Kinshasa says Ebola daily cases have fallen from 120 to 80. WHO says it is too early to call a peak

Congo's health minister said this week that daily Ebola infections have dropped from about 120 at the mid-August peak to about 80, and that ten of 62 affected health zones have recorded no new case for more than 22 days. Samuel Roger Kamba called the figures a sign that the response is bending the curve. United Nations officials in Geneva on Tuesday refused to call a peak. Julien Harneis, the U.N. special Ebola coordinator, said the outbreak remains "a deadly and a massive epidemic." Olivier le Polain of the World Health Organization said, "It's too early to confidently say we have passed a peak."

Government figures published Tuesday listed 7,258 confirmed cases and 3,510 deaths in seven provinces, with 1,726 recoveries and 905 people in isolation or hospital. The case-fatality rate stood at 48.4 percent. The outbreak, declared on May 15 and caused by the Bundibugyo strain, has touched Ituri, North Kivu, South Kivu, Haut-Uele, Bas-Uele, Tshopo and Sud-Ubangi. Transmission remains steady in Ituri, North Kivu and Haut-Uele. Tshopo and Bas-Uele have reported sporadic cases. Dieudonne Mwamba Kazadi, head of the National Institute of Public Health, said after a coordination meeting in Kinshasa that the overall curve appears to have turned down after a mid-August high.

The two messages can both be true. A national daily count can fall while a single province accelerates. North Kivu had 1,269 confirmed cases as of September 12, including 555 in the previous 21 days. WHO said most new infections there still sit outside known transmission chains. That is the definition of an epidemic that has not been mapped. Ituri, where the outbreak likely began near Bunia weeks before the official declaration, has cooled. The map is therefore uneven, which is why Geneva will not use the word peak.

Kamba said six zones can now be taken off the list because they have gone more than 42 days without a case, and four more because they have gone more than 21 days. Those are the standard windows epidemiologists use to judge whether a chain has died. He also said daily deaths have fallen from about 50 to about 30. He asked the public not to treat the improvement as permission to drop precautions. Harneis said some areas still show growth and that full control will take months of support.

This is already the second-deadliest Ebola outbreak on record, behind the West African epidemic of 2014 to 2016. Bundibugyo is less famous than Zaire ebolavirus and still kills about half of those it infects in this wave. Vaccines and therapeutics used against Zaire are not a simple match. Response teams are working with isolation, contact tracing, safe burials and chlorine, the older tools, in provinces that also host armed groups and bad roads.

The argument over the word "peak" is an argument about resources. If Kinshasa declares a turning point, donors may slow. If Geneva keeps saying "massive," money and staff stay in the field. Le Polain's sentence was aimed at that audience. Kamba's sentence was aimed at Congolese households that have lived with the virus since May. Both men are reading the same ministry table. They are choosing different sentences from it.

Six zones silent for 42 days is a real gain. Five hundred and fifty-five cases in North Kivu in three weeks is a real failure of tracing. The national drop from 120 to 80 daily cases sits between those facts. It does not settle them. Health workers in Katwa and Bunia will keep washing in chlorine either way. The test of Tuesday's numbers will be the count in three weeks' time, not the press conference in Geneva.

For now the record is this: more than 7,200 confirmed infections, more than 3,500 dead, a government that says the worst day has passed, and a U.N. health agency that will not yet agree. The next situation report will show whether 80 was a step down the slope or a pause on a plateau.