dhis2-devs team mailing list archive

[Bob Jolliffe <bobjolliffe@xxxxxxxxx>]

Hi Lara

There has not to my knowledge been any attempt to make use of snomed-ct
terminologies within dhis2 implementations.  Being a clinical vocabulary it
is actually hard enough to picture a compelling use case as many typical
data elements are not so neatly clinical. Not being an expert on these
things, but I suspect ICD10 would be a better fit, but even then it is very
"disease-oriented".

There is certainly a case for standardizing the coding of dhis2 metadata
and this happens *within* countries when designing minimum datasets.  There
have been occasional efforts to align with WHO efforts for certain common
indicators, for example those related to MDGs and a couple of attempts to
try and set up sort of dhis2 metadata reference implementations to help
country implementations get started with the benefit of what might be
considered best practice behind them, but these have not really taken off
yet.

As Hannes has pointed out, there is a terminology service in Rwanda and
there has been some not very intense discussion about using it to represent
the HMIS dataelements and indicators but not much serious effort.  Besides
being an interesting idea its not really clear what actual problems would
be being solved by doing it.  I guess one major challenge is how to
describe the relationship between an aggregated data element and an EMR
which contains encounter level data potentially coded using snomed, icd10
or what have you.  Simply tagging a dataelement with the same label is in
general not sufficient - it doesn't tell you what actually to count from
the EMR though it potentially gets you into the right ballpark.

HISP South Africa has also been quite active in trying to setup "standard"
health data dictionary type applications, but again I would be surprised if
they have followed a clinical focus on something like snomed-ct.

Coding of dataelements within dhis2 (ie actively using the code field) is a
relatively recent phenomenon in dhis2 so its not yet clear all the
different ways people will come up with for assigning codes.  I can imagine
there might be some benefit will emerge for using snomed-ct codes for a
subset of dataelements but I haven't seen it yet.

Regards
Bob


On 26 March 2014 20:24, Lara Kellett <Lara.Kellett@xxxxxxxxxxxxxxx> wrote:

> Hi Everyone,
>
> We are in the beginning phases of planning a DHIS installation to collect
> aggregated data from our project sites. We are currently deciding upon
> which terminology standards we would like to make use of with the current
> front runner being SNOMED CT. Would be interested to hear of any other DHIS
> users about which standard terminologies have been used in their DHIS
> implementation (and why) and technically how this has been achieved (is it
> just through assigning the code when creating Data Elements or have
> terminology servers etc been integrated).
>
> Thanks,
>
> Lara Kellett
> Médecins Sans Frontières/Doctors Without Borders
> * lara.kellett@xxxxxxxxxxxxxxx <lara.kellett@xxxxxxxxxxxxxxx>*
> * +1.631.480.4969* <%2B1.212.763.5741>
> Skype: lara.kellett
> _______________________________________________
> Mailing list: https://launchpad.net/~dhis2-devs
> Post to     : dhis2-devs@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~dhis2-devs
> More help   : https://help.launchpad.net/ListHelp
>
>