Streptococcus pneumoniae (S. pneumoniae), which can cause invasive and non-invasive diseases, is the main bacterial pathogen for community–acquired pneumonia (CAP) and mucosal diseases such as otitis media (1). S. pneumoniae frequently colonizes the human nasopharynx and is transmitted through respiratory droplets. The cross-sectional point prevalence of nasopharyngeal (NP) carriage in infants and young children, who are the main reservoir of this organism, ranges from 27% to 85% (2). NP carriage rates are particularly high among children in low- and middle-income countries (LMICs), as well as among children in high-income countries who belong to indigenous populations (1, 3). In China, there are limited data regarding S. pneumoniae carriage in healthy children; however, previous studies, based on pediatric inpatients with pneumococcal infections, indicated geographical differences in serotype distribution between northern and southern China (4). Therefore, the present cross-sectional study in China, designed to investigate the S. pneumoniae NP carriage rate, serotype distribution, and antimicrobial susceptibility patterns among healthy children aged 30 days to <60 months, involved one northern city and one southern city.

Until the 13-valent pneumococcal conjugate vaccine (PCV13) licensure in China in October 2016, PCV7 was available in the private market for children less than 5 years of age, although uptake was <10% (5). Likewise, PCV13 has been used in the private market, and it is not included in the national immunization program (NIP). PCV13 administration in children in China is based on the principle of “informed consent and voluntary payment”: from 2016, it was only for children less than 15 months of age (6); after 2023, the indication was expanded to the age of 5 years (7). Recent data show that the vaccination rate of PCV13 was increasing year by year in economically developed cities. For example, in Huangpu District in Shanghai, a study reported that the PCV13-Pfizer (full series) uptake increased from 2016 to 2019, ranging from 1.9% (2016), 15.6% (2017), 33.3% (2018), and 42.7% (2019) (8).

Currently, there are 21.5 million people living in Beijing, of whom about 13.5 million are classified as “local residents.” The annual birth cohort is approximately 170 thousand (9). In Beijing, the uptake of non-national immunization program (non-NIP) vaccines in the private sector, including PCV, varies significantly between years and districts. Data from one survey showed that non-NIP vaccine uptake ranged from 3.2% (influenza) to 95.8% (varicella) among children born between 2001 and 2006 (10). Another survey reported that the uptake of PCV7 in Beijing was approximately 18% among children born between 2011 and 2013 (11).

Shenzhen is a developed city with a population of 11 million, of whom 3.5 million are “local residents.” The annual birth cohort is approximately 71 thousand (12). The uptake for non-NIP vaccines is greater in Shenzhen than in other cities. For example, within one Shenzhen district, pneumococcal vaccine uptake [including PCV7 or the 23-valent pneumococcal polysaccharide vaccine (PPSV23)] among children younger than 7 years of age was estimated at 51% (13).

The present study was performed from January 2018 to November 2021, which covered the early stage of PCV13 availability for children <15 months of age, and the study extended into the first 2 years of the COVID-19 pandemic. From February 2020, measures were enforced which included mandatory self-isolation for individuals presenting with respiratory symptoms or fever, requirements for physical distancing maintaining a separation of over 1.5 m, and restrictions limiting the size of social gatherings.

In this study conducted among Chinese children 30 days to <60 months of age, overall S. pneumoniae NP carriage was 13.1%. These results were similar to the carriage prevalence reported from children younger than 5 years in Taiwan (12.0%, 60/500) (21), but lower than the NP carriage observed from other studies in Ghana (32.6%, 63/193) (22), Ethiopia (43.8%, 311/710) (23), or Turkey (17.8%, 103/580) (24). In a previous meta-analysis performed in China prior to the introduction of PCV7 that included children younger than 5 years of age, NP carriage was 24.4% (25). In an earlier study conducted between 2012 and 2014 by the Beijing Center for Disease Control and Prevention (CDC) among unvaccinated children 2–5 years old, it was reported that among 3,281 children aged 3.1 (±0.8) years, pneumococcal NP carriage was detected in 22% of the samples (25, 26). Potential explanations for the lower NP pneumococcal carriage observed among Chinese children in the current study could be related to: the increasing uptake in PCV13 among Chinese children (27); the fact that the study was conducted during COVID-19 pandemic with all the implemented recommendations in China to prevent viral respiratory transmission (28, 29); and the low number of children per family and the custom that children under 3 years of age are cared for at home generally by grandparents or nannies instead of at day care centers, as the entry for public day nursery is 3 years of age (30, 31).

We found that vaccination with PCV13 reduced pneumococcal carriage. While an indirect protection with PCV13 against PCV13 serotypes is well-established when PCV13 is included in the national immunization program, the results of the current study also suggest that even in the setting of private market uptake, there may have been an indirect benefit of carriage. As reported previously (32), we found that vaccination with PPSV23 did not influence NP carriage, reflecting an inherent limitation of plain polysaccharide vaccine.

Dominant serotypes differed slightly with PCV13 vaccination history: the top serotypes among unvaccinated children were 19F, 23F, 6B, and 6C, whereas serotypes 15A, 6B, 19F, 15A, 23A, and 6C were the most frequent serotypes among children partially vaccinated with PCV7 (3 children) or PCV13; and 23F, 6A, 15C, 15A, and 23A were the most frequent serotypes observed among children fully vaccinated with PCV13. The association between PCV vaccination and lower vaccine serotype NP carriage has been previously described (33). In Greece, for instance, where PCVs have been in the NIP since 2006, initially with PCV7 and since 2010 with PCV13 at an estimated four-dose uptake rate of about 80%, molecular surveillance of pneumococcal carriage following completion of immunization with the PCV13 (3 + 1 schedule) revealed that non-PCV13 serotypes represented 83.8% of total isolates. Serotypes 19A and 3 were the only two PCV13 serotypes that increased in proportion over the time of the study (P < 0.001 and P = 0.012, respectively) (33).

The percentage of non-serotypeable S. pneumoniae (20.7%) isolates observed in our study is consistent with results from studies conducted in other geographical locations (34–36), which helps to support the representativeness of the samples and the validity of the research results.

In the present study, children enrolled in Beijing had lower S. pneumoniae NP carriage compared with children enrolled in Shenzhen, which became most apparent in the 2020–2021 COVID-19 period. As noted, the carriage rate of S. pneumoniae in Beijing decreased significantly in the post-COVID period, but the carriage rate in Shenzhen did not. The decreasing carriage rate of S. pneumoniae in children in Beijing during the COVID-19 pandemic could be related to the continuous and strict prevention and control measures (such as universal masks, distancing, and hand washing). For instance, in Beijing closed kindergartens from February 2020 until the end of this study while Shenzhen did not. These measures contributed to limiting the transmission of COVID-19, but also reduced the spread of other pathogens. From January 2020 to February 2020, the spread of the epidemic in China had been first curbed. Most of the declared infectious disease incidences showed a downward trend in 2020 in China (37). Other settings outside of China did not observe such a COVID-19 pandemic effect. For instance, in a prospective cohort study in Israel, the mean proportion of children younger than 3 years old carrying pneumococcus, which was 44.3 ± 1.9% during 2016–2019, was somewhat reduced in October–December 2020, although the rates during January–February 2021 were not significantly different from those in the pre-COVID period (38). Nonetheless, unexamined interregional factors (e.g., climate, population characteristics, and economic levels) might have contributed to the observed differences in colonization in China during the COVID-19 pandemic and modified S. pneumoniae transmission.

To our knowledge, this present study (2018–2021) is the largest pneumococcal carriage study conducted in healthy children in China. Serotype 19F (11.2%) was the most prevalent S. pneumoniae carriage serotype, followed by 6B (7.0%), 23F (6.4%), 6C (5.3%), and 23A (4.8%). A meta-analysis among Chinese children in the post-PCV7 era (from October 2008 to 2016) reported that serotypes 19F, 6A, and 23F were most often isolated (24). A systematic review and meta-analysis of studies of carriage of S. pneumoniae and other respiratory bacterial pathogens in LMICs reported that serotypes 6A, 6B, 19A, 19F, and 23F were the serotypes most often isolated (35). In our study, the top five circulating serotypes included two non-PCV13 serotypes (6C and 23A). Researchers from Germany (39), Canada (40), Gambia (41) and Israel (42) suggested that increases in the carriage of non-PCV13 serotypes had offset reductions in the carriage prevalence of vaccine-targeted serotypes. Furthermore, the present study focuses on the children living in only two Chinese cities, where 47.7% of children had received at least one dose of PCV13. The present results demonstrated that 63.3% (408/645) of isolates obtained were non-PCV13 serotypes (e.g., 6C, 23A, 15A, 15C, 34, 16F, and 15B). Between 2014 and 2019, Xu et al. (43) reported that 12.5% (2/16) of fatal IPDs were caused by non-PCV13 serotypes (15A and 15B) that were not identified before the introduction of PCV13. Therefore, it is particularly important to monitor the serotype of S. pneumoniae in the nasopharynx of healthy children.

The highest antimicrobial resistance rates were observed for erythromycin (96.6%) or trimethoprim/sulfamethoxazole (68.2%), while resistance to penicillin (based on the breakpoints for non-meningitis isolates) and other beta-lactam agents was low (<5%). The MIC50 and MIC90 against erythromycin was >256 mg/L, which could reflect macrolide overuse in China (44). The high macrolide resistance is consistent with data reported in other Asian countries (45).

A total of 158 (24.5%) isolates were MDR. Vaccine-serotype pneumococcal isolates possessed more MDR than non-vaccine serotypes, which emphasizes the role for PCV immunization in the control of pneumococcal disease associated with antibiotic resistance. National vaccination using PCVs prevent episodes of both S. pneumoniae diseases and carriage, the latter which can inhibit the spread of antibiotic resistance (46). As non-PCV13 serotypes such as 11A and 24F are associated with antimicrobial resistance, further surveillance is needed (47). Of note, the MDR isolates of non-vaccine serotypes in this study increased year by year during the study period. We found that the rate of MDR non-PCV13 serotypes (including NST) in 2018 was 0.0%, followed by an increase towards 9.6% in 2019, 18.1% in 2020, and 18.7% in 2021. These MDR non-PCV13 strains can evade, at the same time, the pressures of vaccine-induced immunity and antimicrobial selection. The potential relationship between antimicrobial selective pressure and serotype replacement is a reminder that antimicrobials should be used properly to ensure that PCVs remains a powerful tool in the fight against antimicrobial resistance in S. pneumoniae.

One limitation of the present study is that the study was conducted in only in two cities with strong economies and wide availability of medical care, which does not reflect the diversity of China. Further investigations could provide more information about S. pneumoniae NP carriage. Other limitations of this study were a relatively modest pre-NIPs sample size. To compensate, we drew upon the pneumococcal NP swab carriage study conducted previous as an additional reference point (25, 26).

This investigation revealed that one-seventh or one-eighth of children <60 months of age carried S. pneumoniae in their nasopharynx in Beijing and Shenzhen, and like previous studies, S. pneumoniae carriage prevalence increased for subjects of older age (25 to <60 months), those living with siblings, attending day nursery, or who were unvaccinated with PCV13. The strict administration of measures to control COVID-19 led to a decrease in S. pneumoniae carriage rates, which was particularly evident in Beijing. Although PCV immunization uptake was limited, it showed a preventive effect (individual protection) on vaccine serotype S. pneumoniae carriage among vaccinated healthy young children in the community, providing evidence that control of S. pneumoniae carriage through vaccination could have a positive effect on antibiotic resistance.