COVID-19 at a Wisconsin Summer School Retreat, 2020: One Student, 116 Cases

Image from the Centers for Disease Control and Prevention

From July 2 to August 11, 2020, a faith-based overnight summer school retreat for boys in grades 9–11 in Wisconsin brought together 152 students, counselors and staff from 21 states and territories and two other countries. It became a case study in how fast SARS-CoV-2 spreads in a congregate setting.

The precautions — and the gaps

Before arriving, every attendee had to show either a positive antibody test in the past 3 months or a negative RT-PCR test within 7 days of travel, self-quarantine at home for 7 days, and wear a mask while traveling.

At the retreat, students and counselors did not have to wear masks or keep apart, and mixed freely.

• Classes met in outdoor pavilions, about 20 students per class seated less than 6 feet apart; teachers wore masks and kept their distance.
• The 127 students slept in dormitories (four to six per room) and yurts (eight per room), in tightly spaced bunks with shared bathrooms and common areas; the 21 counselors (aged 17–24) roomed with them.

How it spread

1. July 2: students met at a regional hub and rode three buses to the retreat.
2. July 3: a ninth grader — negative on RT-PCR less than a week earlier — developed sore throat, cough and chills, and tested positive on July 5. A family member later tested positive about a week after he left home. He was isolated, and 11 close contacts, including four roommates, were quarantined together.
3. July 7: the 11 contacts were released after negative rapid antigen results that public health could not verify.
4. July 4–7: six of those contacts and 18 other students developed mild symptoms. They got masks, but there was no contact tracing and no isolation.
5. July 13: a second student, one of the original 11 contacts, tested positive; the state health department was notified on July 15. Its advice — symptom monitoring, isolation and quarantine — was overwhelmed by the number of sick attendees.

Cases at the retreat by date, with key events. Image from the Centers for Disease Control and Prevention

The toll

Testing: the health department ran RT-PCR on 148 attendees on July 28 — by then no one had fallen ill since July 20 — and antibody tests on August 5–6.

• 116 of 152 (76%) had COVID-19: 78 confirmed by RT-PCR and 38 probable (symptoms but a late negative test).
• Leaving out the 24 attendees who came with prior positive antibodies, the attack rate was 91% (116 of 128).
• At least one confirmed case occurred in every dormitory room and yurt; rates didn't differ between students and counselors, dormitories and yurts, or grades.
• Genome sequencing of six positive samples found them nearly identical — a common source.
• Illness was mild to moderate: no hospitalizations or deaths, and only 1 of 78 confirmed cases had no symptoms.
• The four staff all tested negative by RT-PCR, though one met the probable-case definition.

Prior antibodies may have helped: none of the 24 attendees with earlier positive antibody tests tested positive at the retreat — a hint of some protection, though how much and for how long was unknown.

Response: after July 28, the 36 remaining susceptible attendees were quarantined from the rest. Outdoor classes and recreation continued, and everyone was cleared to travel home on August 11 — avoiding travel while in isolation or quarantine likely prevented spread to families and communities.

What would have helped

Pre-travel testing and quarantine and outdoor classes weren't enough on their own. A full mitigation plan for overnight camps and residential schools should include:

• a 14-day pre-arrival quarantine, and testing and symptom screening before and after arrival;
• cohorting attendees, including for 14 days after arrival;
• symptom monitoring and early isolation and quarantine;
• physical distancing and masks;
• enhanced hygiene, cleaning and disinfection;
• as much outdoor programming as possible;
• early contact with public health authorities.

Limitations

RT-PCR testing came after the outbreak, likely undercounting confirmed cases; not everyone had baseline antibody results, so some later positives may reflect earlier infections; dates of earlier illness were unknown; and the probable-case definition was adapted for this investigation. Some mild symptoms may have been allergies, travel fatigue or other viruses.

Sources

Based on "COVID-19 Outbreak at an Overnight Summer School Retreat — Wisconsin, July–August 2020," by Ian W. Pray, Suzanne N. Gibbons-Burgener, Avi Z. Rosenberg, Devlin Cole, Shmuel Borenstein, Allen Bateman, Eric Pevzner and Ryan P. Westergaard, Morbidity and Mortality Weekly Report 69(43), Centers for Disease Control and Prevention; a work of the United States government in the public domain.