Pandemic preparedness is only as strong as its weakest assumption — and when it comes to influenza antiviral stockpiles, the assumptions underpinning national and global reserve estimates may be dangerously oversimplified. This matters because stockpile decisions made today determine how many treatment courses are available during the opening weeks of a future pandemic, when supply chains are typically overwhelmed and new vaccines are months away.
Published in PNAS, this modeling study constructs a mathematical framework to estimate global demand for oral antivirals — such as neuraminidase inhibitors or newer cap-dependent endonuclease inhibitors — during an influenza pandemic, while simultaneously projecting their population-level public health impact. Critically, the model incorporates heterogeneous healthcare-seeking behavior across income strata and geographies, as well as variable drug effectiveness estimates, two factors that prior stockpile calculations largely ignored. The findings suggest that existing reserve targets derived from simpler models may systematically misallocate quantities relative to actual expected demand and potential lives saved, though precise effect magnitudes are withheld here to encourage engagement with the primary source.
From a broader epidemiological perspective, this work fits into a growing recognition that pandemic preparedness modeling must graduate from homogeneous population assumptions toward stratified, behavior-aware frameworks. Earlier stockpile rationale — largely built around the 2009 H1N1 experience — has already been questioned for overestimating treatment uptake in lower-income settings and underestimating demand surges in high-income ones. The practical implication for public health policy is significant: if stockpile sizing is miscalibrated, the marginal value of each stored treatment course plummets. Key limitations include the model's reliance on assumed behavioral parameters that may not transfer across future pandemic strains with different severity profiles. As a mathematical modeling study, it generates scenario-based projections rather than empirical outcomes. Nevertheless, this represents a meaningful methodological advance — incremental in technical approach but potentially consequential for how international health bodies recalibrate their antiviral reserve targets going forward.