Guide · Uganda

HMIS 105 monthly report form, and what gets it sent back

Every health facility in Uganda compiles HMIS 105 at the end of the month and submits it to the Ministry of Health through DHIS2. In most small facilities it costs someone one to two days with a calculator, a pile of registers and a ruler. It does not have to.

Where to get the HMIS 105 form

Three places, in the order that actually works:

1. Your District Health Office. This is how most facilities get it, and the fastest. The DHO supplies the printed HMIS registers and report forms to facilities in the district, and can tell you which version is current — which matters, because the form has been revised and an old one will be queried.
2. The Ministry of Health Resource Centre. HMIS 105 is published in the Ministry's Health Management Information System manual, alongside the other HMIS forms and the instructions for completing them. Ask for the current HMIS tools; the Resource Centre is the office that owns them.
3. The Ministry's own website, health.go.ug, when it is reachable. Availability has been patchy, so we are not linking to a specific file that may not resolve — if you need it today, the district office is quicker than hunting for a PDF.

Whichever copy you get, check it is the current revision before you photocopy fifty of them. A facility filling last year's form correctly still gets the report sent back.

What gets a report sent back

Having the form is the easy half. The reason HMIS 105 eats two days in most facilities is rework — the report goes up, something does not add up, and it comes back. These are the usual causes, in roughly the order we see them:

1. Last year's form. The tools are revised. A facility filling an old revision correctly still gets it returned, and this is the one that wastes a whole cycle because nobody suspects it. Check the revision with your DHO before you photocopy fifty.
2. Disaggregated cells that do not sum to their total. If the age and sex breakdown adds to 47 and the total line says 52, the whole section is in question — not just the five. This is the single most common arithmetic fault, and it is almost always a transcription slip, not a counting one.
3. The same figure differing between two sections. A number that appears in more than one place has to be the same number in both. Compiling the sections on different days is how they drift.
4. Blanks where zero was meant. An empty cell reads as "not reported", which is a different claim from "none happened" and invites a query. If it was none, write nought.
5. Totals that contradict attendances. Diagnoses recorded exceeding the people seen, or a referral count larger than the visits it came from, are the checks anybody reviewing will run first.
6. A month's register with a gap in it. If a day is missing from the register, the report cannot be right and will not reconcile against the next month either. Better found now than in the query.
7. Late. A report submitted after the district's deadline is a problem even when every figure in it is correct, and it moves you to the bottom of the queue for the query when one comes.
8. Nobody's name on it. A report that does not say who compiled it is a report nobody can be asked about, and the question comes back to the in-charge instead of the person with the registers open.

The cheapest check there is: before it leaves the facility, add up each disaggregation and compare it with its own total line. That one pass catches most of what comes back.

What HMIS 105 actually is

HMIS 105 is the Health Unit Outpatient Monthly Report — one form, around twelve sections, summarising a month of outpatient work. It is not a new collection of data. Every figure on it already exists somewhere in the facility's registers; the form just asks you to add them up.

Between them, the sections cover:

• Outpatient attendance, referrals and diagnoses — broken down by age and sex, which is where most of the counting time goes.
• Maternal and child health — antenatal care, family planning, immunisation.
• HIV testing services.
• Laboratory — tests done, by type.
• Stock status of essential medicines and supplies.
• A financial summary.

Where each number is supposed to come from

This is the part that decides whether the report takes an hour or two days, and it is settled long before month end:

• Attendance and diagnoses → the OPD register
• ANC, family planning, immunisation → the MCH registers
• Tests → the laboratory register
• HIV testing → the HIV testing register
• Medicines → the stock cards and dispensing log

If a register was filled in properly on the day, HMIS 105 is arithmetic. If it was filled in from memory at 6pm, HMIS 105 is archaeology.

Why it goes wrong

Four failures account for most of the pain, and none of them are laziness:

1. Age and sex disaggregation. A diagnosis column is easy to count. The same diagnosis split into age bands and by sex means re-reading every line. Hand-tallying this is where errors enter.
2. The same patient counted twice. A patient who came, was sent to the laboratory, and came back to the clinician can appear as two attendances unless the register makes the return visit obvious.
3. Diagnoses written freehand. "Malaria", "malaria?", "clinical malaria" and "RDT+ve malaria" are one line on the form and four in the register.
4. Stock figures reconstructed at month end. The medicines section needs the position on the day, not a guess a fortnight later.

What actually makes it faster

In order of how much time each one gives back:

• Record the diagnosis from a fixed list, not free text. This single change removes most of the tallying ambiguity. It costs nothing and works on paper — a printed list beside the register is enough.
• Capture age and sex at registration, once, so no one has to infer them from a name at month end.
• Update the stock card at the moment of dispensing, not in a catch-up session. The medicines section then reads off the card.
• Tally weekly, not monthly. Four small counts beat one large one, and errors are found while the day is still rememberable.

Where software helps, and where it does not

Be clear about the limit first: CareBoon does not connect to DHIS2 and does not submit HMIS 105 for you. Someone at your facility still enters and submits the report. Any vendor telling you otherwise is worth a second question.

What a system removes is the counting. In CareBoon a patient is registered once with age and sex, the clinician picks a diagnosis from your own catalogue rather than typing prose, laboratory tests are recorded as they are resulted, and stock moves when a prescription is dispensed rather than when someone remembers. At month end those totals already exist. You read them off instead of counting them.

The same discipline that makes HMIS 105 quick also makes the rest of the month quicker, which is the honest argument for it — nobody buys software to file a form faster, but everybody notices getting two days back.

Questions we get asked

Related: why insurance claims get paid late · how clinics lose money to expired drugs · what a small clinic actually needs · moving off paper · what happens when the internet goes down · what clinic software costs · the controlled drugs register · an NDA inspection · selling on credit · how much to order