Andes Virus Transmission Shows Uneven Spread and a Real Outbreak Risk
Historical contact tracing data point to a familiar but unsettling pattern in Andes virus infection: most case-patients do not pass the virus onward, yet a minority account for most secondary spread. In this analysis of past person-to-person clusters in Argentina, 23.4% of cases generated 80% of transmission events. The finding places Andes virus in the same broad epidemiologic family as other pathogens prone to superspreading, even though its average transmission appears lower than viruses such as SARS-CoV, SARS-CoV-2, Middle East respiratory syndrome coronavirus, and Ebola virus.
The practical implication is straightforward. A rare but consequential transmission pattern can still create a sizable cluster when infected people remain in close contact long enough, especially in enclosed settings where delays in recognition allow repeated exposures.
Historical Clusters Reveal Strong Transmission Heterogeneity
Investigators reviewed historical outbreak reports and extracted contact tracing data from clusters in Argentina, the setting most relevant to the seed case in the 2026 cruise ship outbreak. Five studies met eligibility criteria because they provided detailed contact histories and viral genetic evidence linking infections to specific lineages or strains. Across those reports, 88 case-clusters were identified, and 50 of them, or 56.8%, produced no secondary person-to-person transmission at all.
This is the signature of an infection with substantial heterogeneity. The pooled effective reproduction number, Rt, was 0.74 with a 95% credible interval of 0.28 to 1.29, while the dispersion parameter, k, was 0.64 with a 95% credible interval of 0.36 to 1.16. A lower k value means transmission is more uneven, with more dependence on a small number of highly infectious or highly exposed individuals. In other words, the average case is not the whole story; the distribution of infectiousness is doing much of the epidemiologic work.
Long Serial Interval Complicates Early Detection
Among the clusters with secondary spread, 76 primary-secondary transmission pairs had usable symptom onset dates. Those pairs produced a pooled mean serial interval of 21.92 days, with a standard deviation of 6.98 days. A spread pattern that unfolds over roughly three weeks gives public health teams a narrow window for rapid case finding, especially when the initial symptoms are nonspecific and easily mistaken for other viral syndromes.
A long serial interval also creates an optical illusion during the early phase of an outbreak. Few cases may be visible at first, not because transmission has stopped, but because the next generation has not yet declared itself clinically.
The analysis did observe little variation in serial interval estimates across lineages or strains, although the estimate is conditional on reported clusters and may not capture the full population-level distribution of infections. One lineage, ANDV Cent BsAs, contributed only a single transmission pair, a reminder that the data remain thin for some genetic lineages.
Cruise Ship Simulations Suggest a Low but Nonzero Risk of Expansion
The 2026 outbreak aboard the MV Hondius cruise ship gave the model a concrete real-world test case. Using the pooled estimates for Rt, k, and serial interval, the investigators ran branching process simulations under a plausible range of initial reproduction numbers for the outbreak. They assumed that effective control measures began after the World Health Organization notification on May 2, 2026, and that Rt fell after that point.
Within 10 weeks after symptom onset of the seed case, 95% of the simulated outbreak trajectories stayed below 15 cumulative cases. Even so, the probability of sustained transmission after one generation reached 20.9% for historical ANDV clusters and 28.6% under the assumed cruise ship reproduction numbers. Those probabilities are not high in absolute terms, but they are large enough to matter in a confined group of close contacts where case isolation, medical evacuation, and contact tracing must work quickly to halt spread.
The model also makes clear why delayed response changes the risk profile. When control measures arrive late, each generation has more time to seed the next, and a virus with a long serial interval can quietly extend its reach before the outbreak is recognized as more than a handful of isolated illnesses.
Why Andes Virus Can Spread Without Becoming Routine Community Transmission
Andes virus remains historically self-limited in person-to-person spread, but that description should not be mistaken for safety in all settings. Hantaviruses are rodentborne zoonotic RNA viruses, and human infection usually begins with environmental exposure to aerosolized rodent excreta. Once person-to-person spread occurs, however, the biology shifts from zoonotic spillover to close-contact transmission, where household exposure, shared travel, and prolonged proximity can sustain a chain.
The pattern seen here is biologically plausible for a respiratory or salivary route that depends on repeated close contact rather than casual community spread. A virus with this profile can simmer below the radar until a household, social gathering, or travel cohort concentrates enough exposed people in a short period of time.
Recognition Still Depends on Clinical Suspicion
The study’s cautionary point is not only statistical; it is clinical. Early Andes virus illness is nonspecific, which makes timely identification difficult unless travel history, contact history, and exposure history are actively pursued. In this outbreak context, systematic risk assessment becomes essential for symptomatic persons and for their contacts, because conventional syndromic screening can miss the relevant window.
The investigators also noted that only eight cases had occurred within roughly one mean serial interval before the initial WHO notification, leaving conventional methods too weak to estimate the initial reproduction number with confidence. That uncertainty is exactly where structured epidemiology helps: contact tracing data can support outbreak reconstruction when raw case counts are still too sparse to tell the full story.
Surveillance Remains the Main Defense
The study is limited by self-reported contact histories, publicly available data, possible misclassification from unreported rodent exposures, and recall bias in symptom onset dates. Subclinical or asymptomatic infections may also have gone undetected, and no clear evidence currently supports transmission from asymptomatic or presymptomatic case-patients. In the simulation framework, control measures were anchored to the seed case’s symptom onset date, which does not fully account for the biological delay introduced by incubation.
Even with those constraints, the message is clinically concrete: Andes virus can produce a sizable cluster in a rodent-free close-contact setting, even if most introductions remain self-limited. Because no vaccine or antiviral treatment is currently available, surveillance, rapid isolation, and contact tracing remain the only reliable tools for preventing a small cluster from becoming a larger outbreak.