Sub-national heterogeneity in the time-varying reproduction number during the 2026 Bundibugyo virus disease outbreak in the Democratic Republic of the Congo: a hierarchical Bayesian analysis
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
National-level estimates of the time-varying reproduction number (Rt) for the 2026 Bundibugyo virus disease (BDBV) outbreak in the Democratic Republic of the Congo (DRC) have converged on a value close to the epidemic threshold since early August 2026, consistent with independent joint Bayesian renewal-model estimates. A single national Rt, however, can obscure divergent sub-national epidemic trajectories, particularly across a five-province outbreak in which provinces range from a declining original epicentre to recently-seeded fronts. We estimated Rt at national, provincial, and — where case volume allowed — health-zone level, using both a standard sliding-window (Cori) estimator and a hierarchical Bayesian renewal model with partial pooling across spatial units, fitted by Hamiltonian Monte Carlo (No-U-Turn Sampler). Provincial estimates diverged materially from the national trend: as of the week of 6 August 2026, Ituri — the outbreak’s original epicentre — had a hierarchical median Rt of 0.91 (95% credible interval [CrI] 0.68–1.24), while Nord-Kivu (1.23, [0.87–1.72]) and Haut-Uele (1.79, [1.24–2.67]) remained above threshold. Health-zone disaggregation, feasible only in Ituri and Nord-Kivu given case volume, showed this provincial picture itself masked further heterogeneity: in Ituri, the zone where the outbreak began (Mongbwalu) had clearly declined (Rt 0.36, [0.17–0.74]) while the two largest zones by cumulative case count (Bunia, Rwampara) remained at or above threshold; in Nord-Kivu, elevated transmission was concentrated in a single zone (Katwa, Rt 1.39, [0.85–1.99]) while a comparably-sized zone (Butembo) had already declined (0.64, [0.23–1.38]). An initial disagreement between the sliding-window and hierarchical provincial estimates was traced to a data-reconstruction artefact (forward-filling, rather than interpolating, multi-day gaps in health-zone reporting) rather than a genuine methods disagreement, and resolved once corrected. The hierarchical model’s dispersion structure, calibration, and sensitivity to the generation-interval assumption were each checked explicitly; a shared (non-province-specific) dispersion parameter was retained on the basis of negligible predictive difference (PSIS-LOO), the model achieved 95.0% pooled 95% posterior-predictive interval coverage, and the province ranking was unchanged across a generation-interval sensitivity grid (Spearman ρ = 1.0). Aggregation masks meaningful heterogeneity in transmission intensity at every spatial resolution examined; response prioritisation based on a single national or even provincial Rt risks directing attention away from the specific zones where transmission remains supercritical.
Author Summary
During an Ebola outbreak, public health teams track a number called the reproduction number, or Rt — roughly, how many new people each infected person goes on to infect. When Rt is above 1, an outbreak is growing; below 1, it is shrinking. For the 2026 Bundibugyo virus disease outbreak in the Democratic Republic of the Congo, the national Rt has hovered close to 1 since early August 2026. We show that this single national number hides very different realities in different places. The province where the outbreak started, Ituri, has an estimated Rt below 1 overall — but even within Ituri, the specific area where the outbreak began has clearly declined while the province’s two largest towns by case count have not. In Nord-Kivu province, where the overall Rt remains above 1, we find this is driven by a single health zone rather than widespread growth. We used a statistical technique called hierarchical Bayesian modelling, which allows data-poor areas to “borrow strength” from data-rich ones rather than producing wild, uninterpretable estimates. We also found and corrected a data-processing error that had made a simpler method appear to disagree with our more careful one — a reminder that methodological rigour includes checking the data pipeline, not just the statistical model. Our results suggest that outbreak response resources should be targeted at the specific health zones still driving transmission, which a national or even provincial figure alone would not reveal.
