It looks like I'm starting to get a few hits on this blog, which encourages me to keep posting. I assume all of the hits are from the link on the In-Hospital Defibrillation site.
I have just started again to track time intervals in the first minutes of resuscitation attempts at my hospital. It seems to me that it's really quite easy to do, and it puzzles me that no one apparently is doing it. (More about the method later.) I have talked at various times with various people involved in the National Registry of Cardiopulmonary Resuscitation (NRCPR) about problems with their reporting of time-interval data, so far with little effect. The basic problem is that they are reporting time intervals based on handwritten records which are quite clearly invalid. For example, over 25% of time intervals from arrest to first defibrillation are reported as "0 minutes," or instantaneous--nice trick if you can manage it. In my view, this lack of rigor is preventing the NRCPR from fulfilling its most important task: compiling good aggregate statistics showing definitively that there are serious delays in defibrillation in hospitals. Instead, the “data” reported so far seem to show that everything is just fine: see Peberdy MA, Kaye W, Ornato JP, Larkin GL, Nadkarni V, Mancini ME, Berg RA, Nichol G, Truitt TL. Cardiopulmonary resuscitation of adults in the hospital: A report of 14,720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation. Resuscitation 2003;58:297-308.
5 comments:
If using a digital wall clock, wouldn't 0-minutes encompass 1-59 seconds? If, as reported by NRCPR, approximately half of all events in a hospital occur in an ICU (and many more in the ED), a response of 0-minutes (1-59 seconds) for various procedures doesn't seem all that unrealistic, does it?
Thanks for the comment/question. This is the same basic "rounding down" argument used by NRCPR on their web site at one time to defend the validity of 0-minute time intervals. The 59-second window would apply only for codes that started right at 0 seconds, and would be shorter in the vast majority of cases: it could just as easily result in ADDING a minute. Say you accurately record the start time to the minute for a code that starts at 9:42:55, and CPR starts 10 seconds later (this is a thought experiment, not the real world). The start time for CPR would be 9:43 and the apparent elapsed time would be one minute.
I really don't think such speculation has much to do with the numbers NRCPR gets; I think it's more haphazard than that. In the face of uncertainty from multiple unsychronized clocks, I think there is probably some effect of reporting artifically short response times as well.
Kaye and Mancini looked at this after I made some noise about it in the NRCPR user group meeting at ECCU 2002 (my hospital is not a member; I attended at my own expense and out of my own interest). K & M say for the sake of argument that maybe the 0-minute times for ICUs might not be totally bogus, but they point out that 0-minute time intervals are reported for first defibrillation, etc., in unmonitored arrests where unit nurses are not authorized to defibrillate. The reference is: Kaye W, Mancini ME, Truitt TL. When minutes count—the fallacy of accurate time documentation during in-hospital resuscitation. Resuscitation 2005;65(3):285-290. I can send you a .pdf file of the article if you want to e-mail me at jastewart2@comcast.net and give me an address to use.
Furthermore, I think any reported interval of less than a minute for first defibrillation is bogus, ICU or not, if the arrest is sudden and the defibrillator is not right at the bedside.
John, I am also a RN working in a 400+ bed hospital entering data inot NRCPR. HOw would you propose that we obtain an exact time interval from arrest to CPR, epi, defib etc. I cant get the nurses to start CPR or place defib pads..now I want them to punch a clock. Could you give me a clearer picture as to how you think this might be accomplished
Thanks for the question, nurse from CT. I have an unpublished manuscript describing what I think is a fairly simple method for getting some of the time intervals, most importantly time from arrest/code alarm to first defibrillation. Write me at jastewart2@comcast.net and I'll send you a copy (anyone else, also). I have done some work on a more comprehensive PDA-based system for recording all major interventions, and I can tell you more about that also.
Your blog keeps getting better and better! Your older articles are not as good as newer ones you have a lot more creativity and originality now keep it up!
Post a Comment