Graeme McCrory wrote:
> 55 year old female smoker asks for a screening CXR as her
> younger sister
> (heavier smoker) has just been diagnosed lung ca.
>
> Should I order a CXR in an otherwise healthy person?
> What is the reliability, validity, predictive value etc. of the test?
> How often should it be repeated?
super dooper flak jacket donned ; hate to do this chaps and chapesses
but i love a wind up; see bottom line below if mathematical masturbation
not your thaaang
working the 2x2 table is easy
some assumptions first;
prevalence of lung cancer in population of 55yr old female smokers (your
client) = 1%
this is a guess and could be wildly wrong; so what?
sensitivity and specificity of cxr in detecting the true positives and
the true negatives respectively? lets guess 85%, why not?
then, and you can play with the maths yourself if you like, i have
calculated the negative predictive value to be 99.8% (tells you that if
the cxr is negative there's a 99.8% chance its right or only changes
your pre test prediction of a 99% chance of not having a lung cancer by
.8%); and the positive predictive value to be 5.4%, which means that
even if the cxr shows a lung cancer there's a 91.5% chance its wrong or
changes your pre test prediction of 1% chance of lung cancer to 5.4% ie
by 4.4%.
BOTTOM LINE (not shouting just capitals for ahmed)
the chance of getting a positive result is 15.7% (which will be wrong
94.5% of the time.....) now to my mind this means that there is a
significant chance of obtaining a positive result that is going to
needlessly scare to death this unfortunate lady; (and it will make her
smoke more)and is an excellent reason for not doing the cxr.
however, yes i do appreciate we're not talking about screening this
population, but the chances are there for the individual too.and yes the
assumptions may be wrong but i think these numbers are quite scary; as
GPs i think we tend to ignore these things partly because as individuals
we deal with relatively small numbers of these situations, so see
relatively few of the disasters and when the _wrong_ (in both senses)
result appears we just shrug and say we did our best whereas in fact we
did the patient a disservice.
medico-legally we're home and dry of course
also inclined to agree with Hotch's views given how desperately
addictive the stuff is
ok, fire away, :-))
cheers
owen dempsey
GP
West Yorks
UK
'best hand held calculator in the north'
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|