High Rates of SARS-CoV-2 Family Transmission in Children of Healthcare Workers During the First Pandemic Wave in Madrid, Spain

Serologic Study

Abstract

Background: 

Data on SARS-CoV-2 transmission among children living with healthcare workers (HCWs) are scarce.

Methods: 

A cross-sectional study was performed at a tertiary Hospital in Madrid, including children of HCW who suffered from SARS-CoV-2 infection between March and May 2020. Children underwent enzyme-linked immunosorbent serological study for detecting SARS-CoV-2 antibodies: VIRCELL IgG assay.

Results: 

One hundred thirteen children from 69 HCWs with confirmed SARS-CoV-2 infection were recruited: 47 children had positive IgG (41.6%). Children secondary attack rate was 43.7% (25% if both parents have had asymptomatic infection; 39.5% if one parent was symptomatic; and 47% when both parents had symptoms). Having a positive sibling was associated with a positive IgG result (odds ratio = 12.2; 95% confidence interval: 4.4–33.7, P < 0.001). Median age was higher in IgG positive children (P = 0.022). Children who referred anosmia presented higher IgG titles (P < 0.04).

Conclusions: 

We observed a very high SARS-CoV-2 transmission in children of HCW during the first pandemic wave, especially when both parents were symptomatic. Having a positive sibling was associated with seroconversion, supporting the important role of family clusters in the transmission of SARS-CoV-2.

The role of children in the transmission of SARS-CoV-2 and their susceptibility to have mild or asymptomatic infections has been object of intense debate. Previous studies have demonstrated lower seropositive rates in children 5–9 years old compared with adults.¹ A seroprevalence study performed in Spain after the first wave showed a global seroprevalence of 5%, being 1%–3% in children, with an age-dependent pattern for antibody detection (from 1.1% in infants to 3.8% in adolescents).² A multicenter study performed in the United Kingdom has recently demonstrated high seroprevalence rates (6.9%) in healthcare workers’ (HCWs) offsprings, although the authors did not observe differences in the rates of seropositivity depending on children’s age.³

Although the information available is still scarce, studies analyzing the transmission dynamics within the family suggest that most children are secondary cases, infected in the context of infection of a family member, usually an adult.^3,4

In Spain, the prevalence of infection was extremely high among HCW, compared with other countries. During the first pandemic wave, up to 26.3% of the reported cases in our country were HCW.⁵ In a context of school closures and strict lockdown as part of a state of alarm, we addressed family transmission dynamics in this high-risk environment by means of a serologic retrospective study in HCW children, immediately after the first pandemic wave.

PATIENTS AND METHODS

A cross-sectional study was carried out at a tertiary University Pediatric Hospital in Madrid, including children of HCW who suffered from SARS-CoV-2 infection [confirmed by polymerase chain reaction (PCR) and/or serology] between March and May 2020, during the national lockdown. Children were recruited if they had cohabited with their parents during the lockdown. The study was offered to all the HCW of the hospital through electronic communication.

The Ethics Committee approved the study (PI-4324) and an informed consent/assent form was signed by parents/legal guardians and by all patients older than 12 years of age.

Epidemiologic and clinical data were collected through a brief survey specifically designed for this study as well as microbiologic results (PCR and/or serology).

All participants underwent a serologic study by enzyme-linked immunosorbent assay (ELISA) for detection of serum IgG antibodies against SARS-CoV-2 (VIRCELL COVID-19 ELISA IgG). Results were given measuring the optical density cutoff points for IgG, considering positive values >6.

Secondary attack rate (SAR) in children was defined as the proportion of infected children following exposure to a primary symptomatic adult index case. Children were considered infected if they had a positive SARS-CoV-2 PCR after the exposure, or if they presented a positive SARS-CoV-2 IgG, determined at least 2–4 weeks after the exposure.

Qualitative data were expressed as absolute and relative frequencies and quantitative data as median and interquartile range (IQR). Categorical variables were compared using χ² and Fisher exact tests, and continuous variables with Student t test or nonparametric tests as appropriate. A 2-tailed value of P <0.05 was considered statistically significant. All analyses were performed using the Statistical Package for the Social Sciences, version 21.0 (IBM Corp., CA).

RESULTS

Sixty-nine HCWs with SARS-CoV-2 infection and their household members were recruited (63 family clusters and 113 children). Infection of HCW was confirmed in 31% by PCR, 29% by serology and 38% using both methods. In 10 families (15.8%), both parents had confirmed infection. The median number of children per family was 2 (range 1–3). The mean age of the children was 8.4 (SD 4.5 years), 40.7% were boys.

A total of 73 (64.6%) children reported symptoms compatible with SARS-CoV-2 infection, although only in 5 of these children PCR was performed, confirming the infection in one. While all of them reported mild symptoms, only 28 (24.8%) had fever. Children presented symptoms at a median of 9 days (IQR 3–14.5) after their parents, except in 9 cases from 6 families (7.9%), in which children developed symptoms 12 days (IQR 6.5–18) before their parents.

Forty-seven of 113 children (41.6%) had positive IgG. The serologic study was carried out at a median of 11.2 weeks (IQR 9–12.28) after the index case developed symptoms compatible with COVID-19.

The SAR among children was calculated excluding those 9 children who presented with symptoms before their parents, as well as 8 children whose parents had not been clinically symptomatic. Children SAR in our cohort was 43.7% (42/96). The proportion of children who became infected increased depending on whether one parent or both parents were symptomatic (SAR 39.5%; 17/43 and 47%; 25/53, respectively)

In relation to the number of cohabitants with confirmed infection, 70% of children whose both parents had confirmed infection had positive IgG serology versus 36% of children with only one parent infected (P = 0.046). Having a positive sibling was associated with a positive IgG result (odds ratio = 12.2, 95% confidence interval: 4.4–33.7, P < 0.001), with 77.8% of children with a positive sibling obtaining a positive IgG result. Equally, 77.8% of the siblings of negative children were IgG negative (Fig. 1).

FIGURE 1.: Description of IgG results from 113 children of 63 SARS-CoV-2 infected healthcare workers (H). Grey color corresponds to negative IgG results and black color to positive ones. The absence of a family member inside a given household cluster is shown with a crossed out box. *Infected adult (confirmed by PCR and/or serology). ◊ Symptomatic.

Median age was higher in IgG positive children 9.7 years (6.1–13.2) versus 7.2 (4.2–11.3) (P = 0.022). Among the symptoms reported by children, only anosmia, which was present in four cases, was associated with a positive IgG (P = 0.016); all patients reporting anosmia seroconverted. A total of 15/46 children with positive IgG were asymptomatic (32.6%).

Children with anosmia presented higher IgG titers [30.7 (IQR = 10.4–81.3) vs 4.5 (IQR = 2.2–22.2), P < 0.04].

DISCUSSION

In this study, we observed a high SARS-CoV-2 family transmission rate among children of HCW during the first pandemic wave, especially in cases were both parents were symptomatic. Seroconversion was observed in >40% of children, increasing to 70% if both parents were symptomatic. Despite the high transmission rate, children had mild disease or were asymptomatic.

Transmission dynamics of SARS-CoV-2 in families suggests that most children are infected secondary to the infection of an adult from the family.^4,6,7 Household contact represents the highest risk for SARS-CoV-2 infection,⁸ and compared with other contacts, intrafamiliar transmission is up to 10 times higher.^9,10

Recent systematic reviews and metanalysis report heterogeneous SAR, with a pooled rate of 27%,^8,9 being lower in children, ranging between 4% and 30%.^8,9 In our cohort, we have observed a surprisingly high SAR, especially in children with both symptomatic parents. Although scarce data are available regarding children in high-risk environments, high seropositive rates among family members living with symptomatic front-line essential workers have been described.¹¹ The higher seroconversion rate in our cohort might be related to children´s exposures to higher viral loads, especially when both parents were symptomatic, as suggested by Bird et al¹² Our findings are in line with previous reports suggesting a higher infection rate among children when the index case is a symptomatic HCW.^8,12 Having a positive sibling was associated with seroconversion, showing family clusters. Genetic predisposition could play a role in SARS-CoV-2 transmission.¹³ Polymorphisms of angiotensin-converting enzyme 2 could influence infectivity and pathogenesis of SARS-CoV-2.¹³ These data suggest a genetic predisposition to the infection, and a special susceptibility among individuals from the same family, reflected in our family clusters. Nevertheless, the type of confinement measures established in each household, as well as its level of compliance, is likely to have affected the rate of transmission in each cluster.

Older children presented with higher rate of seroconversion. Numerous studies have highlighted lower rates of seroconversion in younger children.^1,2,6,7,10 This patient group is known to have a lower angiotensin-converting enzyme 2 expression in their nasal epithelium,¹⁰ as well as a particularly active innate immune response at the upper respiratory tract,^14–16 which is believed to be protective against SARS-CoV-2 infection. These facts could explain the lower rates of seroconversion observed in our younger children, although other causes might play a role in this low SARS-CoV-2 transmission, such as immunizations or the interference of other respiratory viruses.^15,17

Although up to 70% of children referred symptoms, seroconversion was only confirmed in 40% of our cohort. A recent seroprevalence multicenter observational cohort in HCW children from the United Kingdom showed a higher prevalence of asymptomatic infection (50%) compared with our cohort (32%).³ In the UK study, children from HCW with or without confirmed SARS-CoV-2 infection were recruited, and serology was taken across different areas in a 3-month period.³ It is plausible that not all subjects should have had a similar exposure level. In this regard, in our cohort, only children of HCW with confirmed infection were included during a short recruitment period, just immediately after the lockdown finished reducing external sources of exposure.

As the infection was not confirmed by PCR in most of the children included in our study, we are not able to confirm if all these children were infected. Furthermore, a negative SARS-CoV-2 antibody result does not rule out SARS-CoV-2 infection.^18,19 Patients with mild symptoms could have a weaker immune response, with lower IgG titers.²⁰ In addition, some patients might locally control their infection through their innate immunity, without generating antibodies.^16,21 Innate immune mechanisms, pre-existing cross-reactive immunity mediated by previous exposure to other human coronavirus and T-cell immunity could be major determinants of protective immunity and clinical outcomes differences beyond antibody responses in SARS-CoV-2 infected individuals and between different age groups.^14–17

We evaluated possible correlations between antibody detection in children and symptoms. Like the UK report, anosmia was statistically correlated with seroconversion, but not other symptoms, including gastrointestinal symptoms or fever.³

Interestingly, children with anosmia presented with higher IgG titles. Anosmia has been related with better clinical outcome and patients with loss of smell are 10 times less likely to be hospital admitted.²² People with SARS-CoV-2 infection and signs of olfactory dysfunctions could present a greater local immune response and may represent those individuals with faster and stronger immune response against the infection.^22,23

Our study has several limitations. Because of the retrospective design, we were unable to guarantee that the HWC was the index case in the household. Serology was positive in many asymptomatic children. Thus, judging whether adults or children infection came first is impossible. However, Spain went into a national lockdown on March 14, 2020 until June 21, 2020, with extreme social-distancing. Schools were closed on March 11, 2020, and children were not allowed to leave their homes, even for a short walk. Considering these circumstances, it is unlikely that children were infected outside their household and similarly, very unlikely that children have infected their parents, being health-workers attending COVID-19 patients. Some adult index cases´ infections were confirmed only with a positive IgG (29%), and not by PCR. SARS-CoV-2 infection in children was not confirmed by PCR. A recent metanalysis has reported good sensitivity and specificity of ELISA serologic test,²⁴ especially performed after 4 weeks of the infection. Finally, the results of the applied antibody assays in this study are quantitative only; currently, different approaches are under investigation to measure the functionality and efficacy of antibodies generated during or post SARS-CoV-2 infection.²⁵

In summary, in this study, high rates of SARS-CoV-2 infection among children living in high-risk environments were observed, most of them presenting with mild disease or indeed being asymptomatic. Further studies are needed to elucidate the role of children in the dynamics of family transmission of SARS-CoV-2.