Data use at local level to support planning: addressing the denominator problem today at 14:00 (CEST)
You can post your questions ahead of, during, or after the session. Our speakers will check this thread for questions, and select some for responding to in the session, or follow up after the session has ended. Feel free to respond to other questions or add to them if you have something to follow up with.
Calculating local populations to give some sort of accuracy is a big issue. The use of crude population proportions are not always helpful - 4% of ANC 1/Live births/0-11 months pop is a standard used often. Creating a org Hierarchy that takes the big hospitals into a region of its own and then allocating a pop figure to them for estimating delivery coverage and BCG coverage also causes chaos. What seems simple ends up with a much bigger headache
From @philip27:
There is a challenge in adding new facilities and distributing populations for new areas, how can DHIS2 help us at the local level?
From @ksiliadin:
@philip27
Villages list per Facility catchment area is primarily a national administration issue. When a new health facility is created, It is up to MoH to clearly indicate which villages fall into its catchment area. Once this list is available, target population share is obvious.
Should we have villages in DHIS2? We are heading up to there with CHW catchement areas.
From @Jornbraa:
To be able to organise village lists, with population and other attributes and linked to facilities /catchment areas are important . We’ll try to address this in a new initiative to develop a ‘facility profile’ - it might be too *chaotic’ to organise the villages in the general org.unit hierarchy
So yes, we need to include villages in DHIS2 - exactly how (for full country applications), is being debated
From @Randy_Wilson:
Yes… it may be too heavy to include villages in the orgunit hierarchy.
… village listing more for local use - there are too many issues of villages with the same names, different spellings, etc…
From @Jornbraa:
To be able to organise village lists, with population and other attributes and
So yes, we need to include villages in DHIS2 - exactly how (for full country
From @Jornbaa:
standardized ‘Master Village List’ will be more complex than the already very complex ‘MFL’, but some countries have @ksiliadin village lists
From @ Federica Maurizio UNFPA APRO:
For estimating populations baselines related to fertility, for example expected number of pregnancies in a given year, would you use the district or provincial crude birth rate estimated by national surveys like DHS? or would the Statistics office also provide annual revisions of estimated Crude Birth Rates in the districts/provinces?
From @souleymaneya1:
@federica The DHS Survey data is not use the estimation. The statistical office provide annual revisions of estimated Crude Birth rates in the districts/provineces
From @Norah_Stoops :
Population growth rates are often misleading with there being a difference between growth in cities and rural areas
Population figures are usually the last thing to be entered into DHIS2 as they are not always seen as important and they are such a football between different authorities and no one wanting to take responsibility
From @Randy_Wilson:
do most countries store national target age groups - target population percentages - in the system as constants to be used across all facilities for use in indicator calculations? What other options are practical - especially when the percentages (e.g. % women of reproductive age) change over time.
From @Jornbraa
if not % stored in the system, it is most places used (e.g. 4% for expected pregnancies’) for calculating the target population to then include in the DHIS2
From @Edem_Kossi:
What is done in some countries is to simply to store the processed data
From @seidhus:
Here in Ethiopia it has been more than ten years we had the last census. We see that estimates until the district coming from the Central Statistical Authority and estimates coming from the community level Health Posts (the lowest level) don’t add up. Though the community estimates may be close to the correct value, they are not official. So there’s a disconnect.
From @philip27:
Yes, Koffi is very right in the response, but the issues are for real time reporting from remote communities where I am entering data from new hot spots which are not in the data base. Some times getting the National Administrators to address the new areas issues take long time, may not present the case burden per specific location. I am implementing HIV Program for Key Populations in Liberia and because of clients movement, we always placing cases under the nearest hot spots/communities in the dhis2 instead of their actual location of a particular client.
From @Edem_Kossi:
The reality is that some times they keep applying the growth rate for many years and then the next census will reveal that the population is decreasing in some villages
Recording of the session can be found here: