AW: AW: [HSF] Intent to treat and crossover OT For Tea´s comments and Axel

Fri Feb 16 18:05:31 EST 2007

So, Tea would say 200% mortality (from cardiologist and surgeon)...
But God saved his life, I put him on ECMO on Sunday and he got out of it 
today, unbelievable.
Ich hätte Dich gerne kennengelernt, aber musste hier operieren. 
Roberto

-----Ursprüngliche Nachricht-----
Von: openheart-l-bounces at lists.hsforum.com
[mailto:openheart-l-bounces at lists.hsforum.com] Im Auftrag von prof.dr.axel
laczkovics
Gesendet: Freitag, 16. Februar 2007 11:52
An: OpenHeart-L at lists.hsforum.com
Betreff: Re: AW: [HSF] Intent to treat and crossover

honestly speaking this cannot be a problem:

for the cardiologist it is a failure of intervention, conversion to 
surgery and - if they document their cases correctly -  a 30d mortality.

for you (or any surgeon,  who accepted the case) it is a mortality and 
in summary you can show,  that this was an emergency, EU-score  XXL, 
and dead.

therefore as soon as  you  accept  a patient, he/she  is YOUR patient. 
how can anybody see this ?problem different?

(but just returning from the DGTHG in hamburg, i aggree, that 
statistics, ob better documentation is a hairly thing.)

ciao, axel laczkovics

> Tea,
> I agree that intention to treat must be in action and not theoretically
> thinking.
> Think this case: patient with acute MI, the cardiologists loose a lot 
> of
> time trying to recanalize the artery. They fail, and send you the 
> patient in
> cardiogenic shock. You have to assume the case. He dies.
> The original intention to treat in action was PTCA. Is their mortality.
> Or: they made crossover, and it´s your mortality.
> How do you see this problem?
> Roberto
> -----Ursprüngliche Nachricht-----
> Von: openheart-l-bounces at lists.hsforum.com
> [mailto:openheart-l-bounces at lists.hsforum.com] Im Auftrag von Tea Acuff
> Gesendet: Freitag, 16. Februar 2007 05:41
> An: OpenHeart-L
> Betreff: [HSF] Intent to treat
>
> This intent to treat analysis has been setting on my mind, and this 
> week I
> had the real world experience of not treating as I had intended. 
> (That's
> right, it did not work out as planned.) Playing with numbers 
> (statistics) is
> one thing, but do they actually mean anything real? Like so much we are
> taught, we take "intent to treat" at face value. Unfortunately, 
> however,
> intent is not really something, it is "something over there". Let me 
> come at
> this idea from several directions.
>
> Straight up "intent" is not measurable, and may be one of the hardest 
> things
> to decipher from observed actions. "Intent to treat" is supposed to be 
> used
> to hold accountable irregularities in deviations from the study. In 
> fact it
> allows the introduction of bias while ignoring that it is actually
> occurring. The "no I don't mean that data", but I mean the "other data"
> evidence. I would propose based on my recent experience, that an 
> "ideal"
> technique is one which is always acted upon, not one that was 
> intended. Hal
> points to off pump. What about mitral repair? Is my intending to 
> repair a
> valve the same as Hal's intent? I doubt it , and I (and my patient) 
> might be
> better off if more often I just intend to replace it the first time 
> right.
> Or Hal did an on pump. The only possible way intent can mean anything, 
> is if
> intent is an actual action not a thought. Perhaps we should call it the
> analysis by could and couldn't, not intent.
>
> Pursuing this idea of intent as action, we must be careful what we 
> call an
> action. In the formality of a study this is theoretically the
> "randomization".  However this definition has no real world meaning. 
> To use
> a ridiculous example (mind experiment) to show this folly, image an
> "experiment" comparing "surgery" for Type I aortic dissections verses
> medical therapy (as if surgery doesn't also include medical therapy). 
> If we
> include a nuclear medicine viability scan either before or after the 
> intent
> point to assess cardiac ischemia which takes 24 hours, we will likely 
> have
> much confusion in the analysis of the results compared to present 
> standard
> therapy for emergency surgery. (Almost everyone dies and/or is crossed
> over.) Intent must be the intent in action, not theoretically thinking 
> of
> it.
>
> On a strictly numbers basis we are using the model of a "crossover" in 
> what
> are usually inappropiate analyses. It is unlikely that the 
> distributions on
> crossover are bipolar or two tailed. One would be unlikely (even if
> theoretically or occasionally possible) to convert in real time an on 
> pump
> to an off pump, or a CABG to a PCI, or yes, Hal, a replacement to a 
> repair.
> Treating these as equal possibilities does damage to the more robust 
> therapy
> and hides the deficiency of the possible inferior arm. I am not a 
> numbers
> guy, but this is like saying adding and multiplying are statistically 
> the
> same. They both make things larger. (As long as we stay with small 
> numbers
> like 1,2, and 3, and maybe 4 with rare exceptions.) Notice that this 
> quickly
> fails at the upper tail.
>
> The strongest argument against "intent to treat" is that it fits 
> squarely in
> my assertion that our evidence is not about the experiments but the
> experimenters, that is, it is a study more of the doctor's behavior to
> therapy and not the behavior of the patient to the therapy. Nothing is
> learned about the parameters (if any) that guide crossover, which is 
> really
> what the the whole stupid experiment was supposed to be about. For 
> that we
> are left with the valuable explanation of "intent to treat" Yeah, that 
> will
> help the next time I am looking at what Giuseppe imaginatively calls 
> the
> "first time" again and again. And thus we have what we called in the 
> old
> days with Tony, a tautology. We trade not knowing what to do, for the 
> intent
> to do the right thing even if we don't what we will do yet. Nice to 
> have
> evidence that we read the papers and are trying. And most simply, as
> everyone knows that follows HSF, if I am right then almost everybody 
> else is
> either wrong or they don't
>  understand what I "intended" to say.
>
> tea
> _______________________________________________
> OpenHeart-L mailing list
>
> Send postings to:
>  OpenHeart-L at lists.hsforum.com
>
> To UNSUBSCRIBE, to CHANGE email address, or to view archives:
> http://mmp.cjp.com/mailman/listinfo/openheart-l
>
> All messages transmitted by the OpenHeart-L are subject to the 
> policies and
> disclaimers posted at:
> http://www.hsforum.com/listdisclaim
> -----------------------------------------
>
> _______________________________________________
> OpenHeart-L mailing list
>
> Send postings to:
>  OpenHeart-L at lists.hsforum.com
>
> To UNSUBSCRIBE, to CHANGE email address, or to view archives:
> http://mmp.cjp.com/mailman/listinfo/openheart-l
>
> All messages transmitted by the OpenHeart-L are subject to the 
> policies and
> disclaimers posted at:
> http://www.hsforum.com/listdisclaim
> -----------------------------------------
>

_______________________________________________
OpenHeart-L mailing list

Send postings to:
 OpenHeart-L at lists.hsforum.com

To UNSUBSCRIBE, to CHANGE email address, or to view archives:
http://mmp.cjp.com/mailman/listinfo/openheart-l

All messages transmitted by the OpenHeart-L are subject to the policies and 
disclaimers posted at:
http://www.hsforum.com/listdisclaim
-----------------------------------------