16JUN

Most Ebola deaths never reach a ward

3 min read

10:26UTC

Africa CDC placed 683 of 1,033 reported Bundibugyo deaths in the community, outside any treatment or isolation centre. Two fatal cases in three never touch the care system.

← Back to Bundibugyo's fork stays open Jump to analysis ↓

ScienceDeveloping

Key takeaway

Two-thirds of Bundibugyo deaths happen at home, where relatives handle the most infectious bodies.

Africa CDC, the African Union's continental public health agency, recorded 683 of 1,033 reported deaths in DR Congo's Bundibugyo Ebola outbreak, 66% of the total, as having occurred in the community rather than inside a treatment or isolation centre. The finding sits in situation report 71 on Bundibugyo virus disease (BVD), whose case tables run to 27 July with a flash update appended by 30 July ¹. Two fatal cases in three never reached the system built to treat them.

Bundibugyo ebolavirus has been spreading through DRC's eastern provinces since May, and the community-death share gives physical shape to two things already on the record. Jean Kaseya, Africa CDC's director-general, estimated on 16 July that true infections run two to four times the confirmed count , and WHO found that more than 80% of new cases were arising outside any known contact chain . Neither finding could name the patients it implied. Situation report 71 counts 683 of them.

Ebola makes a body most infectious at the moment it stops living, and a body that never reaches a centre is washed and dressed by relatives before burial. Each of those 683 deaths is therefore also a plausible transmission event, which is why the number matters more than its share of the mortality table. Contact tracers reached 78.9% of listed contacts, 13,882 of 17,588, against the roughly 95% coverage epidemiologists treat as the threshold for breaking chains ². A 16-point shortfall in tracing coverage is what produces a death register like this one.

Deep Analysis

In plain English

When someone dies from Ebola at home instead of in a hospital or treatment centre, health workers usually only find out after the fact, if at all. That matters because a person is most infectious to others right around the time of death, and in many communities the body is washed and prepared for burial by family members. Two out of every three Bundibugyo Ebola deaths recorded so far happened this way, outside any medical facility. Each one is a moment where the virus could spread to whoever handled the body, without any health worker there to stop it or even record it.

Deep Analysis

Root Causes

The 66% figure is a case-finding failure with a specific mechanical cause: contact tracing has to reach a household before a death occurs, and Ituri's active ADF and CODECO conflict zones restrict where response teams can safely travel, particularly around Djugu Territory. A death in a contested area may go unreported to any surveillance system for days.

Families in outbreak-affected communities who have already lost members to prior interventions have reason to distrust health teams who arrive only to remove and cremate bodies; reporting a death voluntarily, rather than waiting to be found, requires trust the response has not yet built at scale.

Escalation

Direction: worsening. Africa CDC's own ensemble models project 342-431 further cases in the next seven days and 940-1,368 within twenty-one, built on a trend this community-death figure helps explain: the response is not intercepting cases before death, so its case-finding lags the true epidemic rather than leading it.

What could happen next?

Meaning
A 66% community-death share indicates the epidemic's true size and reach exceed what confirmed case counts capture, consistent with Kaseya's two-to-four-times estimate.
Immediate · Assessed
Consequence
Every unwitnessed death in a household is a probable unrecorded transmission event, which will continue feeding new chains the response only detects retrospectively.
Short term · Assessed
Risk
If community deaths continue at this share, isolation-bed and testing capacity will keep chasing a caseload that is already several steps ahead of it.
Medium term · Suggested

Source Landscape

This story draws on neutral-leaning sources

Primary parallel: During the 2014-16 West Africa epidemic, unsafe burial practices, washing and touching a body at peak infectiousness, were identified by WHO and CDC as the single largest driver of new chains in Guinea, Sierra Leone and Liberia, responsible for an estimated one in five new infections. The mechanism is structurally identical here: a death outside clinical observation removes the response's only chance to intervene before the body itself becomes infectious.

Counter-parallel: The 2018-20 Kivu Ebola epidemic, fought against Zaire ebolavirus in the same eastern DRC provinces, drove burial-linked transmission down through anthropologist-led safe and dignified burial teams that negotiated directly with families and religious leaders rather than imposing protocol by decree. That programme took over a year to mature; the current Bundibugyo response is eleven weeks old.

Consensus view: Africa CDC's Jean Kaseya has argued since mid-July that WHO's confirmed count understates true Bundibugyo infections by a factor of two to four ; the sitrep 71 finding that 683 of 1,033 recorded deaths, 66%, happened in the community rather than a treatment centre is the first hard number that gives that estimate physical shape. Each of those deaths represents a chain the response never saw coming.

Counter-view: Medecins Sans Frontieres' field teams in Ituri caution against reading the 66% figure as pure case-finding failure. Some share of community deaths reflects patients WHO were identified and referred but died in transit or awaiting transport across a province with unpaved roads that become near-impassable in the wet season, a logistics gap rather than a surveillance blind spot.

Key tension: Whether the response's core problem is that it cannot find these patients at all, or that it finds them too late to move them, because the two diagnoses point to different fixes: more contact tracers versus more ambulances.

Sources:Africa CDC

Mentions:Africa CDC →Bundibugyo ebolavirus →Ituri Province →South Kivu →Jean Kaseya →Sierra Leone →Zaire ebolavirus →African Union →DR Congo →Médecins Sans Frontières →World Health Organization →

First Reported In

Update #12 · Two-thirds of Ebola deaths never reach a ward

Africa CDC· 31 Jul 2026

Read original →

Causes and effects

Caused by

Africa CDC: true cases up to 4x tally

The community-death share gives physical shape to Kaseya's estimate that true infections run two to four times the confirmed count.

Occurred 16 Jul 2026

Read story →

80% of Ebola cases now untraced

Community deaths outside care mirror WHO's finding that most new cases evade known contact chains.

Occurred 20 Jul 2026

Read story →

This Event

Most Ebola deaths never reach a ward

Case-finding is missing most of the infections that kill, and every death at home ends in a burial performed at the point of peak infectiousness.

Led to

Ebola case fatality crosses 40% in DRC

The same care-access gap driving community deaths lifts case-fatality as untreated cases skew fatal.

Occurred 27 Jul 2026

Read story →

Different Perspectives

Chinese health authorities

China reported three human H9N2 cases to WHO between 10 and 16 July: two four-year-olds recovered, a 44-year-old man remains in intensive care, the first severity signal in a cluster that has produced 173 cases and 2 deaths globally since 2015. The subtype is distinct from the H5N1 series this topic otherwise tracks.

US public health authorities

CIDRAP and Dr Andrew Racine, president of the American Academy of Pediatrics, call 2026's 2,318 confirmed measles cases by 24 July the worst year since 1991's 2,126. Zero deaths against three in 2025 and a 7% hospitalisation rate, down from 11%, show record case counts have not brought record severity.

CEPI

CEPI put an rVSV-platform vaccine candidate specific to Bundibugyo ebolavirus into trial-production on 31 July, developed by Hilleman Laboratories with MSD manufacturing and IAVI's reference virus, funded up to $8.5m. Eleven weeks into the third-largest Ebola outbreak on record, this is the first species-matched candidate; the licensed vaccine covers only Zaire ebolavirus.

Médecins Sans Frontières

MSF reports sexual-violence care in Ituri is almost non-existent, with 7 of 14 health centres closed or relocated by fighting, as Panzi Hospital logs 646 new survivors in six months, an 85% rise. Withdrawn US funding for post-exposure prophylaxis has left clinics across three provinces without PEP kits for months, a crisis running alongside Ebola, not caused by it.

DRC health workers at Elikya Ebola Treatment Centre

Staff at DRC's Elikya Ebola Treatment Centre walked out on 25 July over unpaid wages, the second such action after Rwampara, while unconfirmed suspects occupy up to 64% of the country's isolation beds awaiting lab results. Workers are absorbing a response that has grown faster than its payroll and its labs.

Uganda's Ministry of Health

Uganda's ministry says it discharged its last confirmed patient on 16 June, completed every contact's 21-day follow-up, and declared the country Ebola-free on 28 July, exactly 42 days later. It names Hon. Dr Chris Baryomunsi, Minister of Health, as overseeing a 20-case, 2-death outbreak it now considers closed.