When a pandemic causes school closures it has wide-ranging impacts beyond public health: our logic model can help in decision-making

James Thomas, Alison O’Mara-Eves, Dylan Kneale and Rebecca Rees.

The closure of schools has been a recommended intervention in response to pandemics because of its potential for reducing the transmission of infection among children, school staff, and those they contact. Previous evidence has shown that closing schools can have the intended effect of reducing infection rates, although factors such as the timing and length of the closures are likely to be important.

The current crisis, however, has highlighted that existing evidence and debates are insufficient. They have been largely focused on the impacts on transmission and health services, with less consideration of other downstream effects.

That is why a group of social scientists has come together to explore all possible outcomes. Here we describe our approach to presenting a logical way to consider the impact of school closures on individuals, families, education and health systems, and the broader economy. This is covered in detail in our paper published today by F1000Research and we now seek feedbackon this systems-based logic model.

Why we thought a logic model was needed

In this current pandemic we have seen an unprecedented number of school students (more than 90% of the world’s student population, according to UNESCO) affected by school closures for potentially unprecedented lengths of time. There are likely to be unintended impacts that need to be considered and, where possible, mitigated.

We thought that it would be useful to integrate formally the impacts of pandemic-related school closures that have been considered within existing research literature into a comprehensive visual representation (a logic model). This could help to ensure that a broad range of possible impacts are considered in decision-making, in impact mitigation strategies, and in future research. By presenting a systems-based logic model, we hope to open up the conversation about what needs to be considered in decisions about reopening schools—and closing schools in future pandemics or subsequent waves of the current one.

The logic model of school closure impacts

The development of this systems-based logic model followed a systematic plan in order to ensure that: a) it represents the state of current knowledge and thinking on the subject; and b) that it is rooted in empirical research and not solely the prior opinions of the authors. We ultimately consulted 177 papers to develop the model, which contains more than 100 concepts.

The graphical depiction of the logic model can be found in the paper. An interactive version can be found on the Miro board, here and below is a screenshot of the model.

Concepts and relationships that were identified in the research literature are coloured in bold with purple connecting lines. Concepts identified through team discussion and their engagement with wider literature (e.g., stories in the media) are depicted in italics with green connectors.

These concepts are organised into seven higher level domains: children’s health and wellbeing, children’s education, impacts on teachers and other school staff, the school organisation, considerations for parents and families, public health considerations, and broader economic impacts. In addition, we identified many factors that might change the impact that school closures have on a given outcome. These effect moderators were organised under five headings:

• pandemic related factors,
• closure related factors,
• child related factors,
• social / political factors, and
• environmental factors.

The final component in the logic model is the presence of the ethical principles that apply throughout school closure processes and which are embedded in decision-making.

We present this model as an aid to thinking about what might be important, rather than suggesting that school closures will necessarily lead to any or all of the proposed impacts. We have not made any attempt to quantify the strength of the relationships between concepts, for reasons including the extreme complexity of the relationships; the idiosyncratic nature of pandemics and the closure implementations; the moderating factors; and the individual responses to and experiences of pandemic policy interventions.

To start quantifying the relationships, we would first need to conduct a systematic review or further research to explore the strength of the relationships for each factor or group of factors. This could be done, but it is a big job. Our logic model could help to identify priority areas for in-depth analysis in the context of the current pandemic, making this task more efficient.

Second, we would need the full contextual information of a given pandemic scenario. The relative importance of each relationship is likely to be heavily influenced by the specifics of a given pandemic and school closure scenario: critical factors in one pandemic might be fairly negligible in another. Our logic model would make a good starting point for determining what will be more or less relevant in a given context.

We searched the research literature and then supplemented the model with additional hypothesised factors noted in the media, policy documents, or anecdotal evidence (e.g., the experience of the researchers). We acknowledge that, just as the original collation from research evidence was incomplete and needed supplementation from the research team, the model that we now present is likely to also have several gaps. We need your help to fill these gaps.

We have submitted this paper to F1000Research because of the time-sensitive nature of this work. If you would like to help us develop this model further, you can by commenting on this blog or on the article page – or you can contact us directly. If we are to develop the model to feed into discussions about schools reopening and – possibly – decisions relating to future waves of this pandemic, then we need your help to fill these gaps now.