At 14:38 19/05/05 +0100, Ted Harding wrote (in small part):
>... I've tested the water somewhat crudely by making a numerical
>estimation of the above sequential effect. This was done by simulation,
>rather than tackling it anayltically. 10,000 virtual mothers ....
>followed for 30 "years" through "ages" from 16 to 45, probability 0.08 of
>a birth in each year of age (simple Bernoulli Births) so as to ensure an
>expectation of 2.4 children per mother.
>.... From the above table I adopted a conservative probability of 1/10
>for a caesarean at any birth for the first time, all subsequent births (if
>any) caesarean. ....
>Results: [snipped]
>As can be seen (allowing for the random variation due to the small
>simulation size of 10000), this exhibits a steady and approximately
>linear rise from about 10% to about 30%. This is very comparable to
>the increase with age in Stephen McKay's table (above), for total
>(planned + emergency) caesareans.
Interesting - and, I suppose not surprising. We are gradually 'getting
blood out of the stone' in very small drops, so you could if you so wished
refine that simulation a bit. The Sentinel Audit tells us (2001) that the
probability of 'a Caesarian Section (CS) after a CS' is about 67%, so you
could use that rather than 100%. I'd also suggest that you could probably
increase that figure considerably (maybe your 100% would not be that far
off) for two or more previous CS.
>The point of presenting this is not to in any sense "prove" that
>the increase in caesarean rate with maternal age is due to this
>effect -- far from it -- but very much to demonstrate that, on
>the basis of simple age-independent assumptions which are numerically
>realistic, just about the entire increase *is capable* of arising
>from the sequential dependency ("once a caesaren, always a caesarean",
>as Alison has put it, in saying that this is no longer as much the
>case as it used to be).
Sure, and if you decrease that probability from 1.0 to 0.67 or thereabouts,
to reflect actual current (recent) practice, you will be left with some
space for all the other known causes of age-dependence of CS - notably
foetal distress (partly due to less efficient placentas in older mothers)
and 'failure to progress' (older reproductive systems not quite so good!).
>Nevertheless, the fact that the age-dependency is capable of being
>entirely explained (and even more plausibly partially explained in
>a proportion which we cannot ascertain) by the knock-on effect
>emphasises one sure point.
>It is only with good data, available at the level of detail which
>enables such effects to be ascertained as fact rather than
>speculation, that we can embark on sound discussions of such
>apparent relationships.
Whilst I certainly agree that the stone has not yet given up anything like
enough blood to allow us to seriously investigate any of this maternity
data, as I wrote earlier today, I don't really think that the reasons FOR
the age-dependency of CS rate is actually important in terms of the mains
issues we've been discussing. No-one is disputing the age-dependence of CS
rates and, indeed, we already know the main reasons for it. All we are
really looking for is adequate data over time to enable us to see to what
extent the overall change in CS rates can be explained in terms of the
concurrent change in pattern of maternal ages.
Kind Regards,
John
----------------------------------------------------------------
Dr John Whittington, Voice: +44 (0) 1296 730225
Mediscience Services Fax: +44 (0) 1296 738893
Twyford Manor, Twyford, E-mail: [log in to unmask]
Buckingham MK18 4EL, UK [log in to unmask]
----------------------------------------------------------------
******************************************************
Please note that if you press the 'Reply' button your
message will go only to the sender of this message.
If you want to reply to the whole list, use your mailer's
'Reply-to-All' button to send your message automatically
to [log in to unmask]
*******************************************************
|