MORTALITY FORUM 2013-04-30 Q1 Update 2

[mortality-forum]

MORTALITY FORUM 2013-04-30 Q1 Update 2

Question from Sam Rubin, Israel
We would like to know what is the current practice in other countries,
regarding the compilation of death certificates.
Do certifiers use a finite list to choose from or are they free to
formulate the conditions as they see fit.

Dr S.Rubin
Senior Nosologist, Central Bureau of Statistics
Jerusalem Israel
********************************
From Ruy Laurenti, Brazilian Classification Centre

In Brazil the doctors enters the causes of death in the Medical part
of the death certificates(Part I--a,b,c,d andPart II) according to the
medical/clinical diagnosis.The doctors don´t use a finite list(???)In
my oppinion it not correct to use a finite list.
**********************
From Christine Fowler/Devindra Awmee, New Zealand

In New Zealand the certifying doctor enters the causes of death in
free text on the medical certificate of cause of death.

Although our Births, Deaths and Marriages (BDM) Registry would prefer
the causes of death be restricted to a finite list of causes with
standard descriptors we consider that this would result in a
significant loss of specificity in our national cause of death
statistics.
**********************
From Colin Fischbacher, Scotland

In the UK, doctors are free to formulate the conditions as they see
fit, though they must describe the cause(s) of death to the best of
their knowledge following the ICD guidance. I find it hard to
understand how a finite list could be used without reducing diagnostic
accuracy, unless it was a very long list indeed. If the intention was
to continue using ICD10 then the list would have to be at least as
long as the ICD10 classification, but in practice it would be much
longer, as for most codes more than one clinical term can map to the
code.
********************
From Luis Manuel Torres Palacios. Mexican Collaborating Center for
FIC (CEMECE)

In Mexico certifiers don´t use a finite list to choose. They are free
to formulate the conditions as they see fit.
Our principal problem is that they apply the correct procedures to
register the causes of death like are described in vol. 2 of the ICD.
We are giving training for the correct filling of death certificate.
********************
From Lars Age Johansson, Sweden

In Sweden as well certifiers can enter any text they like. AS you
might know, there was an update in 2003 to Volume 2 of ICD-10 (section
4.1.3) according to which “automated systems must not include lists or
other prompts to guide the certifier as these necessarily limit the
range of diagnoses”. What we had in mind at that time was, however,
pick lists showing the 20 or 30 most common conditions, and of course
that kind of pick lists would have limited the range of diagnoses very
much indeed.
Since then other methods have been developed that might in fact
contribute to better certification but still not limit the range of
expression. In Denmark, for example, they have been using an
electronic certificate linked to a big database of medical expressions
(35,000-40,000 expressions) for some years now. The physician starts
typing a diagnosis, a search engine finds the most similar expressions
in the database and then the physician clicks one of them which is
then entered on the certificate. The advantages are that there are
never spelling errors, and automated coding is much easier to perform.
There is no evidence that introducing this system in any way changed
the statistical trends.
We are considering introducing something similar in Sweden, although
we would allow the certifier to enter expressions that are not in the
database – it should be a bit more difficult than fetching something
from the database, but not impossible. If you don’t have that
possibility, you would need a very efficient service for quickly
adding new terms to the database!