----------
From: Simon Carley <[log in to unmask]>
To: [log in to unmask]
Subject: size is important when assessing Paramedics!
Date: 02 September 1998 19:11
Everyone is talking about the lack of studies available that show the
effectiveness of paramedic (or prehospital in general) outcomes. It may
interest people to know that you can look at the power of studies in this
area to see if any are really big enough to answer the questions posed.
The
example below is from a forthcoming paper but it seemed topical and useful
to the debate. To my knowledge the number of studies with patient groups
of
the number shown is very small indeed!
Food for thought when one reads many of the studies quoted in this thread
where patient numbers may be as small as 100 per group!!
EXAMPLE
A study is planned to investigate the impact of a new type of paramedic
with extended skills for the management of out of hospital VF arrest. We
know from previous studies that the survival rate is in the region of 10%
at present8.
We need to specify the following.
1. We are interested in finding out if the new paramedics can increase
survival by 5%.
2. We will perform a Chi squared analysis on the data
3. As the outcome measured is live or dead at hospital discharge then
issues of variability do not concern us.
4. We are willing to accept a 5% possibility of getting a type 1 error.
(finding a difference if there isn't one)
5. We want to 80% certain of detecting a difference of 5% or more if it
truly exists (i.e. 80% certain of not making a type 2 error).
Using a nomogram method4,9 we can calculate that the number of patients
required would be in the region of 3500 (or 1750 in each group). To
calculate the exact number we could use one of a number of statistical
computer programs.
Simon Carley
Anaesthetics / Intensive Care
Stepping Hill Hospital
Stockport
England
[log in to unmask]
..4 Practical Statistics for the Medical Researcher. Altman DG (Ed).
Chapman
and Hall, London 1992.
8Mitchell RG, Guly UM, Rainer TH, Robertson CE. Can the full range of
paramedic skills improve survival from out of hospital cardiac arrests? J
Accid Emerg Med 1997;14:274-277
9How large a sample. Chapter 3 in Statistics in practice. Ed. Gore SM,
Altman DG. BMJ Publishing. London. 1982.
----------
> > From: Bjarkoy <[log in to unmask]>
> > To: [log in to unmask]
> > Subject: Re: Peter Holden & Paramedics
> > Date: 01 September 1998 23:48
> >
> > Mike replies
> >
> > J Trauma 1984 Jan 24(1):8-13
> > Pre-hospital advanced life support: benefits in trauma
> > Jacobs, Sinclair, Beiser & D'Agostino
> >
> > Found this on pub med within 7 minutes of looking - I am sure there are
> > more if I look beyond my nose.
> >
> > Mike Bajrkoy
> > paramedic
> > Sussex
> >
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|