dhis2-users team mailing list archive

[Jim Grace <jimgrace@xxxxxxxxx>]

Hi All,

My PEPFAR partner organization FACES in Kenya is configuring our own DHIS2 instance to collect data among the clinics we support. One of the standard government variables we collect and report each month is "ever enrolled in [HIV/AIDS] care". This means the cumulative number enrolled since the start of the service at each facility, which is often several years ago. It equals the sum of all the "enrolled this month" numbers going back through time. For example if "ever enrolled" is 4000 in January 2013, and "enrolled this month" is 100 in February 2013, then "ever enrolled" must be 4100 in February.

I am not (yet) seeing a good way to compute this in DHIS, so I'm asking advice. When we first started collecting this data from clinics in our current system of spreadsheets, we asked each clinic to compute the cumulative total and enter it for each month. We had a lot of errors this way, and we realized that it would be better calculating this in our tool instead of asking the data clerks to do this. So in our spreadsheets we just compute it by taking the previous month's "ever enrolled" and adding this month's "enrolled this month". Our records don't always go back to the start of care at each clinic, so we usually have an initial, hand-entered "ever enrolled" to get things started on the month before we enter real data for that clinic.

Now we are trying to convert our system of spreadsheets to DHIS, and I don't see a good way to do this. I've tried creating a Report Table with "Include cumulative" checked, but it only gives me cumulative numbers within the range of report months. What we really need is to get the numbers for one or more sites, for one or more months, and to have "ever enrolled" as one of the numbers that is reported per site, per month. The best options I can think of are:

1. Ask each site to compute this number each month and enter it as a data element.

2. Export the data from DHIS into our own software and have our software compute this number.

Are there any better options?

Cheers,
Jim