Public Health · Data Analysis

Mapping healthcare access gaps in Nairobi County

A facility-level analysis of Nairobi's health infrastructure against Universal Health Coverage and SDG 3 targets - identifying where the county is underserved and what it would take to close the gap.

Role Independent Analysis
Tools Python, Jupyter, pandas, matplotlib
Data 942 facility records, 2019 census

Context

SDG Target 3.8 calls for universal access to quality healthcare. I wanted to know how close Nairobi County actually is to that — not at the county level, where averages hide a lot, but region by region, facility by facility.

I combined the county's health facility registry with 2019 census sub-county population data, mapped every facility to one of ten regions, and benchmarked each region against WHO and county target ratios for facility density, hospital beds, and four specific services: inpatient/home-based care, childhood illness management, HIV treatment, and sexual and reproductive health.

At a glance

900+
registered facilities across Nairobi
19
closed facilities — untapped capacity
70%
of facilities are clinics or dispensaries
Facility types (top 10) Medical Clinic 448 Dispensary 190 Health Centre 88 VCT Centre 54 Other Hospital 40 Nursing Home 25 Maternity Home 14 Laboratory 13 Medical Centre 12 Dental Clinic 10
Medical clinics and dispensaries dominate; smaller specialty facilities make up the long tail.
Operational status Operational — 97% Closed — 2% Pending — 1%
19 closed, 6 pending opening — 25 facilities of capacity sitting idle.

Where the gaps are

Facility count alone doesn't tell you much — a region can look "covered" and still be underserved once you account for population. Once I normalised by population, three regions stood out as consistently behind:

  • Mathare, Kasarani, and Embakasi all sit under 2 facilities per 10,000 people, the lowest density in the county.
  • Dagoretti, Makadara, Embakasi, and Kasarani fall short of the recommended 18 hospital beds per 10,000 people.
  • Mathare has no cots in the region — a specific, fixable gap in neonatal and paediatric care.

Facilities and beds needed to reach target ratios

To reach 2 facilities / 10k population and 18 beds / 10k population
Region Population New facilities needed New beds/cots needed
Mathare 207,000 26 53 (cots)
Kasarani 1,408,000 98 2,074
Embakasi 989,000 53 909
Dagoretti 435,000 — 44
Makadara 190,000 — 128

Service-specific gaps (regions below county average)

Region Service Current Target Gap
Kasarani Family planning 51 88 37
Kasarani Inpatient/home-based care 46 / 79 — 33 (HBC)
Kasarani Childhood illness (C-IMCI) 7 24 17
Embakasi HIV treatment (ART) 12 25 13
Embakasi Childhood illness (C-IMCI) 9 17 8
Mathare Family planning 5 13 8

Recommendations

  • Quick win Reopen the 19 closed facilities before building anything new — especially the one already equipped with 50 beds and 20 cots that operates 24 hours. This is capacity the county already paid for.
  • Primary care Expedite the 6 facilities pending opening to strengthen outpatient access immediately.
  • Kasarani Priority region for new investment — it's short on facilities, beds, family planning, home-based care, and childhood illness services simultaneously.
  • Mathare Smallest population of the underserved regions but the only one with zero cots — a targeted, low-cost fix rather than a large capital project.

What I'd do next

If I extended this, I'd want to weight recommendations by estimated cost per facility/bed rather than treating every gap as equally urgent — reopening an existing facility and building a new one aren't the same lift. I'd also like to bring in travel-time data instead of straight population ratios, since a region can hit its facility-per-capita target and still leave people an hour from the nearest one.

Tools & sources

Python pandas matplotlib Jupyter PowerPoint