Effect of Pneumococcal Conjugate Vaccine on Pneumonia Incidence Rates among Children 2–59 Months of Age, Mongolia, 2015–2021

Starting in June 2016, the 13-valent pneumococcal conjugate vaccine (PCV13) was introduced into the routine immunization program of Mongolia by using a 2+1 dosing schedule, phased by district. We used prospective hospital surveillance to evaluate the vaccine’s effect on pneumonia incidence rates among children 2–59 months of age over a 6-year period. Of 17,607 children with pneumonia, overall adjusted incidence rate ratios showed decreased primary endpoint pneumonia, very severe pneumonia, and probable pneumococcal pneumonia until June 2021. Results excluding and including the COVID-19 pandemic period were similar. Pneumonia declined in 3 districts that introduced PCV13 with catch-up campaigns but not in the 1 district that did not. After PCV13 introduction, vaccine-type pneumococcal carriage prevalence decreased by 44% and nonvaccine-type carriage increased by 49%. After PCV13 introduction in Mongolia, the incidence of more specific pneumonia endpoints declined in children 2–59 months of age; additional benefits were conferred by catch-up campaigns.

(Bayanzurkh) was privatised (2).Other private hospitals were not included in the surveillance programme as nearly all children are treated in the public sector for pneumonia.Hospitalisations for children are fully funded by the government (2,3).
Two standardised questionnaires collected information on demographic variables, presenting symptoms and signs, previous medication, immunisation history, treatment received, and risk factors.Blood samples, nasopharyngeal swabs and chest x-rays were collected for all enrolled cases who consented.Dedicated study staff monitored patient enrolment by clinical hospital staff to ensure that no eligible patients were missed.Participants who were missed by clinical staff were enrolled retrospectively.If participants were enrolled more than 72 hours after admission nasopharyngeal swabs were not collected (1).

Case definitions and study outcomes
The enrolment case definition for clinical pneumonia and the specific pneumonia endpoints (study outcomes) are detailed below: 1) All clinical pneumonia surveillance case definition (1) Cough or difficulty breathing, with one of the following: • an elevated respiratory rate (≥50 bpm for all ages) • oxygen saturation <90% • a clinical diagnosis of severe pneumonia 2) WHO-defined primary endpoint pneumonia (4): • End-point consolidation (dense or fluffy opacity that occupies a portion or whole of a lobe or the entire lung that may or may not contain air bronchograms) OR • Pleural effusion that is in the lateral pleural space and associated with pulmonary parenchymal infiltrate or if the effusion obliterated enough of the hemithorax to obscure an opacity.Pneumonia with a positive blood or pleural fluid culture.

Sample collection and laboratory procedures
We adhered to the WHO recommended methods for nasopharyngeal sample collection, handling and transport (8).A flocked, nylon swab was placed in 1 ml skim milk tryptone glucose glycerol media (STGG) immediately following collection.Swabs were stored in a fridge and transported to the National Center for Communicable Diseases where they were aliquoted and stored at ultra-low temperature within 8 hours of collection.Samples were shipped to the Murdoch Children's Research Institute (Parkville, Australia) on dry ice and stored at ultra-low temperature until testing.Nasopharyngeal swabs were tested for pneumococci using lytA realtime quantitative PCR (qPCR) and samples that were lytA qPCR positive (Ct value < 35) or equivocal (Ct value 35-40) were cultured on horse blood agar containing 5 µg/ml of gentamicin (Oxoid) (9).DNA was extracted from the harvested α-haemolytic growth (10) followed by molecular serotyping by DNA microarray as previously described (11; C. von Mollendorf, unpub.data, https://doi.org/10.2139/ssrn.4488943).Microarray was performed using Senti-SPv1.5 microarrays (BUGS Bioscience) and analysed using Senti-NET, a custom web-based software (BUGS Bioscience).A total of 1000 cases per year were tested for pneumococci, including all cases with PEP (as this was the primary objective), and a random sample of remaining severe and non-severe cases.

Statistical analysis
To control for seasonal and long-term patterns we included an indicator variable for each elapsed calendar month (time elapsed) over the study period in the main model.We also explored three other options: fitting a spline function of time, Fourier terms and calendar month with a continuous time variable.The time elapsed variable was selected to control for seasonality as it resulted in improved model fit as measured by the Akaike's Information Criterion (AIC).
No indicator variables were used to adjust for the impacts of the COVID-19 pandemic, as schools were closed from February 2020 to the end of the surveillance period with no significant reopening.We controlled for the impact of the COVID-19 pandemic by restricting to the prepandemic period (April 2015-Feb 2020) and then comparing results from the restricted model to a model including the total period (April 2015-June 2021).Model fit for all final models were evaluated using the AIC.

Table 1 .
Characteristics of 17,607 children 2-59 mo of age enrolled in pneumonia surveillance project from 4 study districts in Ulaanbaatar, Mongolia, April 2015 -June 2021 000₮ per person/per month.¶Children were considered PCV13 vaccinated if they have received at least two doses when administered at less than 12 mo of age or at least one dose when administered at greater than or equal to 12 mo of age.##Number of children who died during hospital stay.**Severe pneumonia defined according to WHO integrated management of childhood illness 2005 case definition.^Very severe pneumonia included severe cases complicated by empyema, intensive care unit admission, persistent severe disease post-discharge, hypoxia or death.^^Probable pneumococcal pneumonia was defined as elevated C-reactive protein with either PEP or high pneumococcal nasopharyngeal carriage (either high density carriage of any serotype greater than 1 × log10 GE/mL, or any carriage of serotypes 1 or 5).

Table 2 .
Characteristics of children 2-59 months of age hospitalised with clinical pneumonia in pre-and post-PCV period, April 2015 to June 2021 Hypoxic defined as an oxygen saturation <90%.**WHO defined primary end point pneumonia.***Severe pneumonia defined according to WHO integrated management of childhood illness 2005 case definition.^Very severe pneumonia included severe cases complicated by empyema, intensive care unit admission, persistent severe disease post-discharge.hypoxia or death.^^Probable pneumococcal pneumonia was defined as elevated C-reactive protein with either PEP or high pneumococcal nasopharyngeal carriage (either high density carriage of any serotype greater than 1 × log10 GE/mL, or any carriage of serotypes 1 or 5). ||

Table 5 .
Crude incidence rates by year for hospitalised clinical pneumonia by district and diagnosis for children aged 2-59 months, April 2015 to June 2021 WHO defined primary end point pneumonia.‡Severepneumonia defined according to WHO integrated management of childhood illness 2005 case definition.§Very severe pneumonia included severe cases complicated by empyema, intensive care unit admission, persistent severe disease post-discharge, hypoxia or death.¶Hypoxic pneumonia defined as an oxygen saturation <90%.|| Probable pneumococcal pneumonia included PEP or high pneumococcal nasopharyngeal carriage with a C-reactive protein ≥40 mg/dL. *