Summary
A preprint simulation estimates that faster, strategically placed PCR laboratories could avert more future deaths in the modeled 2026 Bundibugyo Ebola epidemic. The projections depend on assumed epidemic-control conditions and are not results from a laboratory deployment.
A simulation study in a medRxiv preprint estimates that placing PCR laboratories quickly in strategic health zones could reduce future deaths in the 2026 Bundibugyo Ebola epidemic in eastern Democratic Republic of the Congo. In the model, 35 laboratories deployed immediately averted a median 49% of future deaths; a 14-day deployment delay reduced that estimate to 38%, and a 60-day delay to 11%. These are modeled outcomes, not measured effects of laboratories deployed during an epidemic.
The authors say delayed recognition and multi-day transport of samples contributed to the epidemic’s spread. Their proposed approach is to put diagnostic capacity in frontier health zones, where faster confirmation could support earlier isolation and contact tracing as infections move between areas.
A spatial model tests laboratory placement
The researchers fitted a Bayesian spatial branching-process model to public situation reports through 2 August 2026. This kind of model represents transmission as chains of infections that can spread between locations. It incorporated movement estimates from Flowminder and an anchor for infections not yet observed, as well as assumptions about transmission, PCR access and turnaround, clinical recognition, isolation, contact tracing, safe burial and local behaviour change.
Because the observed epidemic history could not determine how future control efforts would unfold, the researchers assumed a national response ramp that produced a West-Africa-scale status-quo burden. They then replayed 2,000 posterior epidemic trees, testing scenarios with 5–100 zone-level PCR laboratories, deployment delays of 0–180 days and different levels of postmortem rapid diagnostic test (RDT) coverage. The authors report that the model reproduced case, death, latent-burden and spatial targets.
Speed and location shaped the projections
For 35 laboratories, the median share of future deaths averted fell from 49% with immediate deployment to 38% after 14 days and 11% after 60 days. The corresponding 90% prediction intervals were 30–65%, 23–53% and 5–20%. In these scenarios, the projected effect was therefore highly sensitive to how quickly diagnostic capacity became available.
Where laboratories were placed also mattered. With a 30-day delay, model-informed placement averted an estimated 18–33% of future deaths across scenarios with 20–100 laboratories, compared with 1–8% under random placement. The model also found that postmortem RDTs could complement live-case detection: with 20 laboratories, a 14-day delay and 75% RDT coverage, the projected share of future deaths averted rose from 26% to 41%.
The findings make the case for treating laboratory location and deployment time as part of epidemic response planning, rather than focusing only on the total number of laboratories. In practice, the authors’ proposed spatial fence also depends on the capacity to staff testing and act on results through isolation, tracing and safe burial. The estimates come from a preprint simulation whose projections depend on its assumptions about future control; they are not evidence from a prospective deployment.