Determinants of school absences due to respiratory tract infections among children during COVID-19 pandemic: A cross-sectional study of the Sentinel Schools Network

Title: Determinants of school absences due to respiratory tract infections among children during COVID-19 pandemic: a cross-sectional study of the Sentinel Schools Network.Antoni Soriano-Arandes1,●,○, Andreu Colom-Cadena2,3,●, Anna Bordas2,3, Fabiana Ganem2,3,4, Lucia Alonso3,5, Marcos Montoro2,3, Mireia Gascon6,7,8, Maria Subirana6,7,8, Ariadna Mas9, Jordi Sunyer6,7,8, Pere Soler-Palacin1, Jordi Casabona2,3,4,8 ; on behalf of Sentinel School Network of Catalonia*Paediatric Infectious Diseases and Immunodeficiencies Unit. Children’s Hospital. Vall d’Hebron Barcelona Hospital Campus, Barcelona, Catalonia, Spain.Centre d’Estudis Epidemiològics sobre les ITS i Sida de Catalunya (CEEISCAT). Badalona, Spain.Institut d’Investigació Germans Trias i Pujol (IGTP), Badalona, Spain.Departament de Pediatria, d’Obstetrícia i Ginecologia i de Medicina Preventiva i de Salut Publica, Universitat Autònoma de Barcelona, Bellaterra, Spain.Fundació Lluita contra les infeccions, Badalona, Spain.ISGlobal, Barcelona. Spain.Universitat Pompeu Fabra (UPF), Barcelona, SpainCentro de Investigación Biomédica en Red de Epidemiología y Salud Pública (CIBERESP) Instituto de Salud Carlos III, Madrid, Spain.Direcció Assistencial d’Atenció Primària i Comunitària, Institut Català de la Salut, Barcelona, Catalonia, Spain● Contributed equally to this work with: Antoni Soriano-Arandes, Andreu Colom-Cadena○ Corresponding author: Antoni Soriano-Arandes,toni.soriano@vallhebron.cat. ORCID: 0000-0001-9613-7228Keywords: COVID-19, respiratory infections, ventilation, school, CO2, Indoor air qualityTo the editor:Different public health measures were implemented during the COVID-19 pandemic to maintain the schools open1,2. Regarding COVID-19 vaccination, at the beginning of this study, more than 92% and 40% of population older than 12 years and children aged 5-11 years in Spain was fully vaccinated against COVID-19, respectively3.Ventilation was proposed as one of the main strategies to reduce aerosol transmission for SARS-CoV-2 during the pandemic. A study from Germany assessed the efficiency of air purifiers in reducing aerosols in high-school classrooms4. However, no clinical endpoints were investigated and, therefore, a very low certainty of evidence was graded by a Cochrane review5. In an Italian study, the authors concluded that ventilation reduced the likelihood of SARS-CoV-2 infection in classrooms by 80%6. Moreover, a survey study from the US reported that elementary schools with face masks and ventilation strategies in place had lower SARS-CoV-2 incidence rates7. In a Lancet Task Force Commission review8, the authors concluded that improving building ventilation systems may carry benefits beyond protection from COVID-19. However, there is scarce evidence that demonstrate the association of epidemiological and environmental risk factors, including CO2 concentrations, with the incidence of respiratory tract infections (RTI) in children who are attending schools.The present study is part of COVID-19 Sentinel Schools Network of Catalonia (CSSNC), a project including 23 schools and 5687 students, teachers and other school staff with the aim to monitor SARS-CoV-2 and other respiratory viruses, their determinants and preventive measures in Catalonia9. The objectives of this study are to assess the potential association of demographic and epidemiological factors, including the indoor environmental conditions (CO2levels) in the classrooms, with students’ school absence due to RTI.We conducted a cross-sectional study in two different periods, the first from April 19 to June 21, 2022, and the second from November 8 to December 21, 2022. The main outcome of the study was the absence of the child from attending the school in person due to RTI or other non-respiratory medical causes. The study population was composed of 253 students (4-11 years) belonging to 20 classrooms, 11 in spring 2022 and 9 in autumn 2022, attending 4 and 16 classrooms of preschool (4 to 5 years-old) and primary school (6 to 11 years-old) stages, respectively. A written informed consent was previously obtained from parents or guardians of children.Absences from school attendance were notified by the tutor/teacher to the study researchers. We proceed to collect epidemiological and clinical/diagnostic data within the first 48 hours through a case report form deposited on the digital platform REDCap©. All the absences (cases) were followed-up through telephone calls made by health professionals until their return to the school. RTI was registered according to the symptoms described during the telephonic interview and confirmed through the computerised health record programme of the Health Ministry (eCAP).Sensors to monitor the CO2 concentrations (DIOXCARE DX700 PDF, Smartcare Services, Spain) were installed in classrooms, recording data every 10 minutes. For the analysis we only used the measurements taken during the time in which the students were in the classroom. We obtained a median [IQR] of CO2 levels which was used to conduct our analyses. All the participating schools received an operational protocol to install the sensor correctly and download data weekly.We calculated the RTI incidence per week and for the total study period. We performed a descriptive analysis of all survey variables, stratifying them by respiratory-related absence or absence due to other causes. Additionally, we ran a univariate logistic model to study the association between the type of absence and each described variable to obtain the corresponding odds ratios, 95% confidence intervals and p-values.Furthermore, we conducted a Latent Class Analysis (LCA) to explore potential groups of students with similar symptomatology. We started exploring the optimal number of latent classes, trying two to five classes, and we selected the best model using the entropy criterion, which indicates the accuracy of the latent classes, combined with other goodness of fit criteria such as BIC, cAIC and likelihood ratio. All these measures suggested that the optimal number of classes was two. However, we also examined the rest of number of classes to see if their classification had more clinical significance. All statistical analyses were performed in R (version 4.2.2).We registered one-hundred and five school absences during the study, 43 in the first study period (spring 2022), and 62 in the last one (autumn 2022) (Figure 1S ). Among these, we could obtain data related to the diagnosis in 98 (93%) cases, the rest of them were lost during the follow-up of the absence. Seventy-one absences were respiratory-related and 27 were due to other causes, mainly with gastrointestinal symptoms (66.7%). Among the absences due to RTI, the most were upper RTI (56/71, 78.9%), and only six (8.5%) were confirmed infections, 2 caused by influenza virus and 4 due to SARS-CoV-2 infection. These results represent a RTI incidence of 15.9 and 33.1 cases per 100 population in the first and second study period, respectively. The maximum weekly RTI incidence was of 10.2 cases per 100 population in December 12-18, 2022.The clinical, epidemiological and environmental characteristics of study sample are summarized in table 1 , categorized by type of absence (respiratory versus non-respiratory). We found a statistically significant association with absences due to RTI when someone else at home had respiratory symptoms (OR=9.12, CI 95%=2.54-33.39). We found a positive association between higher median levels of CO2at class and respiratory-related absences (OR=1.2, CI 95%=0.98-1.46). Moreover, there were more respiratory absences in autumn (OR=2.4, CI 95%=0.97-5.94). However, these last two associations did not reach statistical significance at a level of 0.05; although their p-values were lower than 0.1.No other epidemiological risk factors were associated with RTI incidence, such as household floor level, number of people living at home, living with smokers, having any comorbidity or being vaccinated against COVID-19.Finally, in Figure 1 , we present the symptomatology of absences belonging to each of the two latent classes. We can see that in the first cluster, the most frequent symptoms were cough, nasal congestion and fever, whereas in the second one, fever, gastrointestinal symptoms, and fatigue predominated.Our findings confirm that the RTI incidence during the study period was very high in children attending in the CNSSC schools, and the main medical cause of school absence. Although differences on school absence due to RTI were observed between spring and autumn, they were not statistically significant and in any case they may be due to the respiratory viruses’ seasonal pattern. In fact, the most important and significantly associated risk factor for RTI was the presence of someone else at home with respiratory symptoms, suggesting that households could be the main setting for initiating of the transmission of RTI.We observed a slightly association between RTI and median of CO2 levels in classrooms (p=0.07), which is an indicator of the degree of ventilation. However, we cannot exclude other potential factors such as rainfall, ambient temperature, or air pollutants (e.g. PM2.5, NO2, etc.) influencing on this outcome, as suggested by other authors10. To our knowledge, previous studies assessed the CO2concentration as a proxy of ventilation to evaluate the risk transmission of SARS-CoV-2 in schools11,12, but they did not analyse the association between CO2 median values and RTI incidence.Finally, we studied the symptomatology associated with the school absences through a LCA. The best approach to differentiate RTI from other causes was using two latent classes, and the most frequent symptoms were cough, nasal congestion and fever.The major strength of this study is our extensive data collection on clinical, epidemiological and environmental factors related to the school and also to the households of the participants. However, there were limitations such as possible incomplete reporting of RTI or insufficient sample size to determine small effect sizes. CO2 concentration was only measured in a selection of classrooms per school, so it may not be representative for the entire study period and school.In conclusion, RTI incidence was very high during the study period being the most important and significantly associated factor with RTI to have anyone else at home with respiratory symptoms. This suggests that households and not schools could be the key epidemiological factor for initiating the transmission of RTI to the children. Improving household preventive measures could reduce childhood RTI. In the LCA, the most frequent symptoms associated with RTI were cough, nasal congestion and fever. Although we found a slightly association between RTI and reduced ventilation we cannot exclude other potential factors influencing on this outcome. The study has been crucial to assess the feasibility and potential utility of collecting both school absence and morbidity data for further developing a systematic monitoring system.