In a filovirus outbreak, tracing beats vaccinating

Design

stochastic transmission model on a clustered household-community contact network with Bayesian renewal estimation, probabilistic sensitivity analysis

Population

simulated Bundibugyo virus outbreak calibrated to 2026 Democratic Republic of the Congo data, 10,000 replicates per strategy over 90 days

Primary outcome

reduction in mortality and incidence, total doses, and doses per death averted

Effect

enhanced detection and tracing alone -81.6% mortality (73.1-87.7); ring vaccination added 10.5 points (6.2-15.6) at 35.1 doses per death averted; 80% community coverage -86.6% at up to 111.4 doses per death averted

rVSV-ZEBOV protects against Ebola virus but its effect against Bundibugyo virus is unproven. This stochastic transmission model, built on a clustered household and community contact network and calibrated to 2026 Bundibugyo outbreak data from the Democratic Republic of the Congo, asked what a partially cross-protective vaccine would add under realistic operational constraints, assuming 45% vaccine effectiveness including post-exposure protection.

The dominant finding is not about vaccine. Moving from base operations (30% case detection, 30% contact tracing) to enhanced operations (70% and 80%) reduced expected mortality by 81.6% (95% uncertainty interval 73.1-87.7) with no vaccination at all. Reactive ring vaccination out to contacts of contacts, added to base operations, reduced mortality by 24.6% and needed 35.1 doses per death averted. Added on top of enhanced operations, it contributed a further 10.5 percentage points (6.2-15.6). Community vaccination reduced mortality by 44.7% at 20% coverage rising to 86.6% at 80%, at 53.8 to 111.4 doses per death averted.

This is a simulation, and its conclusions inherit its assumptions — particularly the 45% effectiveness figure, which is hypothetical. What survives that caveat is the ordering. Case finding, isolation and tracing did most of the work, and vaccination added least where operations were already strong. In an outbreak where the vaccine is uncertain and the doses are scarce, the investment that matters is the one that is least photogenic.

• Case detection, isolation and contact tracing carried most of the modelled benefit
• Ring vaccination adds least where surveillance is already strong — and most where it is weak
• Community vaccination averted the most deaths but at up to 111 doses per death
• Vaccine effectiveness of 45% is assumed, not measured; the ranking is more robust than the numbers
• Dose efficiency is the metric to argue about when supply is the constraint

The statistics, in plain English

Uncertainty intervals here come from 10,000 matched stochastic replicates per strategy, so they describe variation within the model rather than uncertainty about whether the model is right — the latter is much larger and is not quantified. Doses per death averted is the useful comparison because it makes the trade-off explicit: ring vaccination at 35.1 doses per death is three times more dose-efficient than community vaccination at 111.4, while community vaccination averts more deaths in total. Effectiveness was fixed at 45% in the base case, so every absolute figure scales with an assumption that no trial has tested.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.