Please refer to the attached file. Summary This incident analysis, covering the period from 1 September to 31 October 2026, shows a fall in the average volume of documented incidents, concentrated in October alone, together with a reshaping of friction around health facilities and schools. As the two months under review are compared with the single month of August, volumes are expressed as a weekly rate. We documented, at the Crisis Analysis Team (CAT), 130 incidents linked to the Ebola response. The volume divides into 81 incidents in September and 49 in October, a fall of 40 % from one month to the next. Expressed as a rate, this moves from 18. 9 incidents a week in September to 11. 1 in October. Across the period as a whole it stands at 14. 9 a week, against 16. 5 in August: the decline therefore becomes clear only in October, September having in fact exceeded the August level. We recorded 31 casualties, 4 people killed and 27 injured, with no abductions. Set against the number of incidents, this represents 24 casualties per 100 incidents, against 45 per 100 in August. The ratio relates casualties to events rather than to violent incidents: the great majority of incidents produce no casualties at all, 24 of the 130 produce at least one, and those twenty-four produce 1. 3 casualties each on average. The fall in frequency is not in itself sufficient to establish any easing of operational risk. Ituri remains the province most affected, with 62 incidents, followed by North Kivu with 54 and Tshopo with 12, but the balance reverses in October, 24 incidents in North Kivu against 22 in Ituri, for the first time since our monitoring began. Three systems of risk structure the operating environment of the response, and they do not yield to the same remedies. The first is distrust of health facilities and their staff, which predominates in Mandima, Beni and Butembo. The second is the rumour that schoolchildren are to be vaccinated by force, absent from our earlier monitoring, which is addressed through communication with schools rather than through security measures. The third is internal protest linked to allowance arrears, which falls to administrative and contractual management. To these are added two cross-cutting registers, the enforcement of health measures by armed personnel and the accusation of deliberate transmission of the disease, which now falls on residents as much as on responders. The forms of confrontation are shifting towards places of care. Attacks on and tensions around Ebola treatment centres (ETCs), hospitals, health posts and transit centres become the leading motive of the period, 21 incidents and 16 % of the total, including thirteen cases of buildings destroyed or damaged, nine of them by fire. This register describes a dipping course rather than a constant level: 18 incidents in July, when safe and dignified burials (SDBs) still ranked ahead of it, 7 in August and 21 across the two months under review. Friction is no longer directed first at the team on the move but at the building that receives patients, and opposition often precedes commissioning, at Drodro, Mandima, Mutwanga and Bafwasende. SDBs fall from 17 to 15 incidents and prevention work from 18 to 14, but their form hardens, with bodies recovered by force and burials carried out by families themselves. Eleven incidents involve refusal of care, transfer or follow-up, against two in August: this register produces neither casualties nor material damage, but it bears directly on epidemiological surveillance. Twenty-five incidents implicate an armed actor or a security force, 19 % of the total. The four deaths of the period fall under distinct registers: a response worker shot dead during an armed raid on a checkpoint at Beni, a nurse killed at Bafwasende under an accusation of spreading the disease, a resident killed by police fire during an SDB at Beni, and a local political figure killed at Butembo after an awareness-raising broadcast. In August, the two deaths we documented arose from a failure of care and from a road accident, and none was attributable to an attack. A further six deaths, caused by the disease itself, appear among contextual events and are not counted as incident casualties. The vaccination rumour is the new feature of the period. Sixteen incidents, 12 % of the total, disrupt schooling in ten health zones across Ituri and North Kivu, together with one Ituri location not further specified, through parents forcing their way in, children withdrawn, stones thrown and lessons interrupted. Fourteen of them turn explicitly on a vaccination rumour, most often with no vaccination campaign scheduled at all. The word vaccine appeared nowhere in our August monitoring. Community feedback constitutes, for the first time, a perception corpus in its own right: 118 items in all, 32 in September and 86 in October, against none before September. This feedback is voiced by communities and gathered from them by the partners of the IERDC consortium through the CARM mechanism; it is distinct from the CPT signals described in the methodology, which are indicators derived from the incidents themselves. Community feedback is reported speech, what a person says, fears or believes, whereas an incident is an act that occurred; the two corpora are not additive. It bears first on ETCs and the quality of care, 52 items, then on vaccination and schooling, denial of the disease and suspicion of profiteering. It documents registers that incident monitoring does not capture, the fear of reporting a sick relative, the perceived quality of care and resort to self-medication, and it reopens the register of rumour, reduced to a single incident over the period. Its coverage follows the consortium’s area of operation. The geography of priorities is being redrawn. Each zone receives a pressure score that adds its incidents, rumours and accusations and counts its attacks twice; above 20, the zone is classed as critical. Mandima becomes the leading centre of incidents, with 19 incidents and a score of 44, although the zone recorded no incident at all in August. Beni and Butembo remain critical, at 26 and 23 respectively. Mambasa, the Lubero axes, Musienene and Bunia reach the high level, as does an aggregate of incidents located in Ituri without a specified zone. Katwa falls from 12 to 7 and no longer ranks among the principal centres, while Bunia and Mongbwalu, the centres of our earliest analyses, drop out of the top of the ranking. The number of critical or high zones rises from four in August to eight, across 33 health zones affected over the period. This geographical footprint describes the evolution of incident monitoring; it is not a measure of epidemiological spread. These developments call for a differentiated response: protection of local health facilities and prior consultation on the siting of new ones, communication with schools ahead of any vaccination campaign, treatment of allowance arrears as a first-order operational risk, prior mediation with families around SDBs, and regulation of the enforcement of health measures by armed personnel. For the months ahead, five points to watch: the trajectory of Mandima, preparation for any vaccination campaign, protection of local health facilities, settlement of arrears and systematic use of community feedback. These findings should be read in the light of the limits of collection: the list of incidents reflects the information actually reported to us and is not an exhaustive census.
DR Congo: Crisis Analysis Teams – Ebola-related incident analysis September-October 2026
RELATED ARTICLES
