The Democratic Republic of Congo's health ministry put the fastest-growing Ebola outbreak on record past 1,500 dead on July 30: 3,442 cases and 1,521 deaths, data as of July 28 — a mortality rise of roughly 50 percent in about a week. [1] The ministry's update counts 797 patients in isolation, 615 recovered, and contact tracing at about 78 percent against a 95 percent target. [1]
The New York Times' framing the same day made the outbreak the second-largest ever, overtaking the 2018-20 eastern Congo epidemic's 3,317 cases in just over two months — a pace that epidemic took a year to reach. [3]
Both the number and the superlative are real, and both arrive inside a caveat the paper carried in the July 29 account of the Congo Ebola ledger passing 3,262 confirmed as response access shrank: this is a cumulative ledger with backfill, not an incidence curve. WHO's July 22 reconciliation note stands — a chunk of every week's rise is earlier cases being found and added, not new onset — and WHO's own modeling runs two to four times the reported scale. [1] [4]
The arithmetic of a record
A 50 percent mortality rise in a week is the kind of figure that writes its own headline. The slower reading is more informative. Some of the dead being counted this week died weeks ago, in villages the response could not reach; the ledger is catching up to them now. Some of this week's dead are not yet in the ledger at all. Access loss makes the count lag in both directions at once — backfilled upward, uncounted downward — and the 1,521 is the residue of those two errors partially canceling. [1]
That is why the superlative and the caveat must travel in the same sentence. Second-largest-ever describes the ministry's cumulative file with precision. It describes the epidemic's true size with considerably less. If WHO's modeling band is right, the outbreak passed the 2018-20 epidemic's true scale some time ago; the record-keeping simply has not. [1] [3]
The response numbers inside the toll
The operational figures are the ones with verbs attached. Seven hundred ninety-seven isolated patients is a capacity statement — beds, staff, chlorine, and safe burials holding, for now, against the inflow. Six hundred fifteen recovered is the outbreak's one unambiguously good column. And tracing at 78 percent against a 95 percent target is the outbreak's binding constraint expressed as a single gap: better than three contacts in four followed, which in Ebola arithmetic is another way of saying one chain in four running free. [1]
The Boston Herald's account of the milestone corroborates the ministry record and little else — appropriately, because the ministry record is what exists. [2] No district-level ledger separating the isolated 797 from community deaths has been published. The ministry's file is the only count, and it is simultaneously the outbreak's map and its blind spot.
What the record frame costs a reader
A reader who takes only the superlative learns that this is the second-worst Ebola outbreak in history and learns nothing about why the number can be both accurate and wrong. A reader who takes only the modeling band learns the official count is fiction, which it is not — it is a census conducted through a keyhole, exact about what it sees. [1] The July 30 record asks both readers to hold two facts: the fastest growth ever recorded, and a measurement system whose own custodians say it is running behind the disease in both directions.
The frozen ledgers stay frozen: 3,200 cases and 1,405 deaths on July 28, 3,262 and 1,437 on July 29, and now 3,442 and 1,521 on July 30. [1] Each succeeded the last; none revised it. Uganda's declaration and Frankfurt's discharge remain separate jurisdictional files, non-additive — an outbreak declared over in one country and a patient discharged in another neither add to nor subtract from Congo's count.
The questions the next updates must answer are the method questions. How much of the week's mortality rise is backfill versus new onset, on WHO's reconciliation method — the one variable that would tell readers whether the epidemic is accelerating or merely being seen more completely? Does tracing close toward 95 percent as access shrinks, or does the gap widen with the geography? And when does the ministry publish a district-level ledger that separates the isolated 797 from deaths in communities the response cannot reach? [1] Those three answers would convert a superlative into a surveillance picture. Until they arrive, the milestone stands with its caveat welded to it.
The next number belongs to the next edition. This one belongs to the 1,521 — counted, at last, whether or not the counting is done.
-- KENJI NAKAMURA, Tokyo