Impact of pneumococcal conjugate vaccine 13 introduction on severe lower respiratory tract infections associated with respiratory syncytial virus or influenza virus in hospitalized children in Ulaanbaatar, Mongolia

Abstract

Objectives: Limited data indicate a beneficial effect of pneumococcal conjugate vaccines (PCVs) on respiratory syncytial virus (RSV) and influenza infections in young children. We evaluated the impact of 13-valent PCV (PCV13) introduction on the incidence of severe lower respiratory tract infections (LRTIs) associated with RSV or influenza in hospitalized children.

Methods: Our study was restricted to children aged <2 years with arterial oxygen saturation <93% and children with radiologically confirmed pneumonia nested in a pneumonia surveillance project in four districts of Ulaanbaatar city, Mongolia. We tested nasopharyngeal swabs collected on admission for RSV and influenza using quantitative reverse transcription-polymerase chain reaction. The impact of PCV13 on the incidence of LRTI outcomes associated with RSV or with influenza for the period April 2015-March 2020 was estimated. Incidence rate ratios comparing pre- and post-vaccine periods were estimated for each outcome for each district using negative binomial models and for all districts combined with a mixed-effects negative binomial model. Adjusted models accounted for seasonality. Sensitivity analyses were conducted to assess the robustness of our findings.

Results: Among 5577 tested cases, the adjusted incidence rate ratios showed a trend toward a reduction in RSV-associated outcomes: all LRTIs (0.77, 95% confidence interval [CI] 0.44-1.36), severe LRTIs (0.88, 95% CI 0.48-1.62), very severe LRTIs (0.76, 95% CI 0.42-1.38), and radiologically confirmed pneumonia (0.66, 95% CI 0.32-1.38) but inconsistent trends in outcomes associated with influenza.

Conclusions: No significant reductions were observed in any outcomes associated with RSV and influenza after PCV introduction.

Keywords: Influenza virus; Lower respiratory tract infections; Pneumococcal conjugate vaccine; Respiratory syncytial virus.

Figure 1

IRRs of outcomes associated with RSV infection: LRTIs (a), severe LRTIs (b), very severe LRTIs (c) and X-ray confirmed pneumonia (d) comparing pre- and post- PCV13 introduction. Comparing pre- and post- PCV13 introduction. Model 1 (M1- unadjusted for seasonality): - For each district SBD, SKD, BZ and CHD (district specific, unadjusted model): crude IRRs with each outcome were calculated using monthly count data and negative binomial regression models. The natural logarithm of the population denominators was included as an offset and a dummy variable was included to represent PCV13 introduction period in these models. - For overall estimates of all four districts (all overall districts, unadjusted model): unadjusted IRRs with each outcome were calculated by using mixed effects negative binomial regression models with a fixed effect for PCV13 introduction variable and a random effect for district. Model 2 (M2- adjusted for seasonality) - For each district SBD, SKD, BZ and CHD (district specific, adjusted model for seasonality): adjusted IRRs with each outcome were calculated for individual district using a negative binomial model that included chronological time period as a continuous variable to adjust for underlying secular trends and calendar month as categorical variable to adjust for seasonality. - For overall estimates of all four districts (All overall districts, adjusted model): adjusted IRRs with each outcome were calculated by using mixed effects negative binomial regression models with a fixed effect for PCV13 introduction variable, a random effect for district and additional fixed effect variables for chronological time period as a continuous variable to adjust for underlying secular trends and calendar month as categorical variable to adjust for seasonality. BZ, Bayanzurkh; CHD, Chingeltei; IRR, incidence rate ratio; LRTI, lower respiratory tract infections; PCV, pneumococcal conjugate vaccine; RSV, respiratory syncytial virus; SBD,Sukhbaatar; SKD, Songinokhairkhan.

Figure 2

IRRs of outcomes associated with influenza infection: LRTIs (a), severe LRTIs (b), very severe LRTIs (c) and X-ray confirmed pneumonia (d) comparing pre- and post- PCV13 introduction. Model 1 (M1-unadjusted for seasonality): - For each district SBD, SKD, BZ and CHD (district specific, unadjusted model): crude IRRs with each outcome were calculated using monthly count data and negative binomial regression models. The natural logarithm of the population denominators was included as an offset and a dummy variable was included to represent PCV13 introduction period in these models. - For overall estimates of all four districts (All overall districts, unadjusted model): unadjusted IRRs with each outcome were calculated by using mixed effects negative binomial regression models with a fixed effect for PCV13 introduction variable and a random effect for district. Model 2 (M2- adjusted for seasonality) - For each district SBD, SKD, BZ and CHD (district specific, adjusted model for seasonality): adjusted IRRs with each outcome were calculated for individual district using a negative binomial model that included chronological time period as a continuous variable to adjust for underlying secular trends and calendar month as categorical variable to adjust for seasonality. - For overall estimates of all four districts (all overall districts, adjusted model): adjusted IRRs with each outcome were calculated by using mixed effects negative binomial regression models with a fixed effect for PCV13 introduction variable, a random effect for district and additional fixed effect variables for chronological time period as a continuous variable to adjust for underlying secular trends and calendar month as categorical variable to adjust for seasonality. BZ, Bayanzurkh; CHD, Chingeltei; IRR, incidence rate ratio; LRTI, lower respiratory tract infections; PCV, pneumococcal conjugate vaccine; RSV, respiratory syncytial virus; SBD, Sukhbaataar; SKD, Songinokhairkhan.