AOWA · Anesthesia knowledge for West Africa

AI assistants built on reality.

Today's AI was trained on hospitals with everything. We're mapping what anesthesia teams in Ghana actually have, so tomorrow's tools give answers that fit the room you work in.

Confident answers. Wrong hospital.

Ask an AI tool or an international guideline about anesthesia and it will assume steady oxygen, full monitoring, an intensive care unit, blood on demand and a specialist on site. Many hospitals work without some of these every day. The people who know what each hospital really has are the people doing the work, and that knowledge has never been written down in a form a machine can learn from or a person can check.

Same question. Two kinds of AI.

Illustration of where this is going. The assistant is not built yet.

Built on assumptions

What monitoring will I usually have at a district hospital?
You should have pulse oximetry, capnography, ECG, blood pressure and temperature monitoring, with ICU backup available.

Sounds right. Is it true here? No way to check.

Built on reality

What monitoring will I usually have at a district hospital?

Pulse oximeter: reported available, often with probe shortages. Source: clinician survey.

Capnography: we don't know yet. Not enough evidence.

ICU backup: not observed at this level. Source: expert interview.

Every answer shows its source. Unknowns are flagged, never guessed.

First the map. Then the assistant.

  1. 01Now

    The map

    What teaching, regional and district hospitals have for anesthesia: staff, equipment, monitoring, medicines, blood and services. Every fact sourced. Every gap counted.

  2. 02Next

    Readiness insight

    A clear picture of what each hospital level has and lacks, so funders and planners put resources where they're missing.

  3. 03Future

    The assistant

    AI that answers only from checked facts, cites every one, and says 'we don't know' when the map does. It informs; clinicians decide.

You are the ground truth.

Most of the map is still blank. We know almost nothing yet about regional hospitals, or about monitoring equipment at any level. No dataset can fill that in. You can. Fifteen minutes of what you know, including 'I don't know', becomes part of the foundation every future tool is built on.

“A small map nobody can fault beats a big one nobody can check.”

Two surveys. Pick one, or both.

Medicine availability

Which anesthesia and pain medicines on Ghana's Essential Medicines List are usually in stock. About 10 minutes.

Hospital resources

Staffing, monitoring, oxygen, equipment, blood, referral and records. About 15 to 20 minutes.

Answer for the hospital level you know best from your own work. You can fill in a survey again for a second level.

Built to protect you, too.

AOWA is a research project by Efeeo Impact. Questions: contact@efeeo.com