The subset of 204 households participating in the CoKids [4] study consisted of 459 adults and 349 children who were recruited from the Spaarne Gasthuis (The Netherlands) and prospectively followed between August 2020 and July 2021, covering the second and third wave of the COVID-19 pandemic in the Netherlands (Fig. 1). Vaccination of children had not yet started. Upon onset of respiratory symptoms in any of the household members, a household outbreak phase was initiated and lasted a minimum of 21 days. All household members were tested for SARS-CoV-2 using a combined oropharyngeal and mid-turbinate nasal swab, which was then analysed within 48 h by reverse transcriptase polymerase chain reaction (RT-PCR). Whenever a next household member developed respiratory symptoms, that individual was promptly tested again using RT-PCR. Household outbreaks of respiratory illness were divided and classified as either SARS-CoV-2 positive or negative based on the outcome of the RT-PCR. Symptoms such as fatigue, loss of smell/taste, muscle aches, runny nose, cold shivers, dyspnea, fever, sore throat, headaches, cough, diarrhoea and vomiting were logged on a daily basis for all household members during the outbreak. Data on severity of symptoms were collected using a 5-point investigator-designed scale, with higher severity scores indicating greater symptom severity [4].
Flowchart of questionnaire response between child participants from SARS-CoV-2 positive and negative households
Not all household members reported symptoms or experienced illness during an outbreak.
All households in a confirmed SARS-CoV-2 positive outbreak were matched 1:1 to household controls from SARS-CoV-2 negative outbreaks. Matching was based on timing of the outbreak and household composition (ages and household size). These households completed questionnaires at 6 and 12 months after a household outbreak on the presence and severity of SARS-CoV-2 related symptoms, general well-being/functioning, cognition, persisting symptoms and Quality of Life (QoL), resulting in 30.4% (106/349) of the children receiving a questionnaire. The generic QoL tools used in this study are the TNO-AZL Preschool Children’s Quality of Life questionnaire (TAPQOL) questionnaire for children 0–2 years old (proxy report) and the Paediatric Quality of Life Inventory (PedSQL) questionnaire in all children aged 2–18 years old (proxy report for children < 8 years old). The PedsQL and TAPQOL questionnaires measure Health Related Quality of Life are well translated, validated questionnaires [5,6,7,8]. The TAPQOL questionnaire is used to evaluate the impact of disease on the lives of young children (from 6 months of age) in different domains (physical, social and psychological). Questions such as “Did your child sleep restlessly?” or “Did your child have a poor appetite?” could be answered with 0 “Not at all”, 1 “Rarely”, 2 “Sometimes” or 3 “Often” by parents or caregivers. The PedsQL questionnaire has four domains (Physical, Emotional, social and school functioning). Questions like “In the past week, how often has your child had problems with: playing, sleeping, running” could be answered with 0 “Not at all”, 1 “Rarely”, 2 “Sometimes”, 3 “Often” or 4 “almost always” by parents or caregivers (see Additional pdf files 1, 2, 3, 4 for the corresponding questionnaires used).
We compared presence of (persistent) clinical symptoms, general well-being/functioning, cognition, and Quality of Life (QoL) at 6 and 12Â months between children in a SARS-CoV-2 positive versus SARS-CoV-2 negative respiratory illness household outbreak.
Medians alongside interquartile range (IQR) were computed for skewed distributed variables that are continuous and medians alongside standard deviation (SD) were computed for continuous variables that are normally distributed. Categorical variables were described as numbers with their corresponding percentages.
Furthermore, we calculated the long term symptom rates, as reported in the 6Â months and 12Â months questionnaires, using the chi-square test and described the variety of symptoms in all the different domains.
Two-tailed P < 0.05 was considered statistically significant. Statistical analyses were performed using R software, version 4.0.3 (R Foundation for Statistical Computing).