Key Points

  • The seroprevalence of SARS-CoV-2 in children in July 2022 more than doubled when compared with the seroprevalence in January 2022, similar to the trajectory of adult infection seroprevalence results .

  • The proportion of children having evidence of antibodies to the spike protein varied by age group, with the 0-4 year old having lower seroprevalence compared with 5-12 and 13-17 year olds.

  • The numbers in each age group are small, particularly the 13-17 years olds, so there is some uncertainty around the estimates. By early July, the overall (S+) seroprevalence was was 61.7%, (95% CI 38.8, 80.6), 91.7%, (95% CI 76.6, 97.9) and 96.6%, (95%CI 79, 100) for 0-4 years, 5-12 years, 13-17 years respectively.

  • In early July, approximately 8 in 10 children aged 5-17 years had evidence of previous infection (S+ and N+), with seroprevalence estimates of 81.6% (95% CI 64.9, 91.6) and 82.2% (95% CI 61.7, 93.1) for children aged 5-12 and 13-17 years respectively.

  • The seroprevalence estimates indicate infection at any time point and thus re-infection is not captured in the data. In addition, antibody levels often wane over time. These two features imply that not all infections may be measured using this data.

  • Work is ongoing to improve understanding of what antibody levels mean in terms of protection against COVID-19. Current thinking is that there is no threshold antibody level that offers complete protection against infection, but instead that higher antibody levels are likely to be associated with lower probability of infection.

Background

The National Serosurveillance Programme (NSP) is led by the HPSC Seroepidemiology Unit (SEU), working in partnership with the University College Dublin (UCD) National Virus Reference Laboratory (NVRL) Serosurveillance Unit and the acute-hospital Laboratory Surveillance Network (LSN). It is overseen by a national multi-disciplinary and multi-sectoral Steering Committee.

The NSP conducts systematic sampling of residual specimens with the aim of estimating the proportion of people who have antibodies to SARS-CoV-2 in the general population, either from vaccination or previous infection, and to see if this changes over time. Residual sera specimens are blood samples that were originally collected for clinical testing and are now due to be discarded.

This report provides information on SARS-CoV-2 seroprevalence in children aged up to 17 years.

Methods

There were three collections of data; 20 December 21 to 20 January 22, 30 May 22 to 10 June 22 and 04 July 22 to 16 July 22.

Paediatric specimens are mainly sourced from Children’s Health Ireland (CHI) at Temple Street, with six samples (from children aged 17) sourced from adult hospitals; Cork University Hospital, Letterkenny University Hospital, St. Vincent’s University Hospital and Tallaght University Hospital. Six sample from Tallaght University Hospital were omitted from analysis as they were collected in February and did not align with the collection dates for paediatric specimens. For paediatric specimens sourced from the clinical chemistry laboratory in Children’s Health Ireland (CHI) at Temple Street, the residual sera specimens were obtained from:

  • CHI at Temple Street Emergency Department, General Paediatric Medical Outpatient Clinic, and Phlebotomy Clinic
  • CHI at Connolly Hospital Urgent Care Centre Blood specimens are only regarded as residual 28 days following completion of the tests ordered.

Antibodies to SARS-CoV-2 were assessed using the Abbott SARS-CoV-2 IgG II Quantitative Assay, which detects antibodies to the receptor-binding domain of the S1 subunit of the spike (S) protein. Results are reported in arbitrary units per millilitre (AU)/ml. Specimens with a result ≥50.0 AU/ml are considered positive or reactive (S+). These were converted to the WHO international standard Binding Antibody Units (BAU)/ml using the formula \[ BAU/ml = AU/ml*0.142\] prior to analysis.

All specimens were then tested on the Elecsys Roche Qualitative Assay which qualitatively detects antibodies to SARS-CoV-2 nucleocapsid protein (anti-N). Vaccines currently approved for use in Ireland target the S protein, and it is not expected that individuals will produce an immunological response to N proteins following vaccination. As such, anti-N reactivity was used to determine antibody responses due to infection, and specimens with a cut-off index of ≥ 1.0 were considered positive (S+ and N+).

Seroprevalence is presented by age group, sex, and target antigen (spike or nucleocapsid). The seroprevalence was adjusted for the misclassification or imperfect sensitivity and specificity in the application of the diagnostic testing using the Rogan Gladen-estimator (see Technical Notes for further details). Unless stated otherwise, adjusted results are presented. Quantitative IgG antibodies in BAU/ml are presented by age group and sex.

Denominator population profiles were taken from CSO 2021 population estimates provided by the HSE National Health Intelligence Unit (‘April 2021/H1’). Data were aggregated into the following age groups for analysis: 0-4 years, 5-12 years, and 13-17 years. As seroprevalence is a core surveillance activity for which the HPSC is legally mandated, no individual patient consent was required.

Results

Demographic characteristics

In all, 449 paediatric residual sera specimens were provided for analysis: 222 females and 227 males. The median age of the children was 6 years, the mean age was 7 years, and approximately half were between the ages of 5 and 12 years.

Table 1: Demographic characteristics of paediatric residual specimens, 20 December 2021 – 16 July 2022
Characteristic Number Percentage
Age median age (years) 6
mean age (years) 7
0-4 171 38.1
5-12 224 49.9
13-17 54 12
Sex F 222 49.4
M 227 50.6

S+ seroprevalence results

For the latest collection period, 04 July 22 to 16 July 22, the adjusted overall seropositivity (S+) rate was 85.7%, (95% CI 75.5, 92.3), (table 2, figure 1). Adjustment had a small effect, increasing estimates of seropositivity by less than 1%. While seropositivity over all time periods was higher for females at 68.2%, (95% CI 61.7, 74.1) than for males 64.5%, (95% CI 57.9, 70.6), the confidence intervals were overlapping.

Table 2: Anti-S seroprevalence (S+) for paediatric residual specimens, 20 December 2021 – 16 July 2022
Characteristic Date Number seropositive Percentage CI lower CI upper
Age group 0-4 05 Jan 22 21 28.8 19.5 40.2
05 Jun 22 58 73.3 62.4 82.0
10 Jul 22 11 61.7 38.8 80.6
5-12 05 Jan 22 44 45.7 36.0 55.7
05 Jun 22 82 87.3 78.8 92.9
10 Jul 22 29 91.7 76.6 97.9
13-17 05 Jun 22 29 91.7 76.6 97.9
10 Jul 22 21 96.6 79.0 100.0
total 05 Jan 22 65 38.4 31.3 46.0
05 Jun 22 169 82.5 76.6 87.3
10 Jul 22 61 85.7 75.5 92.3
Sex F 05 Jan 22 34 40.7 30.7 51.6
05 Jun 22 85 82.6 74.0 89.0
10 Jul 22 31 92.2 77.9 98.1
M 05 Jan 22 31 36.2 26.7 46.9
05 Jun 22 84 82.5 73.7 88.9
10 Jul 22 30 79.8 64.3 90.0

Table note: The rows relating to the latest collection period are marked in bold

Fig. 1: Seroprevalence of SARS-CoV-2 by time

Fig. 1: Seroprevalence of SARS-CoV-2 by time

Overall seroprevalence (S+) for the latest collection period 04 July 22 to 16 July 22 was 61.7%, 91.7% and 96.6% for 0-4 years, 5-12 years, 13-17 years respectively (table 2 and figure 2).

Fig. 2: Overall (S+) seroprevalence by age, by time

Fig. 2: Overall (S+) seroprevalence by age, by time

Quantitative Results

The quantitative antibody levels varied by age-group, with the youngest (0-4 year-olds) having a median of 8 BAU/ml over all the collection cycles, while 5-12 year-olds had a median of 35 BAU/ml. The 13-17 year-olds had a median of 846 BAU/ml, which may reflect higher vaccination levels in this age group. These results are presented over time by collection cycle in figure 3 and table 3.

Fig. 3: Distribution of quantitiative antibody levels by age group

Fig. 3: Distribution of quantitiative antibody levels by age group
Table 3: Quantititive antibodies (BAU/ml) in paediatric residual specimens, 19 December 2021 – 16 July 2022
age (years) date count median 25th percentile 75th percentile arithmetic mean geometric mean minimum maximum
0-4 05 Jan 22 73 1 0.4 48 63 3 0 975
05 Jun 22 80 29 5.2 83 115 21 0 1452
10 Jul 22 18 13 0.9 26 58 9 0 581
5-12 05 Jan 22 97 2 0.5 76 68 6 0 704
05 Jun 22 95 57 11.2 633 588 70 0 5680
10 Jul 22 32 340 71.3 928 898 191 0 5075
13-17 05 Jun 22 32 709 244.8 4149 1939 516 3 5680
10 Jul 22 22 1099 73.3 2332 1495 389 1 5680
all 05 Jan 22 170 2 0.5 67 66 4 0 975
05 Jun 22 207 56 9.4 442 614 60 0 5680
10 Jul 22 72 126 17.2 1135 870 109 0 5680

In the sample that has evidence of previous infection with SARS-CoV-2 (S+ and N+), quantitative antibody levels in the 0-4 year age group were clustered around median values of 206, 51, and 25 BAU/ml across the three collection cycles respectively (figure 4). The 5-12 year olds had slightly higher median antibody levels at 109, 82, and 538 BAU/ml across the collection cycles, and the upper range of results extended further from the median values. The 13-17 year age group median antibody levels were the highest at 956, and 1,189 BAU/ml in June and July, with an approximate bimodal distribution of values.

Fig. 4: Quantitative antibody levels, children with prior infection +/- vaccination (S+ and N+)

Fig. 4: Quantitative antibody levels, children with prior infection +/- vaccination (S+ and N+)

In the sample that has evidence of vaccination only (S+ and N-), median quantitative antibody levels in the 5-12 year age group were 21, 1,031, and 104 BAU/ml across the three collection cycles respectively (figure 5). The median antibody levels for the 13-17 year age group were slightly higher at 723, and 1,074 BAU/ml across the collection cycles. Children aged 0-4 years were ineligible for vaccination during the collection periods.

Fig. 5: Quantitative antibody levels, vaccinated only (S+ and N-)

Fig. 5: Quantitative antibody levels, vaccinated only (S+ and N-)

Infection (S+ and N+) Seroprevalence results

The seroprevalence measurement indicating previous infection (S+ and N+), for the period 04 July 22 to 16 July 22 was 82% for children in both age groups, 5-12 and 13-17 years (table 4 and figure 6). It should be noted that as children aged 0-4 years of age were, in general, ineligible for vaccination during the collection periods, overall seropositivity (S+) results are indicative of previous SARS-CoV-2 infection in this age cohort .

Table 4: Infection +/- vaccination seroprevalence (S+ and N+) for paediatric residual specimens, 19 December 2021 – 16 July 2022
Characteristic date Number seropositive (S+ nad N+) Percentage CI lower CI upper
Age group 0-4 05 Jan 22 18 24.6 16.1 35.7
05 Jun 22 51 64.0 53.0 73.7
10 Jul 22 11 61.3 38.7 80.1
5-12 05 Jan 22 38 39.3 30.1 49.3
05 Jun 22 75 79.3 70.0 86.3
10 Jul 22 26 81.6 64.9 91.6
13-17 05 Jun 22 22 69.0 51.6 82.4
10 Jul 22 18 82.2 61.7 93.1
total 05 Jan 22 56 33.0 26.3 40.4
05 Jun 22 148 71.8 65.3 77.6
10 Jul 22 55 76.7 65.7 85.1

Fig. 6: Infection seroprevalence (S+ and N+), 0-17 years

Fig. 6: Infection seroprevalence (S+ and N+), 0-17 years

Acknowledgements

Thank you to Children’s Health Ireland at Temple Street and the other LSN partners who contribute to the NSP:

  1. The Laboratory Serosurveillance Network Letterkenny University Hospital, Cork University Hospital, University Hospital Limerick, Children’s Health Ireland at Temple Street, Galway University Hospital, Tallaght University Hospital, Beaumont Hospital, St James’s Hospital

  2. UCD National Virus Reference Laboratory Serosurveillance unit Kate Browne, Deirdre Burke, Jeff Connell, Cillian De Gascun

  3. Health Protection Surveillance Centre, Seroepidemiology unit Simon Bergin, Jennifer Doyle, Laurin Grabowsky, Eimear Hayes, Derval Igoe, Soracha McKinley, Gail Melia, Liam Murray, Katie O’Brien

This work is under the governance of the National Serosurveillance Programme (NSP) Steering Committee

For further information on the National Serosurveillance Programme see here

For further information on the Health Protection Surveillance Centre see here

Technical notes

Adjustments for sensitivity and specificity of the test

The seroprevalence was adjusted for the misclassification or imperfect sensitivity (the proportion of true positives that have been correctly identified) and specificity (the proportion of true negatives that have been correctly identified) in the application of the diagnostic testing. If the sensitivity and specificity of a test is known, we can get an approximately unbiased estimate of the true prevalence using the Rogan Gladen-estimator . Thus the adjusted prevalence is estimated by:

\[prev_{adj} = (prev + sp - 1)/(se+sp-1) \]

Where \(prev\) is the unadjusted seroprevalence, \(sp\) is the specificity of the test and \(se\) is the sensitivity of the test. The associated confidence interval is similarly adjusted. The confidence interval is approximate because it assumes that the values of the sensitivity and specificity are known rather than estimated. If they are estimated, then this can be taken into account using the methods of Greiner & Gardner.

The Abbott SARS-CoV-2 IgG II Quantitative Assay has a manufacturer’s stated sensitivity of 98.75% (95% CI 93.25, 99.94) at least 15 days post PCR diagnosis, and a specificity of 99.55% (95% CI 99.15, 99.76).The Roche Elecsys Anti‑SARS‑CoV‑2 assay has a manufacturer’s stated sensitivity of 99.5% (95% CI 97.0, 100.0 ) at least 14 days post PCR positive result, and a clinical specificity of 99.80% (95% CI 99.69, 99.88).

Epidemiological weeks

Dates of epidemiological weeks are available at: https://www.hpsc.ie/notifiablediseases/resources/epidemiologicalweeks/

Further information

The Seroepidemiology of COVID-19 in Ireland data hub can be accessed here.

Links to dashboards of other SARS-CoV-2 serosurveillance programmes are available here.