In a message dated 19/07/96 10:18:10, Jon Wilcox writes:
>Okay subscribers.
>Now for a totally new line of thought.... Viruses and arterial disease.
>I recently had a 42 year old male present with acute carotidynia (severe
>carotid artery pain) while suffering from presumed post-viral trigeminal
>neuralgia (his wife had the same postviral neuralgia at the same time!)
>It settled spontaneously in 1 week without any drama, however he had
>lost his father (aged 35) following a ruptured carotid artery aneurysm.
>I was tempted to give him oral Zovirax, though he didn't in fact need
>it.
Common things occur commonly. Given enough subjects, rare things will occur
together.
What proportion of the population has had one of the herpes viruses ( include
VZ, CMV, EB and all Herpes hominis types). Next question, what proportion of
the population would have abnormalities on detailed ultrasonography of the
carotids. I remember seeing artherosclerosois on post mortem in children who
had died of trauma, when I was a junior doctor.
He had a severe genetic load ( how tall is he/was his father, does his arm
span exceed his standing heigh =Marfans). We are all able to report
extraordinary coincidences because, between us, we are bound to see rare
things occuring by chance in the same family GP UK must cover an enourmous
number of patients. I know of a family with two coeliacs and a case of Wests
syndrome, of a family with trisomy 13 and small cell ca lung under 35. All
very rare, but given unlimeted numbers, bound to occur in the same small
groups, the larger the extended family, the more likely.
The leukaemia cluster analogy is best remembered here. Clusters of leukaemia
occur because of the rarity of the disease. If you take a handful of rice,
consisting of 64 grains and throw it at a chess board, you would never expect
there to be one grain in every square. There will be a lot of grains that
slide off and some squares with several grains. the size of the squares (i.e.
sample size) and thus the size of the board detemines the clustering, and
how hard you throw (c.f. the time scale studied) determines the clusters you
get AND the total numbers you get. Think about it.
I hate stats by the way, but understand coincidence and clustering ( I think)
Intersesting research, however, could be done by finding out how many patient
we can survey between us. How many patients are their in each practice
covered by a GPuker? Given good databases and read coding etc. etc. , what
questions can we answer epidemiologically? I know this has been tried with
specific GP systems, but there are still some possibilities. Answers on a
postcard (or what ever the e-mail equivalent is)
Trefor Roscoe Email; [log in to unmask]
Beighton Health Centre Tel 0114 - 269 5061
Queens Road, Beighon Fax 0114 - 269 7186
Sheffield S19 6BJ
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|