Research note · companion to dar-labs · next: Zanzibar hospital capacity & referral spend
Dar es Salaam vs Zanzibar: where blood is drawn, and how far people are from it
The same method that mapped 2,295 likely blood-draw sites in Dar es Salaam was run over Unguja and Pemba on 30 September. It found 430. This page puts the two side by side and proposes one measurable KPI for the board deck: time to blood — how far the median resident is from the nearest place that will take a sample.
1Supply: what exists today
| Site type (Google Places) | Dar es Salaam | Zanzibar |
|---|---|---|
| Laboratories | 177 | 6 |
| Hospitals | 200 | 73 |
| Clinics, health centres, dispensaries | 1,912 | 351 |
| Pharmacies with a lab | 6 | 0 |
| Total | 2,295 | 430 |
| of which high-confidence (name or type says it draws blood) | 1,338 | 351 |
Both sweeps use the same term list (laboratory, blood test, clinic, dispensary, hospital, maabara…) and the same classification rules, so the counts are comparable with each other even though both undercount official registers. Zanzibar: map · all sites (.csv). Dar: map.
Zanzibar by island
| Island | Population 2022 | Sites found | Sites per 100k |
|---|---|---|---|
| Unguja | 1,347,332 | 338 | 25.09 |
| Pemba | 543,441 | 92 | 16.93 |
2Density: sites per head
| Measure | Dar es Salaam pop 5,383,728 | Zanzibar pop 1,889,773 |
|---|---|---|
| Laboratories per 100,000 people | 3.29 | 0.32 |
| Hospitals per 100,000 people | 3.71 | 3.86 |
| Clinics, health centres, dispensaries per 100,000 people | 35.51 | 18.57 |
| All sites per 100,000 people | 42.63 | 22.75 |
3Access: time to blood
For every 1 km population cell (WorldPop 2020) inside each territory we measured the straight-line distance to the nearest site, then weighted by the people living in that cell. The median is the distance a typical resident has to travel. Walking time assumes 5 km/h. This is measurable today, repeatable every quarter, and it moves the moment a new collection point opens.
| Measure | Dar es Salaam | Zanzibar |
|---|---|---|
| Any blood-draw site: population-weighted median distance to nearest (km) | 0.27 | 0.66 |
| Any blood-draw site: median walk at 5 km/h (minutes) | 3.2 | 8.0 |
| Any blood-draw site: 90th percentile distance (km) | 0.71 | 2.21 |
| Any blood-draw site: % of people within 2 km | 99.6 | 87.3 |
| Any blood-draw site: % of people within 5 km | 100.0 | 99.7 |
| Any blood-draw site: % of people more than 10 km away | 0.0 | 0.0 |
| Lab or hospital: population-weighted median distance to nearest (km) | 0.64 | 1.56 |
| Lab or hospital: median walk at 5 km/h (minutes) | 7.7 | 18.7 |
| Lab or hospital: 90th percentile distance (km) | 1.50 | 6.59 |
| Lab or hospital: % of people within 2 km | 96.2 | 58.0 |
| Lab or hospital: % of people within 5 km | 99.6 | 81.7 |
| Lab or hospital: % of people more than 10 km away | 0.0 | 2.6 |
| Laboratory only: population-weighted median distance to nearest (km) | 0.88 | 6.63 |
| Laboratory only: median walk at 5 km/h (minutes) | 10.6 | 79.6 |
| Laboratory only: 90th percentile distance (km) | 2.11 | 29.64 |
| Laboratory only: % of people within 2 km | 88.2 | 16.1 |
| Laboratory only: % of people within 5 km | 99.0 | 43.2 |
| Laboratory only: % of people more than 10 km away | 0.4 | 40.0 |
4The 30 km catchment
A single central laboratory with a courier network can realistically serve samples drawn within about 30 km. Hub assumed at Muhimbili (Dar) and Mnazi Mmoja / Stone Town (Zanzibar).
| Measure | Dar es Salaam | Zanzibar |
|---|---|---|
| People within 30 km of the hub (2022 census, scaled) | 5,325,406 | 1,222,654 |
| Blood-draw sites within 30 km of the hub | 2,264 | 257 |
| Labs + hospitals within 30 km of the hub | 372 | 44 |
| Sites per 100,000 people inside the 30 km catchment | 42.51 | 21.02 |
| Labs + hospitals per 100,000 inside the 30 km catchment | 6.99 | 3.60 |
| Share of the whole market reachable from one hub (%) | 98.9 | 64.7 |
5Proposed KPI for the deck
Time to blood — population-weighted median distance (and walk time) from home to the nearest sample-collection point, plus the share of the population within 2 km. Baseline today: Dar 0.27 km / 3.2 min; Zanzibar 0.66 km / 8.0 min. Counting only labs and hospitals: Dar 0.64 km, Zanzibar 1.56 km.
Why this one. It is the number a Ministry of Health cares about, it is computed from open data plus our own sweep (no lab has to disclose volumes), and every containerised hub or collection point Burjeel opens moves it in a way a board can see. Tests per head — the number everyone wants — cannot be measured without the NHIF claims data or the mystery-shopper round; this KPI is the one we can put a baseline under tonight.
Secondary: labs + hospitals per 100,000 inside the 30 km catchment, and the % of the market reachable from one hub (65% of Zanzibar vs 99% of Dar) — which is the quantitative version of her "whole nation" argument.
6Caveats
- Google Places undercounts both markets; the Tanzania Health Facility Registry and Zanzibar's PHLB register still need merging. Counts are comparable with each other, not with official totals.
- Low-confidence clinic rows are included by place type only; the "high-confidence" line is the safer number.
- WorldPop 2020 counts differ from the 2022 census; percentages are taken on the grid, absolute people are scaled to the census.
- Distances are straight-line. Road/ferry time on Pemba and inter-island is worse than these numbers show.