Ebola Has Reached the Hubs. Can the Chains Be Stopped?

Nairobi and Abuja meet the geographical condition set out in our May analysis. Kenya's chain has so far been traced inside quarantine. Nigeria's case has no confirmed source, and every account of her exposure now on record carries a different implication. In Congo, Ituri's laboratory signal has improved; Nord-Kivu's apparent deterioration rests on a single day's results, while the share of cases first found at death has risen.

EuroOilWatch · 29 October 2026 · OilWatch Network Analysis

A preliminary reading ahead of the 30 October checkpoint, not the checkpoint itself. DRC laboratory and notification data run through 27 October (INSP report 166, published 28 October; report 167 not located at research cut-off). Kenya through 26 October. Nigeria: NCDC report of 27 October (summary table 26 October), supplemented by reporting to 26 October on the evacuation and contact count.

On 29 May, our first analysis of the Bundibugyo Ebola outbreak named Nairobi and any West African capital among the places where a confirmed case would move its assessment upward. Nairobi reported an imported case on 6 October. Abuja confirmed one on 20 October. The geographical condition has now been met in two places.[1][3][4]

That requires a change in the geographical assessment. It does not establish that either city has become a source of sustained transmission. The distinction is the subject of this third article: what happened after the arrivals, and whether the systems receiving them have kept the chains visible.

There is a timing limit on the earlier paper's record. Both events occurred after its original twelve-week test window ended in August. They belong in the continuing assessment, rather than being retrospectively scored as successes inside that window.[1]

Kenya: transmission found inside a known chain

The first Kenyan patient travelled from DRC through Uganda and arrived in Nairobi on 3 October. Eight subsequent cases were linked to that patient: four relatives and four healthcare workers. WHO reported that all eight were already in quarantine when symptoms developed.[3]

ECDC's 28 October update records nine confirmed cases and two deaths in Kenya, all linked to the first case, with the latest confirmation on 20 October and the second death on 23 October. Twenty contacts were released on 26 October after completing 21 days of monitoring; 98 remained in quarantine.[4][18]

These findings show the response tracking the known chain. They are not yet evidence of containment. Four healthcare workers were infected, which places at least part of the transmission in clinical care, most plausibly before the index case was recognised. The 98 contacts still in quarantine have not completed follow-up. Under the standard 42-day rule, the countdown to an end-of-outbreak declaration begins only after the last case has tested negative twice or been buried; with the last confirmation on 20 October, no declaration is possible before 1 December.

For the framework, this is evidence in both directions. Arrival in a named hub meets the geographical warning condition. Finding subsequent cases among quarantined contacts shows the responder system doing the work that could prevent further spread.

Nigeria: thirteen days to confirmation, and no confirmed source

NCDC's 24 October situation report records arrival in Abuja on 3 October, symptom onset on 7 October, admission to a private health facility on 11 October, notification to NCDC on 19 October and laboratory confirmation on 20 October.[12] Contemporaneous reporting of NCDC's account adds that she first presented at a different private facility on 10 October, the day before admission.[13]

That is thirteen elapsed days between reported symptom onset and confirmation, ten after first presentation and nine after admission, across two private facilities. This chronology does not establish when appropriate isolation began or how many people were exposed during each interval. It identifies the period that an investigation of recognition and referral needs to explain.

Contacts rose from 44 on the day of confirmation to 53, then 62.[5][13][14] ECDC reports that the patient and a close contact were medically evacuated to Switzerland on 26 October; NCDC identifies the contact as her spouse.[5][14][17] NCDC's summary table records no positive contact, and expected follow-up completion dates run from 31 October to 13 November.[5] A suspected case reported by Rivers State on 23 October tested negative.[13]

The source of her infection is unresolved, and the accounts on record diverge. She arrived in Abuja from Uganda, via Addis Ababa.[13] Swiss authorities have said she did not visit DRC and believe she was infected in Uganda.[14] Uganda, which was declared Ebola-free on 27 August and has not detected a local case since June, says she had been in eastern DRC before transiting through Uganda; WHO and Africa CDC have described that account as unconfirmed.[15][16]

The distinction matters more than any other open question in this article. Of the accounts now on record, only the Ugandan one places her inside a known transmission area. If it is not confirmed, the remaining explanations require exposure to a chain that no surveillance system had identified, whether in Uganda, in transit or in Nigeria. NCDC has reported completing genomic sequencing.[13] Placement of that sequence against DRC lineages is the most direct available test, and it has not been published.

Uganda has already acted on the inference. On 25 October it barred anyone who has been in DRC from transiting the country, including on chartered flights.[14] Both of this month's hub importations reached their destinations through Uganda. That makes Uganda a corridor as well as a former outbreak country.

The airport is only one part of the system

The reported Nigerian chronology places symptom onset four days after arrival, and Uganda reports that she was asymptomatic on departure.[15] Neither departure nor arrival temperature screening could have been expected to detect her.

ECDC's 23 October threat assessment explains the limitation: exit screening cannot prevent every exported infection because disease can develop after departure. It also says entry screening has not been proved an efficient way to detect imported Ebola cases. Its recommendations emphasise clinical awareness, testing, referral arrangements and preparedness, alongside travel advice.[6]

Our reading is that the critical interval extends beyond the border: from the first symptoms, through the first request for medical help, to recognition, isolation and the tracing of contacts. A transport hub matters because cases can arrive there. Its response capacity determines what follows. The Abuja case also tests that capacity in an atypical setting: a diplomat's household using private care, then a managed evacuation. It is not a straightforward guide to how a less connected patient would move through the same system.

Congo: two provinces, different readings

The latest INSP bulletin reviewed, report 166, records 10,193 confirmed cases and 4,960 deaths in DRC through 27 October, across eight provinces and 66 affected health zones. It was published on 28 October.[7] ECDC lists the provinces as Ituri, Nord-Kivu, Sud-Kivu, Tshopo, Haut-Uélé, Bas-Uélé, Sud-Ubangi and Nord-Ubangi.[17]

For a preliminary look ahead to our 30 October checkpoint, we extracted selected fields from fourteen consecutive daily reports, covering 14–27 October. Pooling positive laboratory results and samples within each seven-day period gives the following comparison.[8]

In Ituri, fewer positive results accompanied more testing, and the result survives the removal of any single day: the highest second-week value is below the lowest first-week value. That is evidence consistent with improvement over these two weeks. Sampling still falls short of the 290-a-day condition our October paper required before applying its formal positivity thresholds; no single day in the fourteen reached 290. We cannot remove that condition now that the results look more encouraging.[2][8] Nor can positivity alone separate a falling infection rate from a change in what is being sampled, such as the mix of alert, contact and post-mortem swabs. Local reporting from Mandima health zone in Ituri describes continued growth there and gives figures that do not match INSP's.[21]

Nord-Kivu's reading is weaker than it first appears. Pooled positivity rose from 12.9% to 14.8%, but 58 of the second week's 225 positives came on 25 October. Without that day, second-week positivity is 12.5%, slightly below the first week. The apparent deterioration therefore depends on one day's results. It is a signal to watch, not a finding, and it does not yet overturn the September improvement described in our second article.[8]

Two national measures move in opposite directions. The proportion of listed contacts seen each day rose from 78.7% across 14–20 October to 86.1% across 21–27 October. Over the same weeks, daily community-death notifications rose from 30.3% to 36.1% of daily new-case notifications, and from 61% to 66% of all reported deaths.[8] Contact follow-up has improved, but a growing share of the outbreak is still being found only at death. That second measure remains inside the band our October paper set, and it is a notification proxy rather than a case-level post-mortem share, but it is moving toward the upper threshold.

This comparison uses reported laboratory and notification counts. The source reports contain inconsistencies between some daily figures and cumulative changes; those are recorded below. They limit the precision of national comparisons and are another reason to keep this assessment provisional.

What the response reports describe

The operational record helps explain why surveillance and access remain central to this story. On 25 October, the INSP bulletin reported that two suspected cases identified at an Ituri checkpoint could not be transferred because no ambulance was available, and that another Ituri checkpoint, at Avakubi, remained out of action because of insecurity.[9] In the same week, WHO's Director-General reported that a nurse had been killed and another health worker severely injured in Tshopo province while responding to an Ebola alert; local reporting places the killing at Bafwasende on 24 October.[20][26]

On 27 October, the national contact-follow-up rate was 85.4%, below the response's stated 95% requirement. The same report recorded Nord-Kivu treatment-centre occupancy at 88.5%, with another 160 patients receiving care outside the centres.[7]

There is also a small but concrete addition to the logistics evidence. The 22 October report recorded exhausted diesel stocks in Bas-Uélé. The following day's report recorded a delivery of 200 litres to the province's incident-management team. On 25 October, a further 140-litre requirement was listed as unmet.[10]

These entries document local fuel constraints and replenishment. They do not measure the resulting effect on testing, patient transport or transmission, and they do not establish that the wider oil crisis caused the shortages. The evidential advance since our second paper is limited to that documented local constraint.

The response also has a new instrument. On 23 October the TUMAINI trial began ring vaccination of high-risk contacts in Bunia, Ituri, starting with Ervebo, a vaccine licensed against the Zaire species whose protection against Bundibugyo is unknown. Two Bundibugyo-specific candidates, from Oxford and Moderna, are under review by WHO's advisers, and CEPI says they will join shortly; WHO expects results in several months, not weeks.[18][19] Vaccination that began on 23 October cannot have affected the laboratory data in this article.

What can be said before 30 October

The geographical component of the May warning has moved upward. Kenya supplies evidence of a response that found its secondary cases inside quarantine. Nigeria's observation period remains open, and its source of infection is the largest unresolved question. Within Congo, Ituri's latest laboratory comparison is more encouraging and robust to single-day effects; Nord-Kivu's deterioration is not; and the growing share of cases first identified at death argues against reading either province as under control.

That combination calls for a more specific assessment of where transmission remains visible and where access is failing. It does not supply a numerical mortality revision or validate the framework as a whole.

For European readers, ECDC continues to assess the risk to the general EU/EEA population as very low. Its assessment rests in part on the expectation that imported cases would be identified, isolated and managed promptly.[6][17] The Nigerian patient was evacuated to Switzerland, which lies outside the EU/EEA; this outbreak has previously produced four evacuated or imported cases inside it, all among healthcare or humanitarian workers.[17] The Nigerian chronology makes recognition and referral a concrete subject for investigation; the Kenyan chain shows what subsequent contact management can achieve.

The next assessment needs the remaining daily data for the checkpoint, the outcome of contact monitoring in both countries, and the result of Nigeria's source investigation, including the sequence placement. Those observations will determine whether this expansion is being contained at the hubs or has already produced chains beyond them.

Data, method and verification record

The article above is a preliminary synthesis, not the completed 30 October adjudication. This record sets out what was extracted, how it was computed, and where the sources disagree.

Access and provenance

Research cut-off: 29 October 2026, approximately 12:30 CET. INSP report 167 (reporting 28 October) was not located. Absence of a located bulletin does not establish that none exists.

INSP's own indexed report 166 corroborated the reporting date, cumulative totals and headline daily figures; its original PDF could not be retrieved from INSP. The fourteen complete reports were therefore read as INSP-authored PDFs hosted by ebola-tracker.org. Report numbers, printed reporting dates and relevant sections were inspected, and mirror summaries were not substituted for the PDF laboratory sections. For report 164, the copy served by insp.cd is byte-identical to the mirror copy (same SHA-256); the other reports were not compared against INSP originals. SHA-256 hashes of the copies read appear at the end of this record so a later review can identify them. Hashes identify copies; they do not authenticate the publisher.

Laboratory extraction

Each row represents one report's daily laboratory counts. Weekly positivity = sum of positive results / sum of new samples tested. Daily sample mean = sum of samples / 7. Printed daily percentages are not averaged.

Weekly arithmetic:

• Ituri, 14–20 October: 265 / 1,541 = 17.1966%; 1,541 / 7 = 220.1429 samples/day.
• Ituri, 21–27 October: 238 / 1,666 = 14.2857%; 1,666 / 7 = 238.0 samples/day.
• Nord-Kivu, 14–20 October: 183 / 1,423 = 12.8602%; 1,423 / 7 = 203.2857 samples/day.
• Nord-Kivu, 21–27 October: 225 / 1,522 = 14.7832%; 1,522 / 7 = 217.4286 samples/day.

Daily maximum for Ituri samples across all fourteen days: 278 (19 October). No day reached 290.

Leave-one-out pooled positivity (drop each day in turn, recompute the week):

Ituri: the ranges do not overlap. Nord-Kivu: the ranges overlap, and removing 25 October reverses the direction (12.53% against 12.86%).

National notifications and listed contacts

Community deaths are the nationally reported daily community-death notifications, used here as a proxy for post-mortem confirmations. They are not the percentage of all deaths occurring outside treatment centres. Contacts listed and seen are daily stocks, not cumulative unique people or unique daily missed contacts.

Weekly summaries:

The October paper defined this measure as the share of confirmed cases found after death, so new cases are the relevant denominator, and on that basis the figures sit inside its 27–39% band. The proxy pairs a day's community deaths with that day's case notifications; community deaths are not necessarily among the same day's cases, so the ratio is not a case-level post-mortem share. The share rose 5.8 points week on week and sits 2.9 points below the upward threshold (2.2 points using the 489 cumulative denominator).

The improvement in contact follow-up is not an artefact of 25 October's missing Ituri data: excluding that day, the second week is 86.5%. It remains below the 95% standard on every day.

The median daily listed-contact stock for 21–27 October is 34,935. The cumulative case increase over the 21 days after 6 October is 10,193 − 8,728 = 1,465. Dividing gives a provisional ratio of 23.8 contacts per recent confirmed case (24.3 using the median excluding 25 October). The ratio divides a stock by a flow and changes with reporting coverage. It reconstructs the broad earlier method; it is not an audited reproduction of its original calculation.

Daily national case notifications exceed the Ituri plus Nord-Kivu laboratory positives by 1–9 each day, consistent with cases from the other six provinces. These were not reconciled province by province.

Provisional checkpoint ledger

The May geographical condition is met in Nairobi and Abuja, after the original test window. Two qualifying locations are two observations against one geographical trigger, not two independent triggers. This pass does not adjudicate all the May downward triggers or rerun the mortality model.

Source inconsistencies and editorial decisions

• Report 160, 21 October: daily notifications state 71 cases and 41 deaths. Published cumulative totals rise by 61 cases and 40 deaths from report 159. Nord-Kivu's daily laboratory positives are 27, while its cumulative count rises by 17. Daily laboratory counts are retained for the laboratory comparison; cumulative totals are retained for cumulative statements. No silent repair was made. Using 17 rather than 27 for that Nord-Kivu day would give 215 weekly positives and 14.1% positivity, against 183 and 12.9% in the preceding week; the leave-one-out test above is the governing robustness check.
• The seven daily case notifications for 21–27 October sum to 499, while the cumulative increase is 489. Daily death notifications sum to 274, while the cumulative increase is 273. The post-mortem proxy uses 499 consistently; using 489 would give 36.8%, still inside the band.
• Report 153, 14 October: the national headline records 22 community deaths; the laboratory breakdown lists 13 in Ituri, 10 in Nord-Kivu and one in Haut-Uélé, totalling 24. The table retains the national 22. Substituting 24 changes the first week's share from 30.3% to 30.7%.
• Report 154 prints Nord-Kivu positivity as 17.8% for 22 / 128. That fraction is 17.2%. Weekly positivity was recalculated from counts.
• Report 162 has an apparent 22 October carry-over in one cumulative-total sentence; its header, reporting date and narrative identify 23 October. Assigned to 23 October.
• Report 165 prints national positivity of 21.4% for 81 / 383. That fraction is 21.1%. Not used.
• The 25 October national listed-contact total omits Ituri. The decrease is not evidence that contacts completed monitoring or disappeared.
• Kenya's early Ministry of Health account and the later WHO chronology differ by one day on the first patient's death. ECDC is used for aggregate deaths.
• Kenya contact denominators differ across sources: 20 released plus 98 in quarantine (118), against 134 of 135 contacts traced in a secondary summary around 19–20 October.[25] Not reconciled; the article uses only the 20/98 split.
• Nigeria's exposure: three accounts are on record (Swiss: infected in Uganda, no DRC visit; Uganda: prior presence in eastern DRC; WHO and Africa CDC: unconfirmed). The article reports all three and assigns none. No airport or flight is treated as an infection source; transit through Addis Ababa is reported as route only.
• Nigeria chronology: the 10 October first presentation comes from ICIR's report of NCDC's statement;[13] NCDC's situation reports and WHO begin the facility chronology at the 11 October admission.
• Nigeria contacts: NCDC's 27 October report gives 62 contacts and describes the rise from 53 as an additional eight; Euronews reports nine more. ECDC's 54 (26 October) is a different count. The article uses NCDC's 62.
• Nigeria evacuation: an early 26 October report placed the husband in isolation; ECDC later reports the case and a close contact evacuated. Retained as evacuated, per ECDC and NCDC.
• Mandima: a local medical officer reports 328 cumulative deaths, and the same report says nearly 400 confirmed cases "would have been" recorded in 1–27 October, a conditional the outlet itself says needs clarification. INSP report 166 gives Mandima 456 confirmed cases and 242 confirmed deaths, both cumulative. Not reconciled; reported only as evidence that sub-provincial reporting diverges.
• Tshopo killing: the WHO announcement and province are well sourced. The location (Bafwasende) and date (Saturday 24 October) come from local reporting by one outlet, published 25 October;[26] Reuters reporting says Friday. Local reporting describes the attackers as young people who accused the victims of spreading the virus through a water source; WHO did not identify perpetrators.
• The stockout, delivery and later fuel requirement concern Bas-Uélé operational reporting. There is no measured effect on infections or deaths and no causal attribution to Hormuz, oil prices or the international diesel market.

INSP PDF fingerprints

SHA-256 hashes of the mirrored PDF copies read.

Sources

1. Jon Kelly, From Hormuz to Bundibugyo: A Second Case for the Compound Cascade Framework, 29 May 2026, with updates of 18 July and 22 August. Used to establish the prior commitment, not as independent proof of current epidemiology.
2. Jon Kelly, The Reported Ebola Curve Is Bending. In Ituri, the Testing Fell First., 2 October 2026. Source of the scheduled checkpoint and precommitted thresholds.
3. WHO, Disease Outbreak News DON622, published 23 October 2026; primarily data through 20 October. Used for the Kenyan chain and the unresolved Nigerian exposure investigation. Its earlier death count for Kenya is superseded by source 4.
4. ECDC, Multi-country event update, 28 October 2026. Kenya data through 26 October. Its DRC totals (10,141 cases, 4,928 deaths) cover 26 October and are superseded here by INSP report 166 for 27 October. Its Nigerian contact count of 54 is superseded here by NCDC's 62.
5. NCDC, Ebola Disease Situation Report, 27 October 2026, pp. 1–2. The summary table is labelled 26 October. Patient timeline and contact count on p. 1; expected monitoring completion dates on p. 2. The exposure source and precise isolation date are not established by this report.
6. ECDC, Threat assessment brief: importation to Kenya and Nigeria, 23 October 2026. Screening limits, preparedness and EU/EEA risk assessment.
7. INSP / DRC Ministry of Health, SitRep 166, reporting 27 October, published 28 October. National totals on p. 1; contact follow-up on p. 5; laboratory data on p. 6; treatment capacity on p. 7. INSP-authored document accessed through a third-party mirror.
8. INSP reports 153–166, dated 14–27 October; individual PDF links in the laboratory extraction table above. Selected-field extraction, not a full audit of every table.
9. INSP, SitRep 164, reporting 25 October, p. 6 for transport limitations. Ituri did not contribute contact-follow-up data to that day's national contact total.
10. INSP, SitRep 161, 22 October, p. 9; SitRep 162, 23 October, p. 10; SitRep 164, 25 October, p. 8. Successive local operational entries, with no quantified attribution to global fuel markets.
11. INSP, SitRep 145, reporting 6 October: cumulative cases 8,728. Used only for the provisional 21-day contact-ratio denominator.
12. NCDC, Ebola Disease Situation Report, 24 October 2026, p. 1, patient chronology. The 27 October report repeats these dates.
13. ICIR Nigeria, Ebola: NCDC identifies more contacts of Abuja patient, Rivers case tests negative, 25 October 2026 (modified 26 October). Secondary report of an NCDC statement.
14. Euronews, Ebola: Uganda restricts travel as Nigerian case evacuated to Europe, 26 October 2026. Evacuation date; Swiss authorities' account of exposure; nine further contacts after 53; Uganda's 25 October transit restrictions.
15. Daily Monitor (Uganda), Uganda remains Ebola-free as Health ministry probes Nigerian traveller infected with virus, 24 October 2026. Ugandan ministry account: asymptomatic on departure; prior presence in eastern DRC asserted; WHO and Africa CDC describe this as unconfirmed.
16. Nile Post via AllAfrica, WHO, Africa CDC declare Uganda Ebola-free after 42-day countdown, 27 August 2026. 20 confirmed cases, 2 deaths.
17. ECDC, Ebola disease: multi-country outbreak 2026, page updated 28 October 2026. Province list; Uganda outbreak ended; Nigeria case and contact evacuated to Switzerland, outside the EU/EEA; four EU/EEA evacuated or imported cases since May; risk levels.
18. UN News, Ebola: 20 contacts released from quarantine in Kenya, amid vaccine 'hope' in DR Congo, 27 October 2026. Kenya release and quarantine figures; TUMAINI trial; Bundibugyo-specific candidates under WHO advisory review.
19. Al Jazeera, WHO announces start of Ebola vaccine clinical trial in DR Congo, 24 October 2026. Bunia location; ring design; participant groups.
20. ABC News (Australia), Ebola outbreak passes 10,000 cases in DRC after nurse killed, another injured, 28 October 2026. Tshopo incident as announced by WHO's Director-General. Perpetrators not identified.
21. Actualite.cd, Ebola en RDC : la seule zone de santé de Mandima en Ituri cumule 328 décès…, 29 October 2026. Local medical-officer figures for Mandima against INSP's; national totals from the 28 October bulletin.
22. Actualite.cd, Ebola dans la Tshopo : un infirmier pris pour cible et tué sur fond de rumeur à Bafwasende, 25 October 2026. Location and date (Saturday 24 October) of the Tshopo killing.
23. Africa CDC, Africa CDC convenes African Ministers of Health as Ebola outbreak spreads across three regions, 21 October 2026. Ministerial meeting set for 28 October; no containment criteria stated.
24. Radio Tamazuj, South Sudan says suspected Ebola cases test negative, 26 June 2026.
25. Actualite.cd, Le cas suspect d'Ebola à Kinshasa était « très probablement » un cas de choléra, 19 August 2026.
26. Joseph Eastman, Pathogen Update: 10-28-2026, 28 October 2026. Secondary aggregation; used only to log the Kenya contact-denominator discrepancy.

This is the third article in the series, following From Hormuz to Bundibugyo (29 May 2026) and The Reported Ebola Curve Is Bending (2 October 2026). Cross-published on ukoilwatch.com, eurooilwatch.com and americasoilwatch.com.