List Archives

Tue Sep 7 23:49:25 BST 2004

Writing about wine is like.....dancing about architecture?
CM Ursic
Sydney

-----Original Message-----
From: trauma-list-bounces at trauma.org
[mailto:trauma-list-bounces at trauma.org]On Behalf Of Lannehoa Yves
Sent: Wednesday, 8 September 2004 7:44 AM
To: 'Trauma & Critical Care mailing list'
Subject: RE : TRISS - does anyone believe it?

>Is an 85 cabernet equivalent to an 85 merlot???

Surely NOT, but it's my opinion
(Cabernet-Sauvignon)

http://www.lepoint.fr/vins/sommaire.html (in French)

R Parker rating is a rating system buid to determine the price of the wine
for export.

So : if you use a rating system (Triss ...) you must know before for what,
why and how it has benn created, in order to use it correctly .........

YL
Le Mans
France

-----Message d'origine-----
De : trauma-list-bounces at trauma.org [mailto:trauma-list-bounces at trauma.org]
De la part de Karim Brohi
Envoyé : mardi 7 septembre 2004 20:29
À : 'Trauma & Critical Care mailing list'
Objet : RE: TRISS - does anyone believe it?

Ian,

All I can tell you is that it won't be cheap, it won't be easy, and it won't
produce a single number at the end of the day.

I've been trying to think of an analogy, and off the top of my head, the
best I can come up with is the Robert Parker Wine Spectator rating system.
http://www.erobertparker.com/.  This guy has made a stack by assigning
scores to wines - 0-100.  The better the score - the better the wine.
Admittedly the score is subjective, but taste tests are controlled and
performed by mulitple judges along defined guidelines.  So it's great.  An
85 wine is better than a 65 wine.

EXCEPT: How do you define something as complex as a wine with a number?  Is
an 85 red equivalent to an 85 white?  Is an 85 cabernet equivalent to an 85
merlot???  Is an 85 merlot really beter than an 83 merlot?  Is an 85
chardonnay better than a 65 chardonnay with fish?  At the beach? And where
did this rating system come from?  Who are the judges?  Well, it turns out
that the judges are predominanlty American, and hence tend to favour
American-style wines (not out of politics, just out of palate).

THE RESULT: Bordeaux wine makers have switched from making their subtle,
complex clarets to making full-bodied whack-you-in-the-head-wth-fruit wines
that favour the American palate to gain better ratings in the Parker system.
(Just one example of the effect such a ratings system has).

It's true TRISS has statistical validity.  Whether this translates into
clinical validity is a matter of some debate.  (GIGO - Garbage In, Garbage
Out).  But by reducing a hospital's performance on hundreds of patients with
different injury characteristics for one outcome variable (mortality) to one
number is not only ridiculous but also fails to identify the weaknesses or
strengths of the system.

So how should one determine the quality of care?

Well, you can audit outcome and process.  Outcome will obviously include
mortality, but consider also disability, cost, recidivism, patient/family
satisfaction.  Process audit will include efficacy, efficiency, utilization
of resources, documentation etc.

But primarily I think we have to compare apples with apples, lychees with
lychees.  Maybe a hospital manages blunt trauma well but is terrible at
penetrating trauma.  Maybe neurosurgery is excellent but the absence of easy
access to angiography leads to an unacceptable high pelvic mortality.   So
perhaps there needs to be some 'index injuries' and/or 'index physiologies'
that can be audited for both outcome and process. 

Ultimately I think only process outcomes can really help you understand a
system.  Cape Town - my survival rates for penetrating cardiac injury - 70%.
My survival rate since (London, San Francisco) - 5%.  The difference: a 1-2
hour prehospital time in Cape Town.

Clearly I have no useful answers for you Ian.  But perhaos having no answers
is better than having the wrong answer?

Karim 

-----Original Message-----
From: trauma-list-bounces at trauma.org [mailto:trauma-list-bounces at trauma.org]
On Behalf Of Ian Civil
Sent: 04 September 2004 00:58
To: 'Trauma & Critical Care mailing list'
Subject: RE: TRISS - does anyone believe it?

Dear Karim,

I was looking for someone to bite on this but I guess no-one did.

Perhaps you might give us the benefit of your wisdom. What is the best way
to determine the quality of trauma care, both absolutely and relatively?

(I have a trauma outcome paper to review for the AustNZJSurg and I am
looking for inspiration)

Ian Civil
Director of Trauma Services
Auckland City Hospital

-----Original Message-----
From:	Karim Brohi [SMTP:karim at trauma.org]
Sent:	Friday, August 27, 2004 5:33 PM
To:	trauma-list at trauma.org
Subject:	TRISS - does anyone believe it?

I've just received an email from someone in Canada requesting information
on
the W & Z statistics, as they are starting to collect TRISS data for their
ICU.

So I told them.  And I asked why?  Z & W only exist to compare different
trauma systems. But then I also have to say that I have absolutely no faith
in the TRISS process whatsoever.  Every time I look at RTS, ISS, RTS+ISS =
TRISS, Ps, Z & W stats I see fudge factor piled upon fudge factor.
Coefficients trying to correct the inadequacies of the ISS/RTS scores.
Coefficients worked out on a specific subset of patients (US or UK), at a
specific date - which are almost certainly not correct for Canada, - even
if
you could generalise 'Canada' into one system with a homogenous set of
outcomes.

Even if you are a fervent believer in this mystic nonsense.  Say your unit
is calculated to have a low Z statistic.  How do you improve your Z
statistic?  Have less trauma deaths.  Brilliant.  Hospital X is worse than
Hospital Y because it's Z statistic is lower.  Why are they worse?  No idea
- they just are.

And yet people keep producing these meaningless statistics - to justify or
refute systems of care.  To churn out another iffy research paper. Because
it's an easy way of assigning a number to the effectiveness of a trauma
system - even if it's a house of cards.

One argument is that it is better than nothing.  Some comparison is better
than no comparison.  But that's just an argument for laziness.  Instead of
robust quality assurance/performance improvement programmes and process of
care audit - plug a few numbers into Excel and put them on a pretty chart.

So, I think TRISS is the biggest hoax on Trauma Care since the introduction
of colloids - and should be retired to a home for statistically-challenged
scoring systems.

Come on then, let me have it.

Karim

--
trauma-list : TRAUMA.ORG
To change your settings or unsubscribe visit:
http://www.trauma.org/traumalist.html

--
trauma-list : TRAUMA.ORG
To change your settings or unsubscribe visit:
http://www.trauma.org/traumalist.html

--
trauma-list : TRAUMA.ORG
To change your settings or unsubscribe visit:
http://www.trauma.org/traumalist.html

--
trauma-list : TRAUMA.ORG
To change your settings or unsubscribe visit:
http://www.trauma.org/traumalist.html

SOUTH EAST HEALTH CONFIDENTIALITY NOTICE

This email,  and the files transmitted with it, are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you are not the intended recipient, you are not permitted to distribute or use this email or any of its attachments in any way. We also request that you advise the sender of
the incorrect addressing.

This email message has been virus-scanned. Although no computer viruses were detected, South East Health accepts no liability for any consequential damage resulting from email containing any computer viruses.

List Archives

For your convenience, find the Trauma List Archive Below

TRISS - does anyone believe it?