Rastreo de contactos del hantavirus en el MV Hondius: cómo 1.500 pasajeros se vuelven una cohorte manejable

Contact tracing is one of those phrases that has been used so often in the last six years that its specifics have drifted out of focus. The work being done right now to bound the 2026 MV Hondius hantavirus cluster is a chance to look at how a serious outbreak team actually does it — what a contact is, how the rings are drawn, what the follow-up windows look like, and what investigators learn from the data as it comes in.

This article is written from the perspective of the operational playbook, not the headline. If you are a journalist trying to understand a tracker number; a public-health student looking at outbreak response for the first time; or a clinician who wants to check that your follow-up window is the right one — this is for you.

The case definition comes first

Nothing useful happens in contact tracing until investigators agree on what counts as a case. For the MV Hondius cluster the joint ECDC / Dutch Inspectorate / Argentine MoH team has published three definitions:

• Confirmed case. Compatible clinical illness (fever ≥38 °C plus thrombocytopenia or progressive respiratory distress) AND laboratory confirmation by RT-PCR for Andes virus OR seroconversion on paired sera.
• Probable case. Compatible clinical illness in a person with an epidemiological link to a confirmed case, pending laboratory result.
• Suspect case. Compatible clinical illness in a person with the relevant exposure window (on board the MV Hondius between 28 March and 7 May 2026, OR a household contact of someone who was), pending clinical and laboratory work-up.

What a contact is — and is not

For Andes virus the team is working from three contact tiers, adapted from the WHO Andes virus contact tracing guideline (rev. 2024):

1. Close contact. Cumulative ≥1 hour within 1 metre of a symptomatic confirmed case, OR direct contact with their respiratory secretions, blood or other body fluids without appropriate PPE. This is the ring that is actively followed.
2. Casual contact. Shared an indoor space with a symptomatic case for less than 1 hour, or was on the same deck of the ship in the relevant window without a documented close interaction.
3. Environmental contact. Spent time in the identified contaminated zone of the vessel (the affected provisioning store and the connected return-air loop) without PPE, regardless of whether they met any case.

The first ring is monitored daily. The second is monitored passively (handed an information sheet, told to call the hotline on symptom onset). The third is contacted once with public-health information and added to the registry but not actively followed.

The follow-up window: 42 days, not 14

This is where Andes virus differs sharply from a respiratory virus most readers will know better. The reported incubation period for ANDV runs from a minimum of about 7 days to as long as 42 days, with a median around 22. The active follow-up window is therefore set at 42 days from last exposure — three to six times the window most people will associate with the word "quarantine" from the COVID-19 era.

Operationally that means a close contact identified on 7 May will be followed through 18 June at the latest. Each follow-up consists of a daily check-in (text-message symptom prompt with a nurse on call for any "yes"), with a clinical assessment at days 14, 28 and 42. This is a substantial commitment on the part of both the public-health system and the contact; compliance has historically been the weakest link in long-window ring-tracing.

The MV Hondius cohort, by the numbers

Cruise ships are an unusually clean cohort to trace. Manifests are complete, cabin assignments are fixed, swipe-card data records who was in which dining room and when, and disembarkation produces a formal record. From the published Day-3 sit-rep:

• 1,512 persons aboard during the index window (1,238 passengers, 274 crew).
• 1,512 initial information notices issued — 100% coverage at disembarkation.
• 347 close contacts identified, primarily crew quarters and the dry-stores team.
• 30+ household and downstream contacts traced across 12 countries (this is the figure on our dashboard).
• 10 confirmed cases (3 deaths) as of the most recent WHO Disease Outbreak News update.

What investigators learn from the data, in real time

The richest signal in a contact-tracing dataset is not the count; it is the shape of the curve. Three patterns are watched especially closely:

1. Onset-to-onset interval. If symptoms in secondary cases cluster about 22 days after onset in the index case, the chain is dominated by the known person-to-person mode. If they cluster much earlier, common-source exposure is more likely than chained transmission.
2. Attack rate by deck and air-handling zone. High attack rates in cabins served by a single air-handling unit — without a clear human-contact link — point to an environmental rather than person-to-person mechanism.
3. Healthcare-worker rate. Healthcare workers were a sentinel population in the original 1996 El Bolsón cluster; their attack rate is watched closely as an early indicator of person-to-person amplification.

What can go wrong

Three failure modes are common enough to plan for:

• Contact loss to follow-up. Six weeks is a long time. Roughly 10–20% of contacts in similar long-window outbreaks are lost to active follow-up; the playbook compensates with passive surveillance through primary care alerts.
• Cross-jurisdictional friction. A 1,500-passenger cohort spanning 12 countries depends on prompt, two-way data exchange between national focal points under the IHR. Delays of 24–72 hours per hand-off are not unusual.
• Stigma and disclosure. Contacts who fear job-loss or social consequences may under-report symptoms. Outbreak teams typically respond with anonymised case reporting and explicit non-disclosure to employers.

How this connects to what you see on this site

The numbers on the dashboard are downstream of the work above. We count confirmed cases because that is what national focal points and WHO publish; we publish a contacts-traced figure because it gives a clearer sense of the operational footprint than confirmed cases alone. If either number changes faster than the other, that tells you something about where the outbreak is right now: a rising contact count with stable cases means tracing is widening; rising cases with stable contacts means the existing rings are producing.

Frequently asked questions

Is the 42-day window evidence-based?

Yes. It comes from prospective follow-up of close contacts in Argentine and Chilean clusters since 1996. The longest well-documented incubation in that dataset is 42 days; cases at the upper bound are rare but real.

Why not just test all contacts now?

Hantavirus RT-PCR is most sensitive in symptomatic patients with viraemia; it is unreliable as a screening test in asymptomatic contacts. A negative test on day 5 does not tell you the contact is in the clear. Symptom-based daily monitoring outperforms screening in this disease.

Is the contact data on the dashboard the full picture?

Close to it, but not exactly. The dashboard reflects what national focal points report up to WHO and ECDC; some jurisdictions aggregate before reporting. Differences of 5–10% between the public number and the operational number are normal in any ongoing outbreak.

References

1. WHO. "Andes virus disease: contact tracing operational guidance." 2024 revision. World Health Organization, Pan American Health Organization regional office.
2. Toro J., et al. "An outbreak of hantavirus pulmonary syndrome, Chile, 1997." Emerging Infectious Diseases, 1998.
3. Wells R. M., et al. "An unusual hantavirus outbreak in southern Argentina: person-to-person transmission?" Emerging Infectious Diseases, 1997.
4. ECDC. "Public-health management of persons exposed to Andes virus — interim guidance." European Centre for Disease Prevention and Control, May 2026.
5. Padula P. J., et al. "Hantavirus pulmonary syndrome outbreak in Argentina: molecular evidence for person-to-person transmission of Andes virus." Virology, 1998.