Unlike infectious diseases, respiratory disease emergencies rose after compulsory school attendance following the French COVID-19 lockdown.

paediatric emergency department

The COVID‐19 pandemic requires exceptional measures, such as national lockdowns. Several studies have shown important reductions in paediatric emergency department (PED) visits during lockdowns, , and an Irish study reported that changes in PED attendance continued after restrictions were eased. Better knowledge about how our PED was used during lockdown and as services started reopening can help decision‐makers choose restrictions and adapt care services. Our aims were to describe the evolution of PED visits since the French lockdown, shortly after COVID‐19 was declared a pandemic, and to compare this to the estimates produced by predictive models that were based on data from before the pandemic.

We conducted a retrospective study from 1 January 2017 to 30 September 2020 in the PED of our French paediatric university hospital, which normally receives 40,000 visits per year. This is in addition to the visits made to the paediatric primary care services provided family doctors and paediatricians, which are funded by the social security system. The national French lockdown was from 17 March to 10 May. Schools began opening on 11 May and were compulsory from 22 June. Masks were compulsory for secondary school children, from the age of 11, when schools reopened. The summer holidays were from 4 July to 31 August. We split each month into two periods: days 1–15 and days 16 to each month end. The discharge diagnoses were reviewed and grouped into trauma, infectious diseases excluding asthma, respiratory diseases, mental health, intoxication and burns. We used SAS version 9.4 (SAS Institute, North Carolina, USA) for the data analysis. The Holt‐Winters seasonal method was used to calculate the number of predicted PED visits for each half‐month until 30 September, based on the evolution of the numbers of observed PED visits from 1 January 2017 to 15 March 2020. All dates refer to 2020 unless otherwise stated. Ethical approval was not required, because the study analysed routinely collected data.

During lockdown, there were 3227 PED visits, which was 60% lower than the predicted number of 7519 visits. They remained below the predictions until 31 July. Figures S1 and S2 show the evolution of the predicted and actual number of PED visits. A higher proportion of boys attended the PED from 16 March to 30 September (57.0%) than from 1 January 2017 to 15 March 2020 (55.3%). The patients’ median age was 4.2 years, with an interquartile range of 1.5–9.1, for the whole study period. Children under one year of age accounted for 17.7% of visits between 1 January 2017 and 15 March and then fell to 15.4%.

From 17 March to 30 September, 1662/14,701 of the children who visited the PED were virus tested and 25 (1.5%) were positive.

The overall proportion of children who were hospitalised was significantly higher than predicted from 16 March to 15 May: The actual versus predicted values, with 95% confidence intervals (CI), for each half‐month during this period were 33.2% versus 22.6% (95% CI 19.7–25.5), 31.4% versus 22.6% (95% CI 19.3–26.0), 36.3% versus 25.6% (95% CI 21.7–29.5) and 28.7% versus 23.6% (95% CI 19.6–27.6). However, hospitalisations then decreased from 16 March to 30 June, for each half‐month during this period: 264 versus 444.6 (95% CI 368.4–520.8), 234 versus 409.7 (95% CI 332.5–486.9), 282 versus 402.7 (95% CI 324.4–480.9), 261 versus 399.0 (95% CI 319.6–478.3), 290 versus 441.2 (95% CI 359.8–522.6), 304 versus 401.5 (95% CI 320.0–482.9) and 325 versus 422.4 (95% CI 339.3–505.6).

The actual percentage of urgent PED visits was higher from 17 March to 10 May (90.5%) than from 11 May to 30 September (86.3%).

The number of PED visits for trauma was significantly lower than predicted from 16 March to 15 June. Visits for infectious disease decreased significantly from 16 March and thereafter, except for 1 to 15 September. The number of PED visits for respiratory disease decreased significantly from 1 April to 30 June. The number of PED visits for mental health only decreased from 16 March to 15 April, when compared with the predicted numbers. PED visits for burns and intoxication were not statistically different from the predicted numbers, except for 16 to 30 April, when PED visits for intoxication decreased.

The important decrease in total PED visits in our hospital during lockdown was due to the decrease in trauma and infectious and respiratory diseases. In the 10 weeks that followed lockdown, the number of PED visits was always below the predicted values. This was explained by the persistent decrease in PED visits for infectious diseases excluding asthma, due to protective measures, such as hand hygiene. Asthma was excluded from infectious diseases, because of difficulties in ensuring that the trigger was infectious, but infection is a common cause of asthma exacerbation. The number of PED visits for respiratory diseases was similar to the predictions from 1 July, which was 10 days after school attendance became compulsory. This may have been because mask wearing was not compulsory for children in primary school. Likewise, the decrease in the number of PED visits for trauma corresponded to the lockdown period. Recreational and sporting activities in school and leisure centres could have been responsible for the increase in trauma after lockdown ended. Emerging literature has suggested an increase in mental health problems since the beginning of the pandemic, but this was not observed by our study. This impact may be seen later, when people realise that the pandemic will impact their life for a long time. Finally, the rate of hospitalisation increased during lockdown, the percentage of non‐urgent visits was low, but the number of hospitalisations stayed below the predicted values. Caregivers probably took children to the PED when it was really necessary, but this fear of infection also risked delays when children were very sick.

PED visits for trauma and respiratory diseases were lower than predicted when French children were in lockdown. However, predicted and actual respiratory problems were similar 10 days after school attendance became compulsory. We suggest that this was because children in primary school were not required to wear masks.

CONFLICT OF INTEREST

None.

Fig S1

Click here for additional data file.

Fig S2

Click here for additional data file.