The Urban Capacity Paradox
A clinic can be close to home and still be out of reach.
Ask how well a place is served by health care and the usual answer is a distance. Is there a clinic within 5 kilometres? That is the World Health Organization’s yardstick, and by it Nigeria looks well served: a national audit found that 98.5% of built-up areas lie within the limit (Aladejana and Ebijuoworih, 2024). The figure is true. The problem is that it is also a poor guide to what a patient finds on arrival.
I came across the gap while mapping health care in Lokoja, the capital of Kogi State, where the Niger and Benue rivers meet. The local government area has 10 wards, 112 active clinics and, by satellite-based estimates, 308,138 residents. The data came from GRID3 (clinic locations), WorldPop (population) and OpenStreetMap (roads).
One Clinic, 6,576 People
Lokoja C is a peri-urban ward of 0.7 square kilometres, with a single clinic. Nobody there lives far from it, so any distance test marks the ward as fully served. Yet 6,576 people rely on that one clinic, the heaviest load in the area.
Is that a lot? The table gives some sense of scale:
| Ward | Area (km²) | Residents | Clinics | Residents per Clinic |
|---|---|---|---|---|
| Lokoja C | 0.7 | 6,576 | 1 | 6,576 |
| Kakanda | 714.9 | 32,165 | 7 | 4,595 |
| Oworo | 1,408.2 | 118,789 | 29 | 4,096 |
| Lokoja A | 32.1 | 75,768 | 40 | 1,894 |
| Eggan | 29.4 | 3,584 | 4 | 896 |
| Whole area (LGA) | 3,187.0 | 308,138 | 112 | 2,751 |
The average for the whole area is 2,751 residents per clinic, so Lokoja C carries about 2.4 times its share. Its load is higher than that of Kakanda, a remote ward where people travel far for care, and more than seven times that of Eggan.
Why It Is a Paradox
Lokoja C does well on one measure and badly on another. I scored every ward from 0 (least deprived) to 1 (most deprived) on three things: distance to the nearest clinic, residents per clinic, and how few clinics there are per square kilometre.
Lokoja C scores 0.04 on distance, close to the best possible result. On residents per clinic it scores 1.00, the worst in the area. Add the scarcity score of 0.84 and its overall index is 0.63. That makes it the second most deprived ward, almost level with Oworo (0.62), a rural ward of 1,408 square kilometres.
Kakanda shows the opposite pattern. It scores 1.00 on distance and 0.65 on crowding. Two wards therefore rank high for opposite reasons. I call Kakanda’s case the Rural Physical Archetype, since distance and bad roads cut people off. Lokoja C is the Urban Congestion Archetype, since people can reach the clinic but it has too many to see.
The difference is an old one. Penchansky and Thomas (1981) separate accessibility, meaning whether a patient can get to a service, from accommodation, meaning whether the service can take the patient. A distance test measures only the first. If we must have a guide to the second, we have to consider residents per clinic.
Should the Finding Be Believed?
Four doubts come to mind:
- Could residents simply use clinics in other wards? Some probably do. The study counts clinics inside each ward boundary and cannot see where people actually go. This is a real limit, and only a survey of patients could settle it.
- Is the count of one clinic right? This is the weakest point. Lists built from public records can miss small or unregistered clinics. If one clinic had been missed, the load would fall to about 3,300 residents per clinic. That is still above the area average, but no longer the highest. A field check of the facility register in Lokoja C should come first.
- Is the ward truly crowded, or is it only a ratio? Only a ratio. I did not observe waiting times or staff numbers. The ratio stands in for crowding, and it is a fair stand-in, but it is not a measurement of queues.
- Does the ranking depend on how the index is built? In part. When distance counts for half the score, Lokoja C drops from second to fourth place. When crowding counts for half, it returns to second. Kakanda stays first and the two best-served wards stay at the bottom under every weighting, but the middle of the table moves. The paradox is solid as a statement about load. Its exact rank depends on how much weight a planner gives to crowding, and that is a policy choice, not a technical one.
What Follows for Policy
Because the two kinds of deprivation have different causes, they need different remedies.
Kakanda needs better roads, boats on the river, and health posts closer to people. Lokoja C needs more from the clinic it already has. My recommendations were to upgrade it to a 24-hour comprehensive health centre and to add nurses and community health workers. Success can be tracked with one number: residents per clinic in Lokoja C.
When floods come into the picture, the stakes must be raised. Lokoja sits at a river confluence, and in October 2022 flooding coincided with more than 6,000 suspected cholera cases in the wider region (Komolafe, 2023). A clinic that is full in a normal week has no room to spare when floods push demand up. I should be plain that this is an inference. The cholera figures were not reported by ward, so they cannot show whether cases clustered in Lokoja C.
A Test That Travels
The check behind this finding is cheap. Divide a ward’s population by its number of clinics, then compare the result with its neighbours. Both inputs are open data for most of Nigeria. Rapid growth has outpaced primary care in many cities, Lagos among them (Obubu et al., 2023), so Lokoja C is unlikely to be alone.
Distance tells you whether people can reach a clinic. It says nothing about whether the clinic can receive them.