Hi Arthur,
If you have seen some of my mails over the past few months (years
actually), I have been a big critic of the implementation of data
element categories. It has been raised before that the disaggregation
is not an intrinsic property of the data element. There could be a
valid reason why the same data element is disaggregated differently in
some situations than others, and we should not have to tell the M&E
folks they cannot do it, simply because DHIS has a slightly different
(and I think incorrect) model. And of course, we all know that the
disaggregations can change over time. We know the M&E folks love to
create different disaggregations, so it is not really up to us to tell
them whether they are right or not, but rather try and model what is
actually happening.
In your case, I think you can create two separate data elements "IPD"
and "OPD" and apply two separate category combinations, and then
follow this up with an indicator which would combine the two data
elements to get the total.
The other possibility would be to create a single element and
disaggregate by all possible combinations of categories, and then use
the "Grey field management" or "Custom forms" to determine which of
the data element/category options actually get displayed to the user.
The problem arises however in the category model, that each of the
categories must add up to a total. So, if you have age groups which
intersect, say 0-1 and 0-5, and data is entered in both of them (which
should not be your case, as you would either enter IPD or OPD age
disaggregations for a single facility), the totals are going to get
messed up. We have used this approach in Zambia to deal with when
category options change over time, but in a very specific situation.
Previously we have 0-5 and 5+, but this was changed to Under 1, 1-5,
and 5+. However (as you well know) there was a switch at a certain
point in time when the disaggregations changed, and therefore, there
should never be any overlapping data. In our situation, we know we
should never have data for the same period and facility with both 0-1
and 0-5, so application of the category model (whereby it assumes that
each of the category options add up to a total), will work, but it is
risky.
The current category implementation works in many situations, but the
association between data element, instead of dataset, is probably one
which needs to be revisited.
Regards,
jason
···
On Sat, Sep 17, 2011 at 7:34 PM, Arthur Heywood <arthurheywood@gmail.com> wrote:
Hi Jason and Andrew
Your advice is the right and logical one BUT .... for some reason best known
to the Tanzania HMIS ... we have DIFFERENT age groups for inpatients and
outpatients ......
We database developers have allowed ourselves to be talked into different
age categories for IPD and OPD "in order to please everybody" "so they will
all use our database" ... and have landed up with a proper camel ( horse
designed by a committee) with everyones bits and pieces tagged on
By accepting that we are stuck with this confusion for the lifetime of our
database ..... and that is what we need to find a solution to!
Regards
Arthur
On 16 September 2011 15:59, Muhire Andrew <muhireandrew@yahoo.com> wrote:
Dear Arther, i think you have to consider data element category since you
are talking of OPD and IPD.This can differ depending on their categories.
(Make sure they all have same category eg Age is the same,Gender the same).
Because most of the data element are the same but differ when you see their
categories.
________________________________
Muhire Andrew
HMIS/Ministry of Health
andrew.muhire@moh.gov.rw
God is my provider.
________________________________
From: Masoud Mahundi <mmahundi97@gmail.com>
To: udsm_hmis_team@googlegroups.com
Cc: "dhis2-users@lists.launchpad.net" <dhis2-users@lists.launchpad.net>;
Muhire Andrew <muhireandrew@yahoo.com>; Ola Hodne Titlestad
<olati@ifi.uio.no>; Jason Pickering <jason.p.pickering@gmail.com>
Sent: Friday, September 16, 2011 2:50 PM
Subject: Re: [udsm_hmis_team] Re: [Dhis2-users] [Dhis2-devs] Is it
practical for one data element to exit in two dataset?(eg District Hospital
and Health center)?
Sound good, doc. We can discuss that Monday morning
On 16 Sep 2011 15:47, "Arthur Heywood" <arthurheywood@gmail.com> wrote:
> Mahundi et al
> Can we not apply this to OUR Tanzania data elements, which have
> different
> data elements for IPD and OPD , acute and chronic
>
> Working with different data elements with same meaning is driving me MAD
> e.g. 70 (or more) examples such as
>
> - Diabetes Mellitus (ear infection acute, ear infection chronic etc etc
> etc ) in OPD
> - Diabetes Mellitus ( etc) in IPD
>
> *Why can we not have the same disease in two separate datasets??*
>
> Is it not possible to combine these?
>
> Regards
> Arthur
>
> On 16 September 2011 14:22, Jason Pickering >> > <jason.p.pickering@gmail.com>wrote:
>
>> This is totally possible. You may have the same data element being
>> collected on two separate aggregation forms. Instead of creating two
>> data elements to represent the same thing, you just need to place the
>> same data element in two separate datasets. If you have "Total
>> attendance" at both the Hospital and Health center level, then you
>> might have two sepererate datasets (Hospital and PHC), with some data
>> elements which are different, some which are common.
>>
>> Best practice is to try and consolidate as many data elements as you
>> can across the different service levels, as in practice, we have seen
>> that they may be named differently at different service levels, but
>> actually represent the same occurrence.
>>
>> Regards,
>> Jason
>>
>>
>> On Fri, Sep 16, 2011 at 12:19 PM, Muhire Andrew >> >> <muhireandrew@yahoo.com> >> >> wrote:
>> > Dear team,
>> > I can see its possible to link one data element to two or
>> more
>> > different dataset, Did this affect anything? eg one data element exit
>> > in
>> > both District Hospital and Health Center? can this affect the
>> > database?
>> Data
>> > entry? is it practical?
>> >
>> >
>> >
>> > ________________________________
>> >
>> > Muhire Andrew
>> > HMIS/Ministry of Health
>> > andrew.muhire@moh.gov.rw
>> > God is my provider.
>> >
>> > ________________________________
>> > From: Lars Helge Øverland <larshelge@gmail.com>
>> > To: nsanzumuhire venuste <vensanze@yahoo.fr>
>> > Cc: Knut Staring <knutst@gmail.com>; Muhire Andrew <
>> muhireandrew@yahoo.com>;
>> > "dhis2-users@lists.launchpad.net" <dhis2-users@lists.launchpad.net>;
>> > Ola
>> > Hodne Titlestad <olati@ifi.uio.no>; "dhis2-devs@lists.launchpad.net"
>> > <dhis2-devs@lists.launchpad.net>
>> > Sent: Thursday, September 15, 2011 9:30 AM
>> > Subject: Re: [Dhis2-users] Re : Dear Dhis2 users how can i delete?
>> >
>> > Hi,
>> > most operations are cached at the database layer so when doing manual
>> > database operations outside dhis you will need to either restart the
>> > application or clear the cache by going to "adminstration" -> "cache
>> > statistics" and clicking "clear cache".
>> >
>> > Lars
>> >
>> >
>> > _______________________________________________
>> > Mailing list: DHIS 2 developers in Launchpad
>> > Post to : dhis2-devs@lists.launchpad.net
>> > Unsubscribe : DHIS 2 developers in Launchpad
>> > More help : ListHelp - Launchpad Help
>> >
>> >
>>
>> _______________________________________________
>> Mailing list: DHIS 2 Users in Launchpad
>> Post to : dhis2-users@lists.launchpad.net
>> Unsubscribe : DHIS 2 Users in Launchpad
>> More help : ListHelp - Launchpad Help
>>
>
>
>
> --
> *Quality is not an act .... It is a Habit* (Aristotle)
> Skype arthur_heywood_za
> Tanzania +255-773-669393
> SA +27-769405494
>
> --
> You received this message because you are subscribed to the Google
> Groups "UDSM_HMIS_Team" group.
> To post to this group, send email to udsm_hmis_team@googlegroups.com
> To unsubscribe from this group, send email to
> udsm_hmis_team+unsubscribe@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/udsm_hmis_team?hl=en
_______________________________________________
Mailing list: DHIS 2 Users in Launchpad
Post to : dhis2-users@lists.launchpad.net
Unsubscribe : DHIS 2 Users in Launchpad
More help : ListHelp - Launchpad Help
--
Quality is not an act .... It is a Habit (Aristotle)
Skype arthur_heywood_za
Tanzania +255-773-669393
SA +27-769405494
_______________________________________________
Mailing list: DHIS 2 Users in Launchpad
Post to : dhis2-users@lists.launchpad.net
Unsubscribe : DHIS 2 Users in Launchpad
More help : ListHelp - Launchpad Help