A large COVID-19 outbreak in a high school 10 days after schools’ reopening, Israel, May 2020

Chen Stein‐Zamir, Nitza Abramson, Hanna Shoob, Erez Libal, Menachem Bitan, Tanya Cardash, Refael Cayam, Ian Miskin•View original

OverviewBalancedwilliam voice

A school reopened on a Monday. Ten days later, one hundred fifty-three students tested positive for COVID-19. That's the whole story, compressed. But the how — the specific, reconstructable chain of conditions that made it happen — is what Stein-Zamir and colleagues spent weeks documenting. This is what makes this paper worth your attention if you've ever wondered what "reopening safely" actually requires. Israel moved fast in March 2020. On the 13th, the government declared full closure of all schools. Two months later, after a limited partial reopening for younger grades in small groups, all school classes reopened on May 17th. The Ministry of Health had requirements in place: daily health reports, hygiene measures, facemasks, social distancing, and minimal mixing between classes. On paper, this was a managed return. The school at the center of the study — a Jerusalem-area regional public high school the investigators call School 1 — reopened the following Monday, May 18th. It serves one thousand one hundred ninety students aged 12 to 18 and employs one hundred sixty-two staff members. Students returned to their previous classrooms and received instructions on the prescribed preventive procedures. What they returned to was a large, dense, busy institution. Classes ran six days a week, totaling thirty-eight to forty hours. Most students traveled 20 to 45 minutes each way. Many had two to four hours of extracurricular activities per week. This was the operational reality as two months of closure gave way to full in-person schooling. Then, three days in, a heatwave arrived. Temperatures climbed to 40 degrees Celsius from May 19th through the 21st. The Ministry of Health responded by exempting schoolchildren from wearing facemasks for those three days. That decision — reasonable in isolation, catastrophic in context — removed the one mitigation layer that might have slowed what was already building. The first notified case, a student the investigators call Student A, was recorded on May 26th. Student B appeared the next day. Both had attended school during the heatwave days and both reported mild symptoms — loss of smell, loss of taste, fever, and headache — and they were from different grades. They were not epidemiologically linked to each other. Two unrelated cases appearing within twenty-four hours, nine days after reopening, told the district health office everything it needed to know. They declared an outbreak, closed the school, and ordered mass testing. The testing campaign ran over a long holiday weekend, from May 28th through 30th, as a coordinated operation involving the school, four Health Funds, Magen David Adom, the municipality, and the district health office. Of the school community, one thousand one hundred sixty-one of one thousand one hundred sixty-four students and one hundred fifty-one of one hundred fifty-two staff members were tested — a near-total sweep. The results revealed one hundred fifty-three confirmed student cases and twenty-five confirmed staff cases. Among tested students, the attack rate — the share of the exposed group who were infected — was thirteen point two percent. Among staff, it was sixteen point six percent. One student visited an emergency room. There were no hospitalizations. Now here's what the case counts alone don't tell you: where in the school transmission concentrated most intensely. The junior grades were hit hardest. Ninth grade had sixty-one cases out of one hundred eighty-seven students tested — a thirty-two point six percent attack rate. Eighth grade: thirty-four of one hundred ninety-seven, or seventeen point three percent. Seventh grade: forty of one hundred ninety-seven, or twenty point three percent. Within those grades, the clustering was sharp. One ninth-grade class had twenty cases. Two others had thirteen each. One seventh-grade class had fourteen. Several teachers threaded through all of it — four teachers taught all four of the most heavily affected classes, two taught three of them, and one taught two. The virus moved along the actual social architecture of the school. When investigators conducted the environmental assessment, the physical explanation came into focus. Classrooms held thirty-five to thirty-eight students in rooms of thirty-nine to forty-nine square meters. Do that math: one point one to one point three square meters per student — below the one point five square meter standard cited in the report. Physical distancing, the investigators wrote plainly, was "not possible." Air conditioning ran continuously, with each class served by its own unit. Closed rooms, dense occupancy, recirculating air, and three days without masks during the hottest stretch of the school year. The European Centre for Disease Prevention and Control had already warned against what they called the "three Cs" — closed spaces with poor ventilation, crowded places, and close-contact settings. School 1, during the heatwave, embodied all three simultaneously. The investigators are careful not to point to a single cause. It was the convergence that did it. For context: Israel's secondary school classes average twenty-nine students nationally, against an OECD average of twenty-three. School 1 was running at thirty-five to thirty-eight. That gap mattered. The outbreak didn't stay inside the building. By mid-June 2020, eighty-seven additional confirmed cases had appeared among close contacts of the school cases — siblings, parents, other family members, friends from other schools, and participants in after-school sports and dance classes. Transmission had radiated into households and community settings. In total, Stein-Zamir and colleagues estimate roughly two hundred sixty persons were infected, tracing back to that single school. And then there's the population-level signal, which is where this paper shifts from a case study to something broader. The investigators looked at confirmed COVID-19 cases in the Jerusalem district across the weeks of 2020, using the week after schools reopened as the dividing line. Before May 24th — weeks nine through twenty-one — 10 to 19 year-olds accounted for nineteen point eight percent of confirmed cases in Jerusalem: nine hundred thirty-eight out of four thousand seven hundred forty-seven. After May 24th — weeks twenty-two through twenty-five — that share jumped to forty point nine percent: three hundred sixteen out of seven hundred seventy-two cases. Nearly doubled. When the investigators compared Jerusalem to the rest of the country over the same period, the divergence was clear. In Jerusalem, 10 to 19 year-olds made up twenty-two point six percent of confirmed cases. In all other districts combined, it was thirteen point nine percent. The school outbreak wasn't just a local event. It bent the age curve of an entire district's epidemic. Most of the student cases were mild or asymptomatic — forty-three percent of confirmed student cases reported any symptoms at all, compared to seventy-six percent of staff. The authors note reports from New York State and French surveillance linking SARS-CoV-2 to multisystem inflammatory syndrome in children, or MIS-C, though the clinical picture at this school was mostly mild. That context matters for how we weigh the policy tradeoffs. But it doesn't change what the transmission data show. The recommendations Stein-Zamir and colleagues close with are specific, not vague. Keep study groups small. Minimize mixing between classes, activities, and transport. Make sure teachers and parents model mask use, hand hygiene, and distancing — not just students. Require anyone with symptoms to stay home. Consider remote or outdoor learning when conditions allow. And prioritize ventilation: increase air exchange with outdoor air and decrease recirculation inside. That last point connects directly to what failed here — air conditioning systems running in sealed, packed rooms during a period of no masking. The deeper lesson this paper offers is about the gap between policy on paper and conditions on the ground. The Ministry of Health's reopening requirements were real. Masks were mandated. Distancing was instructed. But those measures encountered a school with classes of thirty-five to thirty-eight students in rooms designed for fewer, a six-day week in a building where junior and senior grades shared common spaces, and a heatwave that triggered an exemption at precisely the wrong moment. The conditions that made this outbreak possible aren't unique to this school or this country. Dense classrooms, inadequate ventilation, and the pressure to relax precautions during extreme heat — those are features of school systems everywhere. Stein-Zamir and colleagues documented one outbreak in one Jerusalem high school in May 2020. What they actually mapped was the blueprint for how transmission finds its way through a building when the structural conditions are right and the mitigations slip. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.

This was the operational reality as two months of closure gave way to full in-person schooling. Then, three days in, a heatwave arrived. Temperatures climbed to 40 degrees Celsius from May 19th through the 21st. The Ministry of Health responded by exempting schoolchildren from wearing facemasks for those three days. That decision — reasonable in isolation, catastrophic in context — removed the one mitigation layer that might have slowed what was already building. The first notified case, a student the investigators call Student A, was recorded on May 26th. Student B appeared the next day. Both had attended school during the heatwave days and both reported mild symptoms — loss of smell, loss of taste, fever, and headache — and they were from different grades. They were not epidemiologically linked to each other. Two unrelated cases appearing within twenty-four hours, nine days after reopening, told the district health office everything it needed to know. They declared an outbreak, closed the school, and ordered mass testing.

The testing campaign ran over a long holiday weekend, from May 28th through 30th, as a coordinated operation involving the school, four Health Funds, Magen David Adom, the municipality, and the district health office. Of the school community, one thousand one hundred sixty-one of one thousand one hundred sixty-four students and one hundred fifty-one of one hundred fifty-two staff members were tested — a near-total sweep. The results revealed one hundred fifty-three confirmed student cases and twenty-five confirmed staff cases. Among tested students, the attack rate — the share of the exposed group who were infected — was thirteen point two percent. Among staff, it was sixteen point six percent. One student visited an emergency room. There were no hospitalizations. Now here's what the case counts alone don't tell you: where in the school transmission concentrated most intensely. The junior grades were hit hardest. Ninth grade had sixty-one cases out of one hundred eighty-seven students tested — a thirty-two point six percent attack rate. Eighth grade: thirty-four of one hundred ninety-seven, or seventeen point three percent. Seventh grade: forty of one hundred ninety-seven, or twenty point three percent. Within those grades, the clustering was sharp. One ninth-grade class had twenty cases. Two others had thirteen each. One seventh-grade class had fourteen.

Several teachers threaded through all of it — four teachers taught all four of the most heavily affected classes, two taught three of them, and one taught two. The virus moved along the actual social architecture of the school. When investigators conducted the environmental assessment, the physical explanation came into focus. Classrooms held thirty-five to thirty-eight students in rooms of thirty-nine to forty-nine square meters. Do that math: one point one to one point three square meters per student — below the one point five square meter standard cited in the report. Physical distancing, the investigators wrote plainly, was "not possible." Air conditioning ran continuously, with each class served by its own unit. Closed rooms, dense occupancy, recirculating air, and three days without masks during the hottest stretch of the school year. The European Centre for Disease Prevention and Control had already warned against what they called the "three Cs" — closed spaces with poor ventilation, crowded places, and close-contact settings. School 1, during the heatwave, embodied all three simultaneously. The investigators are careful not to point to a single cause. It was the convergence that did it. For context: Israel's secondary school classes average twenty-nine students nationally, against an OECD average of twenty-three. School 1 was running at thirty-five to thirty-eight. That gap mattered.

The outbreak didn't stay inside the building. By mid-June 2020, eighty-seven additional confirmed cases had appeared among close contacts of the school cases — siblings, parents, other family members, friends from other schools, and participants in after-school sports and dance classes. Transmission had radiated into households and community settings. In total, Stein-Zamir and colleagues estimate roughly two hundred sixty persons were infected, tracing back to that single school. And then there's the population-level signal, which is where this paper shifts from a case study to something broader. The investigators looked at confirmed COVID-19 cases in the Jerusalem district across the weeks of 2020, using the week after schools reopened as the dividing line. Before May 24th — weeks nine through twenty-one — 10 to 19 year-olds accounted for nineteen point eight percent of confirmed cases in Jerusalem: nine hundred thirty-eight out of four thousand seven hundred forty-seven. After May 24th — weeks twenty-two through twenty-five — that share jumped to forty point nine percent: three hundred sixteen out of seven hundred seventy-two cases. Nearly doubled. When the investigators compared Jerusalem to the rest of the country over the same period, the divergence was clear.

In Jerusalem, 10 to 19 year-olds made up twenty-two point six percent of confirmed cases. In all other districts combined, it was thirteen point nine percent. The school outbreak wasn't just a local event. It bent the age curve of an entire district's epidemic. Most of the student cases were mild or asymptomatic — forty-three percent of confirmed student cases reported any symptoms at all, compared to seventy-six percent of staff. The authors note reports from New York State and French surveillance linking SARS-CoV-2 to multisystem inflammatory syndrome in children, or MIS-C, though the clinical picture at this school was mostly mild. That context matters for how we weigh the policy tradeoffs. But it doesn't change what the transmission data show. The recommendations Stein-Zamir and colleagues close with are specific, not vague. Keep study groups small. Minimize mixing between classes, activities, and transport. Make sure teachers and parents model mask use, hand hygiene, and distancing — not just students. Require anyone with symptoms to stay home. Consider remote or outdoor learning when conditions allow. And prioritize ventilation: increase air exchange with outdoor air and decrease recirculation inside. That last point connects directly to what failed here — air conditioning systems running in sealed, packed rooms during a period of no masking.

The deeper lesson this paper offers is about the gap between policy on paper and conditions on the ground. The Ministry of Health's reopening requirements were real. Masks were mandated. Distancing was instructed. But those measures encountered a school with classes of thirty-five to thirty-eight students in rooms designed for fewer, a six-day week in a building where junior and senior grades shared common spaces, and a heatwave that triggered an exemption at precisely the wrong moment. The conditions that made this outbreak possible aren't unique to this school or this country. Dense classrooms, inadequate ventilation, and the pressure to relax precautions during extreme heat — those are features of school systems everywhere. Stein-Zamir and colleagues documented one outbreak in one Jerusalem high school in May 2020. What they actually mapped was the blueprint for how transmission finds its way through a building when the structural conditions are right and the mitigations slip. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.

Related lectures