012 Share Posted May 24, 2012 I'm not arguing to throw it away or disregard it. I'm saying hold the information back until people have presented their thoughts then present the machine findings. That way you get unbiased opinions on findings which can then be questioned or modified after the machine interpretations are taken into account. I think we're just looking at this from two different perspectives. On the ambulance at 3AM is vastly different from five tracings with no scenarios to go along with them. I'm looking at this as purely an academic exercise. Real life, as we know, is different. 1 Link to comment Share on other sites More sharing options...
Bieber Posted May 24, 2012 Share Posted May 24, 2012 I personally have gotten into the habit of never looking at the machine interpretation. You make some good points, chbare, and maybe taking a look at it more often might not be a terrible idea. I'll have to think about it. I do feel the need to point out, however, that this is how Skynet got started... =) First we're trusting the machines to aid us in interpreting our ECG's, next we're their prey to be hunted down and destroyed. 2 Link to comment Share on other sites More sharing options...
DwayneEMTP Posted May 24, 2012 Share Posted May 24, 2012 I get you Chris. My argument isn't based in logic nearly so much as personal bias. Fool me once, shame on you, fool me twice...man, I hate that. So it's more of a knee jerk reaction for me. I trusted it, it let me down, now I don't trust it any more. I think part of it too is that I'm not strong enough in ECGs any more to be able to read the computer interpretation without fearing having it drag me down an incorrect path that I'm not smart enough to think my way off of. I do read the time values often though. And that moment in the ER was certainly painful enough for me to be one trial learning. I abhor the thought now of saying, "This is my opinion on what's going on", hearing "What makes you say that?" and being left with, "Well, the machine said so..." I guess I'm not making the argument against using the machine interpretation so much as not considering it an ECG deity. 1 Link to comment Share on other sites More sharing options...
Kiwiology Posted May 24, 2012 Share Posted May 24, 2012 Dwayne you hit one of my nails on the head, if the machine interpretation says so perhaps there is a bias towards being "seeing what you want to see" or getting lazy and relying on machine interpretation Also boo Los Angeles Paramedics (both LAFD and LACoFD) who rely on machine interpretation for STEMI activation 1 Link to comment Share on other sites More sharing options...
Bieber Posted May 24, 2012 Share Posted May 24, 2012 Also boo Los Angeles Paramedics (both LAFD and LACoFD) who rely on machine interpretation for STEMI activation Boo! Link to comment Share on other sites More sharing options...
Just Plain Ruff Posted May 24, 2012 Share Posted May 24, 2012 Guys, I think that you can use the machine interpretation as a adjunct to your own interpretation. But don't just use the machine. You should use your own knowledge and then use the machine as a second opinion but don't use it as the source of truth. Like we always say here, look at your patient not the monitor, if the patient looks ok and is talking to you but the monitor shows V-fib then the monitor is wrong. It's the same concept and I think Dwayne is the biggest advocate for that thought process here. Trust your own clinical skills and use the machine interpretation as a backup or a 2nd opinion. That's all I'm saying. Link to comment Share on other sites More sharing options...
Eydawn Posted May 31, 2012 Share Posted May 31, 2012 I personally have gotten into the habit of never looking at the machine interpretation. You make some good points, chbare, and maybe taking a look at it more often might not be a terrible idea. I'll have to think about it. I do feel the need to point out, however, that this is how Skynet got started... =) First we're trusting the machines to aid us in interpreting our ECG's, next we're their prey to be hunted down and destroyed. Don't forget the "smart" cars ready to kill us all ala Robopocalypse... (If you haven't read that, go out and get a copy. It's excellent!) As far as releasing information with hospital info attached, yes, it's often against policy, and though it may be a stretch, it could be just enough identifiable information for some asshat on the interwebs to recognize that the strip belonged to grandma when she went to said hospital on said date and had her EKG run at said time by that nice tech with those initials. Ergo, part of grandma's health information, that is identifiable (by this one jerk), has been made public and it is indeed a breach of HIPAA. Not a harmful one, but one nonetheless that someone could get a hospital severely fined for. That's why they're so strict... it's all down to the $$$... It's so easy to screw it up. Especially in small communities, where news gets around. If you're going to post strips, definitely remove any timestamps or other regional identifiers to reduce the risk of someone recognizing it. Photocopy the strip, white out anything that could be identifiable, and then take it home with you to scan... In any case, I'm glad someone's posting strips! Great thread, technicalities aside... and I agree with looking at it yourself first and then looking to see what the machine thinks it is and comparing the two. It's like having your own robot friend... "hey man, I think this is XYZ because of this... oh, you think it's this? Why do you think that?" (Am I the only one that talks to my diagnostic equipment?) Wendy CO EMT-B Link to comment Share on other sites More sharing options...
Recommended Posts