Data Science in Kenyan Healthcare

McTaba runs coding and data bootcamps in Kenya. This guide is written to be useful whether or not you ever join one.

Picture a clinic register in a sub-county: pencil tallies of malaria tests, immunisations, antenatal visits. Each month someone copies the totals into a summary form, and from there they climb to the county and then the nation. Almost everything in Kenyan health analysis starts as that kind of count. This page shows where data science plugs in, what is allowed, and what to learn.

For the wider career map, read the data science career guide for Kenya. Every use below has a source you can open, and we note where a claim is the organisation's own. If you want a structured route into the field, see Data Science to AI Engineering.

1. KHIS and DHIS2: national reporting

The Kenya Health Information System is built on DHIS2, where facility summaries flow up through sub-county and county levels. A 2025 study in PLOS Global Public Health interviewed 89 stakeholders across 15 counties. They liked the dashboards and scorecards, but found the data inaccurate in places, citing different tools across counties, staff shortages and incomplete reporting. They also found limited analytical skills and heavy reliance on a small pool of records officers (PubMed entry for the study, checked on 3 November 2026). That gap is the opening for a data analyst: cleaning, checking and explaining the numbers. A Tenwek Hospital data analyst ad, for example, listed DHIS2 reporting, Power BI and Excel (Tenwek ad on MyJobMag, checked on 3 November 2026).

2. Research units: prediction and open datasets

A KEMRI-Wellcome Trust team built and internally validated a data-driven model to help frontline workers triage children under five at Mbagathi County Referral Hospital in Nairobi. It predicted admission of 24 hours or longer from eight predictors, with an AUC of 0.88 (Wellcome Open Research paper, checked on 3 November 2026). That is a research result, not a tool in every clinic. In Kilifi, researchers published admission records for 29,226 children from 2002 to 2009, listed on LSHTM Data Compass (Data Compass record, checked on 3 November 2026). Check its terms before you reuse it.

3. Hospital data teams

Aga Khan University Hospital advertised a Manager, Data Engineer for its Data Innovation Office in August 2026, to build secure research data platforms (Aga Khan ad, checked on 3 November 2026).

4. Health startups

Kasha advertised a data scientist role in September 2026 to build risk-scoring models for patients with non-communicable diseases at risk of dropping out of care, plus delivery forecasting for medicines. The ad allows the role to be based in Kenya (Kasha ad, checked on 3 November 2026). It is a hiring notice, so it shows intent, not results.

Health data is among the most protected data in Kenya. The Data Protection Act 2019 treats health status as sensitive personal data, with stricter grounds for processing it and a separate section on personal data relating to health (Clyde & Co summary, checked on 3 November 2026). That summary says health data may generally be processed only by or under the responsibility of a healthcare provider, or someone bound by a duty of confidentiality, with a public health route in some cases. Read the Act's own wording before you act on a summary.

• Guidance. The Office of the Data Protection Commissioner adopted a Guidance Note on the Processing of Health Data in December 2023 (Digital Policy Alert record, checked on 3 November 2026).
• Registration. Clyde & Co report that health organisations cannot use the small-entity exemption from registering with the ODPC, and that the Kenya Medical Practitioners and Dentists Council now asks new facility registrations to include an ODPC certificate (Clyde & Co, checked on 3 November 2026).
• Penalties. The same summary lists administrative fines and criminal penalties for breaches.

What this means for you as a learner: never put patient-level data in a public repository, and never test on a hospital's live records. Use aggregated public data, published datasets with a clear licence, or synthetic records. Our guide to anonymising personal data shows how, and the data quality checklist covers the cleaning that the DHIS2 study found missing. This is general information, not legal advice.

• SQL and spreadsheets to pull and check facility counts. See our SQL guide.
• Statistics for rates, confidence and bias. Health data is full of small counts.
• A dashboard tool. The Tenwek ad names Power BI and Excel, with STATA or SPSS.
• Data quality and governance, since the DHIS2 study's main complaint was accuracy.
• Machine learning, mainly for research and startup roles, with fairness and validation.

Our Data Science to AI Engineering programme starts on 11 January 2027 and costs KES 140,000, or KES 120,000 if paid in full upfront. It teaches Python, SQL, Power BI, statistics and machine learning with fairness checks. Its capstones are financial, not clinical, and no school can promise a job. See Data Science to AI Engineering. Farm data has a similar shape: read data science in Kenyan agriculture, and where data jobs are posted for who is hiring.