UNESCO Chair Global Health and Education Health Education, Prevention, Health Promotion, Children and Young People

What does the evidence tell us about keeping schools open safely

Presentation high-level ministerial meeting UNESCO – 29 March 2021

Presentation by Professor Didier Jourdan, chair holder of the UNESCO Chair and head of the WHO Collaborating Center Global Health and Education, during the high-level ministerial meeting “One year into COVID: Prioritizing education recovery to avoid a generational catastrophe” on 29 March 2021.

Ministers, Ladies & Gentlemen, it’s a great honour to be invited to share with you the latest evidence that can contribute to defining educational policies during the pandemic.

For this ‘one year on’ scientific update, I have selected 3 solid facts to present for your consideration.

  • Solid fact 1: schools should be among the last places to close and first to reopen
  • Solid fact 2: school reopening, with comprehensive infection prevention and control measures in place, and when the community infection levels were low or moderate, did not increase community transmission
  • Solid fact 3: the mechanisms of implementation in schools involve institutional, contextual and personal factors

Of course, all the studies I’ll quote today have limitations that have to be taken into account because studies were at risk of confounding factors from other interventions implemented around the same time, because there are a limited number of implementation studies in schools, and because there are few studies from LMIC. This presentation does not consider the epidemiology of new variants, because robust evidence on their potential impact in school settings is not yet available.

Solid fact 1: schools should be among the last places to close and first to reopen

There are two main reasons why.

First of all, more and more studies confirm what we already strongly suspected: that school closures are detrimental to child health and well-being and educational outcomes, including:

  • Learning loss in core subjects;
  • Increase in the socioeconomic skills gap – for example, a model based on the PISA Programme for International Student Assessment data showed that closure could increase the learning gap by more than 30 percent;
  • Unprecedented declines in college enrolments, which was especially steep for graduates of low-income high schools, -11.4% compared to -2.9 % in privileged areas.

A recent systematic review including 46 studies confirms that school closures not only disrupt the education of students, but also affect their physical, mental and social development and wellbeing. Moreover, closures restrict access to school meals, access to health services, social care and school-based vaccinations, all of which disproportionately impact children from disadvantaged backgrounds, increasing inequalities.

So in case of closure, we must provide additional support to schools in deprived areas – and for children living in vulnerable situations – to limit the impact.

The second reason is that the evidence confirms that closing schools is not the most powerful measure to control transmission:

  • A study of 32 countries showed that the largest effect among interventions was observed for stay-at-home orders and teleworking, closure of non-essential businesses and services, and bans on gatherings of 50 adults or more. As one might expect given the older ages of learners in universities and higher education establishments, closing them had a larger effect than the closure of schools at secondary, primary or preschool levels.
  • Some studies showed that transmission started to drop following other interventions before school closures were implemented, and found no change in the gradient of decline after school closures.

School closures can contribute to a reduction in transmission, but by themselves are insufficient to prevent community transmission of COVID-19 in the absence of other interventions. 

Solid fact 2: school reopening, with comprehensive infection prevention and control measures in place, and when the community infection levels were low or moderate, did not increase community transmission

  • First of all, incidence of COVID-19 in school settings is linked to levels of community transmission. It is also age dependant. Younger children (<10 years of age) are less likely to be infected than adolescents.
  • Significant secondary transmission can – and does – occur in school settings when prevention strategies are not implemented or followed. In Israel, for example, high schools were closed less than two weeks after reopening when two symptomatic students attended in-person learning, leading to 153 cases among students and 25 among staff members. Importantly, prevention strategies were not adhered to – including lifting of a mask requirement because of a heat wave and classroom crowding.
  • But a study of Italian schools, which implemented a comprehensive prevention approach, found that school reopening was not associated with the second wave of COVID-19 in Italy.
  • Comparing US county-level COVID-19 hospitalizations between counties with in-person learning and those without in-person learning found no effect of in-person school reopening on COVID-19 hospitalization rates when baseline hospitalization rates were low or moderate (75% of the counties).
  • A systematic review of observational studies showed most studies reported that school reopening, with extensive infection prevention and control measures in place and when the community infection levels were low or moderate, did not increase community transmission of SARS-CoV-2.

Data tell us the use of multiple strategies – what is called layered prevention – provides greater protection in breaking transmission chains than implementing a single strategy. Measures have to be implemented in a coherent way in classrooms, school yards, sport facilities, buses, canteens and after-school activities.

To conclude, data tell us that school reopening, in areas of low to moderate transmission and with appropriate mitigation measures, was generally not accompanied by increasing community transmission. If large outbreaks occur or transmission in the community cannot be controlled by any other measures, reactive school closures may be considered as a last resort. Bearing in mind the harmful effects, policymakers should take a measured approach before implementing school closures; and should look to reopen schools in times of low transmission, with appropriate mitigation measures.  In this context, and also because of the dynamic of the epidemics with new variants, routine testing is a very efficient way to manage the epidemic.

Solid fact 3: the mechanisms of implementation in schools involve institutional, contextual and personal factors

One year on, how to ensure that schools implement measures remains a key question. In this third section we will describe data related to the way in which infection prevention and control measures are implemented at the school level.

  • A study done by Mary Metcalfe and team in South Africa showed that the level of implementation differed from one region to another in conjunction with socioeconomic factors.
  • A survey targeting primary school principals in the UK showed that measures such as regular handwashing and stopping full-school gatherings were considered easy to implement. The most challenging measures included distancing between individuals, spacing out desks, keeping the same staff assigned to each student group and staggering break times.
  • At a global level, the UNICEF/UNESCO/World Bank survey on National Education Responses to Covid-19 School Closures found that nearly all the governments had produced or endorsed specific health and hygiene guidelines and measures for schools, but only around half of low-income and lower-middle income countries reported having enough resources to implement them, in comparison to 80 per cent of upper-middle-income countries and 95 per cent of high-income countries.

It is clear schools require further financial, human resource and other support for effective implementation of prevention measures. Schools should be supported in the improvement of their infrastructure and associated maintenance.

The success of implementation of measures in schools is totally dependent on the commitment and support of the school teams. The situation is complex:

  • A survey concerning 42 countries showed there is fear among some teachers of transmission in schools which is understandable since data show living or working with children attending in-person school is linked to a higher risk of COVID-19 outcomes.
  • Teachers often feel exhausted and undervalued.
  • And the creation of the measures is perceived to be health-led, with limited understanding and appreciation of the school context.
  • Implementation was facilitated by staff commitment and communication among stakeholders, but hampered by limitations with guidance received, physical environments, resources, parental adherence and balancing health promotion measures with learning.
  • Pre-covid research shows 3 main factors influences the adoption of new practices by teachers. Institutional, personal and contextual. In the field of health, the personal factors are major. The implementation of the guidelines is not a linear process but needs an enactment of the measures in a defined school context.

The implementation strategies must be appropriate to the setting. We must empower teachers and students in schools to build a strong culture of prevention and apply common sense to the implementation of the measures. For example, in some parts of the world it may not be sensible to ventilate classrooms without mosquito nets when the risk of transmission of vector-borne diseases is high, such as malaria or dengue fever. SARS-CoV-2 is not necessarily the greatest risk in that context for schools.

To adapt the measures to the context, local actors in health and education need to co-operate in their schools, and we must do more to break down the silos between the two sectors. To do this effectively they need consistent and timely information, backed up by resources and equipment. Testing and national vaccination strategies should ensure schoolteachers and other professionals working in schools are considered when prioritising access.

Supporting every school to integrate health issues into their management and teaching is a priority. The WHO-UNESCO Health Promoting Schools is an important new initiative to support countries in this critical area.  Such school policies should include the basic elements of:

  • protection (including maintenance of sanitary facilities, ability to contact families in the event of an emergency, traffic plans, preparation for a temporary and localised closure)
  • prevention (promoting a culture of prevention among staff, students, families, school partners, and giving information on the various possible epidemics and infectious diseases)
  • and education (including knowledge and skills related to the body, microbes, vectors, psychosocial skills and critical thinking).

These elements are common to all epidemics and are the subject of long-term work. If these principles are adopted, when future health crises occur specific prevention measures will be implemented against a background of greater understanding.


Reducing transmission in schools is a shared responsibility and needs a combination of effective prevention strategies – implemented with the ability and commitment to adhere to them.  Implementing these strategies will benefit not only our responses for COVID, but also for future possible epidemics, and will contribute to overall health and wellbeing of learners and their communities.

Strong implementation strategies based on educators’ involvement and the provision of technical and pedagogical resources in each school is necessary.

The Covid pandemic is an opportunity to invest in school systems in order to improve their resilience in case of crisis.

Didier Jourdan, Nicola Gray & Goof Buijs

More information about the high-level ministerial meeting and the recordings

The recording of Professor Didier’s presentation (starts at 44:16)

The presentation slides