Non-pharmaceutical interventions for COVID-19 transiently reduced pneumococcal and Haemophilus influenzae carriage in a cross-sectional pediatric cohort in Southampton, UK

Abstract

The Southampton pneumococcal carriage study of children under 5 years old continued during the coronavirus disease 2019 (COVID-19) pandemic. Here, we present data from October 2018 to March 2023 describing prevalence of pneumococci and other pathobionts during the winter seasons before, during, and after the introduction of non-pharmaceutical interventions (NPIs) to prevent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission. Nasopharyngeal swabs were collected from children attending outpatient clinics at a secondary care hospital and community healthcare sites. Pre-NPIs, in 2019/2020, the carriage prevalence of pneumococci at the hospital site was 32% (n = 161 positive/499 participants). During NPIs, this fell to 19% (n = 12/64), although based on fewer participants compared to previous years due to COVID-19 restrictions on health-care attendance. In 2021/2022, after NPIs had eased, prevalence rebounded to 33% (n = 15/46) [compared to NPIs period, χ² (1, N = 110) =2.78, P = 0.09]. Carriage prevalence at community healthcare sites fell significantly from 27% (n = 127/470) in 2019/2020 to 19% during the NPI period (n = 44/228) in 2020/2021 [χ² (1, N = 698) =4.95, P = 0.026]. No rebound was observed in 2021/2022 [19% (n = 56/288)]. However, in a multivariate logistic regression model, neither site had a significantly lower carriage prevalence during the NPI period compared to the post NPI period. A reduction in serotype diversity was observed in 2020/2021. Carriage of Haemophilus influenzae was particularly affected by NPIs with a significant reduction observed. In conclusion, among children under 5 years of age, transient, modest, and statistically non-significant alterations in carriage of both Streptococcus pneumoniae and H. influenzae were associated with SARS-CoV-2 NPIs.IMPORTANCEStreptococcus pneumoniae (the pneumococcus) continues to be a major contributor to global morbidity and mortality. Using our long-running pediatric study, we examined changes in pneumococcal carriage prevalence in nearly 3,000 children under the age of 5 years between the winters of 2018/2019 and 2022/2023. This period coincided with the severe acute respiratory syndrome coronavirus 2 pandemic and, in particular, the implementation of national strategies to limit disease transmission in the UK. We observed a transient reduction of both Streptococcus pneumoniae and Haemophilus influenzae in these populations during this period of non-pharmaceutical interventions. This aligned with the reduction in invasive pneumococcal disease seen in the UK and is therefore a likely contributor to this phenomenon.

Keywords: COVID-19; Haemophilus influenzae; SARS-CoV2; Streptococcus pneumoniae; carriage.

Fig 1

Box and whisker plot showing the comparison of child age between sites (A) and density plot showing distribution across age for each site by recruitment year (B). Sites are colored: orange (Site 1, hospital) and blue (Site 2, community clinics). Boxes show the interquartile range with whiskers denoting minimum and maximum values. Outliers are shown as black points. Individual data points are shown in gray. Children recruited at Site 1 between 2017/2018 and 2022/2023 were significantly (P < 0.001) older than those at Site 2.

Fig 2

Carriage prevalence of S. pneumoniae shown for each year of the study and split by Site 2 (hospital, orange) and Site 2 (community clinics, blue). The dashed lines highlight the pre-NPI and post-NPI periods. Only one period of recruitment was done within the period of NPI lockdowns in the UK, 2020/2021. Carriage prevalence at Site 2 decreased significantly from 27% (n = 127/470) in 2019/2020 to 19% [n = 44/228) in 2020/2021 (χ² (1, N = 698) =4.95, P = 0.026)]. No immediate rebound was observed in 2021/2022, but by 2022/2023, carriage was closer to pre-NPIs. Error bars show 95% CI.

Fig 3

Pneumococcal serotypes isolated (n = 35) shown as a per-year proportion (A) with individual counts for each serotype split by year (B). Data presented are based on a subsample of isolates from each year (n = 50) to allow for meaningful comparisons between pre-NPI years and during-NPI periods when both recruitment and pneumococcal carriage was lower. Only three vaccine-type serotypes were isolated (19F, 3, and 19A) and were isolated each year apart from 2020/2021 when 19A was not observed. Fewer serotypes were seen during the NPI period (n = 19) compared to both pre- and post-NPI periods.

Fig 4

Carriage prevalence of H. influenzae (A), S. aureus (B), M. catarrhalis (C), and non-pneumococci alpha-hemolytic Streptococci (D). Carriage for each pathobiont in each year is shown split by Site 1 (hospital, orange) and Site 2 (community clinics, blue). Error bars show 95% CI. The only significant decrease in carriage was observed for H. influenzae during the NPI period (2020/021) where carriage decreased to 3.1% (n = 2/64) and 5.3% (n = 12/228) for Sites 1 and 2, respectively.