key: cord-287676-qh7zeyyx
authors: Angoulvant, François; Ouldali, Naïm; Yang, David Dawei; Filser, Mathilde; Gajdos, Vincent; Rybak, Alexis; Guedj, Romain; Soussan-Banini, Valérie; Basmaci, Romain; Lefevre-Utile, Alain; Brun-Ney, Dominique; Beaujouan, Laure; Skurnik, David
title: COVID-19 pandemic: Impact caused by school closure and national lockdown on pediatric visits and admissions for viral and non-viral infections, a time series analysis
date: 2020-06-03
journal: Clin Infect Dis
DOI: 10.1093/cid/ciaa710
sha: 
doc_id: 287676
cord_uid: qh7zeyyx

A time series analysis of 871,543 pediatric emergency visits revealed that the COVID-19 lockdown and school closure were associated with a significant decrease in infectious diseases disseminated through airborne or fecal-oral transmissions: common cold, gastro-enteritis, bronchiolitis, acute otitis. No change was found for urinary tract infections.

M a n u s c r i p t

In late December 2019, patients with viral pneumonia due to an unidentified microbial agent were reported in Wuhan, Hubei province, central China. This disease outbreak, COVID-19, then grew substantially and was declared a pandemic by WHO on March 11, 2020 . [1] In 1995, a major French nationwide strike paralyzed France. For 19 days from November 30, 1995 to December 18, people stayed at home, including children who normally went to day-care centers and a significant decrease of bronchiolitis cases was observed. This decline might have been caused by workplace and school absenteeism, and lower attendance of day-care centers. [2] Likewise, during the 2013-2014 measles epidemic, a reduction in contact rate during school vacations was associated with 4900 averted cases in The Netherland. [3] After reaching France on January 24, 2020, a major progression of COVID-19 from February to March lead to public health interventions. Partial lockdown and school closure were initiated early March and a national lockdown was officially started on March 17 th , 2020. [4, 5] No previous public health intervention can be compared to the extent of the lockdown established for the COVID-19 epidemic. We hypothesized that this unusual situation in France would be associated with a sharp decrease in pediatric infectious diseases that usually disseminate through social contacts, with schools at their center. Being able to prevent these infections, responsible for many pediatric hospitalizations, would be an unwanted direct benefit of the lockdown for children, that seem otherwise significantly more protected than the adult from the SARS-CoV-2 infection. [6] Even more, this could open the road for future guidelines to control future major health issues once the COVID-19 pandemic is under control. Therefore, the evolution of several major diseases usually correlated with dissemination through contact, such as gastroenteritis, common cold, and acute otitis media, were investigated before and after the start of the national lockdown. Urinary tract infections, which are not reported to be correlated with contacts in children, were used as control outcome.

We conducted a quasi-experimental interrupted time series analysis based on multicenter prospective French surveillance data for pediatric emergency department (PED) visits and related hospital admissions. The Regional Centre of Observation and Action on Emergencies e-CERVEAU, Agence Régionale de Santé, is an official network of emergency departments dedicated to public health that automatically transmit a summary of anonymized data from all their visits to the regional database. The database has been approved by the French data protection authority. These data include discharge diagnosis coded by the physicians in A c c e p t e d M a n u s c r i p t charge of the patient at the end of the PEDs visits according to ICD-10 th revision and hospital admission or discharge. This study covers 6 PEDs from academic hospitals being part of Assistance Publique -Hôpitaux de Paris, located in and around Paris, gathering 250,000 annual visits that daily transmitted their from January 1 st , 2017 to April 19 th , 2020. We used the e-CERVEAU database for this research. Data are anonymous. Patient informed consent is not required according to current dispositions.

Groups of diagnosis extracted were: gastroenteritis, common cold, bronchiolitis, acute otitis media, considered as infectious diseases thriving through social contact, and urinary tract infection (table S1). Visits were grouped by calendar weeks for each year.

The main outcome was the evolution of the number of hospital admissions following the French decision to close schools and start a lockdown for the whole country. [5] The secondary outcomes were the number of PEDs visits for gastroenteritis, bronchiolitis, common cold, acute otitis media. As recommended to prevent potential confusion, [7, 8] urinary tract infections were analyzed as a control outcome, given that this common pediatric infectious disease is not expected to be impacted by social distancing, although indirect effect such as stress or diet change cannot be excluded. This outcome was already used as a control concerning previous acute respiratory tract infections studies. [9] Statistical analysis Outcomes were analyzed by quasi-poisson regression, accounting for seasonality, secular trend before and after lockdown, and overdispersion of data. [7, 8, 10, 11] Seasonality was taken into account by including harmonic terms (sines and cosines) with 12-months and 6months periods to adjust for the seasonal pattern. [11] The time unit chosen was one week to provide optimal precision to the model. [8] We hypothesized that the intervention would have an immediate impact, meaning after one analysis. Analysis of acute gastroenteritis was performed on data from four hospitals combining 81% of the visits during the study period. All statistical tests were two-sided, and we considered a result as "significant" when the p-value was <0.05. All statistical analysis involved using R v3.6.1 (http://www.R-project.org).

A total of 871,543 PEDs visits in the six participating centers from January 1 st , 2017 to April (Table S2 , Figure 1 and Figure S1 ).

We found a significant decrease of acute gastro-enteritis, common cold and acute otitis media (Table S2, Table S3 ) with a sharp decrease over 70% compared to the expected values (Figures 1b-d) . Decrease of bronchiolitis was also significant (-63.5% 95CI [-101.8 ; -25.9]) (Table S2) .

Consistent both with our hypothesis and the incubation time of these different diseases, a dramatic decrease of overall PEDs visits (-68.5%) and hospitalization (-44.7%) was observed as soon as one week after the start of the lockdown ( Figure S1 ). By contrast, urinary tract In this time series analysis of 871,543 PEDs visits, the number of PEDs visits and admission after the lockdown decreased by -68% and -45%, respectively. We found a significant decrease over 70% of acute gastro-enteritis, common cold, bronchiolitis and acute otitis media compared to the expected values. Unprecedented public health interventions were ordered to reduce the risk of SARS-CoV-2 transmission. [4] Our data suggest that these measures have also a critical impact on the transmission of numerous infectious diseases, more specifically on viral or viral-induced pediatric diseases. This major achievement may also play a critical role in making more health resources with adults admitted in ICU and health care workers available to fight COVID-19 pandemic. [12] In Paris area, children with acute illness could be seen not only in PED but also in private office (general practitioners or pediatrician), general practitioner house calls, and communities' center. However most of these sites of care do not work 24/7 and rarely perform additional tests especially for the younger children.

Our study have limitations; we cannot exclude a change in clinical management such as avoidance of ENT examination because of COVID-19 fear which could have influenced diagnosis coding; we did not collect data regarding severity and so we cannot exclude that reduction in presentations was associated with children presenting later in their illness. While the dramatic decrease in PEDs could be partially due to transportation limitations and a fear of going to the hospital and increase of telemedicine, the stability in the number of urinary tract infections cases that we used as control outcome and the significant decrease in hospital admissions do not favor this hypothesis. Moreover visits to private doctors' offices decrease by 40%. [13] While the role of the children in the dissemination of SARS-CoV-2 is still being discussed, finding that school closure and national lockdown were linked to a dramatic decrease in pediatric emergency visits concerning gastroenteritis, acute otitis media, bronchiolitis, and other viral diseases could be not only an unexpected benefit for the children but could also raise the question of the impact on the health care system of starting lifting the French national lockdown by reopening the schools.

[14] Complementary studies using more granular data such as severity could be useful to better understand lockdown's impact on children's health.

This nation-level quasi-experiment is unprecedented in the modern era. It provides unique evidence which could be key in the post COVID-19 era, to implement new guidelines and new routines in our way of life, and in order to fight past but also potential future infectious diseases threats reaching both children and adults. M a n u s c r i p t Impact of lockdown on weekly pediatric emergency department visits and major pediatric infectious diseases, from January 1 st , 2017 to April 19 th , 2020.

-a: Overall PEDs visit (n=871,543). The black line shows the observed data. The bold red slope shows the model estimates based on observed data (quasi-poisson regression modeling). The bold blue slope shows the expected values without lockdown in the post-intervention period (quasi-poisson regression modeling). The start of the lockdown is indicated by the vertical black arrow.

A c c e p t e d M a n u s c r i p t Figure 1 

World Health Organization. WHO Director-General's opening remarks at the media briefing on COVID-19 -11

Coincidence of public transport strike with bronchiolitis epidemic

The Reduction of Measles Transmission During School Vacations

Adaptation of the National Plan for the Prevention and Fight Against Pandemic Influenza to the 2020 COVID-19 Epidemic in France. Disaster medicine and public health preparedness

Décret n° 2020-260 du 16 mars 2020 portant réglementation des déplacements dans le cadre de la lutte contre la propagation du virus covid-19

MMWR Morbidity and mortality weekly report

Interrupted time series analysis in drug utilization research is increasing: systematic review and recommendations

Segmented regression analysis of interrupted time series studies in medication use research

Impact of Implementing National Guidelines on Antibiotic Prescriptions for Acute Respiratory Tract Infections in Pediatric Emergency Departments: An Interrupted Time Series Analysis. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America

Regression based quasiexperimental approach when randomisation is not an option: interrupted time series analysis

Interrupted time series regression for the evaluation of public health interventions: a tutorial

Paediatric Intensive Care Society. PICS and ICS Joint Position Statement

14. French Government. French Government measures

M a n u s c r i p t A c c e p t e d M a n u s c r i p t