Estimating Infectious Disease Importation Risk during the 2026 FIFA World Cup

Technical Summary: Estimating Infectious Disease Importation Risk during the 2026 FIFA World Cup

Problem Statement
The 2026 FIFA World Cup, scheduled for June 11–July 19, 2026, will host 48 qualified nations across 11 US cities, representing the largest tournament in history with an estimated 1–5 million international visitors. Such mass gatherings create conditions for the accelerated importation of infectious diseases from diverse source populations. Existing frameworks for estimating this risk often rely on aggregate passenger volumes that obscure spatial and source-country heterogeneity, focus on single diseases, or fail to account for event-specific travel surges. The authors identify a need for a transparent, reproducible, and data-driven approach to estimate importation risk in advance to support public health preparedness and surveillance prioritization.

Methodology
The study develops a general Poisson importation framework applied to five diseases (dengue fever, influenza, malaria, measles, and pertussis) across 11 US host venues. The framework integrates multiple publicly available datasets: US Customs and Border Protection (COR) I-94 monthly arrivals, Bureau of Transportation Statistics (BTS) T-100 airport routing data, National Travel and Tourism Office (NTTO) travel projections, and WHO disease surveillance data.

The core of the methodology involves three nested models of increasing resolution:

1. Baseline Model (Model 1): Uses routine June 2024 arrival data without World Cup (WC) adjustments. This establishes a lower-bound reference for routine tourism risk.
2. WC-Adjusted Model (Model 2): Scales travel volumes to June 2026 projections using NTTO forecasts and growth factors ($\phi_c$). It applies a specific growth factor of 1.174 for WC-qualified nations and 1.077 for others to capture the aggregate travel surge.
3. Schedule-Driven Model (Model 3): Decomposes travel into a "WC-fan stream" and a "background stream." The fan stream is routed to specific cities based on match assignments (using official FIFA schedules), while the background stream follows habitual routing patterns derived from BTS T-100 data.

Importation Risk Calculation
The expected number of infectious travelers ($\lambda_{c,v,d}$) arriving from country $c$ to city $v$ for disease $d$ is modeled as:
$$\lambda_{c,v,d} = N_{c,v} \cdot I_{c,d} \cdot p_d$$
Where:

• $N_{c,v}$ is the projected number of travelers.
• $I_{c,d}$ is the country-level disease incidence adjusted for under-reporting ($\rho_d \cdot C_{c,d}/P_c$).
• $p_d$ is the probability of traveling while infectious.

Total importation intensity ($\Lambda_{v,d}$) is the sum of contributions from all source countries. The probability of at least one importation is calculated as $P(\ge 1) = 1 - e^{-\Lambda_{v,d}}$.

Uncertainty Propagation
To address uncertainty in under-reporting factors ($\rho_d$) and travel-while-infectious probabilities ($p_d$), the authors employ Monte Carlo uncertainty propagation. They draw 5,000 independent samples from Uniform distributions calibrated to literature ranges for these parameters. Results are reported as medians with 95% uncertainty intervals (2.5th–97.5th percentiles).

Key Results

• Disease Risk Profile: Dengue fever poses the highest importation risk at most venue cities under the schedule-driven model, with median expected importations ($\Lambda$) exceeding 10 from Brazil alone in high-risk cities like New York, Los Angeles, and Miami. Atlanta is a notable exception where malaria risk exceeds dengue due to direct travel from West and Central African nations. Influenza ranks second at most cities, driven by the Southern Hemisphere winter peak coinciding with the tournament. Pertussis shows broad geographic spread but carries the widest relative uncertainty. Measles consistently shows the lowest probability due to severe prodromal symptoms suppressing travel.
• Source Drivers: Latin America (Brazil, Colombia, Venezuela) dominates dengue and influenza risk. West and Central Africa drives malaria risk. European nations lead pertussis and measles risk. Brazil is identified as the single highest-risk source nation overall.
• Travel Stream Composition: Background tourism (travel occurring regardless of the event) accounts for the dominant share of total importation risk. The World Cup fan increment contributes approximately 8.3% of projected arrivals for qualified nations.
• Model Robustness: City rankings (New York, Los Angeles, Miami, Houston as highest risk) remain robust across all three model tiers. The step from the baseline to the WC-adjusted model produces a near-uniform ~17% uplift in risk, while the schedule-driven model introduces city-specific divergence without altering the overall hierarchy.

Significance and Claims
The paper claims to provide a general, reproducible Poisson importation framework built entirely from publicly available data. Its primary contributions are:

1. Transparency and Reproducibility: Unlike proprietary models relying on commercial booking data, this framework uses only public datasets, allowing any public health team to deploy, update, and adapt the analysis.
2. Granular Attribution: The three-model hierarchy allows for the clean isolation of the World Cup travel surge effect and the added value of schedule-based fan routing.
3. Practical Utility: The framework serves as a transparent baseline complementary to proprietary systems. It can function as a standing surveillance tool for routine importation pressure at US international gateways, independent of specific mass gatherings.
4. Risk Communication: The use of Monte Carlo uncertainty intervals allows decision-makers to distinguish robust findings (e.g., dengue risk) from estimates warranting caution (e.g., pertussis risk, which is presented as a conservative upper bound).

The authors conclude that public health preparedness should not focus exclusively on World Cup fans but must treat the entire June international visitor population as the primary exposure group, as background tourism drives the majority of importation risk. The framework is designed to be extensible to additional diseases (e.g., chikungunya, mumps), cities, and mass gathering events.