The Democratic Republic of the Congo has reported 5,290 confirmed Bundibugyo Ebola cases and 2,516 deaths, with 837 patients hospitalised in isolation. The latest daily report added 81 cases and 40 deaths. One figure inside those numbers has moved in the wrong direction, and it is the one that predicts the rest.

Contact tracing has slipped

82.9 per cent of identified case contacts are now under follow-up in the affected provinces. Earlier this week the figure was around 85 per cent.

That is a small movement and a meaningful one. Contact tracing is the mechanism by which an outbreak without a vaccine is contained: find everyone a confirmed patient was in contact with, watch them for symptoms, isolate them the moment they develop any. Every contact not under follow-up is a possible chain of transmission nobody is watching.

The percentage falls for practical reasons rather than dramatic ones. More cases mean more contacts per tracer. People move between districts, or decline to be listed. Insecurity closes off areas. When the caseload rises faster than the workforce, coverage drops — and the drop shows up in new cases two to three weeks later, because that is roughly the incubation window.

Where the outbreak sits

Ituri province accounts for the overwhelming majority: 4,447 cases and 1,984 deaths, reported from 28 of its 36 health zones. North Kivu has 663 cases and 452 deaths across 12 of 34 health zones. The remaining cases are in Haut Uele and Tshopo.

The health-zone counts are worth reading carefully. Twenty-eight of thirty-six zones affected in Ituri means the outbreak is no longer a cluster that can be ring-fenced; it is distributed across most of a province of more than four million people. North Kivu’s ratio of deaths to cases is strikingly higher than Ituri’s, which usually indicates that cases there are being detected late — people reaching care when it is already too late, and milder cases never being counted at all.

Some 1,152 people have recovered.

The trial that has started

Patient enrolment has begun in a scientific trial to identify the first effective treatments for Bundibugyo virus disease. There are currently none — the licensed Ebola vaccines and therapeutics target the Zaire species, which is why this outbreak has been fought with methods that predate them.

Running a trial during an outbreak is difficult and contested. Randomising patients when nothing else is available raises real ethical objections; so does giving untested compounds to very sick people without knowing whether they help. The compromise designs used in previous epidemics were argued over for years afterwards. What is not in dispute is that outbreaks are the only setting in which such treatments can be tested at all, and that the last Ebola epidemic ended with tools that came out of exactly this process.

Results will not arrive in time to change the current curve.

The regional layer

Five central African countries — the Central African Republic, DRC, Republic of the Congo, South Sudan and Uganda — are strengthening cross-border surveillance and information sharing. The Central African Republic has deployed a mobile laboratory near its borders with DRC and South Sudan.

A mobile laboratory near a border is a specific and sensible investment. Testing capacity is usually concentrated in capitals, which means samples travel for days and results arrive after the patient has either recovered, died or infected others. Moving the laboratory to the border compresses that delay, and delay is the thing that decides whether surveillance works.

Whether the contact-tracing percentage recovers over the coming fortnight is the number to watch. Everything else in these reports follows from it.

Sources

• ECDC — Ebola outbreak in the DRC and Uganda, daily update, 21 August 2026
• WHO — Ebola outbreak, DRC 2026, and the Bundibugyo treatment trial