Routine immunization against Streptococcus pneumoniae and Haemophilus influenzae type B and antibiotic consumption in India: a dynamic modeling analysis

Abstract

Background: Childhood vaccinations can reduce disease burden and associated antibiotic use, in turn reducing the risk of antimicrobial resistance (AMR). We retrospectively estimated the population-level reductions in antibiotic use in India following the introduction of vaccines against Streptococcus pneumoniae and Haemophilius influenzae type B in the national immunization program for children in the mid-2010s and projected future gains to 2028 if vaccination coverage were to be increased.

Methods: Using IndiaSim, a dynamic agent-based microsimulation model (ABM) for India, we simulated the spread of Streptococcus pneumoniae and Haemophilius influenzae type B (Hib) among children to estimate reductions in antibiotic use under the scenarios of: (i) pneumococcal and Hib vaccine coverage levels equivalent to the national coverage of pentavalent diphtheria-pertussis-tetanus third dose (DPT3) compared to a baseline of no vaccination, and (ii) near-universal (90%) coverage of the vaccines compared to pre-COVID national DPT3-level coverage. Model parameters, including national DPT3 coverage rates, were based on data from the National Family Household Survey 2015-2016 and other published sources. We quantified reductions in antibiotic consumption nationally and by state and wealth quintiles.

Findings: We estimate that coverage of S. pneumoniae and Hib vaccines at the same level as DPT3 in India would translate to a 61.4% [95% UI: 43.8-69.5] reduction in attributable antibiotic use compared to a baseline of zero vaccination coverage. Increases in childhood vaccination coverage between 2004 and 2016 have likely reduced attributable antibiotic demand by as much as 93.4% among the poorest quintile. Increasing vaccination coverage by an additional 11 percentage points from 2016 levels results in mortality and antibiotic use across wealth quintiles becoming increasingly similar (p < 0.05), reducing in health inquities. We project that near-universal vaccine coverage would further reduce inequities in antibiotic demand and may eliminate of outbreak-associated antibiotic use from S. pneumoniae and Hib.

Interpretation: Though vaccination has a complex relationship with antibiotic use because both are modulated by socioeconomic factors, increasing vaccinations for S. pneumoniae and Hib may have a significant impact on reducing antibiotic use and improving health outcomes among the poorest individuals.

Funding: The Bill & Melinda Gates Foundation (grant numbers OPP1158136 and OPP1190803).

Keywords: Agent-based modeling; Antibiotic use; Antimicrobial resistance; Haemophilus influenzae type b; Hib; LMICs; PCV; Streptococcus pneumoniae; Vaccinations.

Fig. 1

PCV13 and Hib-containing-pentavalent vaccination have averted antibiotic use. Fully calibrated whole-India agent-based model run from 2004 to 2016 to determine health and antibiotic use impacts of PCV13 and Hib-containing-pentavalent vaccination. Two scenarios are compared: no vaccinations (left), and vaccinations as they increased from first implementation in 2004 to integration into universal immunization program by 2016 (right). Symptomatic infection curves showing per-person incidence rate (i.e., the rate that a given child may become sick by S. pneumoniae or Hib) by wealth quintile for a scenario with no vaccinations (a) and increasing vaccinations from zero to 2016 coverage levels (b) by wealth quintile. Proportion of total daily defined doses (DDDs) of antibiotics in India used for S. pneumoniae and Hib for a scenario with no vaccinations (c) and increasing vaccinations to 2016 coverage levels (d). The red line marks the point at which the modeled equilibration period ends.

Fig. 2

PCV13 and Hib-containing-pentavalent vaccination have improved health equity and equitable distribution of antibiotics. Fully calibrated whole-India agent-based model run from 2004 to 2016 to determine health and antibiotic use impacts of PCV13 and Hib-containing-pentavalent vaccination. Two scenarios are compared: no vaccinations (left), and vaccinations as they increased from first implementation in 2004 to integration into universal immunization program by 2016 (right). Cumulative mortality curves (total proportion of wealth quintile that has succumbed to S. pneumoniae or Hib) by wealth quintile for a scenario with no vaccinations (a) and increasing vaccinations to 2016 coverage levels (b) by wealth quintile. Proportion of infected agents receiving antibiotics for S. pneumoniae and Hib for a scenario with no vaccinations (c) and increasing vaccinations to 2016 coverage levels (d). The red line marks the point at which the equilibration period ends.

Fig. 3

Increasing vaccines to near universal coverage (90%) has health equity benefits. Fully calibrated whole-India agent-based model run from 2016 to 2028 to determine health and antibiotic use impacts of increasing PCV13 and Hib-containing-pentavalent vaccination to near-universal levels. (a) Cumulative mortality curves by wealth quintile (total proportion of wealth quintile that has succumbed to S. pneumoniae or Hib) under a scenario of increased vaccination. (b) Proportion of total India daily defined doses (DDDs) of antibiotics used to treat S. pneumoniae and Hib for the same scenario as in (a). (c) Proportion of infected individuals receiving antibiotics by wealth quintile for the same scenario as in (a).

Fig. 4

Geographic impacts of increasing vaccination to 90%. Fully calibrated whole-India agent-based model run from 2016 to 2028 to determine health and antibiotic use impacts of increasing PCV13 and Hib-containing-pentavalent vaccination to near-universal levels. (a) Steady-state disease prevalence from S. pneumoniae and Hib with 90% vaccination. (b) Reductions in antibiotic use from increasing vaccinations under steady-state prevalence and disease dynamics.